Professional Documents
Culture Documents
Digital Finance Study Guide - 1st Edition (With Reprints) Semester 2, 2022
Digital Finance Study Guide - 1st Edition (With Reprints) Semester 2, 2022
CPA PROGRAM
DIGITAL FINANCE
FIRST EDITION
“FMPrelims_PrintPDF” — 2021/10/18 — 11:36 — page ii — #2
ACKNOWLEDGEMENTS
MODULE 1
Figure 1.1: © Gomber, P., Koch, J-A. & Siering, M., 2017. Digital Finance and FinTech: current
research and future research directions. Journal of Business Economics, 87. 10.1007/s11573-017-0852-x;
Figure 1.4: © Courbe, J, 2016, ‘Financial services technology 2020 and beyond: embracing disruption’,
PWC, p. 33, https://www.pwc.com/gx/en/financial-services/assets/pdf/technology2020-and-beyond.pdf
© 2020 PwC. All rights reserved; Figure 1.8: © ITU; Extracts: © Gomber, P., Koch, J-A & Siering, M,
Digital Finance and FinTech: current research and future research directions. Journal of Business
Economics, 87. 10.1007/s11573-017-0852-x; Examples 1.3, 1.5 and 1.8: © Airwallex 2020.
MODULE 2
Figure 2.13 and case study 2.1: © Little Phil; Table 2.30: © Deloitte 2018, ‘SME digital payments:
New opportunities to optimise’, The Paytech Revolution Series, Deloitte, https://b2bpay.com.au/
wp-content/uploads/2018/03/Deloitte_Payments-Landscape-Report-2018.pdf.
CPA Australia wishes to acknowledge the team from Swinburne University of Technology for authorship
and project management of ‘Future of Money’.
MODULE 3
Figures 3.3 and 3.5: © Kinetic Consulting; Example 3.8: © RapidMotion; Example 3.17: © The YGS
Group.
CPA Australia wishes to acknowledge the team from Swinburne University of Technology for additional
content authorship of ‘Technology and its use in Finance’.
CPA Australia wishes to acknowledge the team from RMIT Online for authorship and project manage-
ment of ‘Technology and its use in Finance’. Acknowledgement and special thanks are extended to:
• Amy Wilson from RMIT Online for project management
• Trinity Ryan, Ross Palmer and Tricia Di Giambattista from The Learning Hook for learning design
and authorship
• David Thurkle at IBM for authorship and subject matter expertise.
MODULE 4
Figure 4.10: © Commonwealth of Australia; Figure 4.37: © Bernard Marr; Figure 4.42: © Common-
wealth Bank of Australia 2020, Annual report, pp. 106, 108. https://www.commbank.com.au/content/
dam/commbank/about-us/shareholders/pdfs/results/fy20/cba-2020-annual-report-print.pdf; Figure 4.43:
© David Elliott, It’s a fact… scientists are the most trusted people in the world, 18 September 2019,
https://www.ipsos.com/sites/default/files/ct/news/documents/2019-09/global_trustworthiness_2019-aust-
ralia.pdf; Example 4.10: © iTnews.com.au; Example 4.12: © Commonwealth of Australia.
CPA Australia wishes to acknowledge the team from Swinburne University of Technology for additional
content authorship of ‘Data Analytics, Interpretation and Visualisation’.
MODULE 5
Figure 5.21: © RegTech; Extract: The Treasury © Commonwealth of Australia 2020.
CPA Australia wishes to acknowledge the team from Swinburne University of Technology for additional
content authorship of ‘Risk Management, Governance and Regulation’.
CPA Australia wishes to acknowledge the teams from RMIT Online for authorship and project
management of Risk Management, Governance and Regulation. Acknowledgement and special thanks
are extended to:
• Amy Wilson from RMIT Online for project management
• Trinity Ryan, Ross Palmer and Tricia Di Giambattista from The Learning Hook for learning design
and authorship
• Adam Stevenson from Airwallex for authorship and subject matter expertise
• Yvonne Sears from ISDefence for course mapping.
ACKNOWLEDGMENTS iii
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page iv — #4
These materials have been designed and prepared for the purpose of individual study and should not be
used as a substitute for professional advice. The materials are not, and are not intended to be, professional
advice. The materials may be updated and amended from time to time. Care has been taken in compiling
these materials, but they may not reflect the most recent developments and have been compiled to give
a general overview only. CPA Australia Ltd and John Wiley and Sons Australia Ltd and the author(s) of
the material expressly exclude themselves from any contractual, tortious or any other form of liability on
whatever basis to any person, whether a participant in this subject or not, for any loss or damage sustained
or for any consequence that may be thought to arise either directly or indirectly from reliance on statements
made in these materials.
Any opinions expressed in the study materials for this subject are those of the author(s) and
not necessarily those of their affiliated organisations, CPA Australia Ltd or its members.
As the supplier of third-party study guide materials to the publisher, CPA Australia is responsible for
the use of any materials for which the intellectual property is owned or controlled by a third party.
iv ACKNOWLEDGMENTS
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page v — #5
BRIEF CONTENTS
Subject Outline xii
Glossary 471
Suggested Answers 483
Index 503
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page vi — #6
CONTENTS
Subject Outline xii 1.8 Current use of FinTech by accounting
and finance professionals 40
MODULE 1 FinTech innovations in retail 40
Innovation incentives in the financial
The Digital Finance services sector 42
Ecosystem 1 1.9 Innovation thinking in FinTech 43
Preview 1 Wealth management 43
Part A: Digital Finance Landscape 2 Payments 44
Introduction 2 Financial markets 45
1.1 Technological evolution 2 Emerging FinTech 48
The network age 2 Innovation roadblocks 49
The transformation of accounting Innovation opportunities 49
and finance 3 Summary 50
Creating a future-ready finance function 4 Part D: Digital Transformation 51
1.2 Impact of technology on the digital 1.10 Organisational strategy 51
finance landscape 5 Strategic management processes 52
Analysing the digital finance landscape Strategic management environments
using the digital finance cube 5 and goals 52
Analysing drivers of innovation using the ten 1.11 Digital transformation strategies 52
types framework 11 What is a digital transformation
1.3 Impact of technology on banking/ strategy? 53
financial markets 14 Digital transformation methodology 53
How funds flow through the traditional Digital transformation in action 55
financial system 15 Value drivers 57
A changing landscape 15 1.12 How digital transformation strategy
A new model of innovation — working can help achieve an organisation’s
together 16 goals 58
Open banking and open marketplaces 16 Benefits of digital transformation 58
Summary 18 Why most companies struggle
Part B: Digital Finance Ecosystem 20 to innovate 59
Introduction 20 The role of accounting and finance
1.4 Digital ecosystem 20 professionals in digital transformation 60
Digital ecosystem components 20 Keeping pace in a fast-moving world 63
The characteristics of a digital Methods and practices 64
ecosystem 22 Technologies and architectures 64
How digital ecosystems create and Financial innovation in organisations —
deliver value 23 locking in the benefits 65
Evaluating a digital ecosystem 25 Summary 69
1.5 Digital financial services Review 70
ecosystem 26 References 70
Goals of digital financial services 27
Requirements for a digital financial services MODULE 2
ecosystem 27
Issues and challenges in the DFS Future of Money 73
ecosystem 30 Preview 73
1.6 The payments ecosystem 32 Part A: Digital Currency 75
Summary 34 Introduction 75
Part C: FinTech (Innovations) 36 2.1 Money and currency 75
Introduction 36 Functions of money 75
1.7 FinTech in business 36 Properties of money 75
The scope of FinTech 36 Differences between money
FinTech vs traditional financial and currency 76
services 37 Types of currency 76
Strategic issues in the FinTech sector 39 Traditional currencies 76
vi CONTENTS
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page vii — #7
CONTENTS vii
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page viii — #8
viii CONTENTS
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page ix — #9
4.5 Data strategy: why a business needs a Risk associated with environmental
data strategy 263 changes 299
Developing a data strategy for 4.11 Applying AI to analytics: benefits, risks
analytics 264 and examples 299
4.6 A framework for designing an effective Benefits 300
data strategy 266 Risks 301
Data strategy frameworks 266 Machine learning 302
The smart strategy board 267 4.12 Case study: Detecting money
Identifying data needs 268 laundering 309
Prioritising data needs and obtaining Summary 313
data 273 Part D: Data Interpretation 315
Analysing data 276 Introduction 315
Communicating to support 4.13 Data interpretation: introduction and
decision making 276 considerations 315
4.7 Data strategy implementation 276 Extracting knowledge from data 315
Data ‘ownership’, organisational politics and Domain knowledge 316
culture 277 Quantitative and qualitative analysis 317
Data management systems 279 Data visualisation 317
Data migration 281 Overcoming cognitive bias 317
Change management 282 Correlation 319
Agility, responsiveness and crisis Causation 321
management 284 Common mistakes in data
Implementing data strategies in small interpretation 324
businesses 285 4.14 Data-informed decision making 325
Summary 285 Data-informed decision-making
Part C: Data Analytics 287 frameworks 325
Introduction 287 Iterative approaches 326
4.8 What is big data and analytics? 287 4.15 Example of a decision-making
Big data 287 framework 326
How is business using big data? 288 Assessing the adequacy of the
Big data implementation 289 insights 327
Skills required by finance Making a decision based on the
professionals 290 information 327
4.9 The data analysis process 292 Summary 329
Step 1: Define the purpose 292 Part E: Data Visualisation 330
Step 2: Obtain the data 292 Introduction 330
Step 3: Explore and clean the data 294 4.16 The role of visualisation 330
Step 4: Reduce the data dimension Exploratory and explanatory
(if necessary) 294 visualisations 330
Step 5: Determine the analysis task 295 4.17 Telling a story with data 333
Step 6: Partition the data (for supervised Step 1: Storytelling 333
tasks) 295 Step 2: Understanding the context 334
Step 7: Choose the analysis Step 3: Choosing the most effective visual
techniques 295 display 335
Step 8: Use algorithms to perform the Step 4: Using cognitive principles to give a
task 295 clear picture of the data 347
Step 9: Interpret the results 295 Visualisation tools 350
Step 10: Deploy the model 297 Dashboards 351
4.10 Risk, warnings and challenges in data Summary 367
analytics 297 Review 368
Data quality 297 References 368
Outliers 297
Missing values 298 MODULE 5
Normalising (standardising) and rescaling Risk Management,
data 298
Oversampling rare events in classification Governance and
tasks 298 Regulation 371
Preview 371
CONTENTS ix
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page x — #10
x CONTENTS
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page xi — #11
CONTENTS xi
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page xii — #12
SUBJECT OUTLINE
INTRODUCTION
The purpose of this subject outline is to:
• provide important information to assist you in your studies
• define the aims, content and structure of the subject
• outline the learning materials and resources provided to support learning
• provide information about the exam and its structure.
The CPA Program is designed around five overarching learning objectives to produce future CPAs
who will be:
• technically skilled and solution driven
• strategic leaders and business partners in a global environment
• aware of the social impacts of accounting and business
• adaptable to change
• able to communicate and collaborate effectively.
SUBJECT DESCRIPTION
Digital Finance
Digital finance usually describes an ecosystem that embodies a range of new finance software, businesses
and products, together with novel forms of customer communication and interaction. Today it plays an
increasingly important role in almost every aspect of our society. The Digital Finance subject explores the
knowledge and skills necessary to deal with accounting and finance in a digital world, equipping candidates
with the tools to apply emerging technologies in a real-world context.
The aims of the Digital Finance subject are as follows.
• Assist candidates in building their career through the attainment of emerging accounting and digital
finance skills that are transferrable across a diverse range of roles.
• Provide candidates with a practical introduction to the key technologies used in digital finance and their
business implications.
• Equip candidates with the knowledge and skills to transform their organisation by converting uncertain-
ties regarding the use of digital finance technologies into impactful opportunities that create efficiencies,
generate insights and value, and support strategic objectives.
SUBJECT OVERVIEW
In today’s rapidly evolving digital landscape, with its onslaught of new and innovative solutions,
accounting and finance professionals must stay ahead of the opportunities and challenges presented by
these technologies and other digital disruptors. It is imperative to learn to recognise and reap the full
benefit of these technologies and the digital finance environment. This subject explores the knowledge
and skills necessary to understand and navigate the complex world of digital finance, and provides the
tools to apply emerging technologies in a real-world context. More specifically, the subject discusses the
topics of:
• the digital finance ecosystem
• the future of money
• technology and its use in finance
• data analytics, interpretation and visualisation
• risk management, governance and regulation.
General Objectives
On completion of this subject, you should be able to:
• examine where we are today, and where we will be in the near future, with technologies used in the
digital finance environment such as artificial intelligence (AI), machine learning (ML), robotic process
automation (RPA), big data, digital payments and digital currencies
• apply your knowledge to translate complex digital finance concepts and terminology to support informed
strategic decision making around the use of key digital finance technologies in business
• use your insights of digital finance technologies to navigate the changing technological environment in
which accounting and finance professionals work
• examine the regulatory and organisational implications of digital finance technologies and how they can
be applied in a range of settings including compliance in organisational contexts.
STUDY GUIDE
Module Descriptions
The subject is divided into five modules. A brief outline of each module is provided below.
Module 1: The Digital Finance Ecosystem
The business environment is characterised by networked systems and increasingly open exchanges of data
between them, enabling new ways for businesses to create value. The finance industry is increasingly turn-
ing towards simplified, fast and reliable solutions using AI, automation and data-driven personalisation.
This module is designed to provide candidates with the knowledge required to understand the current
finance landscape and the forces shaping the future — especially the various types of technology and
innovative uses of technology in finance. It aims to provide the skills candidates will need to evaluate the
opportunities and challenges an organisation will face as it develops and implements a strategic approach
to digital transformation in order to prepare for and succeed in the digital finance ecosystem.
Module 2: Future of Money
This module examines innovations in how value is exchanged. While some forms of digital payment are
well established, new and emerging digital currencies and types of digital payment are leading to more
efficient, immediate and personalised exchanges. In particular, cryptocurrencies, based on blockchain
platforms, are seen as a potentially transformative force in the monetary system. Cryptocurrencies,
perhaps alongside distributed ledger technologies generally, have the potential to decentralise and largely
disintermediate transactions.
This module also describes alternative sources and types of finance, which enable organisations to access
capital in new ways. It concludes with a discussion of the future of banking.
As digital payments, digital currencies, including cryptocurrencies, and alternative finance are expected
to reshape the monetary system and the role of institutions within it, this module aims to provide the
knowledge and skills to evaluate the role of these innovations in solving problems facing the organisation
and meeting its objectives.
Module 3: Technology and its use in Finance
This module examines the potential uses of robotic process automation (RPA) and various types of
artificial intelligence (AI) technology to achieve efficiencies and create value for organisations. These
technologies enable automation of various tasks, increasingly personalised interactions with customers
and other external stakeholders and intelligent interaction and collaboration with internal stakeholders.
The aim of this module is to provide the knowledge to evaluate, choose and promote appropriate digital
technologies to help the business achieve efficiencies and solve problems in creating value for a wide
variety of stakeholders.
Module 4: Data Analytics, Interpretation and Visualisation
This module discusses concepts related to the use of data to support data-informed decision making in
organisations, beginning with the need for data literacy and progressing through data strategy, the analysis
process, interpretation of outputs and visualisation for decision makers.
Data analytics and interpretation provide organisations with the ability to transform their data into
insights that enable efficient, informed and actionable decisions. Finance departments are being relied
upon to demystify complex sets of data and derive insights that shape an organisation’s strategy. Effective
teams use visualisation when communicating with stakeholders to translate the data into easily understood
concepts and demonstrate significant findings. Through this module, candidates will learn the relevant
technologies in the organisation and their role in shaping analytics strategy to maximise the organisation’s
return on its investment in analytics technologies.
Module 5: Risk Management, Governance and Regulation
Digital finance is creating new and changing risks, compliance needs and regulatory challenges. This
module aims to provide candidates with foundational knowledge in the areas of governance, risk
management and compliance as it pertains to the digital finance environment. It also examines regulation
of digital finance and the use of technology to help meet complex and changing regulatory requirements.
With digital transformation sweeping the industry, data security and data privacy are increasingly
vulnerable to threats of data breaches and cyber-attacks. Businesses must proactively consider the risks
involved when using digitised data and processes and employ the necessary safeguards to support the
increasing use of data throughout business processes, transactions and stakeholder interactions.
Recommended proportion
Module of study time (%) Weighting (%)
2. Future of Money 15 15
LEARNING MATERIALS
Module Structure
The study guide is your primary examinable resource and contains all the knowledge you need to learn
and apply to pass the exam. The Digital Finance study guide is divided into five modules, suggested
answers and a glossary, and includes a number of features to help support your learning. These include the
following.
Learning Objectives
A set of learning objectives is included for each module in the study guide. These objectives provide
a framework for the learning materials and identify the main focus of the module. The objectives also
describe what candidates should be able to do after completing the module.
Examples
Examples are included throughout the study materials to demonstrate how concepts are applied to real-
world scenarios.
Revision Questions (and Suggested Answers)
These numbered revision questions provide you with an opportunity to assess your understanding of the
key learning points. These questions are an integral part of your study and should be fully utilised to support
your learning of the module content.
Reflective Questions
These require you to reflect on an issue or an example in the study materials.
They are not numbered, rather they are identified by the CONSIDER THIS heading above the question.
Suggested answers are not provided in the study guide for these reflective questions.
Key Points
The key points feature relates the content covered in the section to the module’s learning objectives.
Review
The review section places the module in context with other modules studied and summarises the main
points of the module.
References
The reference list details all sources cited in the study guide. You are not expected to follow up this
source material.
SUBJECT OUTLINE xv
“FMPrelims_PrintPDF” — 2021/6/1 — 8:18 — page xvi — #16
MODULE 1
PREVIEW
The innovative application of technology has transformed the business world. As increasingly ubiquitous
internet-enabled devices generate, store, process and exchange massive amounts of data at low cost and in
near-real time, a new wave of transformation is underway. This transformation is rapidly changing the way
businesses work, communicate and collaborate with each other, and with their customers, to create value.
In this context, accounting and finance professionals need not only a working understanding of the
current finance landscape, but also the knowledge and skills to evaluate and respond to the opportunities
and challenges associated with the digital finance innovations that are transforming the finance ecosystem.
Part A of this module describes the emergence of the data-enabled ‘Network Age’ and how technologies
and innovations are impacting on finance functions and institutions. This establishes the context in which
accounting and finance professionals will work to evaluate and support the adoption and implementation
of innovations that can help achieve the business’s strategic goals.
Part B of the module explains and illustrates the components, characteristics and function of a digital
ecosystem. This provides an understanding of how a digital ecosystem creates value and thus how to
evaluate innovations to determine their role in helping an organisation achieve its strategic goals. To help
navigate the complexity of the technologies and applications involved, we will describe the goals and
benefits of the digital financial services ecosystem, the enabling environment and infrastructure, and the
issues and challenges encountered during the digital transformation of finance. This part concludes with a
discussion of payments in the digital finance ecosystem.
Part C of the module discusses financial technology (FinTech) applications relevant to the work
of accounting and finance professionals, and to the broader digital financial services ecosystem. We
begin by exploring the role of FinTech businesses in the development of innovative digital finance
solutions. FinTechs leverage technology to create cost-effective niche services to fill needs that are not
optimally served by the traditional financial services institutions. As the sector has developed, FinTechs
and traditional providers increasingly cooperate and collaborate. After discussing the digital finance
applications of FinTech, we discuss emerging areas in FinTech and the opportunities and roadblocks that
must be considered when developing, evaluating, adopting or implementing digital finance innovations.
Part D draws together the tools and knowledge presented in parts A, B and C to enable accounting and
finance professionals to play a valuable role in the evaluation, planning and practical implementation of
digital finance innovations that will digitally transform the business to help achieve its strategic objectives.
By developing the ability to understand and evaluate the impact of technology on accounting and
finance, the components and functioning of the digital finance ecosystem, innovative FinTech, and digital
transformation issues and strategies, this module provides the foundations for evaluating and taking
Pdf_Folio:1
advantage of the specific digital finance technologies and developments discussed in the later modules.
PART A: DIGITAL FINANCE LANDSCAPE
INTRODUCTION
The innovative use of technology has shaped the entire history of human society. Just as the development
of steam power, machinery and factory production lines transformed business and life during the Industrial
Revolution of the seventeenth century, today mobile communications, connected smart devices and
massive data processing power are creating a Network Age in which businesses and consumers create
and consume value in new ways.
The technologies of the Network Age affect the full spectrum of finance-related tasks, including sales,
procurement and payroll. Digital finance embodies a range of new financial software, businesses and
products, together with novel forms of customer communication and interaction. FinTech companies are
responsible for many of the technological innovations, but existing providers have also responded to this
competition by transforming their operations (Gomber, Koch & Siering 2017).
This part of the module describes the current digital finance landscape, provides two frameworks to help
analyse digital finance, and establishes the skillset that accounting and finance professionals need in order
to understand, evaluate and adopt the innovations that are digitally transforming the finance function.
Big Data
The datasets that arise from the pervasive creation and capture of data in the Network Age are commonly
referred to as big data. Big data is characterised in terms of:
• volume — the large size of the datasets created by collecting and storing the data generated by devices,
transactions and interactions
• variety — the breadth and depth of datasets that arise from diverse sources
• velocity — the speed at which new data is generated and accumulated within datasets
• veracity — the variable quality of datasets given the diverse source of data collected.
These are known as the four V’s of big data. Sometimes value is described as the fifth V, to recognise
that the effective use of data is often the key to value creation.
It is the data created and exchanged between devices, individuals and organisations that underpins
most aspects of digital finance, including FinTech, the development of digital ecosystems, automation,
personalisation and analytics. Much of this study guide deals with how to use data to create value for
an organisation by achieving efficiencies and higher productivity, improving customer experience by
understanding and responding better to their needs, and gaining better insight into complex business
problems. With this comes a series of challenges and obligations that require strong data governance.
Pdf_Folio:2
2 Digital Finance
THE TRANSFORMATION OF ACCOUNTING AND FINANCE
The key digital technologies re-shaping the finance function are (Agrawal, Dinneen & Seth 2016):
• automation — the use of technology to replace or complement human labour
• analytics — the application of data analysis technologies to big data to generate and visualise
actionable insights
• artificial intelligence (AI) — the use of technology to intelligently analyse data and respond to the
outcomes.
Later modules will explore these technologies in detail. For now, it is important to understand that these
technologies are often used in combination to create complex and powerful digital capabilities with an
enormous scope of application. For example, they are used to predict what a particular customer is likely
to buy, identify credit fraud in real time, detect insurance claims fraud and automate personalised targeting
of digital marketing. Additionally, there is an increasing use of algorithms to detect patterns in vast volumes
of data and interpret what they mean for the business.
These applications depend on:
• scalable and adaptable platforms that enable access to real-time data
• integrated digital ecosystems in which every component seamlessly interacts to share and process data
• advanced analytics capabilities to generate powerful insights.
Effective implementation of digital technology will mean that accounting and finance teams can move
away from transactional activities and focus on valuable tasks that technology cannot handle alone.
Research (Bock, Iansiti & Lakhani 2017) shows that businesses that have successfully embraced digital
transformation experience better financial performance than those that stuck to traditional practices.
A significant challenge is to maintain organisational performance while transitioning to future-facing
technologies.
While a majority of business leaders believe that their industry will be disrupted by digital models,
they feel ill-equipped to adapt (Bersin 2016). The risk of being left behind means that development of
digital capabilities is critical for businesses and finance professionals. Harvard Business School professor
Karim R. Lakhani, who leads a virtual financial leadership program created specifically for CPA Australia
members (Bock, Iansiti & Lakhani 2017), emphasises that all business types need to understand the digital
world, and stresses the importance of a ‘digital mindset’.
Digital Mindset
A digital mindset involves rethinking entire business and operating models in addition to the adoption of
new technology. This does not mean that finance professionals need to become computer programmers or
data scientists, but they do need a firm foundation in digital literacy.
Digital literacy requires knowing enough about the applications, benefits, costs and limitations of
technologies to be able to assess their value to the organisation and support their effective implementation.
A digital mindset, data literacy and effective evaluation and use of technology should be embedded in the
business and drive decision making.
In a dynamic environment, adaptability and investment in continuous learning is a necessity.
Enabling Skills
The recruitment, development and retention of appropriately skilled staff is one of the greatest challenges
facing business in the technology-disrupted future. As traditional tasks are allocated to machines, staff
will be required to develop the technical skills to administer digitally enabled processes, and the analytical
skills necessary to realise useful insights from the abundance of data they will encounter.
Since business leaders disagree on the type of technology that will drive the greatest change in their
industry (PwC 2019), it is not surprising that it is difficult to precisely predict the required skills in the
future finance landscape. Instead, a premium will be placed on broad and adaptable enabling skills rather
than on narrow task-specific skills. These skills are in high demand, but short supply. Examples include:
• behavioural skills — communication, relationship management and strategic thinking
• data analysis skills — data modelling and visualisation, design thinking and programming
• finance skills — business modelling, process design and financial analysis.
Businesses that are early adopters of new technology will likely have a higher percentage of staff affected
by automation. Those staff who are engaged in transaction-oriented processes are likely to be the first to be
impacted. To retain a greater proportion of their workforce, it is important that businesses offer retraining
opportunities to those with high potential.
Pdf_Folio:3
EXAMPLE 1.1
.......................................................................................................................................................................................
CONSIDER THIS
Reflect on your own level of digital literacy. Consider how much you know about automation, analytics and the role
of digital ecosystems in relation to your organisation and the finance function in particular. What are your strengths
and weaknesses and how could you build on these?
QUESTION 1.1
Outline four digital technologies that are re-shaping the finance function and explain how those
technologies are changing the tasks that finance teams focus on.
Pdf_Folio:4
4 Digital Finance
1.2 IMPACT OF TECHNOLOGY ON THE DIGITAL
FINANCE LANDSCAPE
The impact of technology on the digital finance landscape can be understood in terms the intersection
of digital finance business functions; digital finance technologies and technological concepts; and digital
finance institutions. Innovation in these three elements is driven by the relentless pursuit of competitive or
transient advantage. A competitive advantage exists when a business is able to achieve better performance
than its competitors and defend that advantage over the long term. A transient advantage is temporary,
recognising that some advantages cannot be sustained, as competitors quickly copy or improve on an
innovation.
Technology and innovation are complex and fast changing, with numerous potential applications,
benefits and risks that must be understood and evaluated in order to maximise their value to the business.
The Digital Finance Cube and the Ten Types of Innovation framework provide structured ways to explore
and assess digital finance technologies. We will discuss these now.
lo er
ch o- n
d
rk
o
in
no pe
un el
aly ata
wo
ab er
i
Bi y
Pe icat
ha
m -fi
s
g
s
en rth
tic
an d
ler
m ar
et
kc
te r-t
Fu
ln
co Ne
oc
e
cia
Bl
Digital financing
FinTech companies
Digital investments
business functions
Digital finance
Digital money
Digital payments
Digital insurances ce
nan s
Digital financial advice l fi ion
ta t
i gi stitu
D in
Source: Gomber, Koch & Siering 2019, ‘Digital finance and FinTech: current research and future research directions’, Journal of
Business Economics, 87(5).
It is particularly useful in understanding and evaluating digital finance services and service providers.
The focus is on the most important dimension from a business administration point of view: the digital
finance business function. The second and third dimensions support the business function: technologies
enable the business functions, and digital finance institutions provide these functions as a service.
Each segment inside the Digital Finance Cube represents a specific combination of one business
function, a certain technology and a specific type of institution. Any one institution may operate in just
one segment or may operate across a range of segments.
Pdf_Folio:5
EXAMPLE 1.2
Crowdfunding
Crowdfunding is often used by start-ups, social ventures and creatives as it provides a way to seek funds
based on the idea or ‘story’ attached to the business or initiative rather than purely on the financials.
Since customer demand is often directly reflected in the amount of funding achieved via crowdfunding,
the approach overcomes uncertainty surrounding potential sales. In donation-based crowdfunding,
funders do not receive any compensation for their support, and in reward-based crowdfunding, funders
receive non-financial compensation (e.g. objects, products or services). Crowdlending and crowdinvesting
provide more conventional returns to funders: interest, and equity and profits.
Kickstarter (2020), launched in 2009, has attracted funding pledges of more than AUD4.6 billion for
nearly 500 000 projects, including movies, music, theatre and video games. Most Kickstarter most projects
involve reward-based crowdfunding. The Patreon (2020) crowdfunding platform uses a subscription model
whereby patrons provide ongoing financial support for creatives and artists. This model was particularly
popular during the Coronavirus (COVID-19) shutdown, when many artists were unable to perform or exhibit
Pdf_Folio:6
in public.
6 Digital Finance
Brewdog is a UK-based brewery and hotel chain established in 2007 that has used crowdfunding as
part of its financing mix (Crowdcube 2020; Kirkegaard 2020).
• In 2011, it undertook an equity-based crowdinvesting campaign under the branding ‘Equity for Punks’,
raising GBP2 million from shares sold at GBP23.75 each. In addition to gaining equity in the company,
participants became entitled to discounts on Brewdog products.
• In 2017, the company sold a 22 per cent stake to a privacy equity firm for GBP213 million.
• In 2018, UK brewer Brewdog raised over GBP26 million via a fifth round of crowdinvesting, valuing the
business at GBP1.8 billion.
• Brewdog shifted its crowdfunding campaign from third-party platforms to its own website, under the
‘Equity for Punks Tomorrow’ branding.
• In 2020, Brewdog launched a crowdinvesting campaign offering 100 000 $100 non-voting shares,
representing 10 per cent of the equity in the Australian operation and indicating a valuation of
AUD100 million, even though Australian revenues in 2020 were about AUD5 million, resulting in a loss
of AUD1.7 million. The threshold for the campaign to proceed was set at just $300 000. The prospectus
notes that no dividends will be paid, but equity holders will be entitled to a discount at Brewdog bars
and various other rewards. The raising was intended to fund sustainable business initiatives.
• By 2021, Brewdog had raised more AUD154 million from 145 000 crowdinvestors across the world.
The Brewdog crowdfunding campaigns have attracted some criticism due to crowdinvestors paying a
substantial premium for shares over those offered to corporate investors.
.......................................................................................................................................................................................
CONSIDER THIS
Thinking about your own organisation or clients, could crowdfunding or one of the other platform-based financing
innovations play a valuable role in their financing mix? Why/why not?
Digital Investments
Digital investments support individuals or institutions in making investment decisions and in arranging the
required investment transactions. Digital investments include:
• online brokerage, and mobile and social trading, in the B2C area
• high-frequency and algorithmic B2B trading (Gomber, Koch & Siering 2017).
Online brokers fulfil the same tasks as human brokers but operate independently from time and location,
need fewer staff, can process more orders and operate at lower cost. Some platforms, such as eToro, offer
zero-commission trading and the possibility of fractional investing (enabling the purchase of 10 per cent
of a stock priced at $1000 for $100) (Finder 2020). These platforms may also provide access to multiple
markets and a range of trading tools, such as charts and analytics, that are comparable to professional
trading software.
Mobile trading refers to the trading of securities using mobile devices. Platforms for mobile trading offer
real-time access to financial markets and the possibility to trade independently from location and human
advisors or brokers (Gomber, Koch & Siering 2017).
Social trading refers to securities trading via internet platforms that combine social media networking
with investment strategies. The platforms (e.g. eToro) typically enable a user to observe, ‘follow’ or copy
the strategies of other traders (Gomber, Koch & Siering 2017).
Algorithmic trading automatically initiates and manages orders on electronic trading platforms, enabling
institutional investors to trade at high frequency and without human intervention. Algorithms can enable
large orders to be divided into smaller parcels to minimise the impact of trades on market prices or can be
used by hedge funds (Vardy 2020) to quickly exploit market anomalies.
Digital Money
As we shall see in module 2, there is no universally agreed precise definition of terms like digital money
or digital currency. The Digital Finance Cube can be used to examine any digital innovation in how money
and currency are created, stored, valued, regulated or exchanged. The common theme is that they exist
electronically rather than in physical form and thus may vary from digital representations of conventional
money through to new virtual currencies, including cryptocurrencies such as Bitcoin. Transactions using
these forms of money or currency occur via the exchange of data, primarily using online platforms. New
forms of digital money may operate independently from government-issued currencies, bank accounts and
the transactional detours traditionally necessary for foreign exchange. Thus, digital money has the potential
to transform many aspects of finance.
Cryptocurrencies are considered one of the key potential disruptors. They are decentralised, freely
convertible, virtual currencies that are based on encryption technology and hosted on a platform known as
Pdf_Folio:7
EXAMPLE 1.3
8 Digital Finance
In the 6-month period since launch, Crockd experienced rapid sales and customer growth. As expected,
they soon found the need to transact globally, particularly for:
• sourcing materials from international suppliers
• hiring software developers overseas
• paying for software subscriptions in a foreign currency.
Using Airwallex as their business account gave Crockd access to competitive foreign exchange rates
and cost savings as a scaling business, where preserving cash flow is critical. Airwallex’s foreign exchange
rates were locked in upfront, providing Crockd with a cost-effective rate to start and certainty as they were
scaling up.
With the payments system in place, all transactions managed on a single platform and future terms
and conditions established, Crockd was able to then move all of their attention to customer acquisition
and growth.
Source: Airwallex 2020, ‘Crockd, how an Aussie side hustle accelerated into a global brand’.
.......................................................................................................................................................................................
CONSIDER THIS
Payments are a complicated aspect of almost every business. How could your organisation’s payments process be
improved? Are the solutions you have in mind available from conventional or FinTech payments providers or would
they be created by process changes in your own organisation?
Digital Insurance
In the B2B space, digital insurance platforms provide a range of value-adding services, including in
particular data analytics to enable insurers and their clients to service and monetise their customers more
optimally, including through improved relationship management and customer experience, and better
identification of cross-selling opportunities.
In the consumer space, digital insurance has been enabled by the peer-to-peer concept. Friends and
family form alliances, in which each member pays a certain amount of money to the digital platform.
Part of the payment is used for a premium with an insurance company, and the remainder is stored in an
account available to the rest of the group. In the case of small insurance events, the damage is remedied
using money from the group account. The saves the insurance company high administration expenses for
small insurance cases, which can then be reflected in lower premiums for customers. In large insurance
events, the insurance company takes responsibility for the pay-out. If few claims are made, the members
may be repaid a portion of the money stored in the group account (Gomber, Koch & Siering 2017).
Digital Financial Advice
Review sites and comparison portals evaluate, score and compare products and services. In the finance
sector, these usually focus on financial product reviews or financial product comparisons. Finance-focused
online communities and social media platforms facilitate discussion and sharing of information. Review
sites, comparison portals, online communities and social media enable access to advice and also present a
new channel in which finance service providers must monitor and manage their reputation.
Another development in digital financial advice is the creation of robo-advisers — algorithms that gen-
erate investment proposals based on inputs or data analysis of investment goals, financial background and
risk aversion. Robo-advisors are growing in sophistication, but currently focus on portfolio management
and investment strategies based on established theories (e.g. modern portfolio theory) and limited asset
classes (e.g. equities or exchange-traded funds) (Gomber, Koch & Siering 2017). Robo-advice trades off
cost against the benefits of more personalised and responsive advice. For example, InvestSMART’s robo-
adviser (InvestSMART 2020) assesses risk appetite and investment goals to match clients with one of nine
diversified exchange-traded fund portfolios and then keeps the client advised of investment performance
over time.
Digital Finance Technologies and Technological Concepts
Having established the digital finance business functions, the second dimension of the Digital Finance
Cube can be used to understand the technologies and concepts driving those functions, including:
• blockchain
• social networks
• near-field communication (NFC)
• peer-to-peer technology
• big data analytics
• enablers such as mobile devices, cloud technologies, information security technologies, automation
and AI.
Pdf_Folio:9
10 Digital Finance
FinTech Companies
FinTech companies have emerged either as FinTech start-ups or technology companies without a history
in financial services that have developed FinTech offerings. As their name implies, FinTech companies
integrate technology and financial services. Initially, the FinTech sector targeted niche finance functions
that were not well served by traditional providers. This approach has evolved over time, however, as
FinTech companies have used innovative approaches to successfully win customers from incumbent
providers (Gomber, Koch & Siering 2017). FinTech innovations are described in detail in part C of
this module.
FinTechs are often able to create business models that provide more flexibility, efficiency and innovation
than those of traditional service providers, but face strong challenges from the market power of incumbents
and a regulatory system designed around the industry’s old business models.
Traditional Service Providers
Traditional financial service providers include asset management companies, banks, insurance companies
and brokerage companies. Traditional financial services encompass a broad range of services including:
cash accounts, savings, money management, investment management, money transfer and payments
(e.g. credit cards), portfolio management, financial advice, money loaning and lending (e.g. mortgages,
consumer loans, credits), foreign currency exchange, equity trading, brokerage, insurances, and pen-
sion planning (Gomber, Koch & Siering 2017).
The traditional financial service providers have adopted new technologies and sought to innovate
existing business functions. This is driven by the need to continually improve customer service, the
need to be competitive in the marketplace, and the need to defend existing and future business against
emerging FinTechs.
.......................................................................................................................................................................................
CONSIDER THIS
The Digital Finance Cube was designed to be both highly generalisable and flexible. It provides a framework to
consider the interaction of business functions, technologies and service providers. The dimensions are intended to be
adapted and extended by new elements as digital finance innovation continues. How would you draw a Digital Finance
Cube for your organisation or a client? What are the relevant business functions, technologies and service providers?
QUESTION 1.2
Define ‘digital payments’ and explain how the use of digital payments is central to modern
commerce.
Profit model
Configuration
Network
Structure
Process
Offering
Product performance
Product system
Service
Experience
Channel
Brand
Customer engagement
Source: Keely, Walters, Pikkel & Quinn 2013, Ten Types of Innovation: the discipline of building breakthroughs, John Wiley &
Sons.
12 Digital Finance
a business and may include patented or proprietary approaches that yield long-term advantage. A process
innovation must include a methodology or capability that is substantially different from and superior to
industry norms.
Examples of process innovations include lean production whereby managers reduce waste and cost
throughout a system, process standardisation, which uses common procedures to reduce cost and com-
plexity, and predictive analytics, which model past performance data to predict future outcomes — helping
companies to design, price and guarantee their offerings accordingly.
Product Performance: How the Company Develops Distinguishing Features and Functionality
Product performance innovations address the value, featuresand quality of a company’s offering. This
type of innovation involves both entirely new products as well as updates and line extensions that add
substantial value. Product performance innovations can delight customers and drive growth, but are often
easily imitated and so rarely deliver long-term competitive advantage. Common examples of product
performance innovation include: simplification to make it easy to use an offering; sustainability to provide
offerings that do no harm to the environment; or customisation to tailor a product to an individual’s
specifications.
Product System: How the Company Creates Complementary Products and Services
Product system innovations relate to how individual products and services connect or bundle together to
create a robust and scalable system. This is fostered through interoperability, modularity, integration and
other ways of creating valuable connections between otherwise distinct and disparate offerings. Product
system innovations help build ecosystems that captivate customers and defend against competitors. Product
system innovations can include offerings that the company does not itself own or produce.
Product bundling — taking several related products and selling them in a single package — is a common
example of a product system innovation. Technology companies have used this type of innovation to build
platforms (e.g. app stores, developer kits and APIs) that spur others to develop products and services
for them. Other product system innovations include extensions to existing products, product and service
combinations and complementary offerings.
Service: How the Company Supports and Amplifies the Value of its Offerings
Service innovations ensure and enhance the utility, performance and apparent value of an offering. They
make a product easier to try, use and enjoy; they reveal features and functionality customers might otherwise
overlook; and they fix problems and smooth difficulties encountered by customers. A successful service
innovation can elevate an average product into a compelling experience that will encourage return custom.
Common examples of service innovations include product use enhancements, maintenance plans,
customer support, information and education, warranties and guarantees. The AmazonSmile program
(James 2020a) donates 0.5 per cent of all purchases to a charity of the customer’s choice, creating a
compelling reason for customers to prefer the platform.
Service innovations are increasingly delivered through electronic interfaces, remote communications
and automated technologies. Service can be a prominent part of the customer experience, or an invisible
safety net that customers sense but never see.
Channel: How the Company Delivers its Offerings to Customers and Users
Channel innovations encompass all the ways that a company’s offerings connect with customers and users.
While e-commerce has emerged as a dominant force in recent years, traditional channels such as physical
stores are still important, particularly when creating immersive experiences. Skilled channel innovators
often find multiple complementary ways to bring their products and services to customers to ensure users
can buy what they want, when and how they want it.
Channel innovations are particularly sensitive to industry context and customer habits. Flagship stores
can be an extremely valuable channel innovation, creating signature venues that showcase a company’s
brand and offerings, while short-term pop-up stores may be useful to maximise customer access during
holidays. In contrast, selling directly through e-channels or other means can reduce overhead costs,
maximising margins and cost advantage. Indirect distribution or multi-level marketing are alternative
approaches to recruiting others to promote and/or deliver an offering to the customer.
Brand: How the Company Represents its Offerings and Business
Brand innovations aim to ensure that customers recognise, remember and prefer a company’s offerings
over those of competitors. They are typically the result of carefully crafted strategies that are implemented
across many touchpoints between a company and its customers, including communications, advertising,
Pdf_Folio:13
QUESTION 1.3
Differentiate structure innovations from process innovations and discuss how they relate to
improving company performance.
QUESTION 1.4
Compare and contrast the four main types of experience innovation: service, channel, brand and
customer engagement.
14 Digital Finance
HOW FUNDS FLOW THROUGH THE TRADITIONAL
FINANCIAL SYSTEM
The financial system moves money from lender-savers (whose income exceeds their spending) to borrower-
spenders (whose spending exceeds their income), either directly though financial markets or indirectly
through financial institutions as shown in figure 1.3.
Direct financing
Financial markets
Funds Wholesale markets for the creation and sale of financial Funds
securities, such as shares, bonds and money market
instruments. Large corporations use the financial markets to
sell securities directly to lenders.
Lender-savers Borrower-spenders
• Consumers • Consumers
• Businesses • Businesses
• Government • Government
Financial institutions
Institutions, such as commercial banks, that invest in
financial assets and provide financial services. Financial
Funds institutions collect money from lender-savers in small Funds
amounts, aggregate the funds, and make the loans in larger
amounts to consumers, businesses and government.
Indirect financing
Source: Parrino et al. 2021, Fundamentals of corporate finance 4e, John Wiley and Sons Australia Limited.
A CHANGING LANDSCAPE
Banks are domestic, but the network is global. Banks are structured around paper, but the network is
structured around data. Banks distribute through buildings and humans, but the network distributes through
software and servers. Network Age innovations will not replace banking, but they will diminish and change
it. The historical systems of value exchange are becoming a smaller percentage of trade as new structures
are implemented to allow value to flow.
A series of events and trends beginning around 2007 triggered the start of major changes in the banking
and financial services industries:
• the global financial crisis of 2007 to 2008
• the consequent damaged profitability, reduced trust and changed regulatory environment
• technological advances and trends, such as high levels of smart-phone penetration and the development
of sophisticated APIs that allow different technology systems to exchange data.
Start-ups and technology companies began to challenge established financial institutions by offering
specific, niche services to consumers, businesses and incumbent financial institutions. Banks in response
have focused increasingly on their customers’ requirements in a bid to improve customer experience and
service, and thus defend their existing business. Together, this reflected and reinforced a shift in customer
expectations towards innovative services that provide convenience, low cost and personalisation. Banks
also identified a need to understand the balance between the online and the in-person components of
their business.
Other key trends include:
• financial inclusion — the use of mobile smart phones and mobile wallets to bring financial services to
billions of people who were previously unserved by financial institutions
• social networks — the increasing power of consumers to communicate with each other, moving control
of communications from business to consumer
Pdf_Folio:15
16 Digital Finance
FIGURE 1.4 Current and future bank processes
Current state
Human-centric Enterprise integration
business processes Application A
within the walls
Bank P1 P2 P3
of the business
employees
and complex Legacy core
Customer App B App C App D
legacy core systems
systems
P4
P3
Source: Adapted from Courbe 2016, ‘Financial services technology 2020 and beyond: embracing disruption’, PwC, p. 33.
QUESTION 1.5
What are the main differences between FinTech 1.0 and 2.0 companies, and what impact is
FinTech 2.0 having on the banking industry?
SUMMARY
The Network Age is characterised by a global network of internet-enabled devices and machines that
connect individuals and organisations, and create, process and exchange data at low cost and in near-real
time. Combined with automation, advanced analytics applied to big data, and AI, the Network Age is
bringing about a digital transformation of finance.
To succeed in digital finance, accounting and finance professionals will need a digital mindset and
high levels of digital literacy to be able to understand, evaluate and support the adoption of innovative
applications of technology to create value for the business and help achieve its objectives.
The Digital Finance Cube provides a framework to describe, understand and analyse digital finance
technologies/applications and service providers in relation to a business’s finance functions and its
operational, tactical and strategic objectives. The Ten Types of Innovation framework helps identify and
focus attention on the aspects of a business model that may be improved or transformed through the
development or adoption of digital finance innovations. It also helps to evaluate the potential of available
digital finance solutions to add value to the organisation. Together, the frameworks provide tools to
navigate the business impacts of digital finance technologies and innovations on business models, business
functions, and the traditional and FinTech institutions that provide financial services.
Technology is bringing about a change in the financial services market, in which banks’ role is being
diminished. Services are increasingly built through the combination and integration of components from
different organisations. This is expected to increase as data and banking become more open.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
1.1 Evaluate the complexity of digital transformation and its impact on the work of accounting and
finance professionals to meet specific business needs.
• The technologies associated with the Network Age are driving new ways of creating value.
Automation, analytics, visualisation and near-real time data exchange between individuals and
organisations are driving the digital transformation of accounting and finance.
• Accounting and finance professionals need to adopt a digital mindset and develop digital literacy to
be able to assess the value of various technologies to the organisation, and support their effective
implementation.
• Value creation in the finance system, and within organisations, increasingly occurs through the
sharing of data and the creation of services through the flexible combination and coordination of
component services.
Pdf_Folio:18
18 Digital Finance
1.2 Analyse the complexity associated with the practical implementation of emerging technologies
and their impact on the finance ecosystem.
• Automation, analytics and AI can be used in combination to create complex and powerful digital
capabilities with an enormous scope of application.
• The Digital Finance Cube provides a framework for understanding and evaluating the potential of
technologies/applications and service providers to improve or transform the business’s key finance
functions.
• The Ten Types of Innovation framework provides a structure through which to identify or evaluate
the potential value of innovations across specific aspects of the business model.
1.3 Support the adoption of digital technologies to achieve an organisation’s strategic goals.
• Accounting and finance professionals should support the adoption of technologies by focusing
on value creation for the business entity; using agile funding models; developing broad enabling
skills and a digital mindset; early-adopting advanced analytics and automation technologies; and
establish a digital ecosystem model of service delivery.
Pdf_Folio:19
Software Ecosystems
Software ecosystems are created by the interaction of businesses via a common technological platform
to produce software and services. The interactions involve the exchange of information, resources
and artefacts, and enable developers to build specific solutions to create complementary value (Alves,
Oliveira & Jansen 2017). This is represented in figure 1.6. These independent developers extend and
enrich the platform while sharing costs, risks and benefits with the platform owner. Examples of successful
software platforms include Apple’s iOS and App Store, and the Android operating system. These platforms
establish a framework in which other organisations can create and deliver software solutions to the
platform’s customers. For example, a software developer can produce an app using one of Apple’s
developer kits, and distribute that app via Apple’s App Store to customers who have bought into the Apple
product and service ecosystem.
Pdf_Folio:20
20 Digital Finance
FIGURE 1.5 Typical characteristics and components of a customer-centric digital ecosystem
Software
Data-oriented Data-driven
Customer
centred
Dynamic
Information
Common
technological
platform
Resources Artefacts
Data-Oriented Ecosystems
Data-oriented ecosystems provide organisations with data that is relied on to understand customers
and to make better decisions. As described in part A of the module, various technologies have vastly
increased the volume, variety and velocity of data, and this data can increasingly flow in near-real time
across networks. This has provided the basis for the development of data-oriented ecosystems. Analytic
platforms, along with AI, machine learning and other tools, can integrate many data sources to search
Pdf_Folio:21
Customer Centred
A customer-centric approach involves making the customer the focus of decisions across all aspects
of the business. While many companies claim to have adopted a customer-centric approach, in reality
few strategies extend beyond the customer-facing departments of the company. A customer-centric
personalised approach is not limited to customer service and marketing; it is a holistic approach across
every department.
Dynamic
A digital ecosystem is a collaborative and dynamic environment, replacing linear value chains and the
segmented and siloed departments of traditional business structures. Collaboration is adopted at every
level of the business, eliminating the conventional hierarchy and helping to identify new markets, new
technologies and new business models. Cooperation and collaboration also frame interactions with external
partners in the ecosystem. A digital ecosystem provides ways to scale and leverage innovation and new
technologies. This dynamic approach gives the business the agility to move with future disruption, and to
not be a victim of change.
Data-Driven
The more information a company has about its potential target customers, the better it can attract them with
uniquely appealing offerings, create customer loyalty and maintain customer engagement with products
and services. From social media to customer service interactions, big data can reveal insights into what
consumers are thinking about a brand, what they need and what they want. Big data is also sourced from
processes within the organisation and from other participants in the digital ecosystem, providing insights
into how to improve the efficiency and effectiveness of all aspects of the business. For example, Walmart’s
Data Café uses a constantly refreshed database consisting of 200 billion rows of transactional data —
representing the most recent few weeks of business (Marr 2016). In addition, it draws on 200 external
sources for meteorological data, economic data, telecom data, social media data, fuel prices, events taking
place in the vicinity of their stores and other relevant information.
Automated
Digital ecosystems generate big data. Participants in digital ecosystems need to process and interpret this
data at scale and speed in order to provide consumers with the personalised and seamless experience
they increasingly demand. AI and automation technologies make this kind of data analytics possible,
creating consistent valuable insights into consumer preferences, trends and behaviours, and enabling both
automated and human decision-based responses.
Pdf_Folio:22
22 Digital Finance
HOW DIGITAL ECOSYSTEMS CREATE AND DELIVER VALUE
Digital ecosystems enable enterprises to drive business processes more efficiently. Digital ecosystems
deliver value through:
• increasing sales
• increasing return on investment (ROI)
• opening new revenue sources
• reducing risks
• increasing speed of technology adoption.
Increasing Sales
Data-driven insights into customer preferences and habits enable a business to implement consistent and
relevant messaging across their customer service and digital channels. Businesses are then better positioned
to innovate based on those insights and offer new products and services that are responsive to the ever-
changing demands of consumers. This can translate to increased sales.
Increasing ROI
Big data analytics enable greater understanding of each aspect of the business, its interactions with external
parties, and the nature and needs of those external parties. This enables the business to ensure resources are
deployed optimally and processes work efficiently. For example, the business is able to plot touchpoints
throughout a consumer’s experience and thus optimise products and services to meet their demands.
A 2019 study by McKinsey found that businesses with an ecosystem had higher earnings before interest
and taxes (EBIT) than those without, suggesting that over a five-year period, an ecosystem could capture
an additional 10 per cent in EBIT growth (Fong et al. 2019).
Reducing Risks
While decision making in a traditional business is often made on the basis of observation and informed
guesswork, a digital ecosystem embraces the fluidity of data and data modelling to predict the outcome of
different decision options. This reduces risk, while enabling the company to be more receptive and agile
to change and to more efficiently and effectively deploy resources. The integration of data into all aspects
of a business, both internally and in interactions with external parties, does, however, introduce a range
of risks associated with data security and privacy. Cybersecurity risks can have devastating consequences.
These risks therefore need to be managed so the organisation’s overall risk profile remains aligned with
its risk appetite. This will be discussed in module 5.
Pdf_Folio:23
Pdf_Folio:24
24 Digital Finance
Amazon Pay: Trusted Payment Channel and Less Friction
Amazon Pay provides users with the option to purchase goods and services outside Amazon mar-
ketplace using address and payment method information stored in an Amazon account. Third-party
e-commerce stores add an ‘Amazon Pay’ button to their checkout process using the Amazon Payments
software development kit. This button redirects users to their Amazon account, where they select a
payment option and receive their purchase receipt. Amazon generates revenue from Amazon Pay by
charging 2.9 per cent + USD0.30 per domestic transaction. Users in India can make P2P transfers in
Amazon Pay through a unified payments interface.
Amazon Pay:
• increases customer trust, as payment goes through he Amazon channel
• streamlines payment processes across different platforms, as users are not required to re-enter card
and address information
• increases the likelihood of purchase for smaller platforms or retailers where users do not have
an account
• provides Amazon with valuable data on user spending outside of its shopping platform.
.......................................................................................................................................................................................
CONSIDER THIS
A digital ecosystem is characterised by the extensive flow of data within and between organisations to maximise value
creation. To what extent is your organisation or client integrated into a digital ecosystem? Based on the discussion
above, assess the potential and value of more extensive integration.
Question 2 : Do the current software and data tools enable a consistent customer experience?
Are all interactions accounted for at every touchpoint in the customer experience? How is the flow of data
handled throughout the organisation? Can the digital ecosystem be more efficient at distributing data and
insights to the departments that need it?
Pdf_Folio:25
The questions in figure 1.7 form the basis for an evaluation of a digital ecosystem. Comprehensive
digital ecosystem evaluations provide vital insights into how the organisation is functioning, but they take
time. The insights gained from an evaluation will reveal those areas for cost savings, increased efficiency
and identify opportunities to further drive growth and revenue.
.......................................................................................................................................................................................
CONSIDER THIS
The next section examines the digital financial services ecosystem. As we work through it, keep the evaluation
methodology in figure 1.7 in mind and assess how your organisation or clients could integrate services available
in the digital financial services ecosystem.
EXAMPLE 1.5
26 Digital Finance
system to process that data. Airwallex’s current integration involves providing an hourly bank feed into
Xero for all domestic and international transactions. Over the longer term, Airwallex plans to integrate its
platform with a variety of software and tools. For example:
• invoicing/quoting software — payment gateway integration into invoices so customers can pay via
debit/credit card upon receiving the invoice
• accounting software — expense management integration from Airwallex Borderless Cards issued to
employees/teams (i.e. pushes)
• payroll software — automatically push international batch payroll payments into Airwallex from the
payroll software
• e-commerce platforms — payment gateway integration into e-commerce platforms (e.g. Shopify, Big
Commerce, Neto).
This will create more efficiencies and value for Airwallex’s clients by managing the flow of data between
Airwallex services and the other components of the client’s digital finance ecosystem.
Source: Airwallex 2020.
Source: International Telecommunication Union 2016, ‘The digital financial services ecosystem’, www.itu.int/en/ITU-T/
focusgroups/dfs/Documents/09_2016/FINAL%20ENDORSED%20ITU%20DFS%20Introduction%20Ecosystem%2028%
20April%202016_formatted%20AM.pdf.
Infrastructure Readiness
Infrastructure readiness refers to the availability of technical systems that enable digital financial services.
Components of infrastructure readiness are discussed below.
Payments Systems
Payments systems are required for transactions between and among end users, including consumers,
merchants, businesses and governments. These payments systems may be public, semi-public, or private;
they may be ‘closed-loop’ or ‘open-loop’. Security of payments systems is a requirement of infrastructure
readiness. In addition, a certain degree of payments system interoperability among participants in payments
is required. Interoperable payment systems allow two or more proprietary platforms to interact seamlessly,
enabling the exchange of payment transactions between and among payment service providers and,
consequently, users. Interoperability enables users to make digital payment transactions with any other
user in a convenient, affordable, fast, seamless and secure way, possibly via a single transaction account.
Pdf_Folio:28
28 Digital Finance
Communication Networks
Voice and data communication networks support the exchange of data to and from end users and
providers. Certain levels of communication network quality and security are thus a necessary component
of infrastructure readiness.
Sufficient Energy
Sufficient energy must be available to operate the other infrastructure and any devices used by participants
in a DFS ecosystem. The massive exchange of data and the supporting infrastructure and devices involved
in a DFS ecosystem consumes considerable energy. Some Fintech applications, such as blockchain,
consume very large amounts of energy. Current blockchain designs run on algorithms that can consume
up to 215 kWh per transaction (Omezzine & Schleich 2019) (the equivalent of letting a 25 W incandescent
light bulb burn for a full year). This is mainly because validating and securing transactions on the
blockchain requires huge computing power.
Identity Systems
Identity systems must be available to identify end users and their providers and must feature authentication
systems capable of recognising and validating these identities. Identity systems may be national IDs,
sectorial IDs (e.g. financial industry identifiers, bank account numbers, or mobile phone numbers)
or private sector IDs (e.g. WeChat or PayPal identifiers). Some IDs are biometrically enabled (using
fingerprints or facial recognition) and this is expected to become a significant part of the ecosystem, given
the importance of the mobile phone (Thales 2020; Naiya 2018).
Enabling Environment
The enabling environment refers to:
• laws and regulations — the legal requirements relating to financial services, telecommunications, data,
competition and consumer protection that both enable and constrain the activities of DFS ecosystem
participants
• national policies — the frameworks within which a government seeks to promote the overall wellbeing
of its society; for example, measures to improve financial inclusion
• standard-setting bodies and their standards — the organisations that work with industry to formulate
standards and protocols to ensure interoperability between systems and organisations; for example,
EMV (Europay, Mastercard and Visa) is global standard for chip cards and the associated transaction
authentication technology
• industry groups — the organisations that collectively represent DFS ecosystem participants to promote
their interests; for example, the Mobey Forum is a global, not-for-profit industry association focused on
the sharing of digital finance knowledge between financial institutions to help shape the future of digital
financial services.
Pdf_Folio:29
Digital Divide
Digital platforms within the ecosystem can widen the digital divide by serving only those with access to
smartphones and computers and risk deepening exclusion for those offline as more opportunities move
online. Key issues include the following.
Pdf_Folio:30
30 Digital Finance
• Internet access and affordability. Cost is still the main barrier for internet access. In Africa, the average
cost for 1GB of data is 7.12 per cent of the average monthly salary. In some countries, 1GB costs as
much as 20 per cent of the average salary (A4AI 2019).
• Gender gap. Despite growing internet penetration across the globe, the digital gender gap is widening
in some parts of the world. In Africa, women’s access to digital technology is 25 per cent lower than
men’s according to the International Telecommunication Union (ITU) (ITU 2019).
• Access to devices. In 2020, smartphones accounted for 70 per cent of connections (Cerwall, Jonsson &
Carson 2020). Platforms can widen the digital divide by serving only those with access to smartphones
and computers. Depending on the market, options for feature phones can significantly impact inclusion.
• Gap in digital skills. ITU found that 65 per cent of non-internet users in Africa, Asia and Latin America
did not use the internet because of a lack of skills (ITU 2019). Even when they have access to affordable
handsets, reliable data connectivity and stable power, many people do not know how to access digital
financial services or are not comfortable with technology. Digital literacy is needed to effectively close
the digital skills divide.
• Agent networks. As a result of these gaps, it can be hard for DFS providers to reach the bottom of the
pyramid with financial services and the benefits they bring. In many markets, DFS providers still target
the middle class. However, DFS providers can employ agent networks to close the digital divide and
reach low-income people in remote and rural areas.
Quality of Work
Digital ecosystems can expand income opportunities for people, particularly those of low income.
However, work through digital platforms often lacks the security and benefits of conventional work.
• Most digital platform workers — for example, Uber (Reis & Chand 2020) or Freelancer — are
independent contractors rather than employees. As such, they generally do not have the benefits and
workplace protections offered to traditional employees.
• Digital platform work can drive workers to informality. However, this is more prevalent for developed
markets since workers in emerging markets already have a high level of informal and self-employment.
• Most were not salaried employees before joining the digital platform, so they move from informal to
semi-formal employment rather than from formal to semi-formal employment.
• While digital platform work has benefits, including flexibility, autonomy and potential for increased
income, it also comes with risks that undermine the quality of work.
• Many digital platform workers have inconsistent and/or unpredictable income patterns, which can
present challenges if credit repayment is not flexible.
• Lack of worker protections and benefits, such as paid leave and insurance, result in long hours, high
pressure and on-the-job injuries. Some digital platforms set up partnerships to offer some benefits like
insurance at preferential rates.
Pdf_Folio:31
Competition
Because digital ecosystems need to scale, the space is likely to be dominated by only a few players. Limited
competition can affect fees and prices.
• Digital ecosystem providers rely on scale and network effects, which leads to market concentration.
• Providers that can afford to, may expand into new markets and/or users may dominate the market. They
can leverage their dominance in one sector (e.g. search or e-commerce) to gain dominance in another
sector (e.g. P2P payments, digital wallets). The data and customer relationships they gained in one sector
can be used as a competitive advantage in another area, where their competitors do not necessarily have
access to the same data. For example, Google and Apple preinstall their payment services (Google Pay
and Apple Pay) in their devices.
• A fight for market share can result in a pricing war that is detrimental for producers and suppliers
of services on the platforms, especially for generic services like transport and food delivery, where
platforms control pricing directly.
• Dominant companies can use bundling or tying strategies to restrict competition or customer choice.
• Competition laws and regulation are needed to address competition issues that are specific to the
digital economy.
32 Digital Finance
Within the payment ecosystem, a closed-loop payment network exists as a layer operating between
consumers and merchants. Components of this layer include Amazon Payments, Bitcoin, PayPal, Boku,
Sometrics and Dwolla.
In 2014–15, Google Wallet announced plans to launch a P2P payment service that would allow
free money transfers between users. Apple also launched a platform. These platforms developed into
Google Pay and Apple Pay (discussed in detail in example 1.6). The evolution of these technologies has
given rise to new authentication norms at the device level and positioned Google and Apple as viable
alternatives to banks for payments.
EXAMPLE 1.6
Energizer Effects
Apple Pay take advantage of Apple’s vertical integration efficiency through blending hardware, software
and user-friendly interfaces into a harmonious ecosystem. This quality is specific to Apple Pay and helps
provide a new revenue stream for the entire iOS ecosystem.
Privacy
Apple prioritises its customers’ privacy concerns. It does not store:
• what was bought
• where it was bought
• how much was paid.
Apple also refrains from exchanging credit card information with merchants during transactions, which
further boosts its security mechanism as card information is less likely to be compromised. Apple Pay
uses a device-specific number and unique transaction code to avoid the need to store the user’s card
number on their device or on Apple servers.
Pdf_Folio:33
Consumer Adoption
While many larger merchants have set up to accept Apple Pay, consumers still need to take a card or cash
to most stores, not just their phone. If consumers want to use Apple Pay at their grocery store, service
station, laundromat and street food stalls, then Apple needs to provide incentives for small merchants to
adopt the platform.
Security/Fraud Protection
Apple Pay payments are authorised using Apple’s fingerprint and Face ID technologies. Apple Pay
also uses a tokenisation system and security on the iPhone during the payment transaction to protect
against fraud.
We wanted to take the Apple Pay experience even further. We saw an opportunity to transform another
fundamental method of payment. That’s the credit card. We think Apple’s uniquely positioned to make
the most significant change in the credit card experience in 50 years (cited in Dignan 2019).
Source: Information from Apple 2020, ‘Apple Pay’, accessed June 2020, www.apple.com/apple-pay.
.......................................................................................................................................................................................
CONSIDER THIS
Based on the description of Apple Pay, what role do you think digital payments will play in your business or your
clients’ businesses? Will they be primarily used for B2C or B2B transactions or both?
SUMMARY
An organisation will be part of many digital ecosystems. From the organisation’s own perspective, its
ecosystem includes the integration of organisational departments, systems, technologies, customers and
external partners, enabled by software, ICT and data exchange within and external to the organisation.
Digital ecosystems are customer-centric and value is created and consumed through the relationships
between different ecosystem participants.
The effectiveness of an organisation’s digital ecosystem may be evaluated or improvements designed
using the set of questions and issues presented in figure 1.7. A digital ecosystem seeks to adopt systems,
technologies and cooperative partnerships that create synergies.
Pdf_Folio:34
34 Digital Finance
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
1.1 Evaluate the complexity of digital transformation and its impact on the work of accounting and
finance professionals to meet specific business needs.
• Financial services and finance functions exist within a digital ecosystem which emphasises the
integration of systems, data and services both internally to the organisation and with customers
and external organisations.
• The evaluation of a business’s processes and performance increasingly factors in its interactions
in the ecosystem, which create new ways of adding value, but involve complex interconnectivity.
• Digital ecosystems are based on data. Advanced analytical approaches provide a way to generate
insights from this data that enable customer-facing and back-office performance to be improved.
1.2 Analyse the complexity associated with the practical implementation of emerging technologies
and their impact on the finance ecosystem.
• Figure 1.7 presented a methodology based on questions and issues, including consultation through
an organisation, to evaluate current and potential uses of technologies and integrated systems in
the organisation’s digital ecosystem.
• Access to infrastructure and other enablers and willingness to engage with digital finance tech-
nologies varies greatly between institutions and different societies. An organisation must be aware
of the financial inclusion and exclusion implications of the technologies they use.
1.3 Support the adoption of digital technologies to achieve an organisation’s strategic goals.
• The methodology presented in figure 1.7 can be used to identify weaknesses and potential
improvements in an organisation’s digital ecosystem and thus enable the design of an improved
ecosystem based on the adoption of new technologies and greater integration.
Pdf_Folio:35
Pdf_Folio:36
36 Digital Finance
Current FinTech Solutions in Business
Jack Zhang, the co-founder and CEO of Airwallex, closely monitors the development of FinTech solutions
across the sector. He notes a range of emerging and current FinTech applications, including:
• the New Payments Platform released by the Reserve Bank of Australia to enable flexible, real-time and
data-rich payments (the UK and some other countries have implemented similar platforms to which
financial institutions can subscribe)
• AI enabled analytics to extract data from a variety of information sources to fulfil ‘know-your-customer’
(KYC) obligations under anti–money laundering (AML) legislation and achieve a fast onboarding of a
new consumer or business client
• machine learning enabled analytics to analyse big data to determine a credit score and thus decide
whether to offer credit to businesses or consumers
• increasingly open data access so organisations can, with permission, integrate information from multiple
organisations to better serve a customer or client; for example, transparency within the credit score sector
enables data from all the banks in Australia to be collected to create a holistic view of an individual’s
or organisation’s credit history
• real-time notifications of finance data and events to enable more timely financial management.
There are many more examples, with FinTech’s regularly developing new solutions.
.......................................................................................................................................................................................
CONSIDER THIS
The ability to create new customised services by integrating building blocks via APIs is one of the defining features of
the digital finance ecosystem. How aware are you of services relevant to your work or your clients’ businesses that
can be accessed in this way?
In essence, FinTech 2.0 is deconstructing financial services into their base elements. Some are back
office, some middle office and some are front office. Smart devices, integrated via apps, make the front-
office customer experience exceptional, API plug-and-play interfaces in the middle office make the
connection between front and back office simple and in real time, while machine learning, AI, deep data
analytics, cloud structures alongside blockchain shared ledgers are reconstructing back-office services.
Because most FinTechs focus on niches, businesses often need to subscribe to multiple FinTech
providers to meet their end-to-end digital finance needs. The flexibility to combine service components
from different suppliers is one of the benefits of FinTech, but this must be evaluated against the
extra complexity of dealing with multiple platforms and multiple service providers. Some FinTechs are
building towards a fully integrated, broader set of financial products that fulfil end-to-end business needs.
Example 1.8 explores these issues in relation to start-up electric bicycle company Bolt.
EXAMPLE 1.8
FinTech Mix
Bolt is a global start-up founded in Australia, which offers a subscription service for its electric bicycles.
Bolt is used by thousands of food courier delivery riders riding for Uber Eats, Deliveroo, DoorDash and
many others. The electric bicycles allow delivery riders to maximise delivery hours and their subscription
model provides riders with access to credit without taking on substantial debt.
Bolt uses a combination of FinTechs to meet its financial needs, including the following.
• Stripe to provide the payment gateway on Bolt’s website for collecting weekly subscription payments
from riders through credit/debit card payments.
• Airwallex to provide a global business account to collect investment funds from overseas investors and
to make payments to international electric bicycle suppliers.
• Brex to provide corporate cards for paying bills and expenses (e.g. software subscriptions).
Table 1.2 summarises the expansion of Airwallex’s offerings from a global payments and collections
platform through the addition of new financial products that are fully integrated into its core platform.
Pdf_Folio:38
38 Digital Finance
New products Payment gateway Borderless cards
(launched in 2020) Collect credit card Easily issue and manage
or e-wallet payments corporate cards for
(e.g. Wechat pay) from your team.
customers on your Pay domestic and
website. international bills, with 0%
international transaction fees.
Integrated expense
management
Auto-expense reconciliation
on employee corporate
card purchases into the
accounting software.
It is evident from table 1.2 that the new Airwallex Business Account would meet all of Bolt’s needs with
a single fully integrated solution. An integrated FinTech solution could deliver the following benefits.
• Improve cashflow. Using an integrated FinTech solution as both a business account and payment
gateway (for collecting card payments from customers), reduces the time business need to wait to
‘clear customer payments’. An integrated FinTech solution can clear this instantly as opposed to a
2–3 day waiting period, thus improving business cash flow.
• Simplify reconciliation process. Leveraging a corporate card solution that is integrated into a business
account and an expense management functionality further simplifies employee expense management,
without the need for third-party expense management applications.
• Easier access to credit. Access to cash flow is often a challenge for businesses, especially newer
businesses or start-ups. Using an integrated FinTech solution would help businesses consolidate
revenue and expense data into a single account to facilitate credit assessment.
Additionally, through an advanced user access management system, an integrated FinTech could allow
a business to provide restricted read-only access to external parties, such as lenders to assess suitability
for credit. Bolt will need to decide whether these types of benefits are more appealing than maintaining
its current FinTech mix, also considering switching costs and alternative end-to-end solutions on offer.
Source: Airwallex 2020.
.......................................................................................................................................................................................
CONSIDER THIS
How would you evaluate the relative merits of an all-in-one service, such as that provided by traditional financial
services providers, against a service constructed from components offered by FinTechs via API integration?
Mobile Point-of-Sale
Handling physical cash creates many problems for small businesses operating through pop-up shops or
farmers’ markets. Being able to take card payments using mobile devices has made it easier for retailers to
take payment from any customer. All they need is an internet connection, a smartphone and a card reader.
As NFC becomes standard in smartphones even a card reader becomes unnecessary.
Cross-Border Payments
International payments remain a field where a cheap, standardised option is needed. Most payment systems
remain linked to domestic banks or credit cards where currency transactions come with fees attached. The
growth in mobile wallets varies a lot between countries, but with the growing acceptance of cryptocurrency
people are more likely to have some form of digital wallet app on their computer or smartphone. Once this
is combined with acceptance by retailers international shopping will change significantly.
Checkoutless Stores
Increased automation of the in-store experience is still in its early stages. For many Australians supermarket
self-checkout or petrol station pay-at-pump is all they have experienced. However, as Amazon-go stores
have demonstrated, it is possible to serve customers in otherwise conventional stores without any human
interaction. In the extreme example there are no registers, checkouts or queueing as the shopper scans an
app on the way in, selects their shopping and walks out with the purchase costs automatically calculated
and charged to the shopper’s account.
Lessons for Accountants
What can an accountant take from this? Should they be worried that technology may put them out of work,
or can they see opportunities? It is important to remember that there are some things computers do well
and some things they are less able to manage. Computers are good at repetitive routine tasks, and poor at
creative tasks — at least for now. To remain relevant the accounting profession needs to move away from
anything a computer can easily do and concentrate on adding value by doing what computers cannot. In
a sense any accountant worried about technology making them unemployed should concentrate on being
‘more human’.
Pdf_Folio:40
40 Digital Finance
The changes FinTech is making to the retail sector are mainly to do with automation, processing and
data collection. Cashless payment services, open banking and general automation creates large volumes of
data. While technology is great at creating and recording data it still needs the human touch when the time
comes to analyse it with any degree of judgement. That is where accountants can continue to add value for
businesses.
Finally, accountants are well-established as key advisors to their clients. There is a strong opportunity to
use a digital mindset and digital literacy to help clients (and of course the accountant’s own organisation)
to understand and analyse the potential value of technology-driven innovations in their businesses.
Example 1.9 describes the use of asset tracking software to save costs, reduce environmental waste and
improve marketing team access to marketing assets.
EXAMPLE 1.9
Many companies follow strategies similar to ouvar’s, building new functionality into existing technology
to serve niche markets or are content just to offer one or two tools. While that may limit growth it also
limits competition. Other, larger, providers put out a suite of connected products aiming at medium to large
clients needing tools they can use over different business divisions.
Digital technologies are changing how all businesses work, not just retail. While traditional accounting
compliance and financial forecasting is still important, much of the work involved is being automated. As
just as much of the data is now available in real time, the demand is for real time reporting and analysis
for effective decision making. The challenge for accountants is to use technology to find ways to do their
work faster and more effectively.
To add value to an organisation, an accountant should try to develop a digital mindset. This entails
constantly looking at how new technology can help a business automate and simplify processes, and look
for opportunities to create value for customers by adding useful features to products or through better
service delivery. FinTech is making a wide range of new tools available for accountants.
.......................................................................................................................................................................................
CONSIDER THIS
Automating processes, data collection and analytics should be part of an accountant’s everyday experience.
Are they?
Pdf_Folio:41
Provide analysis of potential reasons for digital implementation fails and lessons learned from
these failures. Read the following article before answering the question: www.cio.com/article/
3211485/why-it-projects-still-fail.html.
Profitability
The first common characteristic is profitability. Financial institutions have historically been profitable
and collectively form a highly profitable segment. Savvy technology entrepreneurs and venture capitalists
are recognising that these market conditions in other industries have historically enabled technology
companies to develop and prosper. On the other hand, participants in the financial sector are partnering
with FinTech companies to reduce inefficiencies in their operations to cut costs and improve profitability.
Basic Necessity
Another common characteristic of financial institutions is that the financial services they offer are a basic
necessity. While one can imagine a world where the delivery vehicle for companies that provide basic
financial services is different and largely digital, it is difficult to imagine a world without financial services
for payments, deposits, lending, borrowing and investing.
Regulation
Regulation is yet another common characteristic of financial institutions and that regulatory burden has
increased over time, particularly since the corporate collapses of the early 2000s and the global financial
crisis of 2008–09. This offers opportunities for technology solutions that can alleviate regulatory/compli-
ance issues. Heightened regulation also tends to limit a traditional financial institution’s ability to innovate,
which increases its need and desire for FinTech partners and solutions.
42 Digital Finance
Equity
A significant proportion of the global population is unbanked or underbanked and applying technology to
the financial services industry is viewed as one way to both expand and improve services to this significant
proportion of the population. For example, smart phones with banking apps have had a revolutionary
impact on small business in parts of Africa where fixed line phones and bank branches are rarely found.
Additionally, financial health and literacy are global issues and a number of FinTech innovations offer
opportunities to improve financial health and literacy around the world.
Evolving Preferences
Millennials are driving change in finance and other industries. Those financial institutions that are well
positioned for Millennials will likely outperform those that are not. Millennials tend to be more comfortable
using digital channels for financial services. As they acquire more financial assets, financial institutions
that can leverage FinTech to meet Millennials’ preferences will garner a tailwind for forward growth.
FinTech has risen on the shoulders of technology developments. The development of mobile computing
and communications helped lay the foundations for today’s FinTech environment. However, this is now
mature technology that has been around for more than a generation. As Millennials gained influence,
consumer preferences have shifted towards using technology as a first means of interacting with financial
service providers and other vendors. When COVID-19 resulted in widespread lockdowns in 2020 many
other consumers and businesses had no choice but to interact digitally.
WEALTH MANAGEMENT
FinTech is well-placed to revolutionise various aspects of the wealth management value chain. Some key
areas are summarised in figure 1.9 and described further below.
Typical go-to-market activities include providing real-time data, news and analytics, market intelligence,
product and service development, client communications and performance monitoring. FinTechs currently
deliver intuitive, technology-enabled multi-channel experiences. This is also an area where social media
insights increasingly play a role through analysing ‘what the market says’.
Client onboarding is often an area of frustration for customers, due to the high number of risk-related
questions and repeated conversations required to set up an account. New entrants are solving these issues by
using customer and behavioural data and gamification techniques to automatically identify the risk profile
of clients, the loss acceptance levels, and to capture information as part of new regulatory requirements.
Incumbents should keep up by exploring opportunities for cloud-based utilities to perform risk functions
such as KYC, AML and surveillance monitoring extremely cost-effectively.
In terms of delivery or distribution, smart algorithms have led to the emergence of robo-advice, which
could potentially address the advice gap as adoption scales. A key opportunity area for incumbent players
will be to use hybrid models and goal-based advice. These solutions use complex algorithms to support
life-stage planning, taking clients’ full-life situation (today and in the future) into account. There is a clear
Pdf_Folio:43
PAYMENTS
Payments is arguably the oldest area of FinTech innovation. It is a core part of traditional financial
services but innovations such as cryptocurrency may fundamentally change financial services. Recent
developments include mobile wallets (e.g. Venmo, Alipay, Paytm), mobile payments (e.g. Square, iZettle,
SumUp), online checkout (e.g. Stripe, Klarna, Adyen), digital currencies (e.g. Bitcoin, ether, zCash) and
payments infrastructure (e.g. Digital Asset Holdings, SETL, Ripple). Example 1.10 describes payments
FinTech developed by Airwallex, which has featured in earlier examples in this module.
EXAMPLE 1.10
Airwallex
Airwallex is an Australian/Hong Kong company specialising in providing more efficient foreign exchange
services to business clients. Their main product is based around the three related services of providing
foreign currency accounts, foreign exchange transactions and international payments. Airwallex’s value
proposition is based on reducing the complexity and cost of these services.
Airwallex has developed a virtual credit card optimised for multi-currency transactions. It enables a
business to create and issue staff members with virtual cards, set up for use in any currency, all linked back
to a single Airwallex account. The use of a single account simplifies accounting and the ability to create
new virtual cards on demand reduces administration. In addition, Airwallex provides greater security — a
user can create a single-use virtual card so that a vendor cannot go on to use the card details fraudulently.
Security is further enhanced by customisable transaction limits for each virtual card. Virtual cards can be
stored in digital wallets and the multi-currency feature helps control transaction fees. The Airwallex offering
also provides detailed and real-time data collection to support analysis of transactions.
So, what are the accounting implications?
• Pricing analytics. More accurate data on international transactions will help price future transactions.
• Forex balances. Companies will find it much easier to keep funds in foreign currencies for future use,
but increasing accounting forex exposure.
Pdf_Folio:44
44 Digital Finance
QUESTION 1.7
Airwallex has developed Fintech solutions for businesses such as digital business banking and
virtual employee cards, foreign currency transfers and accounts. Airwallex also provides detailed
and real-time data collection to support analysis of transactions. How would the work of an
accountant be affected by these technology and support the organisation’s goal? Read the ‘The
Role of FinTech in Modernising Businesses’ article on the CPA Australia website before answering
the question. Search for the article in the search bar at www.cpaaustralia.com.au.
FINANCIAL MARKETS
Various FinTech innovations have been introduced to the financial markets. We discuss some of
these below.
P2P Lending
P2P lending (also known as marketplace lending) uses software to connect people and businesses who
have money, with people and businesses needing money, while mitigating risk through real-time credit
analytics. A number of approaches have been pursued, some of which are described in example 1.11.
EXAMPLE 1.11
.......................................................................................................................................................................................
CONSIDER THIS
Accounting and finance professionals advising SME clients or investors need to be aware of the opportunities
business finance P2P lenders provide, but also need to ensure they maintain a current awareness of variation in
offerings across this changing industry. If this role is part of your current or future work, how aware of the difference
options in P2P lending are you?
The main drivers of innovation in P2P lending have been regulation and a shift from personal to
institutional investment. In the United Kingdom, Zopa was originally regulated under fair trading law, not
financial law. Moving to formal financial regulation helps give the sector credibility. Financial regulation of
P2P lending also provides an interesting distinction between Australian and New Zealand law, as Australia
managed to find room in its existing regulatory structure for P2P lending to be introduced but this was not
possible in New Zealand. P2P lending has only been possible in New Zealand since the Financial Markets
Act 2013 replaced the much stricter Securities Act. The move away from personal lending to institutional
lending is also apparent in both Australia and New Zealand, with Harmoney, the highest profile New
Zealand P2P lender, shutting its website to new personal lenders in 2020.
Pdf_Folio:45
SME Finance
Innovations in SME finance enable investors to lend money directly to SMEs and start-ups, generating
a huge new flow of capital that would otherwise probably not exist. Two of the key innovations involve
invoice financing and crowdfunding.
Invoice Financing
Invoice financing often mixes the ideas of P2P lending with financial factoring. Factoring and discounting
has been used for years, but on the internet can grow from a small, relatively unknown, service to a standard
alternative for SME finance. Factoring and discounting are similar services that enable a small business
to be paid early for its outstanding invoices. In factoring, the company providing the finance collects the
invoice payment from the client and takes a commission for having paid the business early. For example,
if a business is owed $20 000 by a client who normally pays in sixty days, a factoring company will offer
the business $18 000 today and it, in turn, will get $20 000 sixty days later for a $2000, or 10 per cent,
commission. This is a form of peer-to-peer lending where the financing of all the outstanding invoices
comes from investors.
Although few local providers of invoice finance use the P2P model, some are making use of other
forms of FinTech to create new business opportunities. FundTap (FundTap 2020) is a New Zealand based
business offering invoice financing across Australia and New Zealand through an application linked to each
client’s cloud-based accounts. Online invoice financing works because the provider can access accounts
from a wide range of clients, giving companies like FundTap the ability to track payment patterns and use
analytics to quickly evaluate credit risk. This puts them in a strong position to accurately price the offer
for the SME client. The client retains the right to choose which invoices to borrow against. When each
invoice is paid FundTap will use a direct debit from the client’s account for repayment. Technically, online
invoice financing is closer to discounting than factoring as payments pass through the client’s accounts.
Online providers are able to provide finance for much lower value invoices than traditional providers as
their online model is less affected by economies of scale. Discounting smaller invoices also helps grow the
market by opening it to a wider range of SMEs, not just those with large receivables from trade customers.
The development of online invoice financing shows how a range of different developments can come
together to create FinTech innovation. We can see a possible starting point in the development of P2P
lending in the UK, where general lending to SMEs led to the innovation of specific invoice-linked lending
to SMEs. However, that form of P2P lending was not replicated in Australia or New Zealand. Instead
the local development of cloud accounting provider Xero created an incentive for linked application
development. Combining new business ideas in the UK with a cloud framework in New Zealand creates
new variants working for local businesses.
Product Crowdfunding and Crowdfunded Equity
Two crowdfunding models, project crowdfunding and crowd-sourced equity, provide new businesses with
an important source of funds. Project crowdfunding enables start-ups to obtain initial funding from their
potential customers through the pre-sale of goods or services. This provides an alternative to seeking bank
loans. The cost for a start-up to reach its target audience via marketing in the media can also be avoided
through the use of crowdfunding platforms such as Kickstarter, Indiegogo and Pozible.
Traditionally, banks have been reluctant to fund start-up SMEs because they have had no data to analyse,
to assess the risk and return of funding the SME. Project crowdfunding businesses overcome this, as the
funding of a crowdfunded business directly reflects the demand from the marketplace and customer. In
other words, the unknown data the bank is looking for — the demand for the product — is determined and
demonstrated upfront, before the business gets started.
Crowd-sourced equity, better known outside Australia as equity crowdfunding, takes this a step further
by allowing companies to sell shares directly to investors through the crowdfunding platform. Unlike initial
public offerings (IPOs) on the traditional share market, crowd-sourced equity offers are often directed at the
same potential customers a company would target in project crowdfunding. These offers are also exempt
from many of the standard IPO disclosure requirements, making the process far more accessible for small
companies needing to raise funds to grow.
Pdf_Folio:46
46 Digital Finance
The growth of equity crowdfunding in New Zealand indicates the importance of network effects
in platform-based FinTechs. Although network effects in social media suggest a single platform will
dominate, it is possible for more platforms to succeed if they are sufficiently different to each other.
Example 1.12 describes the different approaches of the first two licenced equity crowdfunding platforms
in New Zealand, Snowball Effect and PledgeMe. Crowdfunding is discussed in more detail in module 2.
EXAMPLE 1.12
.......................................................................................................................................................................................
CONSIDER THIS
In what circumstances might you advise a business to consider crowdfunding to raise funds?
Pdf_Folio:47
Neobanking
Neobanking refers to the sector created by emerging bank start-ups. Neobanks vary considerably. For
example, the neobanks established in China — YESBANK and WeBank — are very different to those
in Europe and the United States. The Chinese neobanks grew out of commerce and chat services and
therefore have a different ethos and look to banks like Soon in France, from AXA Group. In Europe there
is considerable activity in the neobank sector, with new bank start-ups from bunq in the Netherlands,
CheBanca! in Italy, Lunar Way in Denmark and N26 in Germany. The United Kingdom probably leads the
market for neobanks, however, as there are around forty new bank start-ups. UK neobanks include Atom,
Fidor, Loot, Metro, Monese, Monzo, Starling, Tandem and Zopa. These banks fall into two categories:
full banking services with a bank licence (e.g. Atom and Starling) and bank front ends such as Loot. All
of these neobanks are competing on the basis of giving users a better digital experience than traditional
financial companies.
The ability of neobanks to enter the market depends on the support of regulators. Neobanking in
Australia is still in its early stages, but Australia has now legislated to provide a regulatory ‘sandbox’ to
enable FinTechs to develop new products or services without having to go through the full and expensive
licencing normally required in the financial sector. Qualifying Australian FinTechs can operate for up to
12 months and have up to 100 retail customers without the usual licences. This provides FinTechs with
an opportunity to develop, but also pressure to move quickly to qualify for and obtain a licence and bring
their products to market.
To gain insight into how this works, consider Singapore, where a lot of pressure to succeed in FinTech
arises from competition with Hong Kong to be the dominant financial market in Asia. During 2019 the
Monetary Authority of Singapore (MAS) announced it would issue up to five new digital banking licences
in 2020 (Faridi 2020). Singapore is a prime target for neobanks which tend to appeal to younger customers
who do not value branch networks as a convenience, but instead see physically visiting a bank as something
best avoided. Neobanks also provide mobile users with detailed account dashboards up front and advanced
risk evaluation behind the scenes. Again for younger customers without much credit history these are
valuable services. Singapore and neighbouring countries have predominantly young demographic profiles
which suit neobanks.
General economic changes also support the neobank model. By avoiding branch networks and legacy
systems, neobanks have lower costs structures than traditional banks, and this translates to lower fees and
higher interest rates for savers. Given near-zero interest rates the slightest saving in fees or increase in
returns makes a huge difference.
EMERGING FINTECH
An understanding of FinTech developments to date provides some basis to examine emergent technology.
Edge Computing
‘Edge’ computing is associated with the IoT — the interconnection of smart devices over the internet.
Current IoT systems are often highly centralised with data and processing concentrated in central servers.
Edge computing involves using cheaper new processors to move storage and processing to ‘the edge’,
that is, to the IoT devices that traditionally only create raw data. This approach is:
• more efficient, as less data needs to be sent between the device and central server
• more reliable, as it is less vulnerable to data loss should connections be lost
• faster, as less data is exchanged and data is processed as needed within the device.
Pdf_Folio:48
48 Digital Finance
In addition, from an accounting perspective, edge computing may result in access to more detailed
data. For example, financial accountants may be able to access better data about asset use to evaluate
fair values; management accountants may find it provides relevant non-financial data for performance
evaluation; auditors will need to deal with data access issues.
Decentralised Finance
Perhaps edge computing will develop into something more challenging to traditional finance. Recent
developments in cryptocurrency and blockchain seek to incorporate the underlying distributed blockchain
structure into financial applications. This is known as decentralised finance (DeFi). DeFi is discussed in
more detail in module 2.
AI and Robo-advice
For many years now, digitisation, automation and AI have been expected by some to lead to computer-
based intelligence that will fundamentally change every aspect of society (Kurzweil 2005). Many small
changes have already occurred, not least in financial services and, in particular, in payments. Concerns
about jobs being lost to automation technologies have been somewhat balanced by new jobs created where
people can focus on higher value work.
FinTech innovators have drawn on automation and AI to provide new services, particularly low-cost
services. The financial services industry is transforming rapidly towards fee-only models and the use
of digital financial advisors. Automated investment services are being introduced as automated advice
systems (robo-advisors) become institutionalised.
FinTechs are experiencing pressure to counter the resurgent competition of incumbents. Some are
providing ‘Robo-as-a-Service’ platforms that enable businesses to access automated advice or automated
process technologies via a subscription rather than having to develop or purchase the technology. Others
are focusing on goal-based investing to achieve specific personalised investment outcomes for customers
rather than simply pursue the highest return. Platforms that can deliver personalised strategies enable a
way to compete on factors other than cost. Some FinTechs are focusing on smart compliance services.
These are discussed in module 5.
INNOVATION ROADBLOCKS
Many of the incentives for innovation covered in section 1.8 can also act to hinder future innovations.
For example, while regulations have been modified to enable FinTech innovation, regulators must remain
fundamentally concerned with protecting investors and the financial system as a whole. FinTech has not
made risks disappear. Indeed, the way FinTech has helped bring financial services to many of the world’s
unbanked or underbanked people means that the effect of future financial failures could be much broader
than in the past. Many FinTechs are limited to wholesale and professional investors, simply because they
cannot operate within the financial regulations designed to protect retail investors.
A lack of creativity can also be a roadblock to innovation. Despite all the discussion about FinTechs
creating a new financial system, to date most FinTech is a variation on a traditional finance model. Terms
like WealthTech, RegTech and InsureTech show how new technology is still rooted in old ideas.
Advanced analytics, AI and machine learning technologies often take place inside ‘black boxes’ — we
cannot review the algorithms they use — and they sometimes, for example, produce racist or sexist outputs
due to biases existing in the raw data. This is contrary to the usual careful documentation processes in
traditional finance and contrary to the expectations and needs of modern society.
Profitability and financial sustainability remain uncertain for many FinTechs. Margins are tight and
competition high. Although the majority of FinTechs are financed through venture capital and private
equity rather than public offers and IPOs, a major market correction would cause serious financial and
reputational damage to the industry and present an obstacle to further development.
INNOVATION OPPORTUNITIES
To finish on a more positive note, there are many current technologies still well short of reaching full
potential. Big data analytics, machine learning and AI all offer the possibility of automated yet personal
service. Blockchain, distributed ledger technology, stablecoins and DeFi possibilities are still just in the
early development stages.
Pdf_Folio:49
.......................................................................................................................................................................................
CONSIDER THIS
Knowledge of digital finance technologies, including automation, AI and analytics, and FinTech solutions are enablers.
How do you convert this enabling knowledge into solutions that help a business achieve its strategic goals?
SUMMARY
Numerous FinTech innovations have been developed by start-ups, BigTech and traditional institutions.
Some FinTech solutions are relatively comprehensive, while others are narrowly focused and intended to
provide a specific service or function within a digital ecosystem that is constructed by integrating multiple
‘building block’ components.
Accounting and finance professionals need to be aware of current and emerging FinTech technologies
and applications across various business functions and segments, including payments, wealth management
and financial markets. Knowledge of FinTech innovations enables the accounting and finance professional
to identify and evaluate them as potential solutions to business needs. This is the key step that links
knowledge and awareness to outcomes and will be the focus of the final part of this module.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
1.1 Evaluate the complexity of digital transformation and its impact on the work of accounting and
finance professionals to meet specific business needs.
• Accounting and finance professionals need to maintain a current knowledge of FinTech innovations
across multiple segments and how these innovations are being applied in business. This knowledge
enables evaluation of the potential of FinTech solutions to help meet specific business needs.
1.2 Analyse the complexity associated with the practical implementation of emerging technologies
and their impact on the finance ecosystem.
• FinTech solutions are often narrowly focused to address a specific niche requirement. This means
a digital finance ecosystem can be built by integrating components from multiple service providers
into a business’s own systems. A business can integrate these components using APIs, and
financial institutions and FinTechs can cooperate to provide services to customers or clients
through a single unified interface.
1.3 Support the adoption of digital technologies to achieve an organisation’s strategic goals.
• Knowledge of FinTech innovations enables evaluation of their potential role in the processes and
activities the organisation undertakes to achieve its strategic goals. This in turn facilitates adoption
of value-adding technologies into the business’s digital finance ecosystem.
Pdf_Folio:50
50 Digital Finance
PART D: DIGITAL TRANSFORMATION
Technologies and applications provide tools to more efficiently and effectively pursue an organisation’s
strategic goals, but digital transformation needs to be driven by the organisation’s people.
Every strategy in an organisation should be aligned with the overall organisational strategy. Digital
strategy is now often so fundamental to the organisation’s overall strategy that it not only aligns with
and supports it, but also serves as one of the major inputs to its development. Technology is often deeply
integrated into the ways an organisation creates and delivers value. As technology and its applications
evolve quickly, strategic management is embracing more flexible approaches.
In the modern business environment, an organisation is unlikely to remain successful against its com-
petitors if it fails to use technology and innovation to achieve efficiencies, respond to changing customer
or client needs and more effectively create value. This establishes an imperative to digitally transform the
business to achieve faster time to market, ongoing innovation, higher quality and operational efficiency.
Because digital technologies are deeply embedded into business processes, a digital transformation strategy
needs to be based on thorough evaluation of business needs and the appropriate technologies. A detailed
implementation plan also needs to be prepared and actioned.
Pdf_Folio:51
52 Digital Finance
WHAT IS A DIGITAL TRANSFORMATION STRATEGY?
The appeal of new technology to entrepreneurs is in the way it can make previously unworkable business
models viable. Consider, for example, the possibilities created by 5G cellular services which make faster
data downloads possible. Video content can be downloaded and watched in real time, automated devices
can communicate and coordinate, augmented reality (AR) can better utilise cloud processing instead of
depending on the limited processing power of mobile devices (Brandon 2019).
Faced with these and similar possibilities it is tempting to start making immediate changes to try to
obtain an advantage. Every new technology over the next decade will have businesspeople looking for
ways to turn that technology into valuable change.
A formal digital transformation methodology is described in the next section, but the basic approach
is to:
• analyse the organisation’s needs within the context of organisational culture
• set suitable objectives
• understand and document risks
• run pilot programs and tests
• ask employees for feedback
• proceed with the implementation
• analyse the effectiveness of the implementation.
Within this approach, the business must avoid:
• the problems of embracing the technology too early (e.g. internal resistance to change, or a lack of
internal stakeholders with adequate training on the new technology)
• implementation plans that cause excessive disruption
• any plan that cannot identify, monitor and evaluate meaningful change within the organisation.
A digital transformation strategy will be effective when the organisation can match the technology with
its culture, risks and business objectives.
While technology strategy is an input to organisational strategy, it is critical to separate business goals
and processes from digital transformation goals. Remember, the aim is to improve the business not to
disrupt to the point where it no longer functions.
Some organisations create a digital roadmap to guide their transformation into an intelligent enterprise
by identifying which digital technology components are a perfect fit for their ecosystem.
Opportunity Assessment
During the opportunity assessment phase, organisations need to identify business problems and objectives,
and then evaluate the potential for technology to solve those problems or contribute to achieving those
objectives. This requires a good understanding of the technologies and their applications. The opportunity
assessment should be consultative to gain insight into processes, challenges and ideas that exist throughout
the organisation. Example activities in this phase include:
• identify the organisation’s strategic objectives
• brainstorm innovation ideas as applicable within the organisation
• engage consultants to provide access to expertise and multiple perspectives
• identify potential technology and applications
• begin to evaluate the technology and applications, including their potential to add value to the
organisation and the risks and costs associated with them.
Pdf_Folio:53
Mobilise stakeholders
Transition
Mobilisation
The mobilisation phase is a crucial step that helps the organisation align its strategy and goals to the
transformation strategies and obtain buy-in from stakeholders, in particular the workforce. The buy-in of
the workforce is crucial for the organisation as teams will need to engage with and learn new ways of
working with technology and technology-enabled processes. A lack of buy-in can create resistance which
can undermine the ability of the organisation to successfully transform (Wyer 2019). Example activities
in this phase include the following.
• Map the current business processes, operations and systems to the target structure.
• Consult with stakeholders and placing value on their input.
• Ensure two‐way communication channels are open between stakeholders and those designing the
transformation.
• Promote and demonstrate the value the digital transformation will create for the organisation and
stakeholders.
Design
The design phase (Wyer 2019) articulates the digital transformation goals into an executable plan for
implementing the change throughout the organisation. With a roadmap in place, resources, including
human resources, can begin to be mobilised towards achieving specific aspects of the plan.
Pdf_Folio:54
54 Digital Finance
The plan will help ensure implementation remains aligned with the overall objectives of the transfor-
mation. Example activities in the design phase include the following.
• Create the implementation project plan.
• Create a detailed transformation roadmap for the development activities.
Test
The implementation of technologies, particularly those that are deeply integrated into systems and
processes, is complicated. Digital transformation is intended to re-shape the organisation and as such the
potential benefits, the risks and the costs are substantial. It is therefore prudent to evaluate and test the
design of the transformation when possible to do so. For example, a prototype of a new technology can be
tested for its integration with the organisation’s other technologies and/or systems. Those who will use or
interact with the new technology should be involved to ensure they have the skills to use it and to check
that it functions as intended. For example, if a manual data entry process is to be automated, test data can
be run through the system and the results can be checked to ensure the system works effectively, that the
data is managed in accordance with organisational policies and that processes are in place to deal with
‘exceptions’ that the technology cannot handle.
Example activities in the test phase include the following.
• Plan to continuously improve the prototype and its implementation.
• Test the technology.
• Test the integration of the technology with other systems.
• Test user acceptance.
Customer Engagement
Customer engagement can be very beneficial for technologies that affect customer experience. For
example, implementing a cloud accounting platform that can integrate with client accounting records will
proceed much more smoothly and is more likely to achieve its intended outcomes if customers are involved
in the design and implementation. Customer engagement allows digital transformation to be driven from
the outside-in by engaging with customers to seek input into product design and improve customer
experiences. Ideally, a customer-centric ecosystem should be based on dynamic market conditions and
the customer journey. Example activities in customer engagement include the following.
• Understand customer needs and priorities.
• Test the product or service with customers, including integration with their systems if applicable.
• Seek and evaluate feedback for improvements.
• Address customer concerns.
Transition
The transition phase sees the launch of the technologies and their full integration into the organisation’s
ecosystem of technologies, tools and processes. Example activities in this phase include the following.
• Transitioning responsibility for the ongoing use of the technology to the business team or shared service
team, depending on the enterprise organisation structure.
• Launching the technology for the transformation.
• Reinforcing objectives and performance measures.
EXAMPLE 1.13
EXAMPLE 1.14
Pdf_Folio:56
56 Digital Finance
the organisation prevented the IT team from quickly serving its business unit customers. Application and
server extension over time had become a significant negative impact on the IT operation’s bottom line.
A cloud platform was identified as a potential way to transform the organisation by removing its
dependence on unintegrated legacy systems.
Approach
The enterprise deployed an industry-leading design for a data centre with network virtualisation. A data
centre environment fully automated by software was set up in North America, Europe and the Asia–Pacific
regions. During the transformation period, more than 850 servers a month were migrated to the cloud.
An agile transformation strategy (including DevOps — the integration of software development and IT
operations) was implemented incrementally and iteratively over multiple sprints, leveraging kanban and
scrum methods for tracking and predicting program outcomes in an optimised sequence.
The build time for servers was reduced from 6 weeks to 90 minutes, and was able to integrate
compliance and regulatory requirements, performance monitoring, predictive analysis and enhanced
security.
Results
The business achieved savings in excess of USD200 million in the IT operations from this transformation
of infrastructure and services to the cloud.
To suit the organisation’s individual needs, 50–60 per cent of workloads were migrated to an
on‐premises private cloud while 20–30 per cent of workloads were migrated to a virtual public cloud
solution. The use of legacy infrastructure was reduced to less than 20 per cent of workloads, which itself
resulted in more than USD110 million in savings.
The new data centre operating model resulted in 25 per cent labour savings.
Conclusion — Cloud-Based Solutions
The pace at which businesses are being disrupted, along with the changing demands and expectations
of customers, are forcing enterprises to challenge their existing business and operational models and
leverage cloud‐based solutions across various functions within the enterprises to streamline and scale
operations and stimulate ideas for new products and services.
Cloud computing is now a matured platform for the enterprise to devise a strategy to move to the cloud
and benefit from the explosion of cloud opportunities available, whether it is a single cloud application or
a complete enterprise transformation.
The journey to the cloud involves setting the cloud strategy, identifying the right solution for the appli-
cation portfolio, and restructuring the infrastructure by adjusting the applications within the ecosystem to
take advantage of what the cloud can deliver.
.......................................................................................................................................................................................
CONSIDER THIS
How would a digitally transformed ‘to-be’ enterprise architecture for your organisation or client’s organisation differ
from its current business model?
VALUE DRIVERS
A value driver is anything that can be done to increase value. For example, from the business’s perspective
a value driver would improve profitability, promote growth, reduce risk or contribute to some other
outcome that aligns with the organisation’s strategic objectives. Value drivers from a customer perspective
are anything that improves the value of a product or service relative to a competitor’s product or service.
58 Digital Finance
The general benefits of digital transformation to business span improved resource management,
employee empowerment, better customer insights, easier globalisation, better internal collaboration and
transparency, as well as helping develop a generally agile and innovative culture.
Digital transformation of the corporate accounting function moves restricted access, centralised record-
keeping to an open, decentralised model with real-time data access. Managing accounts receivable and
invoices is traditionally labour-intensive due to the need to keep records as complete, accurate and up-to-
date as possible. Automated systems not only save processing time, but improve analysis of credit risk,
thus reducing bad debts, and can predict which accounts are likely to be problematic and highlight them
for early intervention. Likewise, better accounts payable management can help a business maximise early
payment discounts and benefit from strong vendor relations.
In audit, there is likely to be continued digitisation as distancing and technology change the way the
work is performed. Provided an organisation’s data is in a suitable, machine-readable form, the benefits
are more than simply better access. Digitalisation of audit processes will help increase security by allowing
a digital trail of when and by whom each file was accessed. Digitisation also has the potential to evolve
auditing from sampling accounts to automated checking of every transaction undertaken.
1. Public relation value versus real results. Often organisations exhibit ‘innovation theatre’ or ‘innovation
optics’ to make it look like they are adopting an innovation mindset and practices, when in reality they
are not internalising or acting on it.
2. Hampered by heritage. There is simply too much legacy infrastructure to be able to adapt and change.
3. No real sense of urgency. Established and successful businesses often fail to perceive a sense of
urgency to change. In the case of banks, for example, what usually drives them is not customers or even
emerging FinTechs, but regulators. This promotes the view that unless it is mandated, why bother?
4. Cannibalisation of existing revenue streams. Some digital innovations pose a threat to existing
revenue streams. Many organisations and their stakeholders resist innovations that risk or harm
established revenue. Often, this means organisations focus on short-term results rather than longer-
term sustainable performance.
5. A lack of experienced innovators. The hierarchical nature of many businesses tends to act to eradicate
innovation before it can be acted on. This prevents an innovation culture from being established and
extinguishes the motivation for employees to drive innovation from within.
6. A culture clash. Many companies have a culture that emphasises risk minimisation, which directly
conflicts with the pursuit of change. In such a culture, once there is emerging resistance to change,
that resistance tends to be quickly adopted throughout the organisation.
7. A lack of ownership and sponsorship. The leaders of organisations attribute their success to past
approaches and their knowledge of the organisation and industry. They often resist engaging with
the perceived risks of technology that they do not understand. A lack of support at the top of the
organisation makes it almost impossible to pursue innovation.
8. ‘Compartmentalised’ innovation. Innovation takes place around the periphery in departmental com-
partments but is never internalised across the enterprise. This undermines the value technology-
enabled innovations can deliver through deep integration.
9. Governance. Regulatory requirements are, by their nature, built around the current and past state
of an industry. They tend to constrain innovation and thus discourage organisations from exploring
different ways of conducting their business.
10. Who else is doing this? While some organisations base their strategy on differentiation, many
organisations prefer to adopt and maintain industry-wide approaches. In essence, ‘If no one else
is doing it, then we’re not either’.
Source: CPA Australia 2021.
Pdf_Folio:59
QUESTION 1.8
EXAMPLE 1.15
Getting Started
The digitisation process may begin by taking an inventory of core use cases relevant to the business and
determining where each digital technology could support an operational objective. The potential value
gained from digitisation of a finance process and the feasibility of doing so should be assessed. Unit leaders
should be engaged to identify and explore the key problems in existing processes, such as slow reporting
Pdf_Folio:60
60 Digital Finance
and incomplete data. Finally, a full systematic review of technology capabilities should be undertaken to
define system requirements and investments.
It is important to have a clear vision of the target state for a digital finance function, and that this is aligned
to the company’s overall business strategy. In addition, chief financial officers (CFOs) need to ensure that
digitisation and transformation adheres to strong cybersecurity standards (Choi, Kaplan & Lung 2017).
The CFO
The high-profile digital transformation or innovation stories that get media attention tend to be about
customer-facing processes improving customer engagement. These are often driven by chief marketing
officers and chief digital officers, but what of the CFO? The finance function is central to organisational
change as it allocates resources. Accounting and finance must take the lead in using digital technologies
and make the most of its resources. Digital technology can enable the CFO to help build competitiveness.
Finance has traditionally been a specialist and isolated function within business. Accountants focused
on recording transactions correctly, ensuring that suitable controls were in place and external reporting
existed for compliance. Controls focused on specific business units with plans, budget and profitability
reports. However, in a networked and collaborative world, finance needs to work closely with all other
company functions to solve problems and explore solutions. Recent developments in digital enterprise
technologies can help the CFO transform the finance function. In the following section we will outline
some of the ways the CFO can become involved in digital transformation.
Access to Data
Fast changing and volatile business environments require fast access to the latest data that can be used
for forward planning. Accounting and finance professionals need a single source of truth — data collected
from multiple, possibly inconsistent sources, need to reconcile. Real-time data at the finest granularity level
also allows analysts to run simulations and predictive analytics and explore multiple options. Automatically
reconciled and compliant data platforms help finance departments do a ‘soft close’ anytime.
Involvement in Day-to-Day Operations
In high-performing organisations CFOs are increasingly involved in operations, providing insight to
business units and using agreed metrics to hold them accountable for results, not simply tracking
transactions. The CEO can focus on evolving strategy and communicating with stakeholders, but the line
between CFOs and chief operating officers (COOs) is blurring to the point where the CFO performs some
functions normally assumed by the COO.
The old accounting work of recording, classifying and summarising financial data has transferred to
machines. Where accountants are needed is providing expert interpretation of financial data in analytics.
Interpretation involves an interpersonal capability, nuance and qualitative judgement that remains outside
the ability of machines. Interpretation is where the finance function can best deliver the value to the
organisation.
An Understanding of Data Interpretation
Accounting and finance professionals cannot assume that data interpretation will necessarily become their
role in a digitalised future; they need to prove they deserve the role by showing how they can use it to
create value. That will be difficult if the finance unit is not making full use of digital transformation itself.
To provide valuable operational advice, the finance function’s operating model must maximise its own
efficiency and agility.
Combining Soft Skills with Technological Solutions
Soft skills are becoming more important as routine work is automated. Communication, collaboration and
teamwork skills dominate in senior management. Successful businesses will use evolutionary business
models, constantly embracing new technologies and adapting to the changing business environment. AI
may suggest a solution, but people with influence and communication skills are required to make it real.
62 Digital Finance
career development. This will help ensure as many stakeholders as possible are prepared to contribute to
transformation at the fast pace required by the modern digital world.
.......................................................................................................................................................................................
CONSIDER THIS
If ‘organisations need to change the role of finance’, how will this be done? Who will create this change? What is your
responsibility in this context?
QUESTION 1.9
Compare and contrast the roles taken by the CFO and the management accountant in
implementing digital transformation.
QUESTION 1.10
Identify the changing roles of accounting and finance professionals in supporting digital
transformation in their organisation. Read the following article before answering the question:
www.genpact.com/downloadable-content/insight/leap-on-new-f-and-a-priorities-and-
expectations.pdf.
FIGURE 1.12 Key questions for the enterprise before proceeding with transformation
Pdf_Folio:63
Internet of Things
Technology has given people the flexibility to use mobile smart devices from almost any location, and this
is further enhanced by wearable devices such as smart watches, smart tags, and smart glasses with AR. To
connect these devices and collect the data they generate, the enterprise needs the network layer supported
by IoT.
Through sensors and other devices, this layer creates a vibrant marketplace by connecting employees,
partners, suppliers, vendors, customers and the general public with the enterprise’s business processes. It
supports internal and external collaboration that can optimise business processes and create new business
opportunities.
Pdf_Folio:64
64 Digital Finance
FINANCIAL INNOVATION IN ORGANISATIONS — LOCKING IN
THE BENEFITS
To digitally transform, an organisation needs to evaluate all the emerging technologies against industry
trends, business relevance, technical feasibility and the overall ecosystem. The enterprise should then
identify and prioritise those emerging technologies. Those that have the highest priority and biggest impact
should be implemented as soon as possible. Those that have broad application should be on the horizon to
be implemented in the next one to three years (short-term deployment). Those that should be considered as
part of the enterprise’s long‐term technology strategy and roadmap should be prioritised to be implemented
in the next three to five years. Figure 1.13 sets out a rough timeframe for technology deployment, though
this should be adapted to the organisation’s individual circumstances. Some of these technologies will be
familiar. Many of them will be discussed in more detail in following modules.
EXAMPLE 1.16
Pdf_Folio:65
Pdf_Folio:66
66 Digital Finance
Christos responded first, ‘Pretty much everything I have seen in this proposal concentrates on digital
finance technologies, and that is just one side of the cube. If Evergreen accepts the project, we really
should think about where the other sides fit in. We will need to arrange financing —should we use traditional
sources of funds or should we extend the digital transformation to use digital finance business functions,
which are viable here? That will also help determine how we deal with the third dimension — are we going
to work with traditional financial institutions or look to the new digital financial service providers?’
At first no one responded. Then Sara spoke up. ‘Well, that is a big question. I think we all need to spend
some time reflecting on these issues before we come back and weigh up the possible answers.’
If work had seemed to move quickly for CIO Cameron O’Dell while preparing the digital transformation
strategy, that was nothing compared to the following months. Since the proposal was adopted, a
seemingly endless stream of meetings has converted the blueprints into detailed plans for implementing
Evergreen Market’s digital transformation.
Cameron had noticed that along with the complexity of planning the implementation, he had also
started to face some resistance and pushback, as people within the business started to realise that their
comfortable and familiar lives were about to be disrupted — and they started to worry about job security.
Cameron also had some influential supporters. One of the biggest, CFO Sara Wong, was encouraging
him to be even more ambitious, so the benefits could be realised as soon as possible. In particular,
Cameron was under pressure to get the in-store technology and mobile phone app up and running across
all stores as soon as possible. Strangely, Sara was also proving to be Cameron’s biggest critic, raising
numerous concerns about the implementation, and forcing Cameron to question every aspect of the
plan. Sara said that it was part of her role to ensure all reasonable scenarios had been considered and
to examine every possible risk to the business and its profits. Cameron suspected that Sara just liked
playing Devil’s Advocate and making him second-guess himself.
When Sara discussed the project with Cameron, she was likely to raise all kinds of risk-management
issues and identify potential problems. Mostly these had an accounting focus, and sometimes looked-
into the details of the technology, but when it came to raising questions Sara almost seemed proud of
identifying issues outside either Cameron’s or Sara’s fields of expertise. For example, Sara was very
concerned about both the legal and ethical implications of customers’ privacy needs. A major reason
for her concern was the use of 5G and IoT. 5G’s speed of data transfer meant a lot more data could be
stolen before the business had a chance to respond to a successful cyber-attack. IoT would open many
more gateways or attack surfaces a hacker could exploit to access the data.
Cameron was impressed by Sara’s work. She did not sound like Cameron’s idea of a traditional accoun-
tant when she started asking about staffing and training. She had clearly seen that the transformation was
not just about the technology. Employee interaction with the new systems would be just as important
as the hardware and applications. Given the scale of the technological change, organisational change
management would be required to change company culture across all team members. When Cameron
raised this with Sara, she just brushed it off lightly, saying that it was all part of her training to consider all
of a project’s costs and benefits, which was more what Cameron expected an accountant would say.
A more mundane question needing serious consideration was whether the company should outsource
its cloud or build its own proprietary server farm. While outsourcing was common practice, Cameron
was aware of Australian companies having problems when offshore cloud services failed to provide full
technical support during Australian business hours. Given the importance of minimising system downtime
to the success of the transformation plan, Cameron was very tempted to push for Evergreen to take on
the capital cost of setting up private servers in each state. Sara agreed that keeping servers private and
distributed across the country would aid security and improve latency statistics, but she was wary of the
added cost.
Sara had also been pushing Cameron to incorporate digital finance business functions into the plan;
something she would refer to as part of the Digital Finance Cube. This had always been on his mind as
part of the bigger picture, part of a longer term plan where the current project was just the first step, but he
had not incorporated it into the initial plan as he thought it would make it too big for the board of directors
to accept. There was room to incorporate digital trade finance into the program where suppliers were
already set up with the required technology. He was willing to allow a wider range of digital payments and
limited use of digital money from customers. However, he was wary of being too much of a market leader
in these developments and being stuck with expensive but worthless technology should the market end
up adopting a different standard.
Cameron’s background had included time in two start-up companies. Although he was tempted to
approach this project as a start-up —given the amount of change he often thought of this creating a
whole new company — he knew the organisational culture would not allow that to happen. One idea he
particularly did not want to import from the start-up environment was ‘failing fast’. In a business as large
and traditional as Evergreen Markets any failure would affect hundreds of staff, thousands of customers
and would simply be too damaging to brush off and walk away from.
Pdf_Folio:67
Pdf_Folio:68
68 Digital Finance
‘When I look at the competition, I have to worry that all this work and expense will bring no net change
to our market share. There is nothing in the transformation program that is not being picked up and
duplicated (in many cases it is just a rough copy using more basic technology) by our competitors.
‘Despite that, our early gains in customers and market share might be sustainable, at least for now. The
key difference seemed to be the way Evergreen Markets integrated all the changes and made sure they
worked together and complemented each other. Our competitors only saw a part of what Evergreen was
doing. By failing to address the holistic nature of the digital transformation our competitors have let us
develop a sustainable competitive advantage.’
.......................................................................................................................................................................................
CONSIDER THIS
Evergreen Markets attributes much of its recent success to the holistic nature of its digital transformation —
and it is considering going further by distributing financial products offered by FinTechs. It believes it may have
established a sustainable competitive advantage. Evaluate the pros and cons of taking such an all-encompassing
approach to digital transformation. Would you have done the same? Why/Why not? Which stakeholders do you think
have benefited from Evergreen’s transformation and how? Do you agree Evergreen has a sustainable competitive
advantage?
SUMMARY
Digital transformation needs to be driven by the organisation’s people and align with the goals detailed in
the organisation’s overall strategy. The pace of change in technology and customer expectations creates
an imperative to adopt technologies to achieve efficiencies and improve the value of the organisation’s
products and services.
A digital transformation strategy should be based on an evaluation of how available technologies can be
used to meet business needs. A detailed implementation plan needs to be prepared and actioned.
Accounting and finance professionals have specific roles to play in planning and implementing digital
transformation. Their work will be heavily impacted by digital technologies, especially automation,
analytics and AI technologies. It is therefore important accounting and finance professionals engage with
digital transformation to ensure it achieves its intended outcomes for the organisation and enables future
accounting and finance work to be effective and efficient and thus create value for the organisation.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
1.1 Evaluate the complexity of digital transformation and its impact on the work of accounting and
finance professionals to meet specific business needs.
• The digital transformation of accounting and finance is affected by numerous technologies. The
main ones are automation, analytics and AI which individually and in combination will re-shape the
work the profession does and how it creates value for businesses.
• Accounting and finance professionals need to engage with digital transformation to ensure it is well
planned and implemented, achieves the intended outcomes and results in effective solutions for
the accounting and finance team and the organisation more broadly.
1.2 Analyse the complexity associated with the practical implementation of emerging technologies
and their impact on the finance ecosystem.
• A digital transformation implementation methodology should be used to design and execute the
digital transformation. Ongoing monitoring and evaluation are required to ensure the expected
outcomes and continuous improvement are achieved.
1.3 Support the adoption of digital technologies to achieve an organisation’s strategic goals.
• Rapid changes in technology and the business landscape have created an imperative for organisa-
tions to digitally transform. While many factors work against innovation in organisations, accounting
and finance professionals should evaluate the potential of technologies and drive or support
adoption of those that will add value to the organisation.
Pdf_Folio:69
REFERENCES
A4AI 2019, ‘2019 Affordability report’, A4AI, October, accessed June, https://a4ai.org/affordability-report/.
Agrawal, A, Dinneen, B & Seth I 2016, ‘Are today’s CFOs ready for tomorrow’s demands on finance?’, McKinsey, 9 December,
accessed June 2020, https://www.mckinsey.com/business-functions/strategy-and-corporate-finance/our-insights/are-todays-cfos-
ready-for-tomorrows-demands-on-finance.
Airwallex 2020, ‘Crockd, how an Aussie side hustle accelerated into a global brand’, Airwallex blog, 7 September, accessed April
2021, www.airwallex.com/blog/crockd-customer-story.
Alves, C, Oliveira, J & Jansen, S 2017, ‘Software ecosystems governance: a systematic literature review and research agenda’,
International Conference on Enterprise Information Systems (ICEIS), Conference Paper, January, accessed June 2020,
https://www.researchgate.net/publication/316284388_Software_Ecosystems_Governance_A_Systematic_Literature_Review_
and_Research_Agenda.
Apple 2020, ‘Apple Pay’, accessed June 2020, https://www.apple.com/apple-pay/.
Bersin, J 2016, ‘Digital leadership is not an optional part of being a CEO’, Harvard Business Review, 1 December, accessed
June 2020, https://hbr.org/2016/12/digital-leadership-is-not-an-optional-part-of-being-a-ceo.
Blystone, D 2020, ‘Search engines that compete with Google’, Investopedia, 5 February, accessed June 2020, https://www.
investopedia.com/articles/investing/071515/search-engines-compete-google.asp.
Bock, R, Iansiti, M & Lakhani, R 2017, ‘What the companies on the right side of the digital business divide have in common’,
31 January, accessed June 2020, https://hbr.org/2017/01/what-the-companies-on-the-right-side-of-the-digital-business-divide-ha
ve-in-common.
Brandon, J 2019, ‘What is a digital transformation strategy?’ TechRadar, 13 November, https://www.techradar.com/news/what-is-a
-digital-transformation-strategy.
Cerwall, P, Jonsson, P & Carson, S (eds) 2020, ‘Ericsson mobility report’, June, accessed June 2020, https://www.ericsson.com/en/
mobility-report/reports.
Choi, J, Kaplan, J & Lung, H 2017, ‘A framework for improving cybersecurity discussions within organisations’, November,
McKinsey.com.
Courbe, J 2016, ‘Financial services technology 2020 and beyond: embracing disruption’, PwC, accessed April 2021, www.pwc.
com/gx/en/financial-services/assets/pdf/technology2020-and-beyond.pdf.
Crowdcube 2020, ‘BrewDog Equity for Punks V’, accessed June 2020, https://www.crowdcube.com/companies/brewdog-plc-1/
pitches/bk2BkZ.
Currencycloud 2020, accessed June 2020, https://www.currencycloud.com/.
Dastin, J & Baertlein, L 2020, ‘Amazon pauses sellers loan repayments amid coronavirus’, Reuters, 26 March, accessed
June 2020, https://www.reuters.com/article/us-health-coronavirus-amazon-exclusive-idUSKBN21C3GM.
DE-CIX 2021, ‘GAIA-X: a dynamic and secure way to cloud and edges: working together to build a new generation digital
ecosystem’, accessed April 2021, https://www.de-cix.net/en/about-de-cix/research-and-development/gaia-x.
Deloitte 2019, ‘Moving to the future with finance’, accessed February 2021, https://www2.deloitte.com/content/dam/Deloitte/us/
Documents/finance-transformation/us-moving-to-the-future-with-finance.pdf.
Pdf_Folio:70
70 Digital Finance
Detrixhe, J 2020, ‘Apple Pay is on pace to account for 10% of all global card transactions’, Quartz, 12 February, accessed
June 2020, https://qz.com/1799912/apple-pay-on-pace-to-account-for-10-percent-of-global-card-transactions/.
Dignan, L 2019, ‘Apple’s next big innovation: becoming a bank via Apple Card’, Tech Republic, 25 March, accessed June 2020,
https://www.techrepublic.com/article/apples-next-big-innovation-becoming-a-bank-via-apple-card/.
Faridi, O 2020, ‘Fintech arrival bank is applying for a full service digital banking license in Singapore’, Crowdfund Insider,
8 March, accessed June 2020, https://www.crowdfundinsider.com/2020/03/158452-fintech-arival-bank-is-applying-for-a-full-
service-digital-banking-license-in-singapore/.
FDIC 2017, ‘FDIC national survey of unbanked and underbanked households’, accessed June 2020, https://www.fdic.gov/
householdsurvey/.
Finder 2020, ‘eToro launches zero brokerage share trading in Australia’, accessed June 2020, https://www.finder.com.au/etoro-
launches-zero-brokerage-share-trading.
Fong, C, Huang, J, Robinson, K & Ungerman, K 2019, ‘Prime day and the broad reach of Amazon’s ecosystem’, 2 August,
accessed June 2020, https://www.mckinsey.com/business-functions/marketing-and-sales/our-insights/prime-day-and-the-
broad-reach-of-amazons-ecosystem.
FundTap 2020, accessed June 2020 https://fundtap.co.nz/.
Gomber, P, Koch, J & Siering, M 2017, ‘Digital finance and FinTech: current research and future research directions’, Working
Paper (January 2017) Forthcoming in the Journal of Business Economics, vol. 87, pp. 537–580; doi:10.1007/s11573-017-
0852-x, accessed June 2020, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2928833.
Google Pay 2020, ‘About’, accessed June 2020, https://pay.google.com/about/.
Grunewald, A & Maclay, C n.d., ‘Lynk: a technology platform connecting informal workers to new clients’, accessed June 2020,
https://solve.mit.edu/challenges/work-of-the-future/solutions/5368.
Hamel, G & Prahalad, CK 1989, ‘Strategic Intent’, Harvard Business Review, May–June, pp. 63–76, cited in Schermerhorn, J,
Davidson, P, Factor, A, Poole, D, Woods, P, Simon, E & McBar, E 2016, Management: Asia–Pacific Edition, 6th edn, John
Wiley & Sons Australia.
Hammond, K 2001, ‘Michael Porter’s Big Ideas’, Fast Company, March, cited in Schermerhorn, J, Davidson, P, /Factor, A, Poole,
D, Woods, P, Simon, E & McBar, E 2016, Management: Asia–Pacific Edition, 6th edn, John Wiley & Sons Australia.
Hitt, M, Ireland, R & Hoskisson, R 1997, Strategic Management: Competitiveness and Globalization (Minneapolis: West, 1997),
p. 5, cited in Schermerhorn, J, Davidson, P, Factor, A, Poole, D, Woods, P, Simon, E & McBar, E 2016, Management: Asia–
Pacific Edition, 6th edn, John Wiley & Sons Australia.
Information and quotes from Stepanek, M 1999, ‘How Fast Is Net Fast?’, Business Week E-Biz, 1 November, pp. EB52–EB54,
cited in Schermerhorn, J, Davidson, P, Factor, A, Poole, D, Woods, P, Simon, E & McBar, E 2016, Management: Asia–Pacific
Edition, 6th edn, John Wiley & Sons Australia.
InvestSMART 2020, accessed June 2020, https://www.investsmart.com.au/.
Iqbal, M 2020, ‘WeChat Revenue and Usage Statistics (2020)’, 13 May, accessed June 2020, https://www.businessofapps.com/
data/wechat-statistics/.
ITU 2019, ‘Statistics’, accessed June, https://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx.
Jacobides, M 2019, ‘Amazon’s ecosystem grows bigger and stronger by the day: should we be worried?’, Forbes, 9 May, accessed
June 2020, https://www.forbes.com/sites/lbsbusinessstrategyreview/2019/05/09/amazons-ecosystem-grows-bigger-and-stronger
-by-the-day-should-we-be-worried/#451bc12b31ea.
Jacobides, M, Lang, N Louw, N & von Szczepanski, K 2019, ‘What Does a Successful Digital Ecosystem Look Like?’, BCG, 26
June, accessed June 2020, https://www.bcg.com/en-au/publications/2019/what-does-successful-digital-ecosystem-look-like.
aspx.
James, A 2020a, ‘Sustainability driven innovation: service innovation’, PennState College of Earth and Mineral Sciences, accessed
June 2020, https://www.e-education.psu.edu/ba850/node/722.
James, A 2020b, ‘Sustainability driven innovation: structure innovation’, PennState College of Earth and Mineral Sciences,
accessed June 2020, https://www.e-education.psu.edu/ba850/node/701.
Kabai, I 2017, ‘Measuring the performance of enterprise architecture’, Information Week, 1 November, accessed June 2020,
https://www.informationweek.com/devops/measuring-the-performance-of-enterprise-architecture/a/d-id/1327872.
Keeley, L, Pikkel, R, Quinn, B & Walters, H 2013, Ten types of innovation: the discipline of building breakthrough, John Wiley &
Sons, Koboken, NJ.
Kickstarter 2020, ‘About us’, accessed June 2020, https://www.kickstarter.com/.
Kirkegaard, M 2020, ‘Brewdog launches $10m Australian equity raise’, BrewsNews, 15 December, accessed February 2021,
https://www.brewsnews.com.au/2020/12/15/brewdog-launches-10m-australian-equity-raise/
Kruppa, M & Murphy, H 2019, ‘’DeFi’ movement promises high interest but high risk’, Financial Times, 30 December, accessed
June 2020, https://www.ft.com/content/16db565a-25a1-11ea-9305-4234e74b0ef3.
Kurzweil, R 2005, The singularity is near: when humans transcend biology, Viking Press, cited in Taaffe, O (ed) 2019, Banking on
Change: the development and future of financial services, p. 199, John Wiley & Sons, Chichester, West Sussex UK.
MarketFinance 2020, ‘Press’, accessed June 2020, https://marketfinance.com/press.
Marketlend 2020, accessed June 2020 https://www.marketlend.com.au/.
Marr, B 2016, Big Data in Practice (use cases), John Wiley & Sons, Chichester, West Sussex, UK.
McKinsey 2017, ‘Financial ecosystems: the next horizon for US banks’, 12 July, accessed June 2020, https://www.mckinsey.com/
industries/financial-services/our-insights/financial-ecosystems-the-next-horizon-for-us-banks.
Miller, C 2019, “Facebook Pay’ takes on Apple Pay with in-app purchases, Messenger transfers, more’, 9 to 5 Mac, 12 November,
accessed June 2020, https://9to5mac.com/2019/11/12/facebook-pay-apple-pay-competitor/.
Mintzberg, HH 1978, ‘Patterns in Strategy Formation’, Management Science, vol. 24, pp. 327–36, cited in Schermerhorn, J,
Davidson, P, /Factor, A, Poole, D, Woods, P, Simon, E & McBar, E 2016, Management: Asia–Pacific Edition, 6th edn,
John Wiley & Sons Australia.
MoneyPlace 2020, accessed June 2020 https://moneyplace.com.au/.
Pdf_Folio:71
Pdf_Folio:72
72 Digital Finance
MODULE 2
FUTURE OF MONEY
LEARNING OBJECTIVES
PREVIEW
As explained in module 1, every business is positioned within an ecosystem of value- driven economic
activities. The monetary system — and overall financial system — is transitioning to technology-enabled
business models that can bring further democratisation, inclusion, efficiency and effectiveness, and open
up new pathways for innovative business models.
Achieving a transient or competitive advantage requires every organisation and professional to under-
stand the reason and rationales behind these changes. This will help evaluate and establish business cases
that adopt appropriate digital technologies and applications within an organisation’s strategic, business,
innovation and financial reality. In this module we explore technologies and applications for digital
currency, digital payments, distributed ledgers, cryptocurrencies and alternative finance. We conclude with
an assessment of the future direction of banking.
Part A of the module describes the functions and characteristics of money to establish a basis for
evaluating the potential applications of a variety of digital currencies. Digital currencies involve new
financial and monetary infrastructure and processes that require evaluation and management in terms
of business strategy, operational strategy, regulation, compliance, accounting, ethics, societal impacts
and finance.
Part B of the module describes the characteristics and applications of a variety of digital payment
technologies. Some of these are well-established, widely adopted and exist as a formalised part of the
financial system. Others are being developed using various financial technologies to overcome weaknesses
of existing systems and to provide personalised, frictionless payment experiences to customers. A
common feature of digital payments is the potential to capture rich data associated with each transaction.
Part C of the module describes the concept of distributed ledgers and the mechanisms that enable
blockchain platforms to provide secure transactions without the need for a central trusted authority and
to create a secure, transparent and immutable ledger. We also discuss smart contracts, which are self-
executing contracts that operate on a blockchain platform and a range of other blockchain applications.
Pdf_Folio:73
Part D examines cryptocurrencies, which are a type of decentralised virtual currency that exists on
blockchain platforms. Cryptocurrencies are seen as a potentially transformative innovation that could
fundamentally change the global financial system by eliminating the role of central authorities and
intermediaries. At present, they are largely unregulated and operate outside the formal financial system.
Part E explores alternative lending and funding innovations that have been enabled by technology.
These are of particular value to start-ups and small to medium-sized enterprises (SMEs), which often
face hurdles accessing required capital through traditional providers.
Part F concludes the module with an assessment of the future direction of banking in response to the
increasing integration of technology and finance.
Pdf_Folio:74
74 Digital Finance
PART A: DIGITAL CURRENCY
INTRODUCTION
The nature of money has evolved over time in accordance with changes in technology and changes in
people’s needs. The interconnected nature of the Network Age, the globalisation of commerce and the
speed and ease with which individuals and organisations want to transact have driven the development of
digital currency technologies and platforms. Some of these are essentially electronic equivalents of familiar
fiat currency and other established financial instruments, but some are radically new — independent from
government or central banks, largely unregulated, hosted on platforms that eliminate the use of traditional
intermediaries, and operating outside the formal global financial system.
To evaluate the potential uses of new forms of money or currency, it is necessary to understand:
• the functions of money and currency in the real economy
• the relative advantages and disadvantages of different types of money and currency for particular uses
• the characteristics of new forms of money and currency
• how money and currency may evolve in the future.
We will discuss these topics in this part of the module. This will provide the knowledge required to
evaluate the potential use of digital currencies in meeting an organisation’s specific objectives.
FUNCTIONS OF MONEY
Money can be defined as an item or a verifiable record that is accepted for payments (goods and services)
and repayment (e.g. debts, taxes) within a specific socio-economic context (e.g. a country). For any item or
verifiable record to be considered as money it needs to fulfil the following functions (Swammy, Thompson
& Loh 2018):
• Medium of exchange (primary functionality). Money intermediates the exchange of goods and services,
overcoming barter’s dependence on the mutual want of goods.
• Measure of value/unit of account (primary functionality). Money measures the market value of goods,
services and any other transaction. Money in this sense is a unit of account. For example, in Australia
the unit of account is the Australian Dollar.
• Standard of deferred payment (secondary functionality). Money operates as a standard of deferred
payment — an accepted way to settle a debt (e.g. salary, loans or rent).
• Store of value (secondary functionality). Money is an asset bearing a relatively stable sustaining value.
To be a store of value, it must be able to be saved in a reliable way; stored in a reliable way; retrieved in
a reliable way; and used as a medium of exchange upon retrieval in the future.
Financial instruments are monetary contracts that take place between parties. Financial instruments
are essential for any business activity to function properly. Any financial instrument that can fulfil the
above-mentioned functions of money is considered to be money. All the available financial instruments
within an economy are referred to as the money supply for a specific economy.
PROPERTIES OF MONEY
To fulfil the primary and secondary functions, money must have the following properties.
• Fungibility. The uniform quality of an object makes the object interchangeable and indistinguishable
so that any one dollar is equal to any other dollar.
• Durability. Money must endure long enough to be utilised in any future transactions.
• Portability. There must be ability to handle and deal in money in small quantities that can be exchanged
or transferred in value.
• Recognisability/cognisability. The value of money must be easy to identify so all parties to a transaction
can verify and accept the authenticity and value of the money.
• Stability. Money must be able to sustain value over time so as to be used for transactions and avoid
transaction costs.
Pdf_Folio:75
Money Currency
Can be used to derive the value of any product Cannot be used to derive the value of a product
or service or service
TYPES OF CURRENCY
Figure 2.1 provides a taxonomy of currency. We will briefly discuss traditional currencies in the next
section, before moving on to a detailed discussion of digital currencies.
Commodity
Commodity-
backed
Traditional
Fiat
Currency Representitive/
fiduciary
Regulated E-money
Digital
Unregulated Virtual
TRADITIONAL CURRENCIES
Traditional currency is any form of money that is in public circulation and used as a medium of
exchange — for payments or repaying debt. Traditional currencies are subject to government approval and
endorsement as legal tender. Traditional currencies can be exchanged against each other and can be either
floating (determined by market supply and demand) or fixed (determined by governmental intervention).
Traditional currencies can be commodities, commodity-backed, fiat or fiduciary.
Pdf_Folio:76
76 Digital Finance
Commodity Currency
Commodity currency is a physical commodity (e.g. coined metals, cowry shells or cocoa beans) used to
represent money (Velde & Weber 2010). Table 2.2 lists the advantages and disadvantages of commodity
currency.
Advantages Disadvantages
*Seigniorage is the profit made by a government when it issues currency — the difference between its face value and its
production cost.
Source: CPA Australia 2021.
Commodity-Backed Currency
Commodity-backed currency can be exchanged on demand for a commodity (e.g. gold), but does not use
the commodity itself as the medium of exchange (Selgin 2015). Historically, national currencies were
back by a country’s reserves of gold. Table 2.3 outlines the advantages and disadvantages of commodity-
backed currency.
Advantages Disadvantages
Fiat Currency
Fiat currency is issued by a government as legal tender and is not backed by physical commodities. The
currency’s value is dependent on confidence in the issuing government and is influenced by supply and
demand (Kiyotaki & Wright 1991; Goldberg 2015). Fiat currency has become the norm in most societies
(Kiyotaki & Wright 1991). Table 2.4 outlines the advantages and disadvantages of fiat currency.
Advantages Disadvantages
• Established • Counterfeit
• More stability • Insecurity
• More growth in terms of supply • High transaction fees (contrary to digital currency)
• Inflation
• Value pegged to government
A country’s national currency is a fiat currency. In principle, central banks are responsible for regulating,
controlling and managing the creation, issuance and distribution of national currencies. There are also
supranational currencies, such as the Euro, and currencies that are pegged to another currency (e.g. the
Bulgarian Lev is pegged to the Euro).
Pdf_Folio:77
Representative/Fiduciary Currency
Fiduciary currency is accepted as a medium of exchange, but is not legal tender. Hence, people have the
right not to accept it as a means of payment. Fiduciary currency is based on the promise provided by
the issuer to exchange it for a commodity or fiat currency upon request. The most widely used fiduciary
currency is the bank cheque. Table 2.5 outlines the advantages and disadvantages of fiduciary currency.
Advantages Disadvantages
• Pegged to fiat currency or a commodity (rarely) • Value depending on confidence, good faith and trust
• Paper certificate • Not an enforceable accepted medium of exchange
• Easy to carry • Prone to counterfeit
• Can be manipulated
.......................................................................................................................................................................................
CONSIDER THIS
Before reading section 2.2 on digital currencies, consider what you already know about them. Choose two or
three digital currencies you are aware of and compare them against the traditional fiat and fiduciary currencies we
have described above. What advantages and disadvantages do you think each type of digital currency might have
compared with traditional fiat and fiduciary currencies? How might this affect their use in business?
Advantages Disadvantages
Pdf_Folio:78
78 Digital Finance
As it is a digital representation of a fiat currency, e-money is widely accepted and used as a payment
system. E-money includes stored-value cards (e.g. debit cards, credit cards, prepaid cards and travel cards)
and various services that are accepted by businesses to process payments for goods and services (e.g.
Square and PayPal).
E-money can be differentiated into hardware- and software-based products.
• Hardware-based products feature a physical element and hardware-based security features, such as a
credit card with an embedded security chip. Transactions are typically performed with a physical device,
such as an EFTPOS terminal.
• Software-based products operate on devices such as computers and smartphones. A transfer typically
requires the device to connect to a payments server via the internet. Examples include PayPal, Afterpay
and Klarna, which are established payment processors for many online retailers.
Regulators around the world have converged towards a set of e-money regulations intended to
protect consumers, but enable innovation and competition (Krueger 2002; Mester 2000; European
Commission n.d.).
Cashless Society
E-money and innovative financial products make transitioning to a cashless society increasingly feasible.
The response to the COVID-19 pandemic has perhaps accelerated this transition, by encouraging cashless
and contactless payments, and increased use of online transactions. Governments, financial services
institutions, policymakers, regulators and associations around the globe are investigating how a cashless
society could operate, along with debating the advantages and disadvantages of such a system. Some
potential advantages and disadvantages/challenges are listed in table 2.7.
Advantages Disadvantages/Challenges
• Reduced crime as there is no physical money to steal • As currency becomes data, cybersecurity becomes
• Reduced money laundering as cashless transactions a threat
are traceable and the parties are usually identifiable • Compromised privacy as cashless transactions are
• More cost-effective, as there is no need to print, traceable and the parties are usually identifiable
handle or store physical money • Reliance on availability and scalability of technology
• Seamless foreign exchange transactions, as they will to enable transactions
not need to involve handling of cash • Reliance on interoperability of systems (including
establishment of standards)
• Fees charged by payment service providers,
particularly if one or a few become dominant
• Potential exclusion of those with limited access to
technology or skill with technology
VIRTUAL CURRENCY
Virtual currencies are created by the developers of online virtual communities or platforms. The functions
of any virtual currency within the platform are fulfilled only upon agreement among the members or
participants and/or according to the terms or rules coded into the platform.
Centralised virtual currencies are controlled by a single administrator (usually the developer of the
platform) who acts as the intermediary in transactions. Decentralised virtual currencies use cryptography
to establish trust between participants and thus eliminate the need for an intermediary.
While virtual currencies have no official role in the real economy, some types of virtual currencies
are accepted as payments for goods and services or can be exchanged with fiat currency. Some virtual
currencies, particularly cryptocurrencies, are seen as having the potential for widespread adoption as an
alternative to conventional currencies and payment systems.
Table 2.8 outlines advantages and disadvantages of virtual currencies.
Advantages Disadvantages
Below, we will discuss centralised virtual currencies. In section 2.3, we will discuss cryptocurrencies,
which are a decentralised virtual currency. Centralised virtual currencies may be categorised according
to how they interact with real economy money, as shown in figure 2.2. Note that any particular virtual
currency may not necessarily fall neatly into a single category.
80 Digital Finance
FIGURE 2.2 Relationship between centralised virtual currencies and real economy money
Convertible (bidirectional)
Closed virtual currency Unidirectional virtual currency
virtual currency
Pdf_Folio:81
Second Life
Second Life is an online digital world in which players can virtually pursue their interests and
dreams — from piloting spaceships to creating content and art, or meeting and interacting with other
people (Rymaszewski 2007). The platform’s operator, Linden Lab, regulates the game, including issuing
and managing the in-game currency, the Linden Dollar.
In its terms of use, Linden Lab specifies that the Linden Dollar is not a currency (Linden Lab 2017).
However, given its interaction with the real economy, the European Central Bank does consider Linden
Dollars to be a convertible virtual currency (European Central Bank 2012). This conclusion was based on
the following.
• Linden Lab’s role as an issuing bank and sole authority and regulator within Second Life is specified in
the platform’s terms of use. Linden Dollars can be exchanged for US Dollars at a floating exchange
rate via the Lindex platform. Like a real-world central bank, Linden Lab intervenes to manage the
exchange rate (Chen 2009). The company can modify the quantity of the money in circulation as it
deems necessary (European Central Bank 2012).
• Players who earn Linden Dollars within the platform can convert them to US Dollars. One user
acted as a property developer within Second Life, buying and subdividing real estate in a series of
transactions that saw their USD9.95 account grow to more than USD1 million over a period of 2.5 years
(Nuttall 2006).
• At the peak of its popularity, leading real-world companies established virtual stores on the Second Life
platform as part of their marketing initiatives (Heuzeroth 2010).
• Some in-game banks offered high interest rates on Linden Dollar savings accounts. This led many users
to convert US Dollars into Linden Dollars in order to be able to earn interest. There was some concern
that aspects of Second Life may have constituted investment scams.
.......................................................................................................................................................................................
CONSIDER THIS
Fiat currencies derive their value from confidence and trust in the issuing government. What ultimately underpins the
value of Linden Dollars or frequent flyer miles? How might this affect a business’s decision to use a virtual currency?
Is there a way to manage any risks you have identified?
Regulators have largely ignored virtual currencies that cannot be exchanged back into conventional
currencies, but bidirectional virtual currencies are attracting increasing regulatory attention, partly due to
their potential use for illegal activities and partly due to their potential impacts on the established global
financial system — they essentially compete with conventional currencies (European Central Bank 2012).
Example 2.2 describes how an unregulated virtual currency platform was extensively used for criminal
activities before being shut down by the US Government.
EXAMPLE 2.2
Liberty Reserve
Liberty Reserve (LR) was a convertible virtual currency (pegged to the US Dollar) that could be converted
back and forth between Euros or US Dollars via unlicensed third-party ‘exchangers’ (Simser 2015). Within
the platform, LRs could be transacted securely and anonymously. The exchangers retained no identifying
data or transaction record and thus there was no traceability to individuals or organisations. The absence
of regulatory oversight saw the platform used extensively for money laundering, the funding of terrorism
and other criminal activities (Mullan 2016; Surowiecki 2013; Trautman 2013).
.......................................................................................................................................................................................
CONSIDER THIS
Since virtual currency technology is still developing, along with regulators’ understanding of the implications and the
necessary compliance regime, the technology is expected to evolve into more mature, scalable and secure business
models that will be less prone to manipulation and malicious actions. How would the current lack of regulation and the
potential for future regulatory changes factor into your evaluation of a virtual currency’s potential uses for legitimate
business purposes?
Pdf_Folio:82
82 Digital Finance
The virtual currencies we have discussed above have all been centralised — a central authority issues
and manages the currency within the virtual platform. Cryptocurrencies, such as Bitcoin, which we will
discuss in the section 2.3, are an example of decentralised convertible virtual currencies that operate
without a central intermediary.
QUESTION 2.1
Outline the issues that airlines, and other businesses with points-based rewards programs, should
consider when allowing the volume of outstanding points to grow?
2.3 CRYPTOCURRENCY
Cryptocurrencies are decentralised virtual currencies. On a cryptocurrency platform, the cryptocurrency
is transferable directly from a payer to a payee on a distributed ledger that is handled and secured by
cryptographic protocols. This eliminates the need for intermediation by a centralised authority.
The cryptocurrencies themselves are usually referred to as coins. As there is no centralised issuing
authority, the mechanism by which coins are created is written into the software itself. These rules are
published. Usually, coins are created and awarded to members of the platform that facilitate its operation by
solving the cryptographic ‘puzzles’ that enable transactions. However, coins can also be issued in exchange
for conventional currency or an established cryptocurrency. This approach is usually used to raise the initial
funds to create a cryptocurrency platform.
As a convertible virtual currency, cryptocurrencies can be converted to conventional currencies and vice
versa via businesses operating as cryptocurrency exchanges. Some businesses also accept cryptocurrency
payments for goods and services. As there is no central authority, the value of a cryptocurrency is
determined by supply and demand. They appear to be gaining increasing acceptance as a legitimate
platform for commercial applications.
Table 2.9 outlines some advantages and disadvantages of cryptocurrencies.
Advantages Disadvantages
The best-known cryptocurrency is Bitcoin (Calvery 2013). Bitcoin’s convertibility with conventional
currencies is described briefly in example 2.3. Cryptocurrencies are considered a potentially transformative
technology. Part D of this module will provide a detailed discussion of their characteristics, applications
and potential future role.
EXAMPLE 2.3
Bitcoin
Bitcoin is the most established cryptocurrency platform. It was first released in 2009. Bitcoin is the name
of both the platform and the currency. Bitcoin is perceived as a potentially transformative force in the
global monetary system and payment infrastructures, since the processing of Bitcoin transactions is not
dependent on intermediaries or centralised clearing and settlement bodies. Disintermediation of financial
transactions would fundamentally change the nature of the global financial system.
Pdf_Folio:83
.......................................................................................................................................................................................
CONSIDER THIS
Over the course of its existence, Bitcoin’s value has proven volatile. In 2010, 1 Bitcoin was valued at USD1, in 2017
its value varied between about USD1000 and USD20 000, in December 2018 it was valued at USD3300 and in early
2021 it reached USD50 000. How would this volatility affect the use of Bitcoin (or a similarly volatile cryptocurrency)
in business? Could a business successfully manage the risks associated with this volatility?
SUMMARY
Money serves as medium of exchange, measure of value/unit of account, standard of deferred payment
and store of value. Currency serves as a representation of money and enables transactions. Traditional
fiat currencies are issued and backed by governments. Fiduciary currencies are contractually linked to fiat
currencies and are also widely accepted.
E-money is a digital form of conventional currency and has replaced physical currency in many
transactions. Virtual currency, on the other hand, is not issued by a government and only has value
through the mutual agreement of the members of the platform on which it exists. Some virtual currencies
may be exchanged for conventional currency. Centralised virtual currencies are managed by the platform
developer. Cryptocurrencies are decentralised virtual currencies and are managed by rules coded into the
software that operates the platforms on which they run. Cryptocurrencies are seen to have potential to
revolutionise the global financial industry by removing the need for trusted intermediaries to facilitate
transactions.
Table 2.10 summarises the differences between e-money, virtual currency and cryptocurrency.
Pdf_Folio:84
84 Digital Finance
Conversion Converted to fiat currency Easily converted as Not all virtual currencies
equivalent based on it is a digitally stored are convertible to fiat
the market price of representation of fiat currency. For convertible
cryptocurrency currency virtual currencies,
conversion rate is fixed
by the issuer
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
2.1 Evaluate the effectiveness of digital currency and payments in meeting the organisation’s
objectives.
• An understanding of the features, advantages and disadvantages of traditional currencies provides
a foundation for evaluating the potential applications of digital currencies.
• E-money represents fiat currency and is well established as a payment system through credit cards
and payment processors such as PayPal.
• A centralised virtual currency is issued and controlled by the developer of the currency’s platform. A
cryptocurrency is issued and controlled by transparent rules embedded in the platform’s software.
• An evaluation of the effectiveness of digital currency in meeting an organisation’s objectives must
consider issues including, but not limited to, how the value of the currency is derived, the volatility
of the currency, the regulations and potential regulatory changes that apply, the mechanisms for
converting the currency and the extent of adoption of the currency.
Pdf_Folio:85
Pdf_Folio:86
86 Digital Finance
TABLE 2.11 Common payment methods for personal and business customers
Direct deposits Customers instruct banks to take funds from a bank Personal
account to complete an online payment.
This is an automated clearing house payment.
Mobile wallets (mobile- Mobile wallets use near-field communication (NFC) to Personal
based digital wallet) connect with a point of sale (POS) terminal.
Examples include Apple Pay, Samsung Pay and
Android Pay.
Prepaid cards Store value that customers can spend at a selected Personal
merchant or range of merchants.
Direct debit service This is ideal for managing payments from customers. Business
(initiator) Very useful for managing cash flow.
This is an Automated Clearing House payment.
Credit cards Credit cards are the most common way for customers to Personal and business
pay both online and during face-to-face transactions.
A payment gateway (POS) is required.
Businesses use credit cards for certain employees to
pay for professional expenses,
Automatic payment This is ideal when a fixed and regular amount between Personal and business
personal accounts, or from one account-holder to
another, needs to be paid.
An example is a monthly rent payment.
One-off payment Payment of an amount to a person or business once Personal and business
(direct credit) This is an automated clearing house payment.
Bill payment Each payment is set up individually and can be Personal and business
authorised for future dates.
For regular payments, e.g. paying for utilities.
Funds transfer This involves moving money between accounts. Personal and business
Direct debit (payer) This is an automatic payment attached to a debit or Personal and business
credit card.
It is ideal for utility or other service payments.
The service provider issues an invoice and the amount
is automatically deducted from the card (or from a
bank account) making sure that the invoice/bill is paid
on time.
Authorisation of service providers to deduct the amount
is required.
This is an automated clearing house payment.
(continued)
Pdf_Folio:87
Cheque A cheque is a form of promissory note, usually issued Personal and business
through a bank cheque book.
Payments can also be processed in fiduciary money;
this is becoming less popular.
Cryptocurrencies These are decentralised and distributed payments Personal and business
outside the traditional financial system.
88 Digital Finance
EMV
EMV is a global standard for payments that deals with the authentication and security of chip-card
transactions. EMV chip-cards are almost impossible to counterfeit.
Merchants need special payment-processing devices to comply with EMV standards. The data that
is transmitted from an EMV card during a transaction is encrypted, so even if a third party intercepts
transaction information, it will be almost impossible for them to use.
NFC
NFC payments are equally secure to EMV payments. NFC allow smartphones and payment readers to
communicate when they are placed within a few centimetres of each other. NFC payments use a mobile
wallet. They are contactless, faster and more convenient than EMV, and thus are becoming more popular.
Contactless payments will be discussed in more detail later in this topic.
.......................................................................................................................................................................................
CONSIDER THIS
The security mechanism of EMV chip-cards only works with EMV processing devices. If a merchant runs a fraudulent
credit card in a credit card reader that is not EMV-enabled, then the issuing bank can refuse any liability for loss of funds
since banks and processing networks have changed their liability policies. Businesses have thus been compelled to
use EMV terminals to process EMV cards. However, EMV payments are already being displaced by NFC payments.
How can a business manage the risk of needing to constantly upgrade payment technologies?
Pdf_Folio:89
Digital payments
Electronic fund
Card-based Digital wallet Hybrid
transfer (EFT)
Smart and
Credit card Direct debit Mobile-based contactless (e.g.
Tap & Go)
Biometerics (e.g.
Platform/online-
Debit card Direct credit smile-to-pay,
based
fingerprint, voice)
Cryptocurrency Crypto-enabled
wallet and stablecoins
CARD-BASED PAYMENTS
Card-based payments (e.g. credit cards, charge cards, debit cards, prepaid cards and stored-value cards)
are considered digital payments because they are usually initiated, authorised, authenticated, cleared and
settled electronically (Bank for International Settlements 2016).
The following parties participate in a card-based payment.
1. The consumer (cardholder) intends to make a purchase and promises to pay the issuing bank the amount
of the purchase.
Pdf_Folio:90
90 Digital Finance
2. The issuing bank (where the consumer holds an account) offers credit (line of credit) and/or debit
(debiting funds from the consumer’s bank account) capability to the consumer and is responsible for
transferring the money of the purchase to the acquiring bank. The issuing bank is liable for the amount
of the purchase to the acquiring bank.
3. The acquiring bank (the bank of the merchant) is responsible for accepting the money for the purchase.
If the merchant is not able to provide the purchased goods or services, the acquiring bank is liable to
the issuing bank for the charges made by the merchant.
4. The merchant is responsible for selling goods and services and is able to accept debit, credit or prepaid
cards as a medium and promise for payment.
The terms and references, processing protocols, fees, obligations and other factors are determined by
the companies that administer the particular card, the banking institutions, and the relevant payment and
financial regulations.
Prepaid or stored-value cards have monetary value stored in the actual card or in the account associated
with the card. Prepaid cards are usually issued under the name of the cardholder, whereas stored-value
cards are in most cases anonymous.
A charge card enables the cardholder to purchase goods or services which are paid by the card issuer
(e.g. a company providing a corporate taxi charge card to an employee to travel to the airport).
.......................................................................................................................................................................................
CONSIDER THIS
In 2018, Mastercard initiated the Green Payment Partnership to encourage the global payments industry to shift
towards more sustainable card materials. What other environmental issues might be associated with different physical
and digital payment methods? Would these factor into your business decisions?
DIGITAL WALLETS
While card-based payments and EFT are widely adopted, digital wallets are a relatively new payment
technology. A digital wallet (or e-wallet) is an electronic device, software program or online service that
a party can use to electronically transact with another party by means of digital currency. A digital wallet
can store:
• information about the user’s payment instruments, including credit and debit cards, prepaid cards and
bank accounts
• identification documentation
• a variety of other information such as boarding passes for flights, concert tickets and loyalty cards.
Users access their digital wallets mainly through the internet and via electronic devices such as
computers and smartphones.
• The digital wallet itself can be used for online purchases.
• Digital wallets operated via mobile devices can be used for NFC-based payments.
The functionality of digital wallets is steadily evolving beyond basic financial transactions. A digital
wallet can serve the purpose of a payment channel, storage of information and transmitter/receiver of
encrypted data. For example, a digital wallet can be used to authenticate the wallet-holder’s personal
credentials including legal identification. Cryptocurrency wallets may also store private keys for cryp-
tocurrency coins.
Business Applications
Both from a merchant (business) and customer/user point of view, it is important to understand the
utilisation of digital wallets, and have a solid understanding of the terms and conditions applicable for
responsible use of both the hardware and software technology.
Pdf_Folio:91
Advantages Disadvantages
• Convenience and comfort. No need for access to • Not all retailers may accept all kinds of digital wallets.
cash, physical cards or account numbers. • Consumers still need to carry a smart device.
• Can hold many different types of cards (e.g. credit, • The device that holds the digital wallet needs to be
debit, loyalty, reward and ID cards, and coupons). electronically charged.
• Information is securely stored within a third-party • Security depends on the device settings and
system, and is protected by password and/or up-to-date software. Responsible management and a
biometric verification. degree of ‘tech-savvy’ is required.
• Widely accepted around the world. • Depending on the merchant, wallet and POS
• Payment initiation and authorisation is required by terminal, additional charges/fees may be incurred.
means of PIN or biometrics provision. • Could lead to irresponsible and uncontrolled
• Can track spending habits and promote better spending (e.g. buy-now-pay-later schemes are easy
financial wellness (e.g. assigning a fixed budget, to access and use).
or breaking down spending by category). • Can be costly for merchants, especially if they
• Cost-effective for retailers: no need for human develop their own digital wallet (development,
tellers/cashiers; technological developments in maintenance and software/hardware costs
POS terminals can reduce transaction costs for may apply).
merchants (e.g. Square). • Can lead to human cashiers/tellers becoming
• Seamless experience for customers/users creates a redundant or unemployed.
competitive advantage for merchants. • System outages can prevent merchants from
• Can support scalability of businesses and provide processing digital wallet payments.
access to new revenue streams (e.g. formerly • Encryption of customer information requires constant
cash-only merchants, such as vendors at updates and maintenance to avoid privacy and
markets, may now accept payments using a security issues.
contactless POS).
Pdf_Folio:92
92 Digital Finance
TABLE 2.13 Comparing mobile wallets — Apply Pay, Google Pay and Samsung Pay
In-person Yes, in major retailers Yes, in major retailers Yes, in major retailers
purchases
Supported cards Major credit card providers Major credit card Major credit card
Loyalty and reward cards providers providers
Loyalty and reward cards Loyalty and reward cards
Ability to directly connect
to PayPal account
Security feature Data never stored on phone, Data encrypted and Data not stored on the
or on Apple servers stored on secure servers, phone
Two-factor authentication not on phones Iris, fingerprint or PIN for
through Touch ID or Face ID Virtual account number authentication of every
Actual account numbers not shared with merchants, transaction
shared with merchants not the actual account
number
.......................................................................................................................................................................................
CONSIDER THIS
For your business or a client’s business, what is the relevance importance of each factor listed in table 2.13?
Mobile
Mobile devices, communication technologies and cloud-based POS terminals enable digital payments to
be conducted anywhere (Bank for International Settlements 2016; Euromonitor 2019; Hinchliffe 2019).
Some mobile POS (mPOS) technologies can process transactions without an active internet connection by
communicating directly between the POS terminal and customer’s device.
Digital payment methods that use mobile technologies have applications in-store and for remote pay-
ments. For example, in-store retail transactions can be decoupled from shop checkouts, while tradespeople
can use remote mobile payment methods to process payments using a smartphone, tablet or other device on
job sites. Both approaches provide a seamless customer experience, and a secure and efficient transaction
for the merchant.
In-store and remote payments can use mobile wallets, described in the previous section, or QR
code payments. A QR code is a two-dimensional barcode that can store data information and is readable
and quickly decoded by a smartphone. QR codes work via certain banking apps. Consumers initiate a
payment by scanning the QR code on the retailer’s POS terminal (Euromonitor 2019). QR codes used in
applications such as boarding passes cannot be edited, but QR codes linked to a webpage or other system
can be updated, and these are useful in many business applications (e.g. enabling customers to scan a
QR code on a café menu to place an order). US-based CVS Pharmacy is rolling out Venmo QR code
technology and PayPal as a payment method across its retail stores. Shoppers will be able to scan a QR
code with their smartphone to pay by means of stored credit or debit cards, PayPal Balance, PayPal Credit,
Venmo Rewards or Venmo Balance.
Pdf_Folio:93
EXAMPLE 2.4
.......................................................................................................................................................................................
CONSIDER THIS
What role should mobile digital payments play in your business or your client’s business?
Pdf_Folio:94
94 Digital Finance
Example 2.5 examines one financial institution’s development of wearable payment devices as a
possible next step in customer convenience.
EXAMPLE 2.5
Buy-Now-Pay-Later
Buy-now-pay-later services like Afterpay allow consumers to buy and receive goods immediately
while paying off the full price in instalments. Example 2.6 describes how Afterpay works.
EXAMPLE 2.6
Afterpay
Afterpay enables consumers to pay for purchases in four fortnightly interest-free instalments. Customers
store credit card information with Afterpay, and Afterpay approves spending limits based on customers’
repayment history. Using Afterpay online is similar to other online wallets. Consumers select Afterpay as
the payment option and log in. The payment arrangement is set up automatically without consumers
needing to do anything else. The process of using Afterpay in a physical store is similar to using
smartphone/smartdevice wallets for in-store payments. The first instalment is automatically charged to
the customer’s Australian credit card. An email receipt is sent to the consumer. Provided the consumer
pays on time, there is no additional cost to them. Late fees are charged if instalments are paid late.
From the merchant’s perspective, Afterpay integrates with various POS platforms. Afterpay settles the
transaction with the merchant. Afterpay therefore assumes any credit or fraud risk. The merchant receives
their funds 48 hours after settlement. Merchants pay a fee and a commission per transaction.
Afterpay has been accused by consumer associations of promoting irresponsible spending and leading
customers into huge debts.
There is debate about whether Afterpay is a payment system or not. Afterpay argues that it is a
customer acquisition channel for merchants, and that it has substantial differences from regular credit
card schemes. The value proposition behind Afterpay’s business model is not grounded on its payment
processing functionality, but rather the connectivity attribute between retailers and consumers (especially
Millennials). Afterpay classifies itself as a platform with a built-in payment functionality similar to Amazon
or Uber Eats.
.......................................................................................................................................................................................
CONSIDER THIS
How would you evaluate Afterpay’s value proposition in terms of the needs of your business or your client’s business?
Pdf_Folio:95
Security
• Digital payment systems store payments information in a secure server or a microchip embedded in the
phone (for Apple Pay).
• The consumer’s payment information is tokenised during the transaction. No actual payment informa-
tion is transmitted.
• The security features protect the merchant from fraudulent claims against charges made using a
stolen card.
Convenience and seamless checkout
• For businesses with an e-commerce site, digital payments often mean consumers’ details are pre-
populated, providing customer convenience and reducing the risk of errors that cause problems in
fulfilling the transaction.
• Businesses can focus on providing a customer value proposition that extends beyond the goods and
services offered to the whole shopping and payment experience.
Increased efficiency, long-term cost-saving
• Digital payments are relatively easy for businesses to adopt because most POS terminals work
with a variety of payment technologies. Cloud-based POS terminals enable upgrades as the
technology evolves.
• Errors when processing manual payments are minimised as information is synchronised automatically
and in real-time with providers.
• Digital payments enable a digitised workflow and easier transaction tracking, creating cost-efficiencies
throughout payment life cycle.
Source: CPA Australia 2021.
Possible costs of implementing digital payments for businesses with a physical presence include:
• the costs associated with properly training staff on the various digital payment methods
• possible loss-of-sales costs in the event of widespread digital systems failure (particularly if there are
no alternative cash-based payment procedures in place)
• installation costs for POS terminals.
Pdf_Folio:96
96 Digital Finance
Example 2.7 describes Airwallex’s Airwallet hybrid solution for businesses that transact internationally.
EXAMPLE 2.7
Airwallet
The conventional way to access credit was to complete information on a paper-based application form and
submit it at a physical bank branch for assessment. Digital finance has largely replaced this approach. For
example, Airwallex’s Airwallet lets its customers set up a bank account, even an international account,
instantly. The account instantly allows receipt of payments in various countries, using a card, wallet or
bank-to-bank transfer.
Upon a payment being made, the recipient receives a notification instantly, instead of having to check.
This provides real-time transparency of all transactions. All transaction data is automatically synchronised
with the business’s accounting systems such as Xero, removing the need for Excel-based reconciliation.
Similarly, a corporate card product replaces the traditional corporate card and its associated paperwork.
For example, using a traditional bank, a paper form would be completed for each employee that will be able
to use the card. Employees need to submit physical receipts, the details of which then are input manually
into the business’s expense management system. With the Airwallex corporate card, the business can
instantly issue a card to an employee, all the know-your-customer onboarding is automatic and digitalised,
so it can be done online. The expense information is automatically synced to the accounting system.
Receipts can be recorded as photos and uploaded to the cardholder management system.
.......................................................................................................................................................................................
CONSIDER THIS
How would you evaluate Airwallet’s value proposition for your business or your client’s business?
QUESTION 2.2
CASHLESS SOCIETY
We briefly discussed the concept of a cashless society in section 2.2. RBA data shows cash transactions
accounted for only 27 per cent of all payments in 2019, compared to 69 per cent in 2007 (Caddy et al. 2020)
and the COVID-19 pandemic is expected to have further reduced the use of cash. Globally, countries have
been slowly phasing out higher value banknotes, and there are serious discussions by governments about
the implementation of digital currencies (Wood & Morris 2020). Moving towards a cashless society will
have major impacts (Kale 2020), and businesses would be required to adopt digital payments (either at will,
or through regulation). Organisations offering digital payments will need to build trust and continuously
innovate to increase their ability to cope with a significant uptake of digital payments by consumers and
businesses. Consumers, especially those who are less tech-savvy and those who are vulnerable (and have
limited access to technology and devices), would need support to learn and adapt to these changes, and
this could create some challenges and barriers.
PERSONALISATION
Financial institutions and services providers, like many businesses, are increasingly utilising customer
data to provide personalised services to customers. For example, the financial services affiliate of Alibaba
in China, Ant Technologies, offers tailored financial services packages and merchant offers to customers
using advanced analytics (Dab et al. 2017). Grab in Asia-Pacific uses personalisation to differentiate itself
from competitors by providing a localised loyalty program, GrabRewards, to customers when they use
Grab’s services (Euromonitor 2019).
Pdf_Folio:97
FRICTIONLESS PAYMENTS
Frictionless payment is the ability to make payments with minimal human-to-human and human-to-
machine interactions. Imagine driving up to a petrol station, filling your vehicle with petrol, getting back
into your vehicle and driving off. A minute later, you receive a notification on your electronic device
informing you that you have been charged for petrol. That is a frictionless payment. BP currently has such
a system, but it requires the customer to confirm some details in the car before they fill up.
Woolworths’ Scan&Go app is another example of how frictionless payments are manifesting in
Australia. The app allows consumers to scan items they have picked up, bag them and make payment for
the items. After making payment, the consumer only needs to scan the QR code at the Scan&Go counter to
finalise the transaction. The whole process requires no human-to-human interaction, and minimal human-
to-machine interaction.
There is a trend towards more frictionless payment solutions, in part due to changes in consumer
demographics, with Gen Y and Gen Z consumers gravitating toward mobile payment applications that
minimise interactions (Accenture 2017b; Riemer et al. 2017). This could result in an increase in the uptake
of FinTech platforms providing digital payments. Traditional institutions are also introducing their own
digital payments platform to meet the expectations of these consumer groups. Gen-Z consumers are likely
to want a more advanced payments system that incorporates big data and artificial intelligence (AI) into
mobile payments (Accenture 2017b).
.......................................................................................................................................................................................
CONSIDER THIS
How frictionless are payments in your business or client’s business? Could digital payments improve the payment
experience? How?
CYBERSECURITY/FRAUD PROTECTION
Digital payments involve the transfer and exchange of electronic information, which is susceptible to
cybersecurity breaches and fraudulent practices. While FinTech and financial institutions innovate to
provide more advanced payments systems, criminals are also attempting to out-innovate these companies.
Managing cybersecurity and fraud risks are therefore important priorities in the digital payments space
(Accenture 2017b).
Organisations are providing additional layers of security to protect themselves and provide consumers
with confidence. For example, when a new device is used to log in to an account or service, a notification is
sent to the user’s nominated primary device (and often two-factor identification is required to gain access).
Biometrics can also be used to provide added security and fraud protection. By using a combination of face
and voice to fulfil KYC/AML measures when accessing online-based banking services, organisations are
providing an additional layer of security to protect customers’ banking information and increase customer
trust in financial services. HSBC in the UK claims that the introduction of voice biometrics at its call
centres prevented almost GBP400 million of customers’ money from fraudulently falling into the hands
of telephone scammers in 2019.
FINANCIAL LITERACY
Financial literacy is the possession of the skills and knowledge to make informed financial decisions.
A lack of financial literacy among consumers will be detrimental to their financial well-being (Panos &
Wilson 2020). Individuals and groups lacking financial literacy may be at risk of financial exclusion, taken
advantage of by unethical individuals or businesses, or make poor decisions. For example, a customer with
poor financial literacy could be led into using buy-now-pay-later services despite being ill-equipped to
make timely repayments given their financial situation. This could potentially cause the customer to face
hardship if they default on their repayments.
98 Digital Finance
a payment-as-a-service mentality that will be driven by interconnectivity, interoperability, omnichannel
and integration of systems, processes and infrastructures.
Open banking is expected to play a central role in the sourcing and aggregation of more data and
the democratisation of payments as a vehicle for value creation. The gradual adoption of 5G mobile
communication technologies will enable faster, safer and optimised mobile and contactless payments that
will support innovative business models that utilise biometrics and other types of data for anti-money
laundering, know-your-customer and identification purposes.
The staging of the API and platform economy is expected to enable new governance and architectural
designs across all stakeholders of the digital payments ecosystem and beyond, embracing more change
in order to serve consumers and corporate customers. More developments are expected in order for the
financial services industry to capture growth. Payment and banking business models become customer-
centric, leveraging on streamlined operations, enhanced data analytics capabilities, along with simplified
cross-border payments and transactions.
All these trends need to be positioned within an organisation’s strategy development, in order to
effectively and efficiently evaluate how the business as a whole is changing, and determine which digital
payment processes may be most suitable to adopt (in line with relevant regulations and requirements).
.......................................................................................................................................................................................
CONSIDER THIS
The 2020 Digital Payments Report by Roy Morgan shows a decline in the use of bill payment services and online
payment platforms by Australians in 2020; a slight increase in contactless payment services in Australia; and growth
in the use of buy-now-pay-later, with 10.2 per cent of Australians now using some form of this service. How do you
expect digital payments to evolve? How would you choose a portfolio of digital payments services for your business
or client’s business?
QUESTION 2.3
Increased use of cryptocurrencies and e-money are moving us towards a cashless society. Assess
the pros and cons of a cashless society for a small retail business.
SUMMARY
The emergence of new and technology-enabled players who are introducing innovative digital payment
methods and platforms is shifting the overall payment landscape for merchants, consumers and financial
institutions. A new banking and payment services ecosystem is being shaped, featuring predominantly
digital payments using cards, EFT, digital wallets and hybrid approaches that blend aspects of all of these.
An evaluation of the role of digital payments to meet an organisation’s objectives needs to consider the
benefits, costs and risks of each potential application in terms of providing a personalised and frictionless
experience to consumers, and achieving efficiencies and creating value for the business.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
2.1 Evaluate the effectiveness of digital currency and payments in meeting the organisation’s
objectives.
• A digital payment refers to an exchange in which value is transferred electronically from one party
to another.
• Digital payments are widely adopted throughout the economy. Consumers and businesses are
transitioning away from cash. Consumers increasingly expect convenient and frictionless payments.
• Digital transformation places a business within a digital ecosystem in which the payment experi-
ence can be a source for competitive advantage.
• Evaluating the potential of specific digital payment options to achieve an organisation’s objectives
requires an understanding of their associated characteristics, advantages, disadvantages, value
proposition and risks.
Pdf_Folio:99
DATABASES
A database is a structure for storing, managing and accessing data. Three different database architectures
are represented in figure 2.5.
Database Database
Address Address Database
book book
Address
MEL PER
Database book
Address Database
book Address
book
SYD
Database Database
Address Address
book book
Centralised databases are located on a single storage device in a distinct location. They are easily
developed and maintained, but anything that leads to a loss of access to the storage device makes the
data inaccessible until the issue is resolved. Centralised databases can be slow, particularly as the volume
of data and number of users grow.
Decentralised databases are distributed over multiple interconnected storage devices in different physical
locations. These provide faster access and are more scalable as the volume of data and number of users
grow. They are more complicated to manage than centralised databases.
Distributed databases reside on all the computers participating in the network (these computers are
called nodes). A special type of algorithm called a consensus algorithm synchronises the data across every
participating computer. Since changes made to the database on one computer are replicated in the database
on every other computer, all participating computers need to be verified and honest. Distributed databases
have high robustness, reliability and availability. They are complicated and costly to set up, but operate
cost-effectively once established. One application of a distributed database is a distributed ledger.
Pdf_Folio:100
Permissionless Ledgers
A distributed ledger without a specific owner is known as a permissionless ledger. Anyone can add data to
a permissionless ledger. Permissionless ledgers can be configured in such a way that existing data-entries
cannot be altered. These distributed ledgers then function as a permanent, public record, where people
can, for example, publicly assign property ownership. Because any data can be added and, once added,
cannot be changed, permissionless ledgers are a target for exploitation. For example, an analysis of the
core ledger of Bitcoin identified multiple links to online child abuse content embedded within the ledger.
As such, possession of the ledger would potentially be illegal in some jurisdictions. This poses a moral
and legal dilemma for users (BBC 2019; Gibbs 2018).
Permissioned Ledger
A distributed ledger with a defined set of owners is known as a permissioned ledger. The integrity of data
added to the ledger is assured through a limited consensus algorithm carried out by trusted parties (e.g.
governments or banks). The consensus system of permissioned ledgers tends to be less complex and faster
than the systems used in permissionless ledgers. In addition, permissioned ledgers create highly verifiable
data sets, leading to a high degree of confidence.
Pdf_Folio:101
Database
Address
book
Bank B Database
Database
Address Address
book book
Bank A Bank C
Source: CPA Australia 2021.
Advantages Disadvantages
The business applications of DLTs are largely in the experimental and exploratory stage, but there is
substantial interest and work by financial services institutions, governments, big technology leaders and
others to determine operational models that could be able to be adopted at a greater scale. Example 2.8
presents a number of recent use cases that utilise different aspects and attributes of DLTs.
EXAMPLE 2.8
Pdf_Folio:102
DLTs can play a pivotal role in accelerating digital transformation. It is important to take into account
that each use case is different, and the advantages and disadvantages of DLTs must be contextualised within
the specific business, technological, compliance, regulatory and policy-driven organisational architectures
and designs. For example, a public-permissionless blockchain network (e.g. Bitcoin blockchain) might be
more transparent and free to join, but its attributes may not be fully appropriate for a specific business
model. A public-permissioned network, enterprise-level private-permissioned network, or consortium-
level network that has both permissioned and permissionless attributes may be more appropriate to a
specific use case. In other words, a business needs to identify its problems and requirements and then
evaluate whether or how a particular DLT may be used to address those problems, fulfil requirements and
improve performance.
CONSENSUS MECHANISMS
Figure 2.7 provides a high-level schematic of how blockchain transactions are processed.
Miners are nodes that specialise in adding new data to the blockchain. They do this via a set of
algorithmic-enabled consensus protocols that ensure the highest level of reliability even among unknown
and potentially unreliable parties (Binance Academy 2020). When a miner adds data (in the form of a
‘block’) to the blockchain, they receive a reward or fee, which is paid in cryptocurrency. Cryptocurrency,
as explained in Part A, is a decentralised virtual currency that exists in a blockchain platform. The mining
process is hugely competitive and requires enormous processing power. It is thus energy-intensive (in fact
it uses so much electricity that its environmental impact is considered a major disadvantage of blockchain).
A variety of consensus mechanisms are used on different blockchains. The two most popular are proof
of work and proof of stake. Proof of work is based on the principle that a difficult mathematical puzzle
has to be solved by miners and the solution has to be proven easily by nodes. The distribution of mining
power across numerous nodes around the world makes it almost impossible for an attacker to manipulate
the network (they would need to possess at least 51 per cent of the computational power of the system). The
larger the network, the more resistant to manipulation it becomes (Tar 2018). Some smaller blockchains
have, however, been successfully manipulated in this way.
Proof of stake is a consensus mechanism in which users who have stake in the system are granted
the opportunity to validate blocks and earn rewards (Vasin 2014). Within a proof of stake consensus
mechanism, blocks are not mined by competing nodes, but rather ‘forged’ by users working together.
They are faster than proof-of-work mechanisms and thus are more scalable. Proof of stake mechanisms
Pdf_Folio:103
Database Database
Address Address
book book
Blockchain network
Database
Address
book Nodes
Database Database
Address Address
5. Confirmation of entry book book
added to address book
Blockchain network
Database
Address
book
New block
Database Database
Address Address
book book
Permissionless Blockchains
Permissionless blockchains are truly decentralised public blockchains. As there is no central authority,
this type of blockchain ensures a maximum level of transparency — every transaction can be read and
traced, resulting in a high level of trust in the system. They are open to anyone and can incentivise users
to participate. On the other hand, permissionless blockchains are slow to operate due to the high level of
security and the complexity of consensus algorithms, and there are scalability issues (Singh 2020). Some
potential use cases of permissionless blockchains involve voting, fundraising and digital identity.
Permissioned Blockchains
Permissioned blockchains have a control layer that grants access to accepted users. For example,
users wishing to participate in Facebook’s cryptocurrency Libra, have to be registered and cleared via
government identification (Singh 2019). Private permissioned blockchains allow for additional governance
structures, making them organised and easier to administer. Permissioned blockchains involve fewer nodes,
making the consensus algorithms operate more efficiently, but they tend to be less secure, as their security
relies on the integrity of the members — a relatively small number of nodes with malicious intentions
could alter the blockchain or introduce censorship. Potential use cases include supply chain management,
banking and payments, and food tracking.
Permissioned blockchains may be either public or private. In a public permissioned blockchain, the
control layer does not interfere with the blockchain decentralisation or authority. In a private permissioned
blockchain, the control layer also governs transactions performed by the participants of the network.
Private blockchains are closed ecosystems that are governed by one organisation that is responsible
for participation, consensus protocol, and shared ledger maintenance. Typically, large centralised organ-
isations which leverage the power of the network for their own internal business operations utilise this
approach. By deploying such blockchains in an isolated business network environment, selected parties
and business partners can be connected to the system and the benefits of an immutable shared database
can be fostered. Further, every participant of the network is known, thus proving to be trusted. As this
type of blockchain is utilised within a trusted environment, the consensus algorithms do not have to be as
complex, thus improving processing speed and scalability. If a new member aspires to join the network, it
has to be granted permissions accordingly. These permissions determine what the new member is allowed
to see and whether the new user can conduct transactions. Private blockchains are often criticised for not
ensuring anonymity or fully utilising the concept of decentralisation. Examples of private blockchains are
Ripple (XRP) and Hyperledger.
Hybrid Blockchains
Particularly in complex environments where the partners and clients involved change constantly, the
utilisation of private blockchains proves difficult, as the business network often changes. In order to tackle
such challenges, hybrid blockchains can be utilised. Hybrid blockchains combine the concepts of private
and public blockchains in order to benefit from key aspects of both concepts (Zilavy 2018). This means
that while some processes are kept public, other processes are kept private.
Applications of such blockchains can, for example, be found across supply chains. While products move
along the supply chain businesses can conduct transactions against a private blockchain, allowing
them to track the products’ current status. This utilises the performance and scalability benefits of private
blockchains. When the product arrives in the supermarket, customers can scan a QR code, accessing the
Pdf_Folio:105
Consortium Blockchains
Consortium blockchains are a special hybrid between private and public blockchains and are often
considered as partly decentralised. They combine the single highly trusted entity-model of private
blockchains with the low-trust environment of public blockchains. This is achieved by a pre-selected set
of nodes controlling the consensus process (Buterin 2015).
An example could be 20 financial institutions agreeing on developing a consortium blockchain. Every
institution participates with its own node. If 15 of them approve a newly created block, it is considered
valid. Meanwhile, reading the blockchain could be a public right, being accessible by anyone. Potentially
this reading right can be restricted, only allowing the public to receive cryptographic proofs of parts of the
blockchain (Buterin 2015).
IBM Food Trust is a blockchain application, aiming to create a safer, smarter and more sustainable food
ecosystem. It does so by digitising data and transactions across the supply chain. Whilst authorised users
can provide data, such as tracking information, test data and certifications to the network, consumers can
publicly access the history of individual food items from farm to store (IBM 2020).
Table 2.15 compares the different types of blockchains.
.......................................................................................................................................................................................
CONSIDER THIS
In order to accommodate new business models and innovations, different variations of blockchain architecture have
been emerging including public permissionless, public permissioned and semi-decentralised. All these variations
blend different characteristics and attributes from the main blockchain types. Does blockchain have potential
application in your business or a client’s business? What type would be most appropriate and why?
QUESTION 2.4
A client who wants to adopt distributed ledger technology (DLT) comes to you for advice on whether
to use a permissioned or permissionless ledger. What factors would you advise your client to
consider in making their choice? Why?
Pdf_Folio:106
EXAMPLE 2.9
Smart contracts can self-execute based on rules, but they are not intelligent tools — they cannot parse the
subjective requirements that feature in many conventional business contracts (Levi & Lipton 2018). Grigg
(2015) suggested smart contracts should be developed in such a way that, in addition to the software rules,
they capture the legal contract behind an issuance, document the relevant information the participating
parties need to know, be human-readable, be efficiently executed and enable flexibility.
A major shortcoming of current smart contracts is their inability to push outputs to external systems (e.g.
executing payments in fiat currencies on traditional payment systems). This can be somewhat overcome
by a blockchain intermediary called an oracle (described later in the module).
At present, smart contracts have a scalability issue — their cryptocurrency-only transaction limitations
mean they have not been significantly adopted by businesses.
ETHEREUM
Ethereum was designed as a general purpose blockchain that would enable the creation of smart contracts.
To improve scalability and network performance, the platform is adopting a number of measures as part
of a transition to Ethereum 2.0 (Muzzy 2020). These are:
• sharding — partitioning the blockchain into separate, but interoperable chains so that multiple transac-
tions can be processed simultaneously
• proof of stake — replacing the proof-of-work consensus mechanism with the faster and more scalable
proof-of-stake mechanism
• interoperability with other blockchains — creating a smart contract execution environment that allows
rules implemented by other blockchains to interact with the Ethereum blockchain.
DISTRIBUTED APPLICATIONS
Distributed applications (DApps) feature a web application interface built over the top of one or more
underlying smart contracts. They thus enable easy interaction with services provided by a blockchain
(Antonopoulos & Wood 2018; Metcalfe 2020). Being connected to a blockchain, DApps are more resilient
than established centralised applications and benefit from higher transparency, as everyone can inspect the
code and any interaction is permanently recorded (Antonopoulos & Wood 2018).
Example 2.10 illustrates how DApps can be utilised in order to transform a physical asset into a digital
asset on a blockchain (a process known as tokenisation).
Pdf_Folio:107
Tokenisation
Superfly is an established international airline which had to ground its entire fleet for months, due to
a global pandemic. Eventually, Superfly starts experiencing significant cash flow issues and identifies
blockchain as a potential technology to facilitate short-term liquidity.
The airline develops a DApp on the Ethereum network. Objects are transformed into digital assets
through tokenisation. Throughout the transformation, tokens are assigned to those digital assets. The
following types of tokens can be generated.
• Security token. The airline tokenises its fleet. Each plane is worth 100 tokens (1 token = 1 per cent of
a plane) and should be traded under the symbol ‘FLY’. By purchasing FLY-tokens, buyers conduct a
direct investment.
• Utility tokens. The airline tokenises future services such as flight tickets. For 100 ‘AIR’, backers can
redeem a ticket from Perth to Sydney. However, AIR is strictly linked to flight tickets and cannot be
redeemed for other services such as meals.
• Currency token. By creating currency tokens, the airline creates its own cryptocurrency, similar to Bitcoin.
Tokens will be discussed further in part D of the module.
.......................................................................................................................................................................................
CONSIDER THIS
Beyond the example above, could tokenisation have application for your business or a client’s business?
ORACLES
We identified earlier that the inability of smart contracts to interact outside the blockchain is a limitation of
their application. This can be partly overcome using oracles. An oracle is a live-feed of data from external
systems that are continuously monitored, treated as a source of truth and are used to trigger execution of a
smart contract when the defined conditions are met.
Oracles can be sensors that obtain data from physical objects or software that obtains data from systems
by means of APIs. The container sensors described in example 2.9 are a type of oracle. Inbound and
outbound oracles initiate an action or transaction in response to specific data and thus can enable the
development of smart contracts.
As third-party services, oracles are not part of a blockchain’s consensus mechanism and thus an
additional layer of trust is required between the contracts and the oracles. Oracles also require network
resources for collecting real-time external real-world data. Despite these requirements, oracles are the
most promising interfacing mechanism connecting external real-world data feeds to blockchains to enable
smart contracts.
EXAMPLE 2.11
Pdf_Folio:108
Today
$$
Insurance
Lawyers
Brokers $
Banks
Alice Bob
Smart contract
Database
Address
book
$
Database Database
Address Address
book book
$
Law and legislation (room for human error) Code (no room for human error)
Terms that can be changed, rewritten and interpreted Terms that cannot be changed
differently
Possibility of terms not being fulfilled or being executed Automatic fulfilment by all network participants
in a poor manner
Signed only at personal meetings of the parties/proxies Can be concluded from anywhere in the world without
personal presence
No guarantees, even in case of non-execution Storage of all data and counterparties on the
Law possibly no help to reimburse the losses blockchain
Conditions can be changed, the term of execution can User decision which information will be available to the
be postponed public
QUESTION 2.5
How are smart contracts relevant to the financial services sector (consider, in particular, financial
planning firms and insurance firms)?
REMITTANCES
A remittance is a fund-transfer transaction, mostly associated with immigrants who send money back to
their home countries. Current remittance schemes depend on third-party services and financial institutions
together with legacy systems. Within cross-border payments, the SWIFT network is responsible for
handling the transmission of messages across the payment chain whereas corresponding banks are
actually responsible for settling down the debits and credits of the accounts. Especially in the process
of international payments, multiple stakeholders are involved including corresponding banks, SWIFT and
central banks. This creates a complicated network of layers of messaging and transactions that attract
processing and messaging fees while at the same being a very lengthy process, making the current systems
very inefficient.
Blockchain provides a new way of processing cross-border payments by removing intermediaries and
correspondent banks, reducing at the same time excessive costs, charges, operational costs and turnaround
time for settlement.
ACCOUNTING/AUDITING
Even though blockchain technology is not yet adopted by accounting functions, there are some potential
use-cases worth exploring. The attributes of transparency and immutability of data recording will stream-
line audit and accounting processes by reducing fraud and improving regulatory compliance. Management
of documentation in terms of recording and timestamping will ensure seamless traceability, reduced errors,
cost and improved efficiency.
Smart contracts can facilitate the automated payment of invoices easing data verification in relation to
financial statements.
Conventional accounting is based on a double-entry system that deploys a centralised approach to
data storage and maintenance. Accounting information is retrieved by accounting professionals upon
request by clients and the regulator. This centralised ledger is directly accessible only by the accounting
professional and the auditor. Within blockchain-enabled accounting, the registration, maintenance and
storage of records is deployed in a distributed ledger that is accessible by all the relevant stakeholders. In
this context, a shift from a double-entry to a triple-entry bookkeeping model is observed, since all relevant
parties have an identical copy of the ledger at all times. Users are authenticated by the usage of private
and public keys. In this frame of reference, blockchain will also facilitate close to near-instant transactions
since real-time processing of transactions is introduced, due to the real-time settlement attribute ensured
via the network. Major accounting firms including KPMG, PwC, EY and Deloitte have been experimenting
with blockchain technology aiming at developing and adopting scalable, efficient and secure blockchain-
enabled services.
TRADE FINANCE
By definition, trade finance encompasses the infrastructure, processes and funding that is required in order
to support international trade supply chains. It is considered as the lubricant and fuel for global trade and
commerce. The current lifecycle is inefficient and primarily paper-based, requiring lengthy times for the
processing of letters of credit, document verification, along with establishing trust among the interrelated
stakeholders.
This also creates issues around vulnerability and exposure to fraud. Blockchain technology is able to
shift the current lifecycle into a digitised landscape, which is more secure and efficient, while at the same
time bringing traceability, transparency, effective governance, stakeholder accountability and reduced
Pdf_Folio:111
QUESTION 2.6
QUESTION 2.7
Outline the KYC challenges that may arise due to the anonymity and privacy features of many
cryptocurrencies.
SUMMARY
Distributed ledgers operate on a decentralised network in which each node holds a copy of the ledger. This
provides high availability, reliability and speed. It enables every participant to view and interact with a
common ledger and thus has business applications such as supply chain management.
A blockchain is a particular type of DLT that achieves very high trust in the absence of a central authority.
It does so by using network consensus mechanisms based on cryptography that make it practically
impossible to change data once it has been added to the blockchain database. Because DLTs are shared by
definition, business applications of blockchain technology have often imposed permissions to restrict who
can access with the ledger and how they can interact with it.
One promising use of blockchain is to host and operate smart contracts. Smart contracts are coded into
software and then self-execute based on the conditions they contain. Various types of oracles exist to enable
smart contracts to send and receive data beyond the confines of the blockchain platform. Smart contracts
are still an emerging application.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
2.1 Evaluate the effectiveness of digital currency and payments in meeting the organisation’s
objectives.
• The blockchain DLT includes a mechanism to prevent the same digital currency being used in more
than one transaction, thus preventing double spending of cryptocurrencies.
2.2 Assess the effectiveness of using cryptocurrency and distributed ledger technology in solving
complex business problems.
• As a shared record, distributed ledgers when implemented using blockchain technology provide
high reliability and availability, instant settlement, a permanent record of transactions, high
transparency and high security.
• Blockchain ledgers have limited scalability, transactions made in error cannot be corrected, and
implementation can be complex.
• Most business applications of distributed ledger technology involve restricting access to the ledger
to trusted parties and may also modify the operation of the platform’s consensus mechanism.
• Blockchains can host smart contracts, which are self-executing contracts that are highly transpar-
ent and cannot be modified.
• The consensus mechanism on blockchain replaces the role of central authorities or intermediaries
in transactions and thus provides the potential to avoid delays and fees associated with the
conventional financial system.
Pdf_Folio:112
2.11 CRYPTOCURRENCIES
Cryptocurrencies may be defined as a form of decentralised, fungible and convertible virtual currency
secured using strong cryptography. Cryptocurrencies use blockchain technology as the underlying system
architecture.
Cryptocurrencies function like a digital token of value, that allow people and organisations to make
direct payments to one another through an online system (RBA n.d.). Cryptocurrencies are primarily used
outside traditional and incumbent financial services systems and they are exclusively exchanged over the
internet using a special type of digital wallet called a cryptocurrency wallet (crypto wallet).
Global developments within the cryptocurrency space have given rise to DeFi and new ways of
conceptualising the notion of money and transactions, but cryptocurrencies remain largely unregulated.
Extensive research and evaluation should be undertaken to properly understand cryptocurrencies before
engaging with their use.
A system is classified as a cryptocurrency if it meets the following six conditions.
1. The system is decentralised and is run by means of a distributed consensus protocol.
2. The system is able to keep an overview of the cryptocurrency units and their ownership by users.
3. The system defines if and how new cryptocurrencies can be created. If yes, under which gover-
nance scheme, what is the mining process, which consensus algorithm will be used, what are the
circumstances of their origin, and how will ownership of cryptocurrency units by users be determined
and confirmed.
4. The system is able to prove a user’s ownership of cryptocurrency units only by means of cryptography.
5. The system can enable transactions that can result in changes in the ownership of cryptocurrency units.
6. If two different transaction instructions are entered at the same time, the system can select one
transaction only, to ensure change of ownership of the cryptocurrency units only occurs once.
Most cryptocurrencies are created as a medium of exchange or to monetise operations of blockchain
transactions. Recently, cryptocurrencies are also being traded for the purpose of investment, or as
collectables. Figure 2.9 provides a taxonomy of cryptocurrencies.
We discussed CBDCs (which may or may not be implemented as cryptocurrency) in section 2.2. We
will now discuss coins, tokens and stablecoins.
COINS
The majority of cryptocurrencies are considered to be ‘coins’ when they have their own blockchain, they
are decentralised, and they can be used in peer-to-peer transactions. Coins are branched into two categories:
currency coins and privacy coins.
Pdf_Folio:113
Cryptocurrency
Central Bank
Digital Currency
Commodity-
Privacy coin Security tokens collateralised
Non-collateralised
Consortium
stablecoin
Currency Coins
Most cryptocurrencies are currency coins. Currency coins have their own blockchain within which they
function as a currency. Bitcoin is currently the primary currency coin. Currency coins that are not Bitcoin,
are collectively known as alternative cryptocurrencies (altcoins).
Bitcoin
Bitcoin is the longest-established and best-known cryptocurrency coin. Despite not being legal tender,
some vendors accept Bitcoin as payment for goods and services. For instance, Bitcoin can be used to pay
for courses at Flinders University or buy a Lamborghini at Scuderia Graziani in Woolloomooloo (Crypto
News Australia 2020). See example 2.3 for more information on Bitcoin.
Alternative Cryptocurrencies (Altcoins)
Cryptocurrency coins other than Bitcoin are known as altcoins. Altcoins often differ from Bitcoin in terms
of protocol, architecture, consensus mechanisms, the reward process, or processing fees (among other
differences). Some target a specific industry (e.g. entertainment) or aim to overcome perceived flaws of
Bitcoin (Bradbury 2020).
Not all altcoins are simply alternative versions of Bitcoin. A number of altcoins use very different
algorithms, and serve very different purposes. For example:
• Litecoin (LTC) is able to process payments much faster than Bitcoin and is cheap to send in small
volumes, thus enabling micropayments (ASIC n.d.).
• Ether (ETH) is the cryptocurrency of the Ethereum blockchain. Ether is intended to enable and monetise
the operation of smart contracts and DApps, described in the previous part of the module (Divine 2020).
Unlike most other cryptocurrencies, there is no limit on the number of Ether that will be created, which
may affect its value through the forces of supply and demand.
• XRP is the cryptocurrency of the Ripple platform, which is a real-time digital payments system designed
for use by banking institutions in order to make faster payments. XRP is also known as the banker’s
coin. Ripple enables fast and cost-effective cross-border transfers between users in any currency,
such as fiat currencies, XRP and other cryptocurrencies. Instead of using a distributed ledger system
where users’ computers in the network are used to validate transactions, Ripple uses a distributed
Pdf_Folio:114
Privacy Coins
Privacy coins facilitate monetary or information transactions between users in full anonymity and without
the creation of an identifiable track. Privacy coins are grounded on the philosophy that it is not necessary
to have individuals or organisations monitor and control financial transactions.
The Monero cryptocurrency platform (XMR) enforces privacy, conceals participants’ information and
ensures that no one can deliberately or accidentally reveal transaction details (Seth 2019). A person cannot
be linked to a transaction, but users can choose who can see their transactions (e.g. tax authorities). Due
to the untraceability of transactions, XMR is the most prominent currency on dark web markets.
TOKENS
Tokens operate on top of an existing cryptocurrency blockchain network — they are dependent on the
blockchain network of another cryptocurrency. Tokens can represent a real asset (tokenisation) and thus
can facilitate access to a product/service or a stock.
Tokens are classified as a fungible and tradable asset or a utility. Tokens are usually released via an
initial coin offering (ICO) process. An ICO is the cryptocurrency equivalent of an initial public offering
(IPO) on the stock market. Investors can participate in an ICO to receive:
• utility tokens — tokens for accessing dedicated products and/or services associated with the token
• security tokens — tokens that represent a stake in a real-world asset.
Utility Tokens
Utility tokens are used on DApps, which are built for the purpose of using smart contracts. Tokens are
used either to purchase things on the DApp or to acquire certain advantages (e.g. discounts, voting rights).
Tokens have a corresponding price that holders can utilise for buying and selling purposes. For a transaction
to be made on a DApp, the transaction needs to be verified and a transaction fee is payable paid using the
cryptocurrency (not with a token) related to the blockchain network that the DApp is built upon (usually
Ether or NEO).
Example 2.12 describes the basic attention token (BAT), a utility token for the digital advertising
industry designed to create a fair, transparent, efficient and mutually beneficial advertising exchange
marketplace for publishers, advertisers and users.
EXAMPLE 2.12
.......................................................................................................................................................................................
CONSIDER THIS
What might be the pros and cons of cryptocurrency coins and tokens for micropayments compared to conventional
and other digital currencies?
Security Tokens
Security tokens are essentially digital securities. A holder of security tokens is considered as the owner
of a real-world asset (i.e. one is not in the blockchain). Security tokens are regulated by the respective
government agency, depending on the jurisdiction. In principle, in order for someone to be able to access
Pdf_Folio:115
Future Outlook
The token industry is still nascent and in light of the developments in the blockchain and in the regulatory
space more innovative usages for tokens are expected to emerge. Tokens could be used for payment
purposes if involved parties use this as a currency, digital asset ownership (tokenisation), rewarding
network participants, or even for accounting purposes. Tokens encourage the democratisation of assets
however, as always, clear regulatory and compliance oversight is required to protect investors, consumers
and the overall system.
STABLECOINS
Stablecoins are backed by real-world assets or benchmarks with the intention of making them less
volatile than other cryptocurrencies while still providing the instant processing, security and privacy of
cryptocurrency payments. As such, investors, users and institutions feel a greater level of security towards
stablecoins compared with other cryptocurrencies. There are various categories of stablecoin:
• fiat-collateralised stablecoins — pegged at a fixed ratio to and backed by reserves that include a large
proportion of one or more fiat currencies
• asset-collateralised stablecoins — backed by one or more commodities (e.g. gold)
• crypto-collateralised stablecoins — linked to one or more other cryptocurrencies by means of smart
contract
• non-collateralised (seigniorage) stablecoins — algorithmically regulate the supply of coins to stabilise
coin prices
• consortium stablecoins — issued by groups to enable instant cross-border payments.
EXAMPLE 2.13
USD Coin
USD Coin is an Ethereum-powered fully collateralised stablecoin that is backed up by fully reserved assets,
which include fiat currency (USD cash) and cash equivalent securities (such as Treasury Bills). USD coin
is redeemable on the equivalent fiat currency in a 1:1 ratio. This is basically a digital dollar architecture
that allows for secure and fast transactions with less counterparty risk, operated within a decentralised,
open, global and ‘less trust’ crypto-powered infrastructure.
The value proposition behind the USD Coin is the fact that it develops an architectural design that
allows fiat currency to interact with smart contracts. This interaction provides developers with a viable
Pdf_Folio:116
.......................................................................................................................................................................................
CONSIDER THIS
How would your evaluation of the potential applications of cryptocurrency be influenced by the existence of a well-
established asset-collateralised or fiat-collateralised stablecoin?
CRYPTOCURRENCY APPLICATIONS
A number of uses currently exist for cryptocurrencies.
• Medium of exchange. A business can accept cryptocurrencies as payment for goods and services and
customers can therefore use them as a form of payment.
• Store of value. Investors can invest in cryptocurrencies. They would be considered a high-risk investment
because the market is partly unregulated, and the market value of most cryptocurrencies fluctuates
massively when compared to other assets.
• Income. Individuals and organisations with sufficient resources can contribute computing power to a
blockchain network to mine cryptocurrency.
These functions depend on regulatory and compliance regimes, which can vary across jurisdictions. It
is also important to be aware of the tax implications of cryptocurrencies. For example, Bitcoin can be
taxed as property, and gains and losses from trading cryptocurrencies can be subject to the capital gains
tax regime.
Cryptocurrency holders are able to exchange/trade cryptocurrencies against other assets including fiat
currencies or other digital currencies via cryptocurrency exchanges (e.g. Binance, Coinbase, Bitstamp,
Huobi and Upbit). Even though cryptocurrency owners can choose to be pseudonymous (Bitcoin), or
anonymous (Monero), cryptocurrency exchanges are required by law to collect personal information and
identification from their customers.
Institutional Players
It is useful to explore how institutional players are engaging with cryptocurrencies in order to understand,
adopt and integrate them as part of their business models. Examples 2.14 and 2.15 discuss strategies used
by institutional players to adopt cryptocurrencies.
EXAMPLE 2.14
MUFG
Mitsubishi United Financial Group (MUFG) is the largest financial institution in Japan and second-largest
in the world (Kelso 2018). With the legalisation of cryptocurrency in Japan in 2017, financial institutions
have been looking for ways to integrate cryptocurrency to overcome the operational banking issues of
digital currencies. MUFG’s plan was to create its own cryptocurrency, MUFG Coin, that would allow its
customers to use it to make everyday transactions as a legal tender. However, due to regulatory constraints,
MUFG launched a mobile payment service that utilises MUFG Coin as an e-money instrument instead
(Ueda 2019).
Pdf_Folio:117
Retail Industry
An increasing number of businesses accept cryptocurrencies as a mode of payment. For example, Microsoft
accepts cryptocurrency as a payment method when purchasing games for its Xbox games console.
Starbucks and Overstock in the United States accept cryptocurrencies as payment for coffee and furniture,
respectively.
The video gaming industry has used virtual currencies for a long time (e.g. Second Life’s Linden
Dollar was discussed in section 2.2). As such, it is expected that game developers will incorporate
cryptocurrencies into the gaming experience (Efrima 2019).
As cryptocurrencies are still a relatively new development, there is a possibility that more businesses
will start adopting cryptocurrencies. As observed in Japan, if adopting cryptocurrencies means a better
payment experience for customers, businesses might see them as a way of improving their value proposition
(Orcutt 2019).
Crypto Derivatives
In finance, derivatives play a very strong role within institutional investment strategies. Derivatives are
effectively contracts between two (or more) parties to buy a specific asset at a future price. Derivatives are
used commonly in asset-trading such as the share market and foreign exchange functions (futures, swaps
and options). In the context of cryptocurrencies, there are developments that lean into the implementation
of crypto derivatives, which could result in making this blockchain-enabled asset class more appealing and
attractive to institutional investors.
Crypto derivatives would allow more parties to invest in cryptocurrencies (such as Bitcoin), but they
also potentially expose cryptocurrencies to even more speculative investing behaviour, thus increasing the
volatility of an already-volatile system. Global regulators are exploring ways to regulate futures trading
on cryptocurrencies more effectively. It will be interesting to see whether the developments surrounding
crypto derivatives will bring implications to the adoption of cryptocurrencies as a medium of exchange.
BitMex, Binance, FTX and Deribit are the most popular crypto derivatives exchanges. The Chicago
Mercantile Exchange (CME) recently launched options on Bitcoin futures, becoming the first established
derivatives exchange to adopt crypto derivatives. This constitutes a key milestone in the adoption of crypto
derivatives that provides institutional investors with trustworthy and compliant mechanisms, and tools, to
access them.
QUESTION 2.8
Explain whether stablecoins or older cryptocurrencies, such as Bitcoin or Ether, are more suitable
as a medium of exchange and as a store of value.
Pdf_Folio:118
Database
Address
book
Database Database
Address Address
book book
Sender Receiver
Blockchain network
Source: CPA Australia 2021.
FUTURE OUTLOOK
Further progress, growth and ecosystem diversification are expected. The overall value proposition of DeFi
is to bring more democratisation and transparency to finance by leveraging technological breakthroughs.
DeFi complements traditional financial services by providing the opportunity for institutions to move real-
world assets on to blockchain networks. This shift can better support financial inclusion, especially for
populations in underbanked and unbanked societies.
QUESTION 2.9
Pdf_Folio:120
BENEFITS OF CRYPTOCURRENCIES
The discussion above has highlighted a number of key benefits cryptocurrencies can bring to businesses,
consumers, systems and society as a whole. It is useful to summarise these in terms of four pillars.
Privacy
Using cryptocurrencies increases anonymity and privacy. In traditional financial systems, privacy is
achieved when central authorities limit information access to financial institutions that process the
transactions and the sender and receiver of the transaction. If cryptocurrencies are used, privacy and
anonymity are achieved as the personal details of the sender and receiver are never revealed. They are,
instead, represented on the ledger by their tokens (Yellin et al. 2013). The public is able to see that
a transaction took place, but they are unable to access any information linking the transaction to any
identifiable real-word parties. The user is afforded the privacy of transactions that cannot be traced back
to them, unless they voluntarily disclose the transactions (Reiff 2020).
Transactional Benefits
Cryptocurrencies provide for lower transaction and banking fees (Reiff 2020) by eliminating intermedi-
aries. Disintermediation also means transactions are settled immediately without the time-lag that is often
associated with transactions processed by financial institutions.
Pdf_Folio:121
RISKS OF CRYPTOCURRENCIES
Despite being secure and private, cryptocurrencies also carry many risks.
Regulatory Risk
Cryptocurrencies are a relatively young financial instrument, not yet subject to coherent regulatory frame-
works (ASIC n.d.). Regulatory risk is probably one of the most significant downsides of cryptocurrencies,
as the system is fertile ground for illegal activity. Numerous grey and black-market transactions are
conducted in Bitcoin and other cryptocurrencies (Disparte 2018). Cryptocurrencies are also increasingly
popular among money launderers. These developments are not surprising because, paradoxically, the
strengths that contribute to cryptocurrencies’ anonymity and security, also enable criminals to operate
below the surface. In 2018, the European Union (EU) included cryptocurrencies and cryptocurrency
exchanges when it adopted the Fourth Anti-Money Laundering Directive (Bigmore 2018). Cryptocur-
rencies were defined in the directive as a digital representation of value not issued by a central authority
(Staff of Global Legal Research Directorate 2018).
Market Risk
Holdings of most cryptocurrencies are concentrated; that is, only a handful of individuals and groups hold
the vast majority of funds (Disparte 2018). These individuals essentially control the supply of cryptocur-
rencies, making them susceptible to value swings, manipulation and ‘civil wars’. This exposes investors
and users to the risk of price volatility and negatively affects the adoption of these cryptocurrencies.
Cryptocurrency holders also face exchange issues. Not all cryptocurrencies have dedicated online
exchanges that allow direct exchange of cryptocurrencies to fiat currencies (Comm 2019). Therefore,
holders of non-mainstream and/or alternative cryptocurrencies face the possibility of not being able
to exchange their cryptocurrency for fiat currency without first converting their holdings into a major
cryptocurrency (usually Bitcoin).
Lastly, when compared to traditional payment processors such as Visa and PayPal, cryptocurrencies lack
arbitration processes in the event of disputes (ASIC n.d.). This effectively means that a user would have
no avenue for appeal if they were somehow cheated during a cryptocurrency transaction (Government of
Singapore 2018).
Fraud Risk
The anonymity and lack of traceability of cryptocurrency transactions enable fraudsters, money launderers
and tax evaders to avoid enforcement by authorities (ASIC n.d.; Disparte 2018).
BitConnect is a good example of fraud risk in cryptocurrency. BitConnect was a Ponzi-scheme,
promising investors a 10 per cent return on their initial investment. BitConnect crashed in 2018 and the
value of BitConnect coin plummeted (Comm 2019). Subsequent investigations by authorities suspected
that BitConnect was linked to money laundering activities and criminal syndicates.
The fraud risks associated with the lack of governmental oversight, and the example of BitConnect,
point to the current weaknesses in the cryptocurrency ecosystem. There is a need for a concerted effort
from national governments to collaboratively regulate cryptocurrencies and mitigate fraudulent activities.
Technology Risk
Technological failure (Disparte 2018) in the cryptocurrency network can lead to lost data. Therefore, users
need to take precautions to avoid data loss, and even worse, cryptocurrency loss. For example, users can
lose their private keys, due to data storage failures, which can result in a permanent loss of access to
their cryptocurrency.
Pdf_Folio:122
QUESTION 2.10
SUMMARY
Cryptocurrencies are a decentralised virtual currency hosted on blockchain platforms. In addition to the
cryptocurrency coins themselves, tokens may be created to digitise real-world assets or represent some
specific right. Stablecoins are pegged to some other benchmark in an attempt to achieve more stability
in the cryptocurrency’s value and thus provide greater confidence for users. Cryptocurrency overcomes
many of the weaknesses of other digital currencies. In particular, because it is based on blockchain,
it includes a robust and transparent system to validate and record transactions. It has therefore been
perceived as a potential way to move towards DeFi, which seeks to eliminate the role of central authorities
and intermediaries (such as the traditional financial institutions) in transactions.
In this part of the module we have examined the potential impact of cryptocurrency in business and the
financial system, including a number of use-cases, and explored associated benefits and risks to enable a
considered evaluation of potential cryptocurrency applications.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
2.2 Assess the effectiveness of using cryptocurrency and distributed ledger technology in solving
complex business problems.
• Cryptocurrencies have robust validation and security mechanisms and thus overcome many of the
problems involved with other virtual currencies.
Pdf_Folio:123
Pdf_Folio:124
Invoice trading
Alternative finance has grown in popularity in recent years. One particular contributor to this rise was
the Global Financial Crisis in 2008, and the impact of this event on the traditional capital markets. While
small to medium-sized enterprises (SMEs) heavily relied on the traditional system, banks in several OECD
countries started scaling back their lending portfolios, resulting in the stagnation of European short-term
money markets (Paulet 2018). It resulted in SMEs experiencing difficulties in obtaining required finance,
which drove them (particularly those in innovation sectors) towards seeking alternative finance providers.
Businesses with above-average growth potential were adversely affected by the inability to access
funding (Thompson, Boschmans & Pissareva 2018).
Pdf_Folio:125
Paperwork and business conducted requiring Minimise paperwork and business conducted online
physical presence
A variety of qualifications and prerequisites (can be More relaxed and less strict qualifications and
very strict) prerequisites
Varied approval rate, usually low Varied approval rate, usually high
High priority given to credit score and credit history Low priority given to credit score and credit history
Strict repayment terms, not flexible Relaxed and flexible repayment terms
It is important to note that these different approaches are generic. A prospective borrower needs to
engage with a lending scheme with careful consideration and appropriate research. Any evaluation should
be made against the borrower’s specific circumstances and objectives.
The flexibility of alternative finance is grounded primarily on the reach of lenders and borrowers, the
terms of conditions they have put in place, and the funding philosophy behind each lending provider.
QUESTION 2.11
Why has there been a trend away from traditional sources of finance towards alternative finance
sources in the past decade?
The platform does not need to have a banking license A platform needs to have a banking license
Loan issued at the risk of the peer-to-peer investor Loan issued at the risk of the balance sheet lender
Attracts fee for services (utilisation of infrastructure) Revenue is earned both from a platform’s fee structure
and from the interest payments accruing from loans
Pdf_Folio:128
No additional costs (acquiring capital) to cover losses Additional costs to cover losses on bad loans
on bad loans
Cost structure deemed more transparent and value add Cost structure considered less transparent
Continuous balancing of capital lending and supply – Funds readily available to be distributed
can result in longer funding periods
Since the landscape of alternative forms of lending is evolving rapidly, and also in line with regulatory
and compliance developments, numerous online lenders are adopting hybrid forms of lending, utilising
both peer-to-peer and balance sheet lending schemes. It is expected that the experimentation and wider
adoption of these schemes will unveil more innovative and sophisticated models, enabling online lenders to
both diversify and specialise across different niche markets. It can also be expected that incumbent banking
institutions will engage more actively in partnerships, mergers, or acquisitions with online lenders, or even
run their own proprietary online lending platforms.
INVOICE FINANCE
Under the invoice finance umbrella two mainstream options can be observed: invoice factoring and
invoice trading.
Invoice Factoring
Invoice factoring involves a contractual agreement with a third-party provider, which is responsible for
managing the entire sales ledger. Such providers usually contact customers directly in order to collect
payments. Even though a number of companies find this convenient, research has shown that such an
arrangement may result in loss of trust and mismanaged approaches, which can put a delicate customer
relationship in jeopardy. Moreover, invoice factoring entails fees and contractual arrangements that can be
inflexible, costly and complicated.
Invoice Trading
Invoice trading, peer-to-peer invoice finance and invoice crowdlending allow businesses to sell
individual outstanding invoices through electronic platforms (e.g. Timelio, MarketFinance) in order to
obtain working capital. This type of pre-financing of outstanding invoices is a form of short-term debt, with
lenders being either individual or institutional investors (Dorfleitner 2017). Invoice trading is considered
to be asset-backed lending because the invoice serves as collateral.
Businesses would normally consider invoice trading as a source of finance for the following reasons:
• improving working capital
• freeing up cash that can be used for short-term liquidity optimisation
• paying expenses
• initiating new investments.
Invoice trading is beneficial to small businesses, as it is easy to apply for, and results in immediate short-
term liquidity without diluting equity. In addition, this type of lending allows small businesses to provide
payment terms to their clients without significantly affecting their working capital (Commercial Capital
n.d.; Reynolds 2018).
Pdf_Folio:129
P2P
Seller Financier Seller Investors
platform
Investors
Flexibility: Tend to dictate what is sold, and usually the Flexibility: Seller retains control of which invoices are
seller has to sell everything. sold, and when they are sold.
Transparency: There can be unanticipated, or Transparency: Usually there are no fees to trade
undisclosed fees. invoices, or these fees are very clear.
Contract: Seller can become locked into payment Contract: Seller can sign up and extract themselves at
contract, over a longer term. any time. They can usually pay as they go.
Set-up speed: Can be slow and may take up to 6 weeks Set-up speed: Usually very fast. Set-up occurs within
to receive funds. 48 hours and cash can be delivered within a day.
Privacy and confidentiality: The sellers debtors tend to Privacy and confidentiality: Usually the process is
be notified of their intentions. private and confidential.
QUESTION 2.12
Explain how the invoice finance business model has evolved with the introduction of electronic
platforms.
INVESTMENT-BASED CROWDFUNDING
Investment-based crowdfunding has significantly grown in popularity in recent years. It is based on
companies offering contractual instruments, equity or debt via an online crowdfunding-platform to a crowd
(Rossi & Vismara 2018). An ownership interest in the business is exchanged (Dehner 2015). In contrast
to other forms of crowdfunding, investment-based crowdfunding campaigns are typically launched by
businesses instead of individuals. The maximum amount of funds to be raised is typically limited (Rossi
& Vismara 2018).
The most common types of investment-based crowdfunding are outlined in the sections that follow.
Equity Crowdfunding
Regulatory changes in some countries have allowed privately owned businesses to raise capital by offering
a specified amount of bond-like shares or equity to investors (Ahlers et al. 2015; Salampasis 2020).
Equity crowdfunding enables early-stage businesses that are not yet publicly traded to raise money by
means of equity, by providing the opportunity to the crowd to become shareholders. Equity crowdfunding
can be conducted in a shorter time-frame and more cost-effectively than an IPO (Ahlers et al. 2015;
Salampasis 2020).
This form of investment can offer potentially high returns for investors, but is highly risky and the
investors typically lack the capability to conduct extensive research on potential investments. The shares
are illiquid as they are not listed on stock exchanges (Ahlers et al. 2015; Salampasis 2020).
As equity crowdfunding involves the sale of a security, legislation in many countries imposes restrictions
on its use, including limits on the participating business’ size, or limits on the funds raised per year.
Australia, for example, requires that the crowdfunding campaigns have to involve accredited intermediaries
(Ahlers et al. 2015; Salampasis 2020). More generally, equity crowdfunding is usually subject to securities
and financial regulations. In Australia, types of equity-based crowdfunding that result in a financial reward
fall under the Corporations Act 2001, as they can amount to a financial product. Other types of equity-
based crowdfunding resulting in ownership or equity interest in exchange for funds are also considered to
fall under the Corporations Act as they can amount to a managed investment scheme. All these activities
are also under the oversight of ASIC (Stevens et al. 2019).
Long-established means of equity funding, such as venture capital, private equity and angel investing
schemes, involve one-to-one relationships with investors. Crowdfunding instead involves an ecosystem of
investors by means of an online platform.
The key attributes of equity crowdfunding can be summarised as follows.
• A predetermined set of terms describing the overall scheme (e.g. how much to sell, how investors will
get rewarded and what will be the price). In order to determine these terms an in-depth analysis and
concise valuation of the venture is required.
• A set of payable fees is required for the online platform (e.g. administration, legal, advisory). These fees
can be significant.
• It provides many people with the opportunity to invest; therefore, on the one hand, there is a breadth of
potential investors, while on the other, the venture does not depend only on fewer wealthier investors.
• The venture must be investment-ready; therefore, a detailed business plan and financial predictions/fore-
cast are required. Key information and accurate data must be readily available for investors to review,
ideally along with relevant marketing materials.
• The online platform will conduct the necessary due diligence, but potential investors also have the right
to conduct their own due diligence and often require additional and more specific information and data
that the business must make available.
• The venture will have to factor in legal costs, in order to manage and comply with all the associated
legal aspects.
• The cohort of investors can be diverse and it is very important to have predetermined conditions around
the purpose of voting rights, the level of control external shareholders will have on the business, the
compensation schemes and other relevant issues.
Pdf_Folio:131
• Reduced complexity and risk • Higher autonomy and lower fees compared to
• Diversified investment portfolio traditional bank loans
• Higher reach – democratisation provides smaller • Acquiring financing at a faster speed
investors with the opportunity to invest in high-end,
prestigious developments
Advantages Disadvantages
• Has the potential to generate higher returns. • Returns are driven by market prices (volatility) and
• There are tax benefits (normally tax is paid after the may not be guaranteed.
shares are sold). • Due diligence on the companies that run property-
• There are no upper limits on the return on investment. based crowdfunding projects might be difficult.
• Holding period can be very lengthy (in most cases
investments are illiquid).
• Extensive due diligence is required for informed
decision-making (potential difficulties in properly
assessing the real estate market).
• Some property-based crowdfunding platforms may
only accept accredited investors.
Pdf_Folio:132
Advantages Disadvantages
• It is safer, compared to equity (debt gets paid first). • Has lower returns due to lower risks.
• The return on investment is usually more predictable • Due diligence on the companies that run property-
and stable (fixed return based on the interest rate). based crowdfunding projects might be difficult.
• There is a shorter holding period (duration is always • Extensive due diligence is required for informed
known). decision-making (potential difficulties in properly
assessing the real estate market).
• Some property-based crowdfunding platforms may
only accept accredited investors.
A blended option called convertible debt is also available, in which start-ups acquire funds from investors
in the form of a loan (lending), which is repaid at a later stage in the form of property shares (equity).
QUESTION 2.13
Outline the main features of equity crowdfunding models for financing entrepreneurial ventures.
NON-INVESTMENT-BASED CROWDFUNDING
Non-investment-based crowdfunding schemes can be either rewards-based crowdfunding or donation-
based crowdfunding.
Reward-Based Crowdfunding
In reward-based crowdfunding, a project attracts the crowd by offering a tangible non-financial reward,
such as a branded t-shirt, or the promise of a product that they will receive in the future (Dehner 2015). In
this context, entrepreneurs incentivise financial donations in return for a service or product. The business
benefits from this form of crowdfunding by not having to dilute its equity. This type of crowdfunding
is very popular within the creative industry where products would probably not qualify for other types
of financing.
This form of crowdfunding also enables businesses to test demand for a product at an early development
stage (Salampasis 2020).
Table 2.24 summarises the advantages and disadvantages of reward-based crowdfunding.
Advantages Disadvantages
• Considered to be one of the most cost-effective ways • Usually generates only very small amounts of
of raising funds funding
• No collateral needed, but a credit check is required • Potentially difficult to reach a funding goal
• A simple process • Idea exposure to competitors can potentially result
• Equity and control of the company are retained. in having the idea stolen
Funds do not need to be repaid
(continued)
Pdf_Folio:133
Advantages Disadvantages
• Opportunity to gain wider exposure that can • If the funding goal is not reached, this can incur
potentially result in onboarding more early adopters financial penalties and bad publicity
while raising brand awareness • May not be suitable for complicated concepts, or
• Opportunity to test the value proposition projects that are hard to explain
In Australia, reward-based crowdfunding does not fall within the Corporations Act and ASIC oversight.
However, due to the promise of rewards, it can be considered to fall under contract law or the Australian
Consumer Law (Stevens et al. 2019).
Reward-based crowdfunding campaigns are provided by a range of crowdfunding platforms. The most
popular platforms are Kickstarter, Indiegogo, Crowdfunder.co.uk and Patreon. Each of these platforms
charges a fee, provides different tools and has different terms and conditions that should be understood
prior to engagement.
Donation-Based Crowdfunding
Donation-based crowdfunding builds upon the intrinsic value of a project, receiving donations from
individuals who do not expect ownership or returns from the project (Dehner 2015). This form of
fundraising is typically utilised by individuals or not-for-profits (Salampasis et al. 2020). Donation-based
crowdfunding is a vehicle to pool funding from the crowd in order to finance a project with a social purpose
or cause, without the expectation of having the funds returned. Examples might include raising funds for
renovating a school, creating a community garden or covering high and unanticipated medical expenses.
Table 2.25 summarises the advantages and disadvantages of donation-based crowdfunding.
Advantages Disadvantages
• Donated funds are received quickly. • A number of crowdfunding platforms can keep a
• A number of crowdfunding platforms do not charge percentage of the donations as a fee.
fees for donation-based campaigns. • Requires marketing and in some cases it may be
• Can be leveraged to reach a broader network of difficult to reach out to the crowd.
people. • It can be considered a ‘crowded’ marketplace where
• Can pool small donations from a large number of numerous venturers may be ‘competing’ for funds.
people (crowd). • A number of crowdfunding platforms may impose
• Builds accountability and transparency in terms of minimum funding targets.
how donations are managed and utilised. • It may be difficult to reach funding targets,
• Offers the opportunity to establish a broader sometimes if funding targets are not reached then
community of social cause advocates and supporters. the funds are not released by the crowdfunding
platform.
• Donations could be used for tax evasion or money
laundering purposes.
Within Australia, donation-based crowdfunding does not fall under the Corporations Act and no ASIC
oversight is provided (Stevens et al. 2019).
• The venture that wishes to raise funds prepares a whitepaper describing the purpose of the project and
the token that will be issued for the purpose of the fundraising process (crowdsale).
• A smart contract required for the token sale is written and added to a tokenisation platform such as
Ethereum.
• A website is developed that includes all the necessary information about the project, the venture and the
ICO campaign.
• The venture may wish to list its token on one or several marketplaces or exchanges. This means that
after the ICO campaign is completed secondary trading can be enabled.
• The venture engages in a marketing and promotional campaign (usually on social media).
• Once the ICO campaign is launched investors can purchase tokens by sending transactions to the token
smart contract.
• Upon completion of the ICO and upon condition that they are listed on an exchange, tokens can become
tradable. At the same time, the venture can pool the crowdsourced funds and use them to fund the
venture.
Pdf_Folio:135
EXAMPLE 2.16
BitDegree
BitDegree launched an ICO campaign in December 2017. BitDegree is a blockchain-enabled online
education platform. The platform has its own token (BDG) that is used for scholarships and to acquire
technical talent. Via the BitDegree platform students are offered online courses, and businesses can recruit
technology talent to support the development of the future workforce. The utility token sale began on
1 December 2017 and ended on 28 December 2017, attracting 12 000 contributors. The total amount
raised was $22 130 000. The ICO token price was 10 000 BDG = 1 ETH. The total token number was
660 000 000. The BitDegree utility token distribution was:
• 66 000 000 tokens for the BitDegree Foundation
• 165 000 000 tokens for the scholarship pool
• 336 600 000 tokens available to the public at the time of the ICO campaign launch
• 66 000 000 tokens for the team of developers
• 13 200 000 tokens for advisors and partners as incentives/rewards.
The accepted currency was Ether (ETH). The token was issued immediately after the ICO. KYC
processes took place.
The fundraising campaign is conducted via the token The fundraising campaign is conducted via the
issuer’s (crypto start-up) website. website of the cryptocurrency exchange.
The crowdsale is managed by the crypto start-up. The crowdsale is managed by the cryptocurrency
exchange.
Pdf_Folio:136
KYC/AML can happen but this is determined on a KYC/AML is conducted by the cryptocurrency
project by project basis. exchange.
The crypto-start-up is required to put together a The cryptocurrency exchange handles the marketing
marketing campaign in order to raise awareness and campaign (cost is factored in the fee agreement).
attract attention (can be costly).
Screening is not required in order to launch a crowdsale. Screening is required and the cryptocurrency
In principle, anyone is able to launch an ICO campaign exchange is responsible for screening the crypto start-
in a country where it is legal. up before it will permit the crypto start-up to utilise its
platform in order to raise funds.
If a crypto start-up wishes to list its tokens after the The cryptocurrency exchange will automatically list the
crowdsale then this needs to happen through an tokens after the crowdsale.
exchange.
Due to the lack of regulatory oversight, numerous ICO campaigns took place without ensuring proper
KYC/AML processes were met. Many crypto start-ups (token issuers) exploited the situation and initiated
ICO campaigns, without promising value propositions, and with unscalable business models. Some crypto
start-ups had not even communicated a proper whitepaper. This resulted, across numerous cases, in the
funding of a range of dubious and sometimes illegitimate business models, and many investors lost their
money. This is the main reason for the widespread failure of ICOs, and the development of more concrete
and proper blockchain-enabled funding schemes such as IEOs. IEOs aim to protect investors by ensuring
proper KYC/AML processes and complying with regulatory requirements. To that respect, from an ICO
process, understanding, and evaluation perspective, it is very important to know whether proper KYC/AML
processes took place, in conjunction with the nature of, and the objective of the organisation seeking to
use this business funding model.
They do not need to be in compliance with regulatory They must be in compliance with regulation (in
requirements. principle securities law).
The true value cannot be easily assessed (highly True value can be assessed easily.
speculative).
QUESTION 2.14
Each of these funding schemes can be used for different purposes and has inherent peculiarities and
conditions. Matching a venture’s characteristics with the appropriate form of funding is the most important
strategic question. Consider examples 2.17–2.19.
Pdf_Folio:139
Business A
Business A has high risk and the return on investment is quite uncertain and unstable. The cash-flow is
weak and the growth rate is expected to be low to moderate. The business has high leverage, meaning
that the business has more debt, compared to equity. Finally, the business’ management is unproven and
there are doubts that it will be able to realise the goals and the value proposition of the business.
Points of Consideration
Taking into account the characteristics of the business, the likely options as a source of funding could
be personal funds, friends and family or other forms of bootstrapping (e.g. credit cards, personal loans,
bartering).
EXAMPLE 2.18
Business B
Business B has low risk and the return on investment seems to be more predictable. The cash-flow looks
strong and the leverage is low. Moreover, the financials are properly audited and managed and the overall
balance sheet and cash flow look strong and healthy. There is no history nor indication of loan repayment
issues. Furthermore, the business’ management team are capable, and there is confidence that they are
managing the business in a sound and effective manner. The owner-manager of the business is trustworthy
and understands the business.
Points of Consideration
Taking into account the characteristics of the business the appropriate umbrella of funding schemes would
be debt financing. The specific form of debt financing will depend on multiple parameters, as already
discussed.
EXAMPLE 2.19
Business C
Business C is expected to offer a high return. The value proposition and the business model leverage on
a unique and highly innovative business idea that has a high and sustainable potential. The business is
expected to experience high growth. The business is active within a very niche market with high barriers to
entry. Moreover, the management team has proven itself and is able to drive further growth of the business.
The owner-manager and the team have a solid understanding of the business, they are trustworthy, and
they share the vision and genuine desire to succeed. Management is capable and has identified all the
potential risks and has put in place a range of different contingency plans to minimise disruption.
Points of Consideration
Taking into account the characteristics of the business, the appropriate umbrella of funding schemes
would be equity financing. However, debt financing might also be an option depending on the owners’
willingness to dilute or not their ownership. Both equity and debt financing have advantages and
disadvantages and dedication to either scheme depends on contextual peculiarities. A blended approach
may also be possible.
SUMMARY
Alternative finance is any form of financing through sources outside the conventional financial system.
It has grown significantly in recent years as technology has driven innovation. This has provided new
opportunities for SMEs and start-ups in particular, who have long encountered difficulties raising finance
through traditional providers.
Alternative lending includes peer-to-peer lending, balance sheet lending and invoice finance. Investment
and non–investment based crowdfunding provide another potential source of funding. Finally, for cryp-
tocurrency businesses, a variety of funding mechanisms have been developed. Initial coin offerings are the
cryptocurrency equivalent of IPOs. Perceived shortcomings of ICOs have, however, sparked a number of
alternative approaches.
The choice of funding, whether mainstream, alternative or a mix must be made with careful considera-
tion for the business’s particular needs and circumstances, including its phase of development, cash flow,
expected growth, existing debt, debt-to-equity ratio, credit and other factors.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
2.3 Support the adoption of alternative finance sources to the organisation’s stakeholders as an
innovative solution to traditional finance requirements.
• Alternative finance business models are changing the way debt and equity funding is implemented.
• Stakeholders need to be aware of the changes in order to be able to make informed decisions.
• Alternative finance sources can be cheaper, less complicated and easier to access, particularly for
SMEs and start-ups that traditionally facing multiple barriers in accessing capital.
• The disintermediation and decentralisation of the financial services value chain increasingly enables
a shift toward decentralised forms of funding and lending.
• Alternative finance sources include peer-to-peer lending, balance sheet lending, invoice finance,
crowdfunding and crypto-based funding.
Pdf_Folio:141
CUSTOMER PREFERENCES
Customer preferences are the major catalyst for innovation within banks. Banks are embracing technologi-
cal developments to connect and cultivate relationships with their customers. Moreover, banks are seeking
to leverage the abundance of data that is becoming available to them, in order to better understand their
customers, and provide more customised products and services that personalise the banking experience,
such as real-time offers on the banking app when customers are walking past certain retail stores.
Customers as a whole are increasingly comfortable with technology, and prefer it when it delivers
convenience, practicality and usability. The increased adoption of digitalisation has reduced face-to-face
interaction, but customers still prefer it when they need to solve complex issues, seek specialist advice and
guidance, or engage with complex products, such as a mortgage (Dallerup et al. 2020). Because of this,
bank–customer interactions are shifting to a blended model.
Customers also increasingly expect seamless integration between businesses and financial services
(PwC 2016b). For example, Australian businesses that accept online payments may now benefit from
providing their overseas-based clients with the option to pay in currencies other than Australian Dollars.
Because of this, businesses now expect their banks to be able to manage all the back-end transactional
services of converting foreign currencies and verifying transactions.
Accenture’s 2017 Global Distribution & Marketing Consumer Study identified three archetypes of
banking consumer with different preferences (Accenture 2017a). These are summarised in table 2.29.
Source: Adapted from Accenture 2017, ‘2017 Financial Services Global Distribution & Marketing Consumer Study’.
Pdf_Folio:142
Technologies
Digital banking technologies provide a bank’s customers with the ability to bank online, and process
contactless and instant payments. These technologies also affect the bank’s ability to digitally streamline
and improve everything from front-end operations to back-end functions.
The key characteristic of these technologies is the fact that they are data driven. They depend on solid
data management and governance schemes, along with interactivity and interoperability capabilities.
A key element, when considering and evaluating technologically led applications, is the fact that behind
a single business model a combination of technological capabilities can co-exist. Due to the fact that all
these technologies are data-driven and data-led, they institutionalise and embody a data-embedded culture
within banking, opening doors for cutting edge products, services and business models.
At the other end of the spectrum though, banking organisations need to realise that these technologies
cannot simply function as ‘plug-and-play’ or off-the-shelf solutions — well considered and planned
integration processes are required to shift from legacy systems and existing organisational architectural
designs. These technologies require appropriate mindsets that embrace innovation and thrive in a multi-
disciplinary setting. Organisations will require talent that can see and operate beyond generic and siloed
proprietary domains.
The most prominent technologies that are spurring change, and have the most potential to shape the
future of banking, include:
• hybrid cloud — enterprise-wide frameworks to enable AI to perform advanced analytics (discussed in
modules 3 and 4) on banks’ big data in order to better understand and engage with customers
• automation — a blend of technologies (be discussed in module 3), is being used to automate tasks
and processes, augment human performance to improve the speed, quality and scale of operations, and
enable new service innovations
• API platforms — the integration of data, processes and services across multiple systems to create new
service experiences for customers (Shevlin 2020)
• instant payments — payment innovations that enable near-real time exchanges to meet business and
customer expectations
• augmented reality — an immersive technology to enhance customer experience, and provide a new way
of capturing information and visualising data
• blockchain — a potentially decentralising and disintermediating transformative force in the financial
system, but also a platform that may enable traditional institutions to improve certain banking functions
and activities such as cross-border payments and digital identity verification
• AI — enabling advanced analysis of big data to improve and implement sound and informed decision-
making across the entire value chain of financial services
• prescriptive security — the use of advanced analytics, AI, big data analysis and real-time monitoring
tools to help organisations foresee and gain earlier visibility of potential cyberthreats
• smart devices and the Internet of Things (IoT) — enabling new ways for institutions and customers to
interact.
Regulatory Framework
Key developments in the regulatory framework include: smart regulation and compliance, regulatory
sandboxes, and the open banking and the Consumer Data Right (CDR) scheme. We will discuss each
of these below.
Smart Regulation and Compliance
Regulation tends to be slower than technological developments. It often fails to keep pace in a proactive,
agile and responsive manner. The exponential growth of technological developments and the high adoption
rate in banking and the other financial services industries signal to regulators and policymakers that
current regulatory frameworks and regimes are not keeping up with the emerging financial technologies
environment.
Regulatory technology (RegTech) aims to instil a data-driven and data-led mindset in the key areas of
regulatory monitoring, reporting and compliance. RegTech aims to implement more granular and effective
Pdf_Folio:143
Barriers to Change
Alongside the enablers of change, it is important to recognise there are barriers to digitally enabled banking
applications. These include:
• cost and resources
• regulation and compliance
• organisational and cultural resistance
• scalability of IT-led business models
• legacy systems and processes.
Source: Adapted from Deloitte 2018, ‘SME digital payments: New opportunities to optimise’,
The Paytech Revolution Series, Deloitte, https://b2bpay.com.au/wp-content/uploads/2018/03/
Deloitte_Payments-Landscape-Report-2018.pdf.
Businesses have increasingly comprehensive access to information about financial products and increas-
ing ability to acquire and utilise financial products that are tailored to fit their needs. This adoption of
new technologies and digitisation will allow these businesses to provide better experiences to their own
customers, and enable them to gain access to international markets. The next step for businesses is to
conduct internal and external assessments of their capabilities and options so they can make informed
decisions on how to best implement these developments into their existing processes.
EXAMPLE 2.20
Ant Technologies
Ant Technologies is a spin-off of China’s Alibaba, offering financial services such as a mobile wallet
(i.e. Alipay), online lending platform (i.e. MyBank), and an investment fund (i.e. Yu’e Bao) (Mansueto
Ventures n.d.). In 2018, Ant Technologies introduced a cross-border remittance network based on its
blockchain system through Alipay in Hong Kong and GCash in the Philippines.
Ant chose to partner with Standard Chartered Bank as the settlement bank for Alipay and GCash. This
collaboration means customers can remit money instantly between Hong Kong and the Philippines at
more competitive exchange rates with a minimal transaction fee (Business Wire 2018).
Ant Financial recently rebranded to Ant Technologies under the scheme of ‘definancialising’ operations,
in order to focus on providing technology services to financial institutions. This scheme could be a
precursor to regulatory arbitrage, as it operates the Alipay mobile payments network and the Tianhong Yu’e
Bao money market fund as a technology provider, not a financial institution. Time and future developments
will determine how BigTech players will behave when trying to meet the digital demands of the new era
of banking.
Pdf_Folio:145
Banking-as-a-Service (BaaS)
Banking-as-a-service (BaaS) is where organisations with a banking licence integrate their digital banking
services directly into the products and services of other non-bank businesses (Dolan 2019). As a key
component of open banking, banks are sharing their APIs with partners and other FinTechs to allow these
parties to develop new services around the banks’ digital banking services. These third-party partners
normally pay a fee to the bank to gain access to its API (Dolan 2019).
Banking-as-a-Platform (BaaP)
Banking-as-a-platform (BaaP) is where banks integrate services from other FinTechs to change the
banks’ existing offerings. In the BaaP model, the customers belong to the bank and the bank integrates
functionalities and services from FinTechs and non-bank partners. For example, a bank might integrate
a third-party robo-advisor or chatbot into its banking app. Banks utilise the BaaP approach to retain
customers who might be tempted to move to FinTechs that offer more personalised products.
Trust
Essentially, trust means believing that someone (or something) is inherently good and honest, and also safe
and reliable (Cambridge Dictionary 2020). Trust is an important competitive advantage in the increasingly
data-driven banking industry (McIntyre 2019).
Trust will determine the level of blended and augmented relationships between consumers and banking
institutions, along with the overall dynamics of banking ecosystems. The interoperability, transferability
and sharing of data will enable new forms of trust, which will not be solely based on long-lasting principles,
but on consent. Consumers’ consent will not be ongoing and permission to use data will not be permanent.
Banking institutions will have to constantly and continuously ‘verify’ the level of trust. This will occur
not by being self-contained providers, but by keeping their promise to deliver on personalised customer
experiences. Trust could become portable and may depend on the most optimal insights generated by the
usage and control of data sharing. Banking providers should not view consumer trust as a taken-for-granted
proprietary asset.
Pdf_Folio:146
QUESTION 2.15
Discuss how RegTech can be used to make financial regulation relevant for FinTech companies.
SUMMARY
The banking industry is undergoing unprecedented change. Banks are seeking to leverage their data to
better understand their customers, and thus provide a personalised banking experience. New technologies
increasingly enable data-driven approaches to create innovative business models and services.
Regulators have introduced a number of changes to support innovation in the banking industry, including
the promotion of open banking and the creation of regulatory sandboxes to support the development
and testing of innovative services (and enable regulators to learn about them). RegTech aims to make
regulatory monitoring, reporting and compliance data driven, with a focus on stakeholders, competition
and opportunity rather than risk and problem solving.
Banks are responding to FinTech by digitally transforming their core offerings; creating new distribution
channels by providing third parties with access to data and services; embedding services into third-party
customer experiences; developing and launching co-branded products and services; and launching new
ventures based on new business models.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
2.1 Evaluate the effectiveness of digital currency and payments in meeting the organisation’s
objectives.
• Banks are responding to financial technology by transforming their offerings and adopting data-
driven business models and services, resulting in increasingly integrated and personalised cus-
tomer experiences, including in the payments space. Businesses shifting towards digital payments
experience reduced payment processing times, more efficient approvals, better access to data and
a range of other benefits.
Pdf_Folio:147
CASE STUDY
Little Phil
Problems and Opportunities
Charities generate nearly AUD1 trillion of donations worldwide per year, with the funds used to support
nominated causes and to administer the charity. Despite the size of the charitable-giving sector, a number
of problems inhibit people from giving, and reduce the impact of funds that are donated. These issues
include the following.
• Expensive intermediary costs. Third-party fundraisers, international money transfers and internal charity
staffing costs can take up the majority of funds raised by a charity. For example, third parties can charge
large up-front fees for securing donations and marketing. These fees can sometimes exceed 80c for
every dollar donated. Therefore, the funds actually reaching the recipients intended by the donor are far
short of funds original given. The use of traditional fundraising tactics (cold calling, door knocking and
street canvassing) may create an aversion to giving. As a consequence of poor fundraising practices,
the public is losing trust in charities, and fewer people are donating.
• Outdated fundraising methods. Traditional fundraising methods do not appeal to Millennials, leading
to a generation gap in giving. Donors are demanding transparency and accountability from charities,
wanting to see the direct impact of their donation. They think, ‘If technology can track an Uber, your
post or a pizza, why can’t donations be tracked?’
• Lack of transparency. With limited exceptions, donors cannot see how and where their donation is being
distributed; that is, the impact of their giving.
• Exploitation by recipients. Many charities do not have the ability to validate beneficiaries’ access to
support from other charities, leading to inefficiencies and undermining the impact of donations by
denying other recipients who have a genuine need. There is no system of data sharing among charities
to validate beneficiaries who may register with multiple charities.
Approach
Little Phil, short for Little Philanthropist, intends to apply a combination of blockchain technology,
cryptocurrency, digital payments and software applications (collectively ‘the Little Phil Platform’) to:
• decrease fundraising and operational costs so more funds are distributed to the cause
• improve transparency, accountability and control over how donors’ funds are used
• provide evidence of the direct impact of donations.
Little Phil conceptualises the problems and the solution as shown in figure 2.13.
A user case scenario for the Little Phil Platform, involving the donation of funds as part of a charity
campaign, is set out in the steps below. The platform can also be used to make direct donations to
Pdf_Folio:149
Pdf_Folio:150
REFERENCES
Accenture 2017a, ‘2017 Financial Services Global Distribution & Marketing Consumer Study’, https://financialservicesblog.
accenture.com/series/2017-financial-services-global-distribution-marketing-consumer-study.
Accenture 2017b, ‘Driving future of payments: 10 mega trends’. Accenture, https://www.accenture.com/au-en/insight-banking-
future-payments-ten-trends.
Ahlers, GKC, et al. 2015, ‘Signaling in equity crowdfunding’, Entrepreneurship Theory and Practice, vol. 39, no. 4,
pp. 955–80.
Alderman, L 2020, ‘Our cash-free future is getting closer’, New York Times, 6 July, accessed July 2020, https://www.nytimes.com/
2020/07/06/business/cashless-transactions.html.
Antonopoulos, AM & Wood, G 2019, Mastering Ethereum: Building Smart Contracts and DApps, O’Reilly Media.
Anwar, H 2018, ‘The ultimate comparison of different types of distributed ledgers: Blockchain vs Hashgraph vs Dag vs
Holochain’, 24 December, accessed June 2020, https://101blockchains.com/blockchain-vs-hashgraph-vs-dag-vs-holochain/#
prettyPhoto.
Artzt, M & Richter, T 2020, Handbook of Blockchain Law: A Guide to Understanding and Resolving the Legal Challenges of
Blockchain Technology, Wolters Kluwer.
ASIC n.d., ‘Cryptocurrencies and ICOs’, Moneysmart.gov.au, accessed July 2020, https://moneysmart.gov.au/investment-
warnings/cryptocurrencies-and-icos.
Assenmacher, K & Krogstrup, S 2018, Monetary policy with negative interest rates, International Monetary Fund,
Washington, DC.
Pdf_Folio:151
Pdf_Folio:152
Pdf_Folio:155
Pdf_Folio:156
PREVIEW
Modules 1 and 2 explored how the near-real time and free flow of data in the global network of
interconnected devices, systems, individuals and organisations has shaped a digital finance ecosystem
that redefines how value is created, delivered and consumed. Data is at the core of almost every business
process, every transaction and every service experience. In this module, we will explore the use of digital
tools that apply data in a wide range of business processes to achieve efficiencies and create value. These
tools have extensive ‘back-office’ and ‘customer-facing’ applications — and many of them increasingly
integrate across the entire business.
Part A explores the concept of automation — the use of technology to replace human labour to achieve
efficiencies and enable the deployment of human resources to focus on higher value work. A balanced
scorecard is explained as a framework in which to evaluate the potential and performance of automation
technologies in achieving the business’s objectives.
Part B of the module focuses on robotic process automation, which is the use of technology to replace
human labour in simple rules-based tasks, often those involving the processing of data. This type of
automation essentially mimics the steps that a human worker performs, but many times faster and without
quality variation.
Part C of the module explores artificial intelligence (AI) — a collection of technologies that exhibit
capabilities normally associated with human intelligence. AI technologies can work with data in sophis-
ticated ways and has many applications in generating business insights, improving customer experience,
automating processes and helping team members work more effectively.
Part D of the module explores a specific type of AI called machine learning (ML) that is able to interact
with data in even more sophisticated ways.
With the benefits, costs, risks and potential applications of these digital tools established throughout
the module, Part E of the module concludes with a discussion of how to select and implement the most
appropriate digital tools to solve business problems and help achieve the business’s objectives.
Pdf_Folio:157
PART A: AUTOMATION
INTRODUCTION
Automation enables more efficient use of resources. It is one of the key technologies underpinning
digital transformation and all of the digital tools we will discuss in this module have some application
in automation.
In this part, we introduce the foundational concepts of automation. We will describe what automation
is and how it works and explore the value and benefits that automation of business processes can provide
to consumers, employees and organisations.
We discuss how to match automation capabilities to business goals using a balanced scorecard, and how
to develop an effective automation strategy and implementation plan.
Data in Automation
Data is used both for analysis and to exchange information between different processes and systems.
Historically, digital systems could only work with highly structured data. Today, technology can work
with data from almost any source and in a wide variety of formats.
It is useful to think of data in terms of its structure.
• Structured data adheres to a pre-defined model, which makes it straightforward to analyse. In a
database, it conforms to a tabular format with a clear relationship between the rows and columns (in the
same manner as a spreadsheet). For example, in a database, a column might be defined ‘SupplierABN’
and the data in that column in each record would be the supplier’s 11-digit Australian Business Number.
• Semi-structured data does not conform to a tabular format like structured data, but does contain tags,
markers or metadata that facilitate processing. For example, an ABN would be enclosed in tags such as
‘<ABN>…</ABN>’. Examples of semi-structured data include JSON and XML formatted documents.
Another example is the inclusion of date and geolocation information in a photo file taken with a
camera phone.
• Unstructured data does not conform to a predefined structure and is not organised in a predictable
manner. Unstructured data makes sense to humans, but is difficult for traditional software algorithms to
process. Common examples of unstructured data include legal documents, audio or video, social media
posts or movie reviews.
Pdf_Folio:158
Automation Technologies
Desktop automation (e.g. Excel macros and auto-emailers) has been in use for many years to handle simple
data and tasks. While automation is useful in managing basic tasks, such as copying and pasting data, it can
also undertake much more complex tasks such as evaluating customer reactions to marketing campaigns
and adjusting outputs accordingly. To achieve these various results, different technologies are required.
These include:
• robotic process automation (RPA)
• artificial intelligence (AI)
• machine learning (ML), which is a particular type of AI.
Note that AI and ML also have numerous other applications and can be used to generate a range of
benefits for the organisation beyond just the efficiencies created from automating business activities. By
understanding the potential applications, benefits, challenges and risks of each technology, it becomes
possible to identify potential solutions to complex business problems and to evaluate how various digital
tools could be chosen, adopted and applied to create value for the organisation.
We will briefly introduce each of these three technologies below. They are discussed in detail in later
parts of the module.
Robotic Process Automation
Robotic process automation (RPA) refers to software robots (bots) that automate simple to medium-
complexity tasks and processes by applying a set of rules known as an algorithm. In essence, RPA bots
interact with other software to perform tasks in the same way a human user would, but much more quickly
and without variations in quality. RPA relies on pre-defined business rules, which means it works well with
structured data, but is unable to handle unstructured data. Examples of RPA software include Blue Prism,
Automation Anywhere and UiPath solutions.
An email bot is an example of an RPA application. An email bot can examine an incoming email’s
sender details, subject line and attachments, and take appropriate action, such as quarantining virus-
infected emails (through the use of simple algorithms that track known threat patterns or search for
obvious suspicious criteria), downloading attachments for an invoice processing application to work on,
or connecting the email to a bill processing system. Another example, new hire onboarding, is outlined in
example 3.1.
EXAMPLE 3.1
An Onboarding Bot
Onboarding refers to the process of bringing a new hire into the organisation. When a new colleague joins
the team, they need a new user account, an email address and phone number, and IT equipment and
systems access specific to their role. In many organisations, this process would involve manual processing
by team members across the organisation, including at least the new hire’s manager, IT staff and HR staff
who are likely to exchange paperwork or emails to organise and coordinate the onboarding process.
As much of the onboarding process involves repetitive and formulaic tasks, it is a good candidate for
RPA. An RPA onboarding bot would begin work when an HR officer enters the new hire’s details into the
organisation’s employee records system. Based on the new hire’s role, the bot could follow a rule-guided
process to, for example, automatically:
• create a user account in the organisation’s IT systems
• assign credentials
• create an email address based on the hire’s name
• allocate a phone number or update the office phone directory based on the hire’s location
• send a request to IT to install a desktop computer or provide a laptop with the relevant software installed
• send a request to the office services team to provide an access card to the building
• send a welcome pack with relevant information and documents to the hire’s new email address.
Pdf_Folio:159
While RPA bots largely replicate what a human software user would do, they can perform the tasks far
more quickly, 24 hours a day, 365 days a year without making errors. This can deliver substantial benefits
to the organisation. For example, Australian Unity, a mutual health and wealth company with more than
one million customers, introduced an RPA solution to extract and review data from government forms.
The project eliminated 22 493 hours of manual work, freeing up skilled labour to focus on higher value
tasks (UiPath n.d.).
.......................................................................................................................................................................................
CONSIDER THIS
What aspects of your work could be handed off to RPA bots? How would you add value to the organisation with the
time this would free up in your day?
Artificial Intelligence
Artificial intelligence (AI) describes a machine that uses algorithms to simulate human intelligence. It is
more sophisticated than RPA in that it is not restricted to applying predefined rules to predefined inputs.
AI is an umbrella term — it includes any technology intended to model human intelligence, for example:
• ML — a type of AI that can train and improve algorithms based upon historical experience
• deep learning — a type of ML that mimics the neural networks in human brains to work with complex
interactions and relationships between data.
The relationship between these three technologies is illustrated in figure 3.1.
Artificial
intelligence
Machine
learning
Deep
learning
AI also includes a range of other specific technologies. For example, natural language processing
(NLP) is a type of AI that applies computational linguistic techniques to text to enable a machine to
communicate with a human through natural language.
AI is valuable for:
• adding intelligence to process automation, thus creating intelligent automation
• processing and analysing unstructured data to derive insights
• enabling interaction between machines and humans, particularly customers and employees.
Apple’s Siri, Google Assistant, Amazon’s Alexa and Microsoft’s Cortana are examples of consumer-
focused AI. They can understand human verbal requests and respond. They can control a smart device,
undertake helpful activities such as listing or dictating messages, searching the web, and making useful
suggestions. An application in identify verification is described in example 3.2.
Pdf_Folio:160
Passport Validation
Tens of thousands of people enter Australia from abroad every day. Australia invests significant resources
to prevent criminal entry into the country and ensure national security. The Australian Border Force uses
AI technology to calculate the risk each traveller presents in terms of criminality and national security. The
traveller’s biometric characteristics such as facial features, iris and fingerprints are collected unobtrusively
by AI technology before departure and on arrival into Australia through an airport. In near-real time, this
information is matched against existing data to confirm the person’s identity and visa or residency status.
Human personnel are alerted when a potential issue is identified.
.......................................................................................................................................................................................
CONSIDER THIS
What aspects of your work could benefit from AI technology? Try to think of at least one example for each of intelligent
automation, analysis of data, and machine–human interaction.
Machine Learning
Machine learning (ML) is a specific type of AI. In ML, the machine is able to learn from data to develop
its own algorithms and thus make predictions. The volume of data ML can analyse and the speed with
which it can explore complex relationships between data enable it to far exceed humans’ ability to analyse
data. There are many different methods within ML. We will explore some of these later in the module.
An example application of ML is in fraud detection when dealing with credit card and online banking
transactions. In this application, the ML platform processes vast amounts of financial data generated
by the usage pattern of each account holder in order to develop a fraud detection model. Each future
transaction can then be examined using this model to distinguish between legitimate behaviour and
fraudulent behaviour, and thus generate an alert in relation to fraudulent transactions or even block the
transaction. This application thus combines elements of analytics and automation. An ML application that
combines analytics, automation and machine–human interaction is described in example 3.3.
EXAMPLE 3.3
Sales Optimisation
Sales and marketing campaigns have traditionally viewed the target market as a whole or divided into a
few target segments. The business treats each customer within a particular segment the same way.
Some businesses are seeking a competitive edge by personalising sales and marketing to individual
customers. This is made possible by ML. ML can analyse customer data to predict their interests and
needs. This provides the organisation with information to better understand their customers. Moreover,
ML can personalise marketing and sales information for each customer, thus automating both the analysis
and the action taken. ML can use the same data to analyse customer feedback, forecast sales and
predict customer churn.
ML-assisted invoice management can handle accounts payable/receivable processing for a large number
of clients. These solutions can alert sales teams in advance about specific deals, or about customers that
are at risk of defaulting. ML can also be used to help companies detect vulnerabilities to cyber attacks or
to data privacy breaches. ML-based email scanners can be implemented to reduce spam phishing emails
reaching human employees, thus, reducing the risk of these emails being inadvertently actioned.
.......................................................................................................................................................................................
CONSIDER THIS
Suggest at least two ways your business or a client’s business could use ML to analyse data to discover insights and
automatically act on those insights.
It should be evident from the examples above that AI and ML are not merely automation technologies;
rather they support a wider range of digital tools that have broader applications across analytics, decision-
making and customer experience. We will look at some of these applications in more detail in later sections
in this module.
Pdf_Folio:161
EXAMPLE 3.4
Evolving Automation
An organisation operating in the Australian property industry explored the potential to automate aspects
of its business. This led to the adoption of an RPA solution from service provider UiPath to automate tasks
within its accounts payable processes.
One formerly manual task being handled by the RPA bot was matching invoices with the associated
purchase order and receiving document to ensure the invoice details were genuine before authorising and
making payment. If the bot identified missing information or a discrepancy between the three documents,
it held the invoice back from the payment process and generated an alert for a team member to investigate.
Automation of this matching process led to significant efficiencies. However, the RPA bot required
structured invoice data, so members of the accounts payable team were spending significant amounts
of time examining invoices, identifying the relevant information and manually keying the data into the
organisation’s enterprise resource planning system (ERP). Besides the time involved, the keying of data
introduced a new error point.
Looking to eliminate the risk of errors and free team members from laborious data entry, the organisation
sought to improve the capabilities of its automation solution. The enhanced solution would need to be
able to identify incoming invoices, automatically extract the relevant information and enter it as structured
data into the enterprise resource planning system.
The solution the organisation adopted involved adding intelligent content management technology
into the accounts payable process. A document capture and optical character recognition solution from
ABBYY was implemented to automatically pick up an invoice as comes into the organisation, regardless
of how it arrives or what format it is in, extract the invoice data and deliver that data to the RPA bot
(Rapidmation 2021).
This intelligent process automation solution eliminated manual data entry. The enhanced solution was
more efficient, reduced errors, reduced fraudulent invoices and required team members only to investigate
alerts generated by the system and provide final approval for payment of certain invoices, enabling them
to focus on higher value work.
.......................................................................................................................................................................................
CONSIDER THIS
Does manual data entry consume resources in your business or a client’s business? Could a solution like the one
described in example 3.4 add value to the organisation and, if so, how?
QUESTION 3.1
Visual Tasks
Unstructured data contains a wealth of information. Historically, though, it has been difficult or impossible
for computers to work with unstructured data. AI, including ML and deep learning, can interpret
unstructured data and create useful output as a result. Below, we will discuss tasks that AI can perform on
visual data.
Facial Recognition
Many technologies now recognise facial features (human or otherwise) in photos, pictures and videos.
They then interpret the characteristics to provide useful information. For example, this functionality can
be used by:
• social media platforms to identify people so they can be tagged in photos
• retailers to determine how happy a customer is during, or following, an interaction in store
• security services to identify particular persons of interest.
Item Recognition
‘Smart’ cameras can identify objects. For example, in the retail environment smart cameras can identify
stock items, visual merchandising assets and so on. This enables use cases such as instant inventory
checks to ensure an appropriate level of stock is on hand in any given area, and visual merchandising
checks to ensure optimal use of shelves and displays. Convenience store chain Amazon Go uses item
recognition technology to register when items are being taken off (or returned to) shelves, and to track
them in ‘virtual trolleys’ (Cheng 2019). Customers can simply take what they like, pay automatically
from their Amazon accounts and walk out of the store. The conventional check-out process is redundant,
improving the customer experience and saving the store costs in labour and stock losses (theft).
Traffic Flow
Some software platforms monitor and analyse the flow of traffic, whether vehicular traffic on roads
or shopper traffic in a mall or store. This enables designers, engineers and store managers to optimise
flow. For example, the insights gained can be used by retail managers to arrange displays and stock to
facilitate better customer search patterns. Shopping centres can provide facilities such as extra seating in
heavily trafficked areas, and optimise rental income by charging more for store locations that have more
customers walking past their doors.
Pdf_Folio:163
Linguistic Tasks
Linguistic tasks involve language-related functions (e.g. interpreting text or understanding the intent in
voice-based searches).
Content Categorisation
Content categorisation is the automatic understanding and interpretation of large volumes of text to identify
themes, duplication or repetition, important content and so on. It can be used to analyse customer reviews
(e.g. positive, negative or neutral), gain new ideas for promotional campaigns, or identify new business
opportunities. It can also consolidate key reports into short, more usable pieces for reporting or planning,
or to complete competitor research or general market research.
Contextual Extraction
Context data are relevant facts about an environment. For a business organisation, this can include
data from customer interactions, social media, weather, news, data from Internet of Things devices and
location data.
Contextual extraction is the mining and processing of relevant information, in a structured manner, from
a broad range of sources. This can play a similar role to the content categorisation function, but there is a
greater sense of focus and relevance in what it picks up and highlights.
For example, Apple’s Siri and Google Assistant provide suggestions based on previous user behaviour,
while smartphones provide inclement weather warnings based on the user’s location. For a business,
seasonal weather patterns could affect sales in certain places, particularly in areas subject to extreme
weather, and this information may be crucial for planning. A company may obtain real-time contextual
data to learn about customers in these areas who may need different products or services at these times,
and the business can respond accordingly to provide better experiences.
Document Summarisation
Document summarisation involves automatically providing summaries of large bodies of text to highlight
the most important points. It can be used for a variety of research functions as well as performance analysis
and reporting. Apple’s Siri is able to fulfil this function.
Speech-to-Text and Text-to-Speech Conversion
Speech-to-text and text-to-speech conversion enables businesses to offer their customers greater con-
venience in performing certain tasks and/or enjoying certain product features. This technology enables
chatbots and voice-based assistants like Alexa and Siri.
Chatbots and Voice-Based Assistants
Chatbots and voice-based assistants enable humans to interact with machines in more human-like ways. A
chatbot is a text-based version of this functionality. Chatbots are often used to improve customer service
and experience across most components of the buying cycle. In the United States, Taco Bell’s chatbot
takes customer orders (Heilpern 2016), while Virgin America’s chatbot enables customers to search and
book flights. Many organisations now support customers on their websites to more easily and effectively
complete tasks with the help of a chatbot, rather than needing to call for support.
Pdf_Folio:164
Prediction Tasks
The visual and linguistic tasks described above involve analysing data and automatically acting on the
basis of that analysis. Prediction tasks go a step further and provide recommendations or automatically
take action based on what the AI predicts will happen in the future. As with the other task categories, AI-
powered prediction enables businesses to be more effective and efficient in their regular activities, while
also generating a greater customer experience.
Recommendation Engines
AI, particularly deep learning, enables organisations to predict what their customers may benefit from and
be looking for. This empowers the organisation to suggest a product, service or experience to a customer
before they ask for it. It may be the next song to listen to, movie to watch, item to purchase, person to
connect with, or research term to consider. For example, Netflix, Spotify, LinkedIn and Google all feature
a recommendation engine.
This functionality allows businesses to offer greater customer value and greater customer experience,
boosting sales and customer satisfaction.
Behaviour Prediction
AI can examine customer and market data to predict the products that customers are likely to purchase
based on other purchases they’ve made and/or actions they’ve taken. Marketers can then design and
optimise campaigns in real time to foster the best possible results. On the flip side, businesses can also
identify ‘at risk’ customers who are unlikely to make a purchase or are unlikely to be a return customer —
and invest in new tactics to address the situation.
In terms of security-related functions, certain technology use cases identify fraud by comparing new
data and behaviour against data, and behaviour from previous instances.
Intelligent Predictive Analytics
Predictive analytics uses historical data to predict potential future scenarios. AI is able to greatly enhance
predictive analytics capabilities by dealing with greater volumes and variety of data.
In healthcare, this functionality can be used to identify patients’ potential health risks based on different
types of diagnostic test results. In retail, it can help to optimise promotional campaign design and customer
Pdf_Folio:165
QUESTION 3.2
Differentiate visual tasks, linguistic tasks and prediction tasks, providing an example of each.
QUESTION 3.3
Explain how the balanced scorecard can be used to identify and assess the benefits of automation.
Pdf_Folio:166
Flexibility
Benefits of leveraging automation
• Ensure processing is not restricted to normal hours
of operation.
• Have the capacity to be switched off or redistributed.
Controls
Benefits of leveraging automation
• Improve audit and risk management.
• Provide improved control of the risk framework.
Data
Sufficient quality of data is a key component of a successful automation strategy. The type and quality
of data the organisation currently collects or can access needs to be assessed to determine whether it will
effectively support the intended automation applications. If the organisation does not currently have the
appropriate data, it needs to find a way to collect or organise access to it as a prerequisite for continuing
with an automation strategy. It may need to adopt further processes to prepare the data for use. In
addition, automation delivers the most benefits when systems are integrated. Work may need to be done
to ensure data is consistent and shared appropriately across the organisation. These issues should inform
development of a data strategy or review and revision of the existing data strategy.
Infrastructure
Automation technologies and the data they rely on require appropriate computing power for storage,
access and processing. The initial and ongoing costs of this infrastructure need to be considered when
developing the automation strategy. An organisation should consider the data storage and processing
solutions available to determine whether the best solution is on- or off-site (the latter may be a cloud-
based solution) or a combination of both. Relevant issues to consider include costs, the availability of
expertise, how cybersecurity risks will be managed and how robust data governance will be achieved.
Organisation
The benefits of automation will not be realised unless the organisation is adequately prepared. Automation
provides the highest ROI and the most benefits for all stakeholders when it is coordinated across multiple
areas of the organisation. Teams need to be willing and able to interact with each other.
Existing processes need to be sufficiently streamlined and robust for the application of automation to
them. Automation applied to an inefficient process will not correct the inefficiencies. Processes should be
optimised before being automated.
Skills
As in all business endeavours, an organisation’s people will have a substantial effect on the success of
automation implementation. Organisations can upskill existing staff members who already have knowledge
of the business’s unique needs, bring in new talent or outsource automation services to specialist external
providers. Perhaps a combination of approaches may be the best fit.
In addition, most automation approaches still ultimately provide outputs that humans use to make
decisions or incorporate into more complex work. Working effectively with technology solutions requires
a cultural shift and a digital mindset. The organisation should invest in preparing and supporting team
members to work in a more automated environment. This will benefit the organisation and its employees.
Importantly, automation often replaces work currently done by team members. Team members should be
given the opportunity to develop their skills in higher value work wherever possible.
.......................................................................................................................................................................................
CONSIDER THIS
Two elements of an effective automation strategy are organisational preparedness, including the ability to cooperate
and coordinate across different teams, and skills, including the ability to help plan and implement automation and then
work in a or automated environment. How would you assess your organisation’s or a client organisation’s readiness
for automated solutions?
SUMMARY
Automation refers to the use of computer and software resources to perform tasks and processes that would
otherwise be done by human resources. Automation is a key technological application enabling digital
transformation. Automation technologies can help an organisation to reduce its operating costs, enhance
productivity and efficiency, achieve high levels of scalability and process reliability, Increase customer
engagement, and finally improve its performance. The balanced scorecard can be used to evaluate and
communicate the benefits of potential automation initiative to a variety of stakeholders.
Beyond well-established desktop automation technologies, robotic process automation (RPA) and
artificial intelligence (AI) — including machine learning (ML) and other types of AI technologies —
have created new opportunities to automate tasks that conventionally required human labour. RPA is the
use of software robots to automate tasks and processes that can be described by pre-defined business
rules. AI algorithms can similar aspects of human intelligence and thus automate activities that require
Pdf_Folio:169
KEY POINTS
3.1 Select appropriate digital tools, using accounting and finance knowledge, that provide
innovative solutions to complex business problems.
• RPA, AI and ML can be used to automate business processes, including those involving intelligent
analysis, decision making and engagement with customers and employees.
• Different automation technologies can be combined to create intelligent automation solutions
across workflows.
• The balanced scorecard provides a framework to evaluate the potential and performance of
automation solutions against a range of financial and non-financial objectives.
3.2 Evaluate the impact of innovative technologies on the work of accounting and finance
professionals to communicate the benefits and risks to key stakeholders in the organisation.
• Automation can reduce the need to allocate human resources to low-value repetitive work, thus
creating cost and time efficiencies, reducing errors, enabling professionals to focus on higher value
work and improving employee morale.
• A balanced scorecard can be used to identify and focus on a mix of financial and non-financial
outcomes for stakeholders. This in turn facilitates communication of these benefits to stakeholders.
3.3 Appraise the suitability of technology in creating efficiencies and solving complex
business problems.
• Automation technologies can deliver efficient processes, facilitate new ways of doing business,
result in more satisfied and engaged customers and improve employee morale and engagement.
• Intelligent automation can create efficiencies in business processes, undertake visual and linguistic
tasks, and perform predictive analysis.
Pdf_Folio:170
Pdf_Folio:171
QUESTION 3.4
What are the main factors to consider when choosing between unattended RPA and attended
RPA?
RPA BPM
Deals with smaller, repetitive tasks Covers the entire enterprise business workflow
Completed by bots that comprise a certain part of a A complete enterprise application system
business process
Level 3
Pattern recognition Intelligent
Pattern-based
Work with unstructured data automation
automation
Self-learning
Limited intelligent decision making
Level 1
Traditional Simple macro application (e.g. Excel) Macro app
automation
Predefined connectors to other applications
Traditional automation — using macros or integrating systems with application programming interface
(API) protocols — operates at Level 1. RPA offers the most benefit in processes that follow structured
rules to complete simple tasks using structured data (Level 2 in figure 3.3). The following section
compares traditional automation and RPA automation.
.......................................................................................................................................................................................
CONSIDER THIS
Which level or levels of automation in figure 3.3 are most relevant to your work? What role does RPA have to play?
Pdf_Folio:173
Does not require any modification in the existing Requires certain customisations in the existing IT
systems or IT infrastructure. infrastructure.
Mimics human actions to automate repetitive, Does not mimic human actions.
rule-based tasks.
Users can preconfigure a large number of bots to Users require different domain-specific programming
perform the allotted tasks without the requirement techniques on physical machines to perform
of physical machines. parallel execution.
Orchestrates multiple applications (e.g. email, Interoperability issues are a challenge due to the
enterprise resource planning and customer limitations of customised APIs.
relationship management) to synchronise data flow
and create automated results.
Flexible to update any business flow. Difficult to maintain or upgrade traditional automation
technology due to hardware and software incompatibility.
RPA can thus be more easily implemented than traditional automation approaches that require cus-
tomised code or work by integrating systems using API. RPA can also be contrasted with spreadsheet
macros that automate the work of completing complicated mathematical calculations, or invoicing
programs that automate the generation of sales invoices based on goods or services provided. The major
difference between these types of automation and RPA is the level of human input required in the process.
Once RPA is initiated, it performs a series of tasks, whereas these other approaches to automation perform
a single task at a time. Figure 3.4 compares how a macro would be used and how an RPA bot would be used
to complete a simple calculation. Note that in the RPA process the human inputs required by the macro
are replicated by the software.
Pdf_Folio:174
EXAMPLE 3.5
An RPA Solution
Xcello Credit has a backlog of customer requests that now average a turnaround time of 3–5 days. A
sales manager and sales assistant at Xcello have identified the possibility of using RPA to automate
some processes to improve turnaround times. They have decided to investigate and perhaps pilot an
RPA system to automate the creation of new line of credit accounts for existing customers.
The sales manager has documented the process, including the pain points, for the vendor as follows.
The Customer Request
A call centre representative has to take a phone call from an existing customer who requests a new line
of credit account. Because the caller is an existing customer, the call centre representative only needs
to record information about the new request and the details of the customer’s property that will serve as
collateral. All the other information will come from Xcello’s customer database. The next step is for a sales
assistant to complete the request.
The Backlog
As the next stage of the process can only be handled manually by the sales assistant, the customer
request sits in a backlog, in many cases for up to 5 days, until the sales assistant has time to process it.
This is inefficient and a suboptimal experience for customers.
The Sales Assistant’s Process
When processing the backlogged requests, the sales assistant uses information from two systems:
• the business process management system, where the details were originally entered
• the credit line confirmation decision-making system.
Using the customer number, the sales assistant copies information from the customer database into
yet another system: the line of credit account system.
The sales assistant then takes the information from the credit line confirmation decision-making
system and uses a calculator on a government website to determine the tax valuation and assessment
of the property used as collateral. Accessing yet another system takes more time, and introduces the
potential for inadvertent errors, even when the sales assistant is working as effectively as possible. About
10 per cent of credit account requests would contain an error at some point in the process.
Finally, the sales assistant copies the information in the line of credit account system and presses the
button to create an account. The system is now ready to generate the documents the customer will need
to sign. It only takes the sales assistant about 20 minutes to complete the process for each customer, but
Pdf_Folio:175
Pdf_Folio:176
$60 000
$49 000
$50 000
$44 950
$30 000
$20 000
$10 150 $7250
$10 000
$5400
$–
US UK AUS Philippines India RPA
Source: Kinetic Consulting Services 2016.
RPA clearly has a strong potential financial benefit. In section 3.2, we introduced a balanced scorecard
as a framework to understand, prioritise, evaluate, communicate and assess the performance against a
range of other benefits for an organisation. Using the scorecard to identify these benefits is a useful step
in developing a comprehensive business case to adopt RPA.
Pdf_Folio:177
Productivity
• Increased productivity through reduced errors (avoiding
employee fatigue and double handling)
• Increased productivity through greatly increased speed
and volume of work performed
Flexibility
Scalability
• Cheap and easy to implement for well-defined processes
• Requires very little computing power to run, so they can
be scaled to higher volume or more widespread processes
with little business investment or risk
EXAMPLE 3.6
Coca-Cola
Coca-Cola is one of the world’s largest beverage companies with complex multinational operations.
It recently consolidated its HR and finance shared services organisations, along with its customer
contact centre.
Pdf_Folio:178
.......................................................................................................................................................................................
CONSIDER THIS
Which segments of the balanced scorecard were addressed by Coca-Cola’s RPA solution and how?
Portal Queries
Organisations often work with suppliers via portals that do not have well-defined integration protocols (or
APIs). RPA can streamline the process of working with supplier portals by connecting to all portals and
automatically gathering or posting relevant information. This can occur either at pre-set times or as part
of a defined workflow.
Data Consistency
Customers’ details can frequently change as they move house, change their name, or even make significant
changes to their financial situation (e.g. taking out a mortgage). RPA can assist by regularly scanning
customer information for accuracy using a designated reference point, such as a bank statement or tax file
number, and update customer records if required.
Regulatory Compliance
Regulatory technology, or RegTech, is a large and growing area that has emerged largely from regulatory
requirements imposed in the finance field. Initially introduced in many organisations to improve internal
processes and reduce compliance costs, RegTech became important as a way to monitor risk in order to
comply with increasing regulatory requirements following the global financial crisis.
While AI and ML have a role to play in RegTech, RPA is the most commonly used type of technology
currently underpinning it. RPA offers benefits in the RegTech area by:
• decreasing errors in data and reporting
• increasing staff efficiency, allowing greater time to investigate and resolve instances of non-compliance
• reducing risk by improving compliance with anti-money laundering and know-your-customer (KYC)
regulations.
Pdf_Folio:180
EXAMPLE 3.7
Pdf_Folio:181
.......................................................................................................................................................................................
CONSIDER THIS
To configure an RPA bot requires every step of a process to be understood and documented. The aspects of the
process that can be automated can then be identified. Choose a familiar manual process from your own work and
list each step. Which steps could be automated and why? Which could not and why?
RPA IN BUSINESS
RPA can be used to target improvements in key processes in a wide range of business functions
beyond finance, in processes that span different parts of the organisation and across different industries.
An accounting and finance professional should be aware of broader applications that are relevant to
business overall.
Pdf_Folio:182
EXAMPLE 3.8
.......................................................................................................................................................................................
CONSIDER THIS
Given the pervasive use of technology in business today, why do you think manual processes such as the one
described at the start of example 3.8 still persist in many organisations?
Business Processes
RPA can exchange data across multiple systems and thus automate processes that involve different parts
of the organisation.
Sales Order Processing
RPA can complete many actions across multiple business portals and systems; some of the most common
applications include the following.
• Incoming customer emails and attachments processing. The company can apply RPA to screen all
incoming emails and classify them. Sales-related emails can be assigned different priority levels to
ensure a response within an appropriate timeframe by the appropriate personnel. Automating email order
processing can increase the speed and accuracy in handling incoming orders as well as empowering the
sales team with a tool that helps them deliver a better customer experience.
• Customer orders processing. With all the required information available, RPA bots can handle customer
orders automatically without human intervention from the early stage of receiving the orders, validating
the correctness of the order line items, updating the order processing system, checking for estimated
time of arrival (ETA), and continuing with the order fulfilment stage. RPA can eliminate manual data
entry effort and reduce human errors and accounting costs. It can also allow for better pricing validation
and sales data control.
• Order fulfilment. Online storefronts can use RPA bots to fetch sales order details from different provider
databases, warehouse stock level can be monitored and shipments can be tracked in real time from
mobile devices. RPA enables streamlined and parallel order processing tasks to achieve improved
accuracy, reduced labour costs, faster delivery of shipments and reduced delivery overheads.
• Customer account reconciliation. Customers commonly request statement reconciliation either by
phone or in writing. This can be handled by the RPA bot to process the request without delay. RPA
can check for duplicate payments to reduce the time spent manually verifying bank statements with
customers and improve overall efficiency by helping customer service personnel, who will then have
more time to handle other more critical requests.
IT Processing
Many IT processes can be automated with RPA, including the following.
• Application system user access control. Modern-day role-based access control mechanisms require IT
personnel to propagate user access rights from one application system to another, based on the role
or roles of the user. RPA bots can perform this task without human intervention. They can assess the
employee’s credentials and grant access and permission to the relevant system resources.
• Retrieving and distributing data to and from multiple data sources. Multinational organisations run their
business with back-end databases located in different locations. RPA bots can process the requests in
batches asynchronously or in real time, depending on the priority of those requests, to ensure better
performance to the whole of the enterprise’s application system.
Pdf_Folio:184
Industries
While some core business functions operate similarly across many different businesses, different industries
have their own particular characteristics and thus opportunities for RPA to add value.
Healthcare
There are many opportunities to leverage the benefits of RPA in the healthcare sector, including deploying
RPA to:
• manage the online and offline complexities of appointment scheduling to allow for patients to make
online bookings
• process finance-related activities such as recording costs of treatments administered, generating bills,
accepting online payments and synchronising data across multiple systems
• handle health insurance coverage and claims and streamline settlement processes.
Insurance
As another industry driven by large amounts of data — and the need for extremely high levels of data
accuracy — insurance companies can use RPA to:
• verify customer information using a variety of systems and use this as the basis of robust quoting and
underwriting processes
• make basic changes to policies and customer details, within given parameters
• remind customers when renewals are due and process these renewals as required
• settle claims in low value, standard cases.
Telecommunications
With the vast majority of people in industrialised countries owning at least one mobile device — if not
more — the data created by the telecommunications industry is extensive. RPA can help by:
• collecting and consolidating customer data from different phone companies and systems
• transferring customer data between different provider and carrier systems, portals and applications
• extracting data on competitors’ plans and pricing, to ensure the company remains competitive.
Government
Governments can also benefit from the use of RPA to cut costs and enhance process efficiencies, including
when it is used to:
• complete verification processes on individuals, such as when they register a property purchase or other
legal transaction
• populate automatic forms and notifications
• integrate legacy systems with newer systems.
Not-for-Profits
RPA can be a valuable technology for not-for-profit organisations. The simplicity and low cost of RPA
implementation makes it an appropriate first step into automation for not-for-profit organisations, which
face many of the same challenges as corporations — in particular, the growing requirement for staff to
complete menial time-consuming processes to ensure compliance and data quality, and generally support
Pdf_Folio:185
DATA SECURITY
As RPA is a software-driven innovation, it creates vulnerabilities in the areas of cybersecurity and data
privacy. By their very nature, RPA software bots are designed to access data across various departments
in an organisation — and even from partner systems outside the organisation.
Although not common, experts have suggested that hackers may be able to modify any session that a
software bot is performing to either affect the operation of the RPA or extract sensitive business data.
Mitigation Strategy
The organisation’s overall digital strategy should include cybersecurity and data management strategies
to ensure data is handled appropriately and secured from theft and malicious or inadvertent misuse. It is
crucial to ensure the organisation’s technology security systems are robust prior to implementing RPA,
investigate whether the RPA solution will introduce any new vulnerabilities, and determine if additional
security is necessary for each RPA process implemented.
TECHNICAL ISSUES
All software has the potential to contain errors in programming or design which can impact on its operation.
When an RPA bot is configured by the users themselves, this risk may be greater than for professionally
configured systems.
Flaws in the implementation of RPA can then compound across huge amounts of data. For example, if a
data field from one portal is copied and then pasted in the incorrect place in the target location, all records
will be erroneous.
These kinds of technical issues can also damage the customer experience if, for example, they result in
incorrect invoices being generated or marketing campaigns being sent to the wrong people.
Mitigation Strategy
Appropriate testing must be conducted, with the nature and extent of testing to be based on the potential
impact of the RPA making an error. More critical processes naturally require a higher level of testing.
It is also advisable to scale the implementation of the RPA into production. A small number of
transactions can be processed and the results closely monitored to determine whether the bot is operating
as expected. Once reliable results are achieved, the volume of processing can be progressively increased
over subsequent days, while still monitoring the outputs. to ensure the RPA is operating as intended.
Pdf_Folio:186
SYSTEM CHANGE
RPA processes are only as stable as the underlying information or systems they access. That is, if a
particular RPA bot sources data from one system to use in a task in another system, this process will
only continue to work for as long as both base systems remain the same. If one of the source data fields
changes, the information the bot accesses is likely to be compromised and therefore unreliable.
Mitigation Strategy
The organisation should put in place mechanisms to ensure any systems changes are made with considera-
tion for the impact they will have on other integrated systems and planned changes should be communicated
to relevant parties throughout the organisation.
The user should regularly check that the RPA bots are functioning as intended and that the data they are
using remains accurate. If any changes have occurred, the bot’s programming should be adjusted to render
it functional again.
PROCESS SELECTION
Although it will not impact on the functional integrity of an RPA bot, the selection of the most suitable
process to automate will make an enormous difference to the outcomes achieved. An RPA implementation
should have a defined set of expected business outcomes and established metrics for determining
performance. A successful implementation should be quickly evident, but applying RPA to unsuitable
processes may not only waste time and money but may in fact magnify any underlying inefficiencies in
the overall workflow — and require substantial time to rectify.
Mitigation Strategy
Process selection should be based on a careful analysis of different processes to identify which are the best
candidates for automation. In addition, the processes and overall workflows should be evaluated to identify
whether they are optimally designed. Business processes should be improved where possible, including
through better coordination between different parts of the organisation, before they are automated.
EMPLOYEE MORALE
Introducing new software or technology, such as RPA, has the potential to positively transform a workplace
by delivering wide-ranging and substantial benefits to numerous stakeholders. However, it is important to
recognise that the introduction of RPA will affect various employees differently.
• Some employees would find their jobs more fulfilling once RPA takes over mundane tasks, but others
may not.
• Employees who are not technologically proficient may be increasingly concerned about training and
upgrading their skills for adapting to RPA and other technologies.
Failing to communicate the inevitable changes to various workflows and processes can result in
confusion or disengagement among employees.
Mitigation Strategy
RPA will inevitably impact on team members. Any business process and workflow changes as a result of
RPA should be clearly communicated to all affected employees, along with the rationale and benefits for
the changes.
.......................................................................................................................................................................................
CONSIDER THIS
How would you communicate plans for an RPA initiative to your colleagues? What would you expect to be
communicated to you if the initiative was being introduced to your work area?
Pdf_Folio:187
Pdf_Folio:188
Create a checklist that can be used to assess the suitability of a process for RPA.
n
atio
orm
tra nsf
ig ital Leadership
s sd
ne
B usi
Agility
Leadership
RPA
Agility Transition capability
stage
Quick wins Effectiveness
Leadership Transition Maturity
stage Efficiency
POC
Evolution
Validation
RPA maturity
Source: CPA Australia 2021.
Pdf_Folio:190
EXAMPLE 3.9
We don’t want our human resources performing mundane tasks. It’s so much more powerful for our
people to spend time with customers, understanding what they want and need, listening to them and
solving problems for them.
Pdf_Folio:191
SUMMARY
RPA is an automation technology that uses software robots to perform low-complexity, high-volume tasks.
RPA bots follow simple algorithms to interact with data, software and systems in the same way a human
user would. RPA is able to process data far more quickly than humans with no variability in quality (and
thus with no errors), enabling humans to focus on more complex and value-adding work. RPA may work
autonomously or as an assistant to a team member, depending on the specific processes involved.
RPA has a wide range of applications in various business functions and across industries. To properly
evaluate its potential, the associated risks and challenges need to be assessed. These are related to data
security, technical issues, system changes, process selection and employee morale.
While RPA is a relatively simple technology, many initial attempts to implement it fail. An RPA strategy
should focus on suitable processes, optimise processes before automating them, include initiatives to
achieve stakeholder buy-in, introduce RPA technology gradually, and scale up once the application of
RPA is proven and accepted.
In practice, an RPA strategy is likely to be informed and shaped by other aspects of the organisation’s
digital strategy, including potentially the use of AI and ML technologies. These technologies are discussed
in the following parts of the module.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
3.1 Select appropriate digital tools, using accounting and finance knowledge, that provide
innovative solutions to complex business problems.
• RPA automates low-complexity, high-volume tasks to create time and cost efficiencies.
• RPA is suitable for the automation of processes that can be described by a set of simple rules; it
is not suitable for processes that involve many decision points.
• RPA can work across multiple systems, but requires structured data.
• Applications of RPA include compiling financial reports, reconciliations, journal entries, cleaning
data, regulatory and compliance processes, basic information analysis, document matching and
many more.
3.2 Evaluate the impact of innovative technologies on the work of accounting and finance
professionals to communicate the benefits and risks to key stakeholders in the organisation.
• A balanced scorecard can be used to identify and focus on a mix of financial and non-financial
outcomes for stakeholders. This in turn facilitates communication of these benefits to stakeholders.
• The use of RPA to automate processes in accounting and finance enables professionals to focus
on more complex, value-adding tasks. This improves employee morale and creates value for
the organisation.
• RPA, once implemented, performs tasks with no variation in quality and thus does not make errors.
• Like any business innovation, a proposal to introduce RPA should be evaluated with regard to its
benefits, risks and challenges. Generally, RPA is easy to implement and low risk.
3.3 Appraise the suitability of technology in creating efficiencies and solving complex
business problems.
• RPA is a useful technology to automate processes to create time and cost efficiencies for an
organisation. It can be combined with other technologies to enhance its capabilities. It enables
team members to focus on more complex issues facing the business.
Pdf_Folio:192
Pdf_Folio:193
A BRIEF HISTORY OF AI
Before we explore current AI capabilities and the benefits they can offer, we will briefly review key points
in the historical development of AI.
Pdf_Folio:194
AI ENABLERS
It appears in recent years as though a ‘tipping point’ has been reached in the application of AI to business
contexts. A number of key factors have fallen into place, enabling AI to deliver demonstrable improvements
when used appropriately.
Data
The Internet of Things (IoT), which refers to any and all devices that have internet connectivity, generates
volumes and varieties of data (known as big data) that far exceed the data collections of the past. As
of 2020, it is believed there are more than 20 billion IoT devices, with this number expected to grow
rapidly. Capturing this data provides organisations with the potential to develop deep and strategically
important insights.
Processing Power
Computer processing power grows exponentially, with hardware, software and internet connec-
tivity now incomparably faster than decades ago. Alongside this, the interest in AI, and specific
advanced types of AI including ML and deep learning, has continued to grow. Recent breakthroughs
in deep learning mean computers can now uncover insights which had previously been thought impossible.
Readiness to Adopt
As with all technologies, if people are not prepared to adopt and use it, then its relevance and value
is hampered.
As a result of living in an increasingly digital and technology-enabled world, customers are far more
comfortable trying new technologies, and looking for ways to maximise the value on offer.
With greater readiness and interest from people (and thus business customers and markets), it is easier
to gain a return on investment (and therefore easier to secure an investment to try something new).
Use Cases
Thanks to a rising tide of technological development, there is now far more relevance and value to be
uncovered through the use of AI. Data has been proven time and again to be invaluable in business
decision-making, so using AI-powered technologies to more effectively evaluate this data makes good
business sense.
As the economy continues moving towards a digital-centric model, having technologies which support
customers in their digital interactions is also appealing to business. For example, an increased volume of
email traffic has created a relevant use case of AI automatically sorting it for us.
This increase in relevance and value in turn drives an increase in willingness to adopt AI technologies,
in a reinforcement loop.
Investment
The previous factors have increased the willingness of investors to back AI. These lead indicators of
success and easier-to-achieve successful use cases provide a clear ROI. Each success story brings a
slightly bigger investment, which brings a slightly bigger return — and the cycle continues.
TYPES OF AI
AI capabilities can be classified into two main categories: weak AI and strong AI.
• Weak AI — also called ‘narrow AI’ or ‘artificial narrow intelligence (ANI)’ — is the simplest form
of AI and refers to any AI instance that can perform a single task effectively. Weak AI aims to mimic
human behaviour in a particular context rather than mimic full human intelligence. Because they are
programmed within a narrow set of constraints, weak AI instances may seem quite intelligent, but their
ability to apply their algorithms outside of their given context is extremely limited. Even with recent
advances in the field and increasing business adoption, all forms of AI in the world at the moment are
considered weak AI.
Pdf_Folio:195
QUESTION 3.6
Differentiate weak from strong AI and discuss which is most likely to be found in Australian
businesses.
3.9 BENEFITS OF AI
Earlier in the module, we described that AI could be used to add intelligence to automation. Integrating
intelligent automation across an enterprise can transform the customer experience, increase productivity
and reduce costs. It enables new business models in which organisations can:
• re-design business processes with intelligent automation at the core
• transform how, where and when the work is done
• minimise human intervention in day-to-day processing, and focus on more critical activities
and decisions.
We have also noted that AI can be used in analytics and in engagement with stakeholders. There is often
no clear delineation between these applications, as AI often works by analysing data and making some
type of automated response, including potentially communicating with employees or customers.
Applications of AI can be:
• Insightful. AI can enable visibility and insight across hundreds of processes, allowing employees to
spend more time collaborating with their colleagues, suppliers or customers.
• Efficient. AI is capable of performing tasks that would once have required intensive labour, or may not
have been possible at all. It enhances efficiency and output while freeing people up to focus on important
tasks that involve critical thinking and emotions.
• Progressive. Automation is an ever-evolving process. Advancements in AI are spawning a new phase
of automation: intelligent automation. This is particularly relevant to workflow, and document ingestion
and generation.
• Personalised. AI is changing the way enterprises operate to optimise processes, personalise customer
and employee experiences, and enable rapid and informed decision-making.
• Evolving. AI continues to evolve and is reaching larger scale use in industry and government. The
advancement in AI technologies and access to broader sets of data in numerous forms (e.g. structured,
unstructured, sentiment) continues to unlock more value for businesses and deliver a great experience
for customers.
• Touchless. AI can perform procedural and manual activities. Delegating these tasks to AI means that
humans no longer have to touch them, unless or until there is a need for a person to be alerted. This frees
up the organisation’s people to:
– focus on where they add the most value to the customer experience
– create new value for the organisation
– enforce the necessary controls and checks to protect the organisation and help maintain quality
products and services in response to targeted alerts from the AI.
The design for a ‘no-touch’ environment is characterised by reduced operational risk, decisions based
on fact or high confidence and a superior experience for customers, colleagues and investors.
AI in Different Industries
Information is a key component of business competitiveness in the twenty-first century, but without the
ability to process and analyse that information, businesses risk falling behind their competitors.
In the previous section we discussed different types of generalised benefits AI can provide businesses.
Below, we will discover how AI can enhance functions across some specific industries.
Finance
AI has countless applications in financial services, including:
• informing customers about their spending behaviour by analysing data from their smartphone check-ins
• providing customers with customised updates using NLP to read text such as news, broker reports and
social media feeds
• making credit decisions through automating the approval process for loan and credit card applications
• forecasting future sales and demand for products and services using historical market data, as well as
external data sources such as social media.
ANZ Bank in Australia moved early in creating a proof of concept that would help predict risk on
home loans (Mason 2017). Another common use case in banking is the performance of KYC processes,
as described in example 3.10.
EXAMPLE 3.10
Improving KYC
Banks and other financial institutions use a KYC process to validate the identify of their customers, both
when they open an account and periodically after that (Thales Group n.d.). KYC is required by regulators
as part of measures to combat money laundering, terrorism financing and other criminal activities. If the
KYC process breaks down, the bank is fined by regulators (globally, banks lose billions of dollars due to
such fines each year) and their reputation is damaged, which in turn means a loss of trust with customers.
AI technologies have potential application to improve efficiency and compliance in KYC processes. The
bank’s business goals mean that a successful AI application would be defined by the ability to:
• reduce the time to check a customer’s identify documents
• ensure regulation compliance
• safeguard against damage to the bank’s reputation.
AI technologies are able to scan identity documents to extract data, thereby improving data entry
efficiency. AI can also access external sources for additional customer identity information, if necessary,
to validate earlier results.
.......................................................................................................................................................................................
CONSIDER THIS
Why does the application described in example 3.10 require an AI solution rather than an RPA solution?
Healthcare
AI diagnostics are widely used in healthcare to:
• recognise the presence of disease in patients
• produce diagnostic interpretations for diseases
• scan digital images to highlight possible diseases
• design treatment plans.
Pdf_Folio:197
AI RISKS
Risks associated with AI technologies can cause consequences at the individual, organisational and
societal levels.
Risks to Society
The risks that AI poses at a societal level can include:
• compromises to national security — through data breaches of a highly sensitive nature or loss of control
over AI-enabled machinery or services
• insecurity and volatility in financial markets — caused by flawed AI algorithms
• manipulation of governmental processes — leading to many outcomes including incorrect
electoral results
• threats to interconnected physical infrastructure such as communications, electricity and water
supplies — due to flawed AI decision-making models.
DATA CHALLENGES
Aside from the risks related to AI, issues around data quality present a significant challenge to successfully
implementing AI. The suggestion that there is not enough high-quality data available for AI to function
properly may seem counterintuitive, given that big data is one of the biggest drivers of the current interest
and success of AI in the business world. However, the issue is not straightforward. There are some key
criteria which need to be met for data to be suitable for use in creating accurate and effective AI models.
Sensitivity
As data becomes more private, and/or there is a greater risk for negative outcomes if the data is used
inappropriately, consumers are becoming increasingly wary as to when, where, how and why they make
such data available. As the instances of data misuse increase, so too does the wariness of consumers.
Availability
Organisations such as Amazon that have collected data for extended periods of time, or even younger
companies that have proactively collected data, have the key ingredient for starting their AI process.
Conversely many organisations may find that, as a result of ignorance, naivety or poor decision-making,
they may have failed to collect the necessary data for AI and data-driven decision-making. Without the
data, they have no AI capability.
Organisations may also find that despite sensible and even extensive data-collection activity, they do not
have the right type(s) of data for their proposed AI use case.
Usability
Many organisations have been collecting potentially very useful data for a long time; however, much of
that data has been recorded in ways that are not useful for AI. For example, governments and councils
have extensive data records about a host of interesting topics relevant to humans, but this information is
often filed in archives that are inaccessible due to bureaucracy or kept in paper form which is impractical
to try to digitise.
Volume
Despite the size and variety of big data, it is not uncommon that the data is insufficient to successfully
complete a particular AI task. According to one Amazon AI team member, machines do not just need
Pdf_Folio:199
Bias
Humans unwittingly exhibit bias in decisions and actions. Therefore, it stands to reason that the data we
collect (and use to train our AI systems) will likewise contain bias in myriad different ways, including along
racial, gender or ethnic lines. For AI to be truly effective, such biases must be identified and eliminated
from data sets and hence AI algorithms — or the resulting recommendations or outputs from AI will be
flawed. Example 3.11 briefly discusses how bias in Amazon’s data led to unethical and poor decisions by
its AI-enabled recruiting solution.
EXAMPLE 3.11
Many other businesses may be making decisions driven by unintentionally biased data, but there are
also ways to counteract this. Some service providers specialise in deploying AI solutions to identify and
counteract hidden bias in companies’ operations.
.......................................................................................................................................................................................
CONSIDER THIS
What type of biases could appear in data that you use to develop business insights? How would you recognise the
presence of the bias?
Appropriate Data
The quality of AI outputs depends on the quality of data inputs. The organisation should ascertain from
the outset what data it needs and how it should source it to support its AI solutions. The types and sources
of data should be directly relevant to the business’s objectives. Often the most relevant data will be internal
data, such as that captured during interactions with customers, but ambitious AI initiatives will likely
need data from multiple sources.
Data should be analysed to detect any hidden bias. One way to achieve this would be to run some of the
data through the AI model and compare it to the results of a person doing the same analysis. Differences
that occur may indicate a bias in either the data itself or the resulting AI model.
Data Preparation
A clear data governance model and the use of high-quality data sources will increase the confidence of team
members in the information and recommendations being provided. Giving team members visibility of the
data lineage (where the data has come from) and the rules that are being applied to make recommendations
or decisions will help build confidence that the AI solution is performing as expected.
Data from multiple sources will often be in a variety of formats. This data will need to be collated and
prepared for use by the AI model. Structured data is more readily usable in AI; if unstructured data is used,
the AI will first need to be trained to read the data before it can use it.
As described above, with the advancement of AI technologies and their sophistication in analysing data
and trends, it is important to regularly review and validate AI models for bias. Otherwise, this could lead
to incorrect recommendations (and hence, inappropriate decisions) if left undetected.
Pdf_Folio:200
Skills
One of the key decisions that affects the success of the AI solution is whether it is designed and deployed
by way of:
• outsourcing to specialist providers
• in-house development
• purchase of an off-the-shelf solution.
The more complex the solution is, the more specialist skills will be required. If these skills do not exist
within the organisation, they could be developed or acquired, but support from specialist services providers
is also a valuable option.
Costs
Costs to be considered when evaluating a potential AI solution include:
• cost of outsourcing specialist skills and services
• costs associated with extending the capabilities of AI post-implementation.
• costs of upgrading the technology
• cost of the ongoing maintenance of the AI models or algorithms
• additional future costs as the solution is scaled up.
To maximise return on investment, AI models and data sets should be leveraged across the organisation
where possible. Some AI solutions may lend themselves to modification or adaptation to solve other
business problems.
Example 3.12 examines Tesla’s development of self-driving vehicles based on cutting-edge AI. Self-
driving vehicles clearly involve many benefits, but also substantial risks.
EXAMPLE 3.12
Driving Innovation
Tesla is developing AI technology to enable vehicles to drive themselves. Since its entry into the automotive
market, Tesla has used AI to collect all possible data analytics from Tesla car owners. This data is sent
to the cloud for analysis with algorithms and software. The collected data has helped Tesla to grow and
improve their self-driving and driver-assist systems.
With the help of AI, and in particular ML, Tesla cars can decide what driving actions are needed in real
time, including identifying hazards such as oncoming traffic, following road signs, finding empty parking
spots and parking automatically.
The AI also gives the cars the ability to form networks with other Tesla vehicles nearby to share local
insights and information.
Tesla uses big data and AI to analyse customer satisfaction. The capture gathers data from its online
customer forum and analyses this to generate insights that inform future products development.
It is evident that driverless vehicles must raise legitimate safety and privacy concerns that will need to
be addressed before mainstream acceptance can be achieved.
• Will data analytics sourced from car owners help or hinder the advancement of autonomous vehicles?
With any technology solution, it is important to consider data privacy issues and take measures to
protect consumers’ personal information (such as medical and financial information) being captured or
stored. In Tesla’s case, this data also holds tremendous value so there it may be a target for cyber-
attack. Researchers at McKinsey and Co estimate that the market for the vehicle-gathered data would
be worth around USD750 billion a year by the end of 2030, providing a potential motivation for Tesla to
sell access to its data.
• Are autonomous cars safe? Like any computer system, AI is not infallible and there has already been
at least one fatality caused by a self-driving car. In 2018, an Uber self-driving vehicle struck and killed
a female pedestrian in Arizona. The Uber vehicle was in autonomous mode, with a human driver at
Pdf_Folio:201
.......................................................................................................................................................................................
CONSIDER THIS
If your organisation or a client adopts an AI solution to analyse financial data to inform decision-making around
investment opportunities, how serious could the consequences of an error be? What mechanisms would be in place
to detect such an error?
QUESTION 3.7
What are some of the likely risks for organisations adopting AI models?
Business strategy
AI strategy
Pdf_Folio:202
EXAMPLE 3.13
.......................................................................................................................................................................................
CONSIDER THIS
Does your organisation have in-house expertise to understand the possibilities of AI? Could the business develop an
effective AI strategy? If not, at what point would you advise the business to seek expert help?
Pdf_Folio:204
A client who wants to integrate AI systems into her business comes to you for advice on how to
create an effective AI strategy. What are the key parts of an effective AI strategy? Why?
SUMMARY
AI refers to technologies that can behave in ways that mimic aspects of human intelligence. AI has three
broad applications:
• automation
• analytics
• employee and customer engagement.
Across these categories, there are numerous business applications. AI can add intelligence to
automation to transform the customer experience, increase productivity and reduce costs. AI can be used
in analytics to generate insights from data that could not be generated using conventional approaches. AI
can engage intelligently engage with stakeholders to respond to requests or proactively provide support or
services. There is often no clear delineation between these applications, as AI often works by analysing
data and making some type of automated response.
AI requires vast volumes of data and substantial processing power. The introduction of AI to an
organisation requires specific skills and acceptance by team members. AI technologies are complicated
and the costs to develop, implement and improve an AI machine over time are significant. For some
organisations, purchasing ready-built solutions or outsourcing a solution can be an effective approach.
An AI strategy will form part of a broader digital strategy. AI solutions must be aligned to and support
the larger business objectives. Clearly defined measures of success must be established to enable proper
development, implementation and evaluation of proposed or actual AI solutions.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
3.1 Select appropriate digital tools, using accounting and finance knowledge, that provide
innovative solutions to complex business problems.
• AI is able to process unstructured data and perform high-complexity analyses and tasks.
• AI can be used to intelligently automate processes, analyse big data to generate insights, and
intelligently engage with stakeholders.
3.2 Evaluate the impact of innovative technologies on the work of accounting and finance
professionals to communicate the benefits and risks to key stakeholders in the organisation.
• The use of AI to analyse big data generates insights that cannot be obtained using conventional
analytics. This provides accounting and finance professionals with information to better support
the organisation’s objectives.
• The use of AI may have a negative impact on employee morale and performance, particularly if job
losses are feared.
• Like any business innovation, a proposal to introduce AI should be evaluated with regard to its
benefits, risks and challenges. AI requires a significant investment of time and resources and
introduces potentially serious risks that must be understood and managed.
3.3 Appraise the suitability of technology in creating efficiencies and solving complex
business problems.
• AI is a useful technology to automate processes to create time and cost efficiencies for an
organisation; to analyse big data to generate insights; and to intelligently engage with customers
and employees.
Pdf_Folio:205
Traditional
programming Supervised ML Unsupervised ML
Pdf_Folio:206
EXAMPLE 3.14
ML in Google Maps
Google Maps is a good example of how a predictive ML model learns from historical data. Every time a
Google Maps user enters a destination, Google Maps adds this to its data set of ‘places the user goes’. It
also collects information about other locations the user goes on the same trip (such as a favourite coffee
shop, where the user often stops for five minutes to pick up a hot drink), the time of day or days of the
week the user goes to each destination, and so on.
Then, a predictive ML model uses that information to help make the user’s interactions with Google
Maps quicker and easier. When the user opens the map to enter a destination, a list of possible
destinations will appear, so that if one of those is correct the user can select it with a single touch instead
of having to key in an address. Additional data can be drawn from the user’s Google calendar to identify
destinations they are likely to want to go in the immediate future based on their upcoming appointments.
With each selection the user makes, the Google Maps algorithm is further refined, so that next time they
wish to select a destination, the list of suggestions will be more closely matched to their likely requirements.
Google Maps also uses outlying data to help improve its ML algorithms. For example, when the user
does not want to go to a destination that is in the list of suggestions provided, they may think that the
predictive ML algorithm has failed them — and to some extent it has, as it currently cannot accurately
predict every single trip the user will take. However, in entering an unpredicted destination the user is also
providing the ML model with additional data to work with. It will analyse this new data and try to improve
its algorithms to refine its predictions for next time.
Pdf_Folio:207
QUESTION 3.9
How does the data used in supervised machine learning and unsupervised machine learning differ?
Deep Learning
Deep learning is a discipline within ML. Like all approaches to ML, deep learning is used as a way of
teaching machines to behave like humans and therefore bolster opportunity and productivity for businesses.
In place of conventional ML algorithms to evaluate data and generate outputs, deep learning uses artificial
‘neural networks’ to mimic the complexity of analysis performed in a human brain’s neural networks.
Neural networks process data through a layered structure of algorithms or decision points. A deep learning
model which uses neural networks therefore has a greater capacity to ‘learn’ or evaluate multifaceted data
than standard ML models. In fact, not only can neural networks process a significantly greater amount of
data than ML models, they actually require a significantly greater amount of data in order to create accurate
outputs. Figure 3.12 illustrates the basic principle by which neural networks function.
Figure 3.12 illustrates that a simple neural network contains connections between multiple nodes,
allowing for powerful processing and analysis of data. It takes in the data, called ‘input data’, trains itself
to recognise patterns in that data and then makes a prediction or an informed decision about the data.
The result is referred to as ‘output data’. The right side of figure 3.12 shows how a deep learning neural
network takes this strategy further by applying multiple layers of analysis to data, exponentially increasing
the patterns and insights that can be uncovered. The level of learning and analysis from a deep learning
neural network can be complex enough to mimic human expertise in some fields, such as market valuations
and loan approvals (see example 3.15).
Pdf_Folio:208
EXAMPLE 3.15
.......................................................................................................................................................................................
CONSIDER THIS
In what aspects or your work do you make decisions like Joan’s? Would any of those decisions benefit from full
consideration of all relevant factors, as could be achieved with deep learning?
QUESTION 3.10
Explain why an AI system can be better at finding patterns in data than even a very experienced
financial analyst.
Pdf_Folio:209
ML and NLP
NLP is a branch of AI that allows computers and machines to understand, manipulate and interpret human
language. In simple terms, NLP uses algorithms to identify and extract data from text and speech and
match it to natural language rules that have been converted into a format that the computer can understand.
When data has been provided in natural language format, the NLP model will use algorithms to extract
the meaning associated with the input and provide the appropriate outputs.
NLP is the driver and enabler of many other technologies and technology use cases. For example, NLP
can be used in combination with ML to streamline customer support by analysing customer phone calls.
This is beneficial to customers and the company as it improves the ability to analyse the calls and resolve
the issues that prompted them. NLP ‘interprets’ what a person says or writes by identifying key words
and phrases and sending requests to an app. It can pair this with other information, such as the person’s
location and map and business data, to respond to requests such as ‘What pizza restaurants are close by?’
Like ML, NLP is not a new field, but increasing computer power, increasing levels of available data
and a growing number of use cases have driven a resurgence of interest and development.
Figure 3.13 shows how the NLP machine understands a human who is asking a question in speech or
writing and uses that understanding to inform its response. Without NLP, the human had to type their
question in a syntax that matched the computer’s predetermined instructions.
(Intellectually) responding
to the user on the basis of
the understood content
Humans communicate in countless ways and, to maximise value, NLP technology must be equally
dynamic in its ability to understand. Current sophisticated NLP makes it possible for computers to read
text, hear speech, interpret it, measure emotional responses and determine which parts are important.
Pdf_Folio:210
EXAMPLE 3.16
Telstra’s Chatbot
Telstra handles more than 50 million calls a year through its customer contact centre. Telstra’s adoption
of AI began with the implementation of one customer-facing chatbot (called Codi), which was initially a
routing bot developed to understand a customer’s needs and direct them to a live chat agent. With the help
of IBM’s Watson virtual assistant technology, Codi has evolved into a ‘digital concierge’ or ‘conversational
agent’ that can offer customers’ answers in real time. This has resulted in a reduction in agent handling
time, a reduction in traffic to chat, messaging and call centres, and increased adoption of digital self-
service tools.
Having identified a range of potential use cases for virtual agents like Codi, Telstra has now expanded
its chatbot network. Around nine chatbots, conversational agents and virtual assistants operate within a
variety of functions, from an ACE bot assisting field service engineers with troubleshooting, to product-
specific bots such as Simmo, tasked to answer questions about pre-paid products.
.......................................................................................................................................................................................
CONSIDER THIS
The virtual assistants from Apple, Amazon, Google and Microsoft are probably the most familiar NLP applications for
most people. Could you envisage a role for an NLP-enabled assistant in your own work?
QUESTION 3.11
Cognitive ML
Cognitive ML focuses on recognising and understanding text, speech, photos, audio and video in a way
that aims to mimic human thoughts. Cognitive ML uses data as models of humans, mostly focusing on the
areas of text, speech and common-sense reasoning to construct effective and self-teaching algorithms to
mimic human outcomes.
For example, NLP uses data to build algorithms that enable chatbots to recognise and mimic not only
how humans use language, but also the tone of happiness, excitement, sadness, anger and other human
emotions that form the non-verbal element of language. Visual recognition uses data to build algorithms
for visual pattern recognition.
Cognitive ML supports businesses through applications such as:
• indexing business documents, emails, images and social media to help users search through them
• processing photos or audio files to measure and control quality
• using virtual agents (chatbots) to communicate in natural language with customers in commercial or
customer service settings, and analysing customers’ emotions and social tendencies.
Predictive ML
Predictive ML characterises and analyses historical and recent data to predict an accurate future outcome.
Predictive ML utilises data analytics techniques and tools to produce predictive outcomes. As the
algorithms learn from new cases, the software improves over time.
Predictive ML is extremely powerful and can help businesses offer better services, be more efficient and
outperform competitors. It is currently predominantly used by larger companies. Applications include:
• forecasting revenues and sales for certain products in combination with weather predictions (e.g. stores
sell more umbrellas when it is raining)
• predicting the probability of customers moving to a competitor (churn rate) based on their contacts with
customer service
• predicting maintenance of engines and wind turbines based on data coming from IoT sensors; an IoT
sensor enables computers to collect data from other internet-enabled devices via cloud technology
• supporting intelligent credit management and collection processes through automated advice to release
or block a sales order
• supporting businesses’ legal functions with advice based on jurisprudence data.
Optimising ML
Optimising ML focuses on applying algorithms to find the most efficient and effective outcome for a
process. Examples of everyday use include automatic colour correction of photographs and GPS navigation
systems that find the optimum route to the user’s destination. For example, Google Maps uses ML to
identify traffic patterns from previous data and therefore consider routes that have been the best in the
past. Additional information from crashes, road works and closures is then analysed by AI algorithms to
calculate the best route.
Optimising ML supports businesses through:
• process optimisation
• logistical optimisation (e.g. in warehouse operations)
• supply chain optimisation (e.g. based on time or cost).
Classifying ML
Classifying ML is an intelligent way to segment and cluster data in large data sets. In some cases, it is
unsupervised, which means its algorithm has not ‘learned’ from historical data but is ‘smart’ enough to
identify patterns in data on its own.
One common use of classifying ML is credit card fraud detection. For example, if a credit card
transaction takes place in Amsterdam and within 10 minutes the same credit card is used for a transaction
in Singapore, the ML algorithm detects an irregularity, blocks the transaction and warns the customer.
Pdf_Folio:212
3.13 BENEFITS OF ML
Five main areas in which businesses can realise benefits from the use of ML are as follows (Raza 2019).
1. Real-time decision making. Businesses can use ML to transform large, unwieldy data sets into actionable
information. The speed at which ML can process data means this information is available in real time,
allowing businesses to respond quickly to changing market conditions.
2. Eliminating manual tasks. Not only can automation replace the need for humans to complete basic
repetitive tasks but, with the addition of ML, machines can now also adjust to variable inputs or changing
internal conditions. Automation can therefore effectively complete tasks which are more complex or
dynamic, further reducing the need for human intervention.
3. Enhancing security and network performance. Cybersecurity threats occur quickly and can cause
widespread damage or data loss within a very short space of time. The speed and processing capacity
of ML means it is uniquely suited to continuous analysis of network security, identifying and reporting
on any breaches or anomalies much faster than a human could.
4. Improving business models and services. The application of technological innovation to existing
products and services can help businesses gain and preserve an advantage over competitors. ML has
been particularly beneficial in this regard in recent years with businesses such as Airbnb using models
heavily leveraging ML capability.
5. Reducing operating expenses. In addition to completing routine tasks, ML can also support business’s
customer support functions through the use of chatbots and other similar technologies. By answering
customer queries quickly and accurately, ML can significantly reduce customer support costs to a
business.
Fraud Detection
Security threats in finance are increasing and ML algorithms are excellent at detecting fraud. For example,
banks can use this technology to monitor thousands of variables in every account transaction in real
time. The algorithm examines each action a cardholder takes and assesses if an attempted activity is
characteristic of that particular user. This model spots fraudulent behaviour with high precision.
Pdf_Folio:213
Underwriting
ML algorithms work well with the underwriting tasks common in finance and insurance. Data scientists
train ML models on thousands of customer profiles, with hundreds of data entries for each customer. A
well-trained system can then perform the same underwriting and credit-scoring tasks with real-world data
inputs, thereby helping human underwriting employees work much faster and more accurately.
Banks and insurance companies tend to retain large amounts of historical consumer data, so they can
use these data sets to train ML models.
Algorithmic Trading
Algorithmic trading uses a computer program to place a trade. The trade, in theory, can generate profits at a
speed and frequency that is impossible for a human trader. In algorithmic trading, ML uses a mathematical
model which monitors the news and trade results in real time and detects patterns that can force stock
prices to go up or down. It can then proactively sell, hold or buy stocks according to its predictions, which
helps make better trading decisions.
ML algorithms can analyse thousands of data sources simultaneously, something that human traders
cannot possibly achieve.
Robo-Advice
A robo-advisor is an online application that provides automated financial guidance and services. Robo-
advisors are commonplace in the financial industry, with two major applications of ML in the advisory
domain.
• Portfolio management. An online wealth management service that uses algorithms and statistics to
allocate, manage and optimise clients’ assets. Users enter their current financial assets and goals, such
as saving a million dollars by the age of 50. A robo-advisor then allocates the current assets across
investment opportunities based on the risk preferences and the user’s desired goals.
• Recommendation of financial products. Many online insurance services use robo-advisors to recommend
personalised insurance plans to a particular user. Customers choose robo-advisors over personal
financial advisors due to lower fees and personalised and calibrated recommendations.
EXAMPLE 3.17
Streamlining Reporting
Goal
A financial institution seeking to enhance user experience established a goal: Improve the experience for
users involved in journal processing and reporting tasks.
The steps to achieving this goal were determined to be:
• automating journal processes
• enabling users to acquire multi-disciplinary skills
• facilitating more accurate continuous reporting.
Challenges
Users were unable to access the required data, prepare the journal and identify errors. In the original
process, various factors made it challenging for the user to complete the required tasks including the
following.
• Accessing the required data
– Users had diverse needs based on their role.
– Users accessed multiple different source systems.
Pdf_Folio:214
QUESTION 3.12
Explain how applying machine learning to unstructured data can assist accountants in their work.
Pdf_Folio:215
Data-related Risks
Not having enough data and/or having ineffective data can present risks to any application of digital finance
technologies. These risks are significant in relation to ML models as the quality of the input data largely
determines the quality of the algorithms created and, in turn, the output data.
Data-related risks in ML include the following.
• Data quality. Assess the data source, how it has been processed, its completeness and its accuracy to
ensure data quality is sufficient.
• Limited data. An effective ML model can be built on a small amount of data, but that model might
not continue to be accurate in the long term unless more information is provided to test and refine the
algorithms created. Therefore, the data set should be expanded over time to extend the capabilities and
learning of the model.
• Homogeneous data. This risk relates to lack of variety in the data. For example, to forecast home prices
in a city it is more effective to use as many (high-quality) different data sets as possible to build the
models. Ideally the different data sets will feature many different types of demographical data points.
Time should then be allocated to ‘feature engineering’ to find the best model inputs for accurate outputs.
Training on homogeneous data can lead to algorithms that do not give accurate outputs when more
diverse data are introduced.
• Data compliance and privacy issues. data privacy laws and ethics constrain how an organisation can —
or should — use the data it holds. The risk of misusing data must be managed.
Earlier in the module, we noted that biased data could results in biased outputs from AI machines.
Significant investment is required to ensure input data is not biased. This becomes much more challenging
with ML, because ML continuously seeks new data, uses it to refine its own algorithms and often
automatically takes action based on its findings. Biases evident in data can quickly skew the outputs.
This can result in poor or even quite damaging decisions being made. Microsoft, for example, launched an
ML-based chatbot on the Twitter platform in 2016, only to withdraw it within a day because it published
thousands of racist and misogynist tweets in response to similar tweets directed at it from a particular
group of users.
.......................................................................................................................................................................................
CONSIDER THIS
If ML processes millions of data points in real time and automatically acts on that input, how can an organisation
ensure the decision-making of the ML is legitimate, ethical and in the best interests of the business’s objectives?
QUESTION 3.13
Discuss the degree to which results produced by each of the automation technologies can be
explained.
Pdf_Folio:217
EXAMPLE 3.18
CBA ML Strategy
How CommBank is Training its Machine Learning Customer Engagement Engine
The Commonwealth Bank of Australia (CBA) has built a Customer Engagement Engine it has touted as
powering customer experience through the use of artificial intelligence (AI) and machine learning.
‘[It] is an advanced system which combines AI, machine learning, and our customer data to continuously
optimise and prioritise across all of the available messages, alerts, conversations, and communications
we can have with our customer at any given time, across all of our channels,’ CBA chief analytics officer
Andrew McMullan said.
‘And all of that in real time.’
Speaking with media at the launch of the bank’s new app, McMullan said CBA runs over 200 machine
learning models on top of 157 billion data points. He said these models are continuously ‘optimising,
prioritising, and returning the next best personal message to our customers as they interact with us across
all channels’.
McMullan said the bank has invested a lot in automated machine learning capability.
‘So maybe there’s a message that we want to test with some of our customers, in one of the assets
within the mobile app, we’ll push that message live and in the background we’ll switch on one of our
advanced machine learning models,’ he explained.
‘As our customers engage with that message, maybe a click through a “yes, please”, or “no, thanks”,
the machine learning model in the background is learning from that to really more accurately predict which
type of customers really enjoy experiences with that particular message.’
When CBA is happy with the performance of the model, McMullan said it would switch that conversation
on and add it to the system to become another message that will be shared with customers.
‘As I navigate my way through the experiences in the mobile app, each of the assets has been
personalised by making a call to the Customer Engagement Engine to say, “What’s the next best message
to determine this particular communication?”’
Providing a scale of the system, McMullan said that each day the bank’s applications make over 20
million calls to the Customer Engagement Engine to return personalised messages.
Over the next 12 months, the bank expects it will have the opportunity to deliver 3 billion personalised
messages to its customers.
‘One of my favourites is our Smart Alerts ... on credit cards. We’re constantly monitoring our customer
data and if a customer hasn’t paid the minimum amount or paid down their credit card balance, three days
Pdf_Folio:218
.......................................................................................................................................................................................
CONSIDER THIS
Two years after the Commonwealth Bank launched the strategy discussed in in example 3.18, it announced a ‘refresh’
of its digital strategy. It had doubled the number of ML algorithms in its Customer Engagement Engine and was
making changes to it every day. In general terms, how can an ML strategy provide direction while enabling ongoing
development?
SUMMARY
ML is a type of AI in which the machine can develop its own algorithms based on patterns it identifies
in data. It refines its algorithms over time based on experience. Thus, the machine ‘learns’. ML requires
large volumes of data and substantial processing power, and is implemented in conjunction with other
AI technologies, often including NLP, which enables it to handle unstructured data and communicate its
outputs with stakeholders in natural language. It may also be combined with RPA so that it can drive
intelligent processes.
ML applications may be categorised as cognitive, predictive, optimising or classifying, but many
applications combine multiple approaches. ML enables the analysis of large datasets to support real-
time decision-making, supports intelligent automation, enhances security and network performance,
provides new ways to improve business models and services, and can create efficiencies that reduce
operating expenses.
Deep learning is a discipline within ML. Deep learning uses artificial ‘neural networks’ to mimic the
complexity of analysis performed by humans by processing data through a layered structure of algorithms
or decision points.
The key challenges in using ML relate to the need for robust data governance, ensuring valid
interpretation of outputs, and ensuring engagement with customers or other users of the technology is
effective and beneficial.
For organisations with valuable use-cases and the required resources and capabilities to use an ML
solution, ML can play an important role within the organisation’s digital strategy. The ML strategy should
ensure the investment can be scaled across the entire organisation over time rather than being restricted to
narrow applications.
Pdf_Folio:219
KEY POINTS
3.1 Select appropriate digital tools, using accounting and finance knowledge, that provide
innovative solutions to complex business problems.
• ML is able to intelligently handle unstructured data.
• ML is able to generate insights not possible through other approaches to data analysis.
• ML integrates with other technologies such as NLP to enable automated responses to the insights
it generates, including interacting with customers and employees.
• ML can generate outputs which predict future behaviour based on past behaviours and thus enable
proactive and pre-emptive decision making.
• ML enables optimisation of processes.
• ML’s ability to undertake classification tasks has a range of applications in areas such as fraud
detection and cybersecurity.
3.2 Evaluate the impact of innovative technologies on the work of accounting and finance
professionals to communicate the benefits and risks to key stakeholders in the organisation.
• ML enables real-time decision-making based on real-time analysis of large volumes of complex
data, including through predictive analysis.
• ML supports intelligent automation, enhances security and network performance, provides
new ways to improve business models and services, and can create efficiencies that reduce
operating expenses.
3.3 Appraise the suitability of technology in creating efficiencies and solving complex
business problems.
• ML can be used to optimise and intelligently automate processes, including engagement with
stakeholders.
• ML can generate insights and predictive analysis from complex data in real time.
Pdf_Folio:220
Therefore, the selection of technologies is inseparable from the understanding of how that technology
will be used to benefit the business. Technology can help organise processes, provide business com-
petitiveness, improve customer service, enhance internal and external communication, increase business
productivity and profitability, and more.
With the rapid development in technologies, this section aims to model the choice of technology against
a particular scenario. The guiding principle should be that the ‘best’ technology is technology that matches
the business’s needs.
Example 3.19 steps through a bank’s launch of a new credit card product to model how technologies
might be chosen for any one particular problem.
EXAMPLE 3.19
Technology Selection
A bank recently launched a new credit card with a market-leading rewards program and a 0 per cent
interest rate on all transferred balances for the first six months. The bank typically receives around 1000
credit card enquiries and applications each month, but since the campaign launch they are now receiving
5000 a month. Call wait times in the bank’s customer contact centre have increased from an average of
5 minutes to 15 minutes, with a much higher drop-out rate than normal.
The bank’s marketing manager and head of credit card processing are concerned that they are missing
out on possible new credit card customers due to the poor customer experience. They brainstormed some
ideas to use automation technologies to help address the problem.
They chose to proceed with an AI-powered virtual agent to capture and answer customer enquiries
about credit cards. By implementing a virtual agent to capture and answer customer queries, the bank’s
customers reported a significantly improved experience: 30 per cent of all customer enquiries and
applications for the new credit card were handled by the virtual agent and average call handling time
was reduced by 50 per cent for customers who interacted with the virtual agent.
Pdf_Folio:221
Automation technologies are most effective when the most relevant capabilities are harnessed. In each
phase of example 3.19, many different technologies could have been deployed, but it was critical for the
bank to appropriately match the technology with the functionality required.
.......................................................................................................................................................................................
CONSIDER THIS
Where could RPA, AI, ML and combined technologies be applied in your business or a client’s business? Try to
develop some specific use cases and the rationale for the choice of technology. Focus on the use cases that would
provide most value to the organisation. Could any of the solutions be applied across multiple problems?
Pdf_Folio:222
Technology strategy: The what The organisation must keep abreast of emerging technologies and determine
how these technologies might be applied to the organisation.
Talent strategy: The who The organisation must acquire or grow talent to ensure staff have the skills
and knowledge required to transform, or have a strong framework to source
specialist skillsets.
Governance strategy: The where The organisation must clearly define the KGIs, responsibilities, account-
abilities, organisation design and KPIs that will be employed across the
enterprise to govern the implementation of digital transformation.
(continued)
Pdf_Folio:223
Processes strategy: The how The organisation must create well-defined business activities, rules and
procedures, and prioritise data and resource ownership, so that process
transformation can be applied throughout the enterprise.
Common Themes
Matt et al. 2015 suggest that all digital transformation strategies have four key objectives in common.
• Technology creation. Application of new technologies to the organisation to improve how important
resources are used, create greater organisational agility, and alter the basis of competition
• Value innovation. Creation of superior value for customers and lower costs for the company in a way
that aligns with the company’s strategic focus
• Structural modification. Creation of dynamic operational relationships between various entities within
(and in relation to) the organisation
• Financial revolution. Disruption of financial institutions and transformation of how financial institutions
generate and utilise insights from business data, in turn producing business model innovations,
enhancing the competitiveness of the company and its workforce, engendering new risk dynamics and
anchoring new challenges to the policymakers.
Example 3.20 provides an example of technology creation leading to financial revolution.
EXAMPLE 3.20
Key Elements
We have discussed key priorities and common themes of digital transformation strategies. We will now
look at three key elements of the strategy:
1. KGIs
2. gap analysis and identifying alternatives
3. road map and KPIs.
Key Goal Indicators
One of the enduring questions regarding digital transformation is how success is measured against the
investment that is made in transformational technologies. One way to do this is through the use of key
goal indicators (KGIs). Establishing KGIs is a critical part of any digital transformation journey, as there
can be a significant gap between the digital transformation vision and its ultimately realisable reality.
Some companies that are extremely successful in the non-digital world fail to execute in the digital
universe, whether through technology failures, misalignment between operational and technical strategy,
or a failure to properly measure and manage risks.
Digital transformation goals can be measured through KGIs, including:
• attraction of new clients
• establishment of new service supply channels
• reduction of cost per client
• increases in productivity
• reductions in risk
• alignment with industry standards.
Pdf_Folio:224
EXAMPLE 3.21
Pdf_Folio:225
Deloitte Implemented AI-enabled document There was a 50% reduction in the time
reviewing process to extract relevant spent in extracting legal documents,
information from various documents. invoices, customer financial statements
and minutes.
Applied IBM Maximo technology to
automate cognitive asset inspection
process.
PwC Applied NLP to process unstructured NLP was able to generate insights from
data. complex agreements, contracts and
minutes.
Developed general ledger AI (GL.ai), to
analyse documents and prepare reports. The GL.ai system, which has been
successfully implemented in 12
countries, is able to speed up the audit
process and generate insights that boost
efficiency.
Pdf_Folio:226
Pdf_Folio:227
Pdf_Folio:228
EXAMPLE 3.22
.......................................................................................................................................................................................
CONSIDER THIS
What aspects of governance, risk management and compliance in your role could be affected by the adoption of
digital technologies? Could any of these prevent the adoption of any particular technology? Could some of your
governance, risk management and compliance work itself be automated?
QUESTION 3.14
Explain how automation can both increase governance risks, and improve the management of
governance risks.
Pdf_Folio:230
• Is the AI initiative well-defined, governed and monitored against your expected business outcomes?
• Are all ethical, moral and safety considerations fully addressed in any responsible AI program?
• Are ethical criteria well-defined and are algorithms developed effectively?
• Are the relevant criteria appropriately tailored for the business?
• Are algorithms subject to continuing review and monitoring for their continued relevance and
effectiveness?
• Is data subject to processing being completely and accurately captured?
• How is data security, privacy and protection maintained?
• In what circumstances should a human override be applied were things to go wrong, and how is this
decision and application controlled and monitored?
Source: CPA Australia 2021.
Müller (2020) highlights some of the key topics of debate when considering the ethics of AI and robotics
and outlines some key concerns. Many of these topics are complex and we only provide a brief overview
of some interesting challenges and considerations below.
Manipulation of Behaviour
Policies are required to limit online traders and service providers from exploiting customers using the
technologies and personal information at their disposal. At the moment, there are calls to stop many online
platforms from abusing personal information to maximise their profits, or influence a person’s behaviour,
by exploiting a users’ personal information and behavioural biases. Some extreme cases may involve
online criminal activities such as deepfake scams, deception, and making false claims to induce or facilitate
addiction to online gambling services. (Shao 2019)
Opacity of AI Systems
There is a need for humans to be able to gain insight and control how a particular pattern was identified by
an automated decision algorithm with due process, accountability, community engagement, and auditing.
It is risky and dangerous for an AI system to draw conclusions without a human having some awareness
or oversight as to how this conclusion has been reached.
.......................................................................................................................................................................................
CONSIDER THIS
If your organisation or client implemented an ML solution to analyse data and provide insights and recommendations,
what would you need in order to have confidence to act on those insights and recommendations?
Pdf_Folio:231
Human–Robot Interaction
AI can be used to manipulate humans into believing and doing things. There is a belief that humanity needs
to prevent AI from working with processes or behaviours that involve deception, threaten human rights
and dignity, or violate the fairness in humanity. A common example is automated web bots or auto-dialled
phone calls that seek to scam users into making fraudulent payments.
Autonomous Systems
There is a need to balance ethical issues in autonomous systems, such as self-driving cars or robot vacuum
cleaners, where the AI exists to serve its owner. It must be determined whether an owned AI may be
influenced to pursue the owner’s self-interest in a manner that could endanger or disadvantage others (e.g.
whether it is possible to instruct a self-driving car to speed at the risk of other members of society).
Machine Ethics
This refers to the principles and techniques required for creating an artificial machine that can follow ideal
ethical principles, and exhibit the mechanisms required to be capable of engaging in moral behaviours
Anderson and (Anderson 2007). There is a need to ensure that the behaviour of AI towards human users
and other AI machines is ethically acceptable. There has also been significant controversy around the use
and scope of technologically enabled autonomous weapons systems, or so called ‘killer robots’ that in some
cases may be able to harm humans based on predefined rules, without the need for a specific human-made
decision or input.
QUESTION 3.15
Discuss the relative risks and benefits of using personal data in RPA, AI and ML applications.
Pdf_Folio:232
EXAMPLE 3.23
.......................................................................................................................................................................................
CONSIDER THIS
How capable is Hannah at:
• selecting appropriate digital tools, using accounting and finance knowledge, that provide innovative solutions to
complex business problems
• evaluating the impact of innovative technologies on the work of accounting and finance professionals to
communicate the benefits and risks to key stakeholders in the organisation
• appraising the suitability of technology in creating efficiencies and solving complex business problems?
Think about how your own abilities compare with Hannah’s.
SUMMARY
Technologies such as RPA, AI and ML are valuable solutions to complex business problems when they
are appropriately applied. Any particular problem will be most effectively and efficiently addressed by
a specific technology or combination of technologies. Understanding the applications, benefits, risks and
challenges of each technology enables an appropriate evaluation against well-defined goals and objectives.
Digital transformation of an organisation requires a cohesive strategy aligned to the overall business
strategy. Each decision should be focused on the value that is created for the organisation and its
stakeholders.
Pdf_Folio:237
KEY POINTS
3.1 Select appropriate digital tools, using accounting and finance knowledge, that provide
innovative solutions to complex business problems.
• Digital tools should be selected for their ability to answer a specific need. A useful guiding principle
is that the ‘best’ technology is the technology that matches the business’s needs.
• Technologies should not be evaluated in isolation, but rather as synergistic parts of an overall digital
strategy.
3.2 Evaluate the impact of innovative technologies on the work of accounting and finance
professionals to communicate the benefits and risks to key stakeholders in the organisation.
• Innovative technologies can achieve process efficiencies, create new insights by analysing data in
powerful ways and engage with employees and customers to enhance their experience.
3.3 Appraise the suitability of technology in creating efficiencies and solving complex
business problems.
• Understanding the applications, benefits, risks and challenges of each technology enables an
appropriate evaluation against well-defined goals and objectives.
Pdf_Folio:238
REFERENCES
Agarwal, D 2018, ‘An introduction to AI at LinkedIn’, LinkedIn, accessed August 2020, https://engineering.linkedin.com/blog/
2018/10/an-introduction-to-ai-at-linkedin/.
Anderson, M & Anderson, SL 2007, ‘Machine ethics: Creating an ethical intelligent agent’, AI Magazine, vol. 28, no. 4, accessed
October 2020, https://www.researchgate.net/publication/220605213_Machine_Ethics_Creating_an_Ethical_Intelligent_Agent.
Barbaschow, A 2019, ‘How CommBank is training its machine learning Customer Engagement Engine’, ZDNet, 2 August,
accessed March 2021, https://www.zdnet.com/article/how-commbank-is-training-its-machine-learning-customer-
engagement-engine/.
Bates, M 2020, ‘AI, machine learning behind new Ellie Mae credit evaluation automation’, Mortgage Orb, accessed October 2020,
https://mortgageorb.com/ai-machine-learning-behind-new-ellie-mae-credit-evaluation-automation/.
Blue Prism n.d., ‘Coca-Cola extends business services capacity and improves performance with RPA’, October 2020, https://www.
blueprism.com/uploads/resources/case-studies/blue-prism-cola-case-study.pdf
Brodsky, L, et. al. 2018, ‘Open banking’s next wave: Perspectives from three fintech CEOs’, McKinsey, accessed October 2020,
https://www.mckinsey.com/industries/financial-services/our-insights/open-bankings-next-wave-perspectives-from-three-fintech-
ceos/.
Chan, R 2020, ‘The Cambridge Analytica whistleblower explains how the firm used Facebook data to sway elections’, Business
Insider, accessed October 2020, https://www.businessinsider.com/cambridge-analytica-whistleblower-christopher-wylie-
facebook-data-2019-10.
Cheng, A 2019, ‘Why Amazon Go may soon change the way we shop’, Forbes, accessed October 2020, https://www.forbes.com/
sites/andriacheng/2019/01/13/why-amazon-go-may-soon-change-the-way-we-want-to-shop/.
Crowe, D 2020, ‘Simpler lending rules for home loans and credit to free up the economy’, Sydney Morning Herald, 24 September,
accessed October 2020, https://www.smh.com.au/politics/federal/simpler-lending-rules-for-home-loans-and-credit-to-free-up-
the-economy-20200924-p55yz9.html/.
Davenport, TH & Ronanki, R 2018, ‘Artificial intelligence for the real world’, Harvard Business Review Magazine, January–
February, accessed March 2021, https://hbr.org/2018/01/artificial-intelligence-for-the-real-world.
Dilmegani, C 2020, ‘15 RPA benefits compiled from top sources [2020 update]’, accessed October 2020, https://research.
aimultiple.com/top-robotic-process-automation-rpa-benefits/#maintaining-an-audit-trail.
Downes, L & Nunes, P 2013, ‘Big bang disruption’, Harvard Business Review Magazine, March, pp. 44–56.
DTA (Digital Transformation Agency) 2018, ‘Digital transformation strategy’, accessed October 2020, https://www.dta.gov.au/
digital-transformation-strategy.
Faggella, D 2020 ‘AI in the accounting Big Four — Comparing Deloitte, PwC, KPMG, and EY’, accessed October 2020,
https://emerj.com/ai-sector-overviews/ai-in-the-accounting-big-four-comparing-deloitte-pwc-kpmg-and-ey/.
Fisher, C 2020, ‘Adobe previews AI-powered “Sky Replacement” tool for Photoshop’, Engadget, accessed October 2020,
https://www.engadget.com/adobe-sensei-ai-sky-replacement-preview-155256593.html/.
Goh, C, Pan, G, Poh Sun, S, Lee, B & Yong, M 2019, Charting the future of accountancy with AI, Research Collection School Of
Accountancy, accessed September 2020, https://ink.library.smu.edu.sg/cgi/viewcontent.cgi?article=2833&context=soa_
research.
Gusher, T 2018, ‘Learning to manage machine learning’, KPMG, accessed March 2021, https://assets.kpmg/content/dam/kpmg/br/
pdf/2018/12/br-learning-to-manage-machine-learning.pdf.
Pdf_Folio:239
Pdf_Folio:241
DATA ANALYTICS,
INTERPRETATION AND
VISUALISATION
LEARNING OBJECTIVES
PREVIEW
In module 1, we described how advances in technology had led to a massive increase in the creation,
capture, processing and sharing of data. This explosive growth in data is a fundamental enabler of the
digital finance ecosystem we described in module 1, the emerging forms of digital currency we described
in module 2 and many of the finance and other business applications of automation, artificial intelligence
and machine learning technologies we described in module 3.
Module 3, in particular, described the use of technology to interact with data in ways that create value for
the organisation and its stakeholders. All of those approaches involved the analysis of data and application
of the findings. In module 3, the focus was on how technology could analyse data and, in many cases,
automatically respond through applications such as chatbots and robo-advisors. In this module we will
examine how data can be analysed to generate insights that can be presented to decision makers to help
them achieve the business’s operational, tactical or strategic goals.
Accounting and finance professionals are experienced data owners, analysts, information creators and
information providers, but the recent massive expansion of data availability and analytical abilities has
created the need for an expanded skill set.
The field of analytics formalises the role of data for decision making in organisations. As such, account-
ing and finance professionals need to attain and encourage data literacy throughout the organisation,
make meaningful input to the organisation’s data strategy, be able to prompt, shape and work with the
output of analytics processes, interpret the meaning of analytics outputs and use appropriate visualisation
and communication tools to ensure others, particularly decision makers, gain actionable insights arising
from analytics.
Part A of the module establishes the concept of data-informed decision making and the need for
accounting and finance professionals to develop data literacy skills. Part B explores how to develop and
implement a robust data strategy to capture, store and process the data required by the organisation’s
analytics applications. Part C steps through the process of analysing big data, including the application of
artificial intelligence technologies. Part D examines the process of interpreting the outputs of analytics
to derive insights and make recommendations to decision makers about possible courses of action.
Pdf_Folio:243
Part E concludes the module by exploring visualisations and other techniques that can be used to effectively
communicate the actionable insights and recommendations that have been developed.
PART A: DATA LITERACY
INTRODUCTION
Data is ubiquitous and cheap to acquire. It is being generated by organisations and individuals at high
velocity and in large volumes. A business that is able to extract relevant insights from data to inform
strategy and operations can obtain a competitive edge over those that do not. Extracting and using these
insights requires the organisation to develop or acquire analytics expertise (e.g. to employ data scientists
or outsource analytics to a specialist provider). Moreover, however, to translate insights into value-adding
actions requires the organisation to adopt a data-informed approach to decision making. This, in turn,
requires data literacy skills throughout the organisation.
ANALYTICS
As explained above, the basis of data-informed decision making is knowledge created through analytics.
The earlier modules in this study guide frequently referred to the value of data. Data is simply a fact or
observation, of little use in its raw state. It is what the organisation is able to do with its data that creates
value. Analytics refers to the use of skills, processes and technologies to derive information from data.
Information has meaning and purpose relevant to an individual’s or organisation’s needs. Knowledge is
created by integrating information with prior experience, skills and opinions. Finally, wisdom is the ability
to use knowledge to make decisions and act on them. Analytics, therefore, is a crucial step in deriving value
from data.
Pdf_Folio:244
Questions What has Why did it happen? What is likely to How can we make it
answered happened? What happen? happen?
is happening?
Data sources Surveys, interviews, Internal and external Historical and real- Historical and real-
financial data data (e.g. weather, time data including time data including
location) transactional data, transactional data,
customer service data feeds and
data and big data big data
(e.g. social media)
Analysis tools/ Metrics reports, data Principal component Data mining, Rules, simulation
techniques mining, aggregation, analysis (PCA), quantitative analysis, analysis,
used clustering and sensitivity analysis, predictive modelling, recommendation
summary statistics data discovery, machine learning engines, artificial
regression analysis algorithms and intelligence and
and data mining artificial intelligence neural networks
.......................................................................................................................................................................................
CONSIDER THIS
Study table 4.1 carefully. Do you understand all of the terms and concepts and their relevance to your work?
Many years ago, companies would use teams of analysts and statisticians to manually work with data
sets to derive insights, which could then be used to drive business decisions. Over time, advancements in
computing power and analytical techniques, combined with the massive growth in data availability, have
led to the growing application of data science in the business world.
Pdf_Folio:245
Data Mining
In business, most analytics now are produced through data mining rather than manual analysis. Data
mining is the use of technologies that incorporate data science principles to produce information from
data (Provost & Fawcett 2013). This information includes meaningful patterns, correlations and trends.
Technologies used in data mining include data management technologies, artificial intelligence (AI),
machine learning (ML) and data visualisation tools. These concepts will be discussed in detail throughout
this module.
Table 4.2 summarises some of the applications of data mining in business and example 4.1 examines
the extensive used of analytics and strong commitment to data-informed decision making at Amazon.
Retail • Customer preference and behaviour analysis (what products should be targeted
at what customers and at what prices)
• Customer relationship management (CRM) and churn (attrition to competitors)
• Workforce management
• Supply chain management
EXAMPLE 4.1
Pdf_Folio:246
.......................................................................................................................................................................................
CONSIDER THIS
Do you agree with Amazon’s principles? To what extent are these principles evident in your own organisation?
EXAMPLE 4.2
.......................................................................................................................................................................................
CONSIDER THIS
Ready access to analytics tools is an important part of developing an analytics culture. What else is required?
It is essential for to understand how data literacy is vital for an organisation that wants to succeed in the
digital economy. As someone who will deal with data and analytics as a core part of their work, the role
of the accounting and finance professional requires developing that literacy as an individual and helping
to drive a data literacy culture in the organisation. This is the next topic we will explore.
QUESTION 4.1
Suggest what the most important driver of a data-literate organisational culture would be and
explain your rationale.
Pdf_Folio:248
Data Personas
Members of an organisation will typically fall into one of the four types of data literacy personas: Data
Aristocrat, Data Knight, Data Dreamer and Data Doubter (Morrow 2018). The personas are ranked from
the most data literate to the least. The role of each persona will include learning, and for the more advanced
categories, will also include mentoring or leadership. Mentoring is essential to creating a data literacy
culture. Those with skills teach others to help them grow, and subsequently they pass on their skills to
others. The role of each data literacy persona in data literacy culture is summarised in table 4.3.
Pdf_Folio:249
develop their use of data in their work, and their skills in data science with the aim of progressing
towards the Data Aristocrat category.
Like Data Aristocrats, these individuals may have varying levels of leadership and mentoring
abilities. Part of their role is to act as data literacy ‘champions’, which means vigorously
supporting the data literacy culture in the organisation and acting as mentors where possible.
Data Dreamer
The Data Dreamer has basic awareness of data and a desire to learn and improve their data
science skill set.
Similar to the other data personas, the individual should continue to learn and develop their skills.
In the case of Data Dreamers, this would include learning fundamental concepts in statistics and
data analysis.
Data Dreamers might be found in roles that are typically less technical, as in human resources
management, where there is some, but not a high level of exposure to data.
Data Doubter
The Data Doubter often has roles that carry responsibilities. These individuals require the support
of the other data personas. Attention to the Data Doubter is crucial to developing a data literacy
culture, as they can hinder the adoption of a data literacy strategy or even derail the process.
The Data Doubter is often sceptical of the value of data and data-informed decision making.
on their extensive experience, or it is a vast change from the work culture they are familiar with.
This data persona may also fear the complexity of the data analytics or decision-making process
or may find the amount of information overwhwlming.
data and analytical approaches to support intuition, experience and organisational memory
Least data in decision making. By doing this, the Data Doubter can potentially be transformed to a Data
literate Dreamer (Morrow 2018).
EXAMPLE 4.3
.......................................................................................................................................................................................
CONSIDER THIS
Bloomberg uses data and analytics to drive innovative product development and to automate processes within the
organisation. It maintains a continuously updated ‘employee of the future’ profile, defining what skills are valuable
and necessary to meet the company’s business objectives. This profile is used to guide internal training and external
recruitment. For internal training, Bloomberg has developed a curriculum that is tailored to its specific needs. The
training program consists of 60 hours of face-to-face learning time, conducted over 10 weeks, supplemented by
self-paced study and assignments. The learners undertake a project involving data analytics related to their everyday
work, facilitating application of their new skills and knowledge. Consider what a data literacy curriculum for your
organisation or a client’s organisation might look like.
QUESTION 4.2
In most organisations the conversion of data analytics into business outcomes is not yet reaching
its potential. Explain why this may be the case.
Pdf_Folio:251
DATA CATEGORIES
As we learned in module 3, data can be categorised with respect to its structure.
• Structured data adheres to a pre-defined model and aligns to the table structure of relational database.
Structured data can be easily sorted, manipulated and analysed. Examples of structured data include
records contained in spreadsheets and SQL databases.
• Semi-structured data contains tags (e.g. XML) or other markers (e.g. commas in a CSV file) to separate
elements and enforce hierarchies of records within the data. This enables a degree of analysis, but semi-
structured data would require mapping and processing before it could fit into a conventional database.
• Unstructured data has no predefined structure or organisational scheme. It is difficult to process or
analyse using conventional software, though advanced analytics technologies can work with it. An
example would be a photograph, voice recording or magazine article.
Understanding the structure of data is important as it significantly influences how it can be stored,
processed and analysed.
In conventional statistics, data is often described as either:
• qualitative data — data distinguished by non-numeric characteristics, such as eye colour or blood
type
• quantitative data — numerical data distinguished by numeric characters that represent counts or
measurements, such as age or shoe size.
These can be further categorised as shown in figure 4.1 and described in more detail below.
Data
Qualitative Quantitative
(categorical) (numerical)
Interval Ratio
Nominal Ordinal
Pdf_Folio:252
Quantitative Data
Quantitative data can be categorised into two sub-types that differentiate between the meanings of the
numerical range of values: interval data and ratio data.
• Interval data is similar to ordinal data, but with the additional property that the difference between any
two data values is measurable. There is, however, no natural zero starting point, that is, a situation where
none of the quantity is present. For example, measuring the number of years from 500 CE to 2020 CE
has a start and end point, but neither point is truly a zero point as years exist both prior and after the
selected range of intervals.
• Ratio data is interval data that includes a natural zero starting point. Zero indicates that none of the
quantity is present. For values of this type, differences and ratios are meaningful. For example, textbooks
will be sold for a range of prices, however, a textbook cannot sensibly be offered for less than $0 (i.e.
free), making $0 a logical zero point.
Quantitative data can further be described as discrete data and continuous data.
• Discrete data has a distinct set of values, which can be counted and belong to a whole numbers set
(integers). A good test to use is asking whether a fraction makes sense for a quantity of the type. For
instance, it does not make sense to record the number of cars in a parking lot as 1.52 cars, so the number
of cars in a parking lot is a discrete data type.
• Continuous data has infinite possible values. The values correspond to some continuous scale that
does not have gaps. In other words, a value belongs to a set of real numbers within a given interval. We
can have decimal places or fractions when recording measurements — for example, 1.344 metres of
wire cable.
Note that some quantities can be described by discrete or continuous data. For instance, if a person’s
age is 24 years and 1 month, this can be treated as:
• Discrete: 24 years
• Continuous: 24.083 years.
PROBABILITY
Knowledge of probability helps us to assess and communicate the degree of uncertainty in forecasting an
outcome or event.
Probability is defined as the chance or likelihood that a particular event will occur. This can be expressed
as a number ranging from 0 to 1, or as a percentage ranging from 0 to 100. When we say that an event has
a probability of 0, this means there is no chance of it occurring. When we say that an event has a 0.5 (or 50
per cent) chance of occurring, this means that there are 50 chances out of 100 of the event occurring. This
probability could also be expressed as a ‘1 in 2’ chance of occurring, or an equal likelihood of the event
occurring compared to not occurring.
The notation for ‘the probability of event A’ is P(A).
Understanding probability is an essential part of data literacy. Probability theory underpins many
important concepts in analytics. Uncertainty and randomness occur in many aspects of financial analysis.
Probability provides a way to measure this uncertainty and randomness. Various analytics, including
inferential, predictive and prescriptive analytics produce results that describe the probability of an event
occurring.
Probability, for example, allows inferences to be made regarding the risk or volatility of a particular
asset or cost, which can help to inform decision making or planning. Probability calculations can be used
in different contexts, such as:
• calculating the probability that a customer will default on a loan
• calculating the probability that a gross profit margin of 54 per cent will be reached in the next month
• calculating the probability that 15 litres of machine oil will be used by factory machinery in a day
• calculating the probability that a sales lead will result in a conversion.
Pdf_Folio:253
Description of outcomes (sample space) Examples of partition for the sample space
Unconditional Probability
Unconditional probability or ‘marginal probability’ is a probability calculation that ignores information
regarding any historical or external events.
Conditional Probability
Conditional probability is the likelihood of an event occurring based on the occurrence of another event.
The probability of event A occurring given event B occurs is expressed as P(A|B).
Total Probability
Total probability is a calculation of the probability of an event from a set of conditional probabilities.
Figure 4.2 shows a step-by-step breakdown of how to calculate total probability using conditional
probabilities.
Pdf_Folio:254
= 0.6
Source: CPA Australia 2021.
Total probability allows us to calculate the probability of an event based on the probabilities of
individual events that are related to it. This can allow us to make more accurate predictions based on com-
bining bits of information. Total probability is used in Bayes’ Theorem, which is a useful predictive method
in analytics.
Bayes’ Theorem
Bayes’ Theorem involves inferring information from probability calculations and is used in many data
analytics techniques. Bayes’ Theorem can be applied in different financial analysis contexts:
• classifying whether a transaction is fraudulent
• predicting whether a customer will default on a loan
• predicting the impact of a systematic interest rate change on the value of an asset
• predicting the impact certain events will have on a company’s net income.
To understand the underlying concept of Bayes’ Theorem, consider the following example.
• There are two tubs containing blue and orange plastic balls. In tub A the probability of picking an
orange ball at random is 3/10 = 0.3. In tub B, the probability that a randomly picked ball is orange is
6/40 = 0.15.
• If we combined the balls from both tubs and a random pick from the pool produced an orange ball, what
is the greater probability: that the ball originally came from tub A or tub B?
• There is a higher probability the orange ball came from tub B even though the probability of picking an
orange ball from tub B is less (0.15) than picking an orange ball from tub A (0.3). Six of the orange balls
were originally from tub B and only 3 from tub A, so if an orange ball is selected from the combined
tub it is more likely to be one of the original tub B balls.
• If we dumped another 1000 blue balls into tub B, the probability of picking a random orange ball from
tub B would drop to about 0.006 or 0.6 per cent. There would, however, still be a higher probability that
a randomly picked orange ball, out of the total population of balls, originally came from tub B — there
were still more orange balls in tub B than tub A.
Bayes’ Theorem uses this principle to make inferences from information. It allows us to use a number
of known conditional and unconditional probabilities (prior information) to calculate new probabilities.
Pdf_Folio:255
There is no ‘correct’ answer to the question we posed above. Different people have different interpretations
of ‘likelihood’.
Joint research between the Korean Accounting Standards Board (KASB) and Australian Accounting
Standards Board (AASB) was conducted to determine the consistency (or inconsistency) of interpreting
‘terms of likelihood’ in International Financial Reporting Standards (IFRS) between accounting profes-
sionals in the two countries. These are verbal terms that attempt to verbally describe or interpret a numerical
probability, for example in a written report (AASB 2016).
The research found interesting interpretation differences between the two countries, as well as between
professionals in the same country. Although the research mapped the most common verbal terms to a range
of numeric values, this activity demonstrates how interpretation can work in reverse: how each person tends
to subjectively evaluate the degree of uncertainty given a numeric probability.
Accounting professionals were asked to provide a range of numeric probabilities that they would assign
to each term of likelihood. The means of the minimum and maximum values of each range were calculated
and are shown in table 4.5.
Pdf_Folio:256
.......................................................................................................................................................................................
CONSIDER THIS
Compare your chosen term of likelihood with these survey results.
Figure 4.3 visualises the difference in interpretation of terms of likelihood between the professionals of
each country, and the variances within each country (Australian results are in gold, Korean in dark blue).
The research concluded that the inconsistencies could result in material differences in the evaluation of
financial statements. The key recommendation was to establish international standards that define a set
of acceptable terms of likelihood and their corresponding probability values to avoid any ambiguity in
reporting. Similar studies have also found the same issues of inconsistency.
Pdf_Folio:257
Virtually certain
Substantially all
Highly probable
Reasonably certain
Reasonably assured
Probable Korea
Likely Australia
Reasonably possible
Possible
Unlikely
Highly unlikely
Extremely unlikely
Remote
0 10 20 30 40 50 60 70 80 90 100 (%)
Source: CPA Australia 2021.
Probability in Practice
There are many financial applications for probability in decision making, such as scenario analysis,
including profitability and investment return analysis. These analytics would use a large amount of
historical data to create probability distributions. Examples 4.4 and 4.5 present simplified examples.
EXAMPLE 4.4
0 0.1
50 0.3
100 0.4
150 0.2
Pdf_Folio:258
Probability of more
Number of cakes ordered than $1000 profit Probability of making a loss
The analysis can also use probability to calculate an expected outcome, typically referred to as
‘expected value’. Expected value is the estimated value of an investment at some point in the future.
It is calculated by:
• taking the probability of each outcome and multiplying it by its corresponding return, then
• summing the products to give the expected return.
For example, we can calculate the expected value for baking 50 cakes. First, we take the probability of
each outcome and multiply by its corresponding return.
Pdf_Folio:259
Stage Description
New A newly received lead, which may include a contact form submission from
the website or an email inquiry.
Contacted A sales team member has contacted the customer, discussed their
requirements and provided some possible solutions.
Each stage is given a ‘weight’, which is a probability that the deal will close as shown in table 4.10.
The idea is that as the opportunity progresses through the various stages of the pipeline, the probability
of closing the deal increases. For example, it is more likely that a deal will be closed if the customer has
already reached price negotiations with the company, than if they have just submitted a website enquiry.
The forecasting concept is similar to expected value, which we considered earlier, as it involves
multiplying the opportunity value with the probability that the deal will close (depending on what stage
it is at). The individual opportunity forecasts are aggregated to form a total.
The probabilities for the stages would typically be created using historical data. Table 4.11 is an example
of a weighted sales pipeline with stage probabilities.
Pdf_Folio:260
Individual Weighted
Instead of assigning fixed probabilities based on the stages of the pipeline, each individual opportunity
can be given its own probability. This probability could be calculated objectively (based on historical data)
or subjectively (based on gut-feel). Table 4.12 shows a sales pipeline with individual probabilities.
Using this method, the forecast relating to the upcoming period for this pipeline is as follows.
(0.8 × $25 000) + (0.6 × $80 000) + (0.5 × $40 000) + (0.3 × $50 000) + (0.1 × $10 000) = $104 000
One important consideration for calculating event probabilities based on historical data is whether the
events are independent or not. If we have a data set that shows 100 out of 1000 new leads resulted in a
closed deal, a probability of 0.1 is valid if those 1000 leads were all independent events.
This might be accurate in a high–sales volume retail business where most transactions are one-off, but
if those 100 deals came from only 10 different clients who are providing repeat business, the 100 deals are
not independent. The data set may have to be reduced, for example, including a returning customer’s first
sale only, which would be safer to assume is an independent event.
SUMMARY
Research suggests businesses that adopt data-informed decision making perform better than businesses
that do not. The basis of data-informed decision making is knowledge created through analytics, which is
the use of skills, processes and technologies to derive information from data.
Data-informed decision making requires, not just high-quality data, but a data literate organisation with
members at all levels who are willing and able to use data to inform their work. Data literacy is the ability to
collect, evaluate and understand data to obtain insights and then to effectively communicate those insights.
Most large companies are encouraging data literacy through the use of analytics to reduce operational
and reporting costs, the use of data to inform decision making throughout the organisation, training in the
applied use of analytics and sharing data and insights with external stakeholders.
Every professional needs to develop data literacy to be able to maximise the value of the organisation’s
data and more effectively contribute to the organisation’s objectives.
Pdf_Folio:261
KEY POINTS
4.1 Evaluate the effectiveness of integrating big data and data analytics to create efficiencies and
valuable insights in solving complex business problems.
• Organisations that adopt data-informed decision making outperform those that do not.
• Descriptive, diagnostic, predictive and prescriptive analytics have numerous applications in busi-
ness, including optimising processes to create efficiencies and providing insight into complex
problems to enable informed decision making.
4.2 Appraise the impact of big data and data analytics on the work of accounting and finance
professionals to determine the organisation’s resource needs.
• Accounting and finance professionals can use analytics to understand many aspects of a
business’s operations as the basis for solving problems and identifying efficiencies by optimising
resource use.
• Accounting and finance professionals, like other members of an organisation, need to build data
literacy through internal and external development activities.
• With data literacy established, the accounting and finance professional can identify the tools and
other resources needed to produce the required analytics outputs.
Pdf_Folio:262
EXAMPLE 4.6
Pdf_Folio:263
.......................................................................................................................................................................................
CONSIDER THIS
Analytics can generate information to be used by decision-makers and information that is automatically actioned by
intelligent automation technologies applied in the business. How integrated would you expect these uses to be?
A data strategy needs to comprehensively reflect the organisation’s data needs. Here we focus on
the aspects most relevant to analytics, though many of these are also common to other applications
involving data.
Business strategy
– Strategic priorities
– Information needs
– Use cases
Implementation
– Data governance
– Systems
– Change management
Pdf_Folio:264
Data Requirements
The data needed is determined by the information needs that have been identified. Once the data needed
is known, the business can determine how to source it. This may involve structured, semi-structured or
unstructured data; internal or external data; existing data or data that must be created. The acquisition
and use of data should be cost effective and efficient, aligned to the information needs that support the
business strategy.
.......................................................................................................................................................................................
CONSIDER THIS
Do you have access to all the data you need to do your work optimally? If not, is it because the organisation does
not have it or is it because you cannot access it efficiently?
Analytics Requirements
Data is only useful in relation to how it can be analysed. There are numerous analytics options, and a
methodical approach should be adopted to ensure the investment in analytics provides the information
needed to support the business strategy.
Implementation: Systems
Technology and infrastructure are required to manage the organisation’s data. This can include database
architecture and software, analytics tools ranging from simple spreadsheet tools through to advanced ML
solutions, and visualisation and reporting platforms.
Communication
Decision-makers are often inundated with information. It is crucial the data strategy provides for clear
communication of the outputs and insights to decision makers.
Pdf_Folio:265
Framework Description
Tableau framework Tableau emphasises the integration of business and IT strategy so that key
(Tableau 2020a, b) data and analytics initiatives are linked with business goals, objectives and
the company’s mission. The framework identifies three critical steps:
• Define clear business objectives.
• Capture the right data to support these objectives.
• Modernise the data architecture to support data volume, variety
and velocity.
Tableau also provides a ‘Blueprint’ designed to help organisations
transform into data-driven organisations in terms of agility, proficiency and
community.
SAS data strategy framework SAS data strategy framework focuses on data management, identifying five
(SAS 2018) core components.
• Identify: Know what data exists and what it can be used for.
• Store: Hold data in a way that enables easy, fast and flexible access
for analysis.
• Provision: Enable data sharing within a set of rules and guidelines.
• Process: Bring data into a format ready for analysis.
• Govern: Establish, communicate and manage data and information policies.
Gartner approach The Gartner approach focuses on how data and analytics technical profes-
(Gartner 2019) sionals can create a data strategy that addresses technology, organisation,
processes and policies. There are eight steps:
1. Preliminary analysis of strategy and preliminary gap analysis.
2. Develop tentative set of data projects.
3. Develop data policies.
4. Develop data processes.
5. Choose appropriate technology.
6. Articulate human and organisational resource and capability requirements.
7. Prioritise projects to create an implementation roadmap.
8. Follow up.
It is evident then that there are commonalities running through most data strategy frameworks, while
some place a greater emphasis on business needs and others on the process and technologies of data
management. We will examine one particular framework — the SMART strategy board — in detail, but it
is important a business adopt (and then adapt) a framework suitable to its unique circumstances and needs,
driven by the objective of obtaining the information needed to make decisions that support the overall
business strategy.
Pdf_Folio:266
Purpose panel
Purpose: What is our purpose? (mission statement)
Resource panel
IT systems and data: Infrastructure: What People and talent: Culture, values,
What are the key IT are the key What are the key leadership: What are
systems and data infrastructure people and talent the key culture and
deliverables? deliverables? deliverables? leadership
(Systems, networks, (Property, machinery, (Recruit, develop, deliverables? (Values,
data sources, etc.) land, etc.) retain, engage, etc.) behaviours, etc.)
The Purpose panel relates to the organisation’s strategy and is therefore determined by the organisation’s
leadership.
The business strategy prompts the questions a business needs to answer. For example, if part of a
business’s strategy is to increase its customer base, it might need information that answers: Who are our
customers? What are the demographics of our most valuable customers?’ Stating the questions facing the
business enables it to identify the data it needs to answer those questions.
For each of the lower panels, the business should:
• engage key personnel from each area to facilitate buy-in
• identify and focus on the key questions
• refine the questions to be as clear and concise as possible
• use the key questions to guide decisions about the data needed to deliver relevant and meaningful
information.
Accounting and finance professionals are the ‘information providers’ in an organisation and are likely to
make a meaningful contribution to most panels. They are familiar with the organisation’s key information
and understand how each part of the business works together to deliver the organisation’s value proposition
and pursue its strategic objectives.
Pdf_Folio:267
Purpose
The purpose panel reflects the corporate strategy. It clearly establishes that the business strategy is what
sets the overall context and direction for the data strategy. The purpose panel details the business’s:
• Mission statement. Why the business exists. A mission statement is a clear, concise statement of purpose
setting out why the organisation exists. It should communicate the business’s intentions powerfully,
providing a road map to guide action and decision making as the organisation strives toward the strategic
goal or objective. It is primarily an internal document designed to motivate stakeholders and define the
key measures of organisational success. As such, it should include the target market, what products or
services the organisation provides to its customers and what makes those products or services valuable
to its customers.
• Vision statement. What the business aims to achieve. A vision statement also defines purpose, but
from the perspective of the business’s objectives, that is, what the business intends to be in the future.
The vision statement gives direction to internal and external stakeholders. Employees can be inspired
to action by a strong and meaningful vision statement; customers may choose a business over its
competition based on its vision statement; and shareholders can be encouraged to invest. The vision
statement gives direction about the business’s values and therefore establishes the behaviours it adheres
to and expects from its stakeholders.
Customer
The customer panel prompts the organisation to consider how much it knows about its target customers and
what it may need to find out in order to achieve its strategic objectives. There are two aspects to consider:
target market and value proposition.
In terms of target market, consider the business’s strategy (including its mission and vision).
• What is its target market?
• Does the business plan to appeal to a particular segment? If so, why and what does the organisation
know about that segment?
• Is the business targeting a particular geographic region or specific demographic? If so, what does it need
to know about those potential customers?
In terms of the business’s value proposition from the customer’s perspective, again consider the
business’s strategy.
• Why are the target customers going to engage with this business?
• Is it quality, price, innovation, service, or something else that the target customers will value?
• What will contribute to customer satisfaction and loyalty?
Consideration of the organisation’s customers in relation to its strategy will trigger questions related to
customers. These questions in turn reveal what data the organisation will need to understand its customers.
There are many customer data source options of varying quality, complexity and expense. For example:
• Sales and transaction data have numerous multifaceted uses. They are a good measure of past sales but
will often need to be combine with other data sets to explain what is observed.
• Survey research data, whether conducted in-house or through an external specialist usually yields
trustworthy data. Often the data is experimental in nature and can include research design, normative
data, mathematical modelling, stimulus controls, statistical controls, historical experience and quality-
assurance standards. Survey research is also relatively inexpensive.
• Eyeball tracking data is one of various technological advances capable of measuring what is attracting
a customer’s attention and what is not. Eyeball tracking is commonly used to understand website
engagement, but also finds use in window displays and institutions such as museums.
There are many others, including search engine data, social media data, experimental data, geospatial
data, sensor data/biometric data, video data and voice data. Example 4.7 explains one retailer’s need to
collect data to solve a business problem.
Pdf_Folio:268
.......................................................................................................................................................................................
CONSIDER THIS
Does your organisation generate or collect all the data it needs as an integrated part of its business processes or
could it need to source other data?
It is of course crucial to get the value proposition right in terms of appealing to customers, as explored
in the customer panel, but being able to deliver that value proposition relies on numerous factors, many of
which are explored in the remaining panels of the strategy board.
Finance
The finance panel prompts the organisation to consider how much it knows about the financial implications
of its strategy and what it may still need to learn.
The finance panel prompts questions such as the following.
• How does the business strategy generate money?
• What is the business model?
• What assumptions have been made about the revenue, profit and growth of the business as it implements
the strategy — and are they reasonable?
• How much will it cost to deliver the value proposition to customers — to produce and deliver the product
and services?
It is evident that all these considerations relate to the value proposition identified in the customer
panel — asking the right questions in the customer panel is crucial to asking the right questions in the
finance panel to support revenue growth, profit and shareholder returns.
Many of the sources of finance data will be second nature to accounting and finance professionals. It
is often the way that finance data is combined with other relevant data that transforms the accounting
and finance role in the era of big data and advanced analytics. The increasingly real-time nature of data
collection and analysis also has important consequences for the accounting and finance function. Finance
data includes the following.
Pdf_Folio:269
EXAMPLE 4.8
.......................................................................................................................................................................................
CONSIDER THIS
Do have access to the data you would need to conduct the analytics project described in example 4.8 for your
organisation?
Operations
The operations panel prompts consideration of what the organisation needs to do internally to deliver its
strategy and what it needs to know in order to do so. There are two components of the operations panel:
partners and core competencies.
All organisations rely on partners to deliver on their value proposition. These include suppliers,
distributors and other intermediaries. Important questions related to the organisation’s partners include
the following.
• Are the current partners the best ones?
• How healthy are the organisation’s relationships with its partners?
• How do current partners compare to other potential partners in contributing to the organisation’s strategic
objectives?
• Could other potential partners bring something new to the organisation?
Pdf_Folio:270
EXAMPLE 4.9
Once the individual elements highlighted in the customer, finance and operations panels are clear,
the organisation must consider how they interact and influence each other. The customer, finance and
operations form the core of the business and they need to work together. Accounting and finance personnel
are often best positioned to understand, analyse and communicate how the organisation’s components work
together and how one might be affected by a change in another part of the organisation. It is crucial therefore
that accounting and finance personnel are engaged with other departments and business units, regularly
interacting and communicating with business and operational personnel. This helps build a culture with
organisational processes and routines that support the business’s objectives.
Pdf_Folio:271
Pdf_Folio:272
Internal semi-
structured
Internal
unstructured
External
structured
External
Difficult/high cost unstructured
Transaction Data
Transaction data is generated every time a customer buys something. This is true online and off and it
provides a powerful insight into what was bought, where it was bought and when. The value of transaction
data can be increased by combining with other data. For example, Walmart cross-referenced transaction
data with weather data and found that when a hurricane warning was issued not only did sales for expected
Pdf_Folio:273
EXAMPLE 4.10
.......................................................................................................................................................................................
CONSIDER THIS
If customer data becomes more open and thus shared among competing businesses, how will an organisation create
a data-based competitive advantage?
Pdf_Folio:275
Differentiate structured from unstructured data, and internal from external data. Using these
classifications, identify which type of data is easiest and least expensive for an organisation to
collect and provide an example.
ANALYSING DATA
The choice of analytical approaches, development and validation of models and the various issues that
need to be considered in the data analytics process are discussed in detail in part C. An organisation’s data
strategy must of course be constructed so that it supports the types of analyses an organisation will perform
to obtain the information it requires. As explained in table 4.1:
• descriptive analytics summarise past or current performance
• diagnostic analytics determine cause and effect
• predictive analytics forecast future performance
• prescriptive analytics determine a course of action.
Descriptive and diagnostic analytics often draw on conventional statistical methods while predictive and
prescriptive analytics often require the application of AI, including ML. The cost of the technology and
other resources to support various types of analysis should be assessed against the value expected to be
generated. As is evident from CBA’s investment in data analytics technologies, described in example 4.10,
the resources required can be enormous — and so can the value created.
Pdf_Folio:277
EXAMPLE 4.11
.......................................................................................................................................................................................
CONSIDER THIS
Has your work been affected by issues around data ownership or the siloing of data with your organisation? Why
has this occurred?
Understanding the impact mismanaged data can have on business performance highlights the need to
make data ownership and data accuracy a high priority. These are part of the concept of data governance.
The other important aspect of data governance relates to complying with legal and ethical obligations.
Most countries have implemented laws and regulations that affect what data an organisation can collect
and what it can do with that data. Laws also impose obligations on organisations to ensure data is secure —
there are substantial penalties for breaching confidentiality around customer data.
As part of their data culture, organisations should also embed ethical standards of conduct. A key ethical
issue is transparency around how the organisation will use the data it collects from customers. Companies
and applications often present extensive terms and conditions to users and customers, but most people do
not read them — and even if they do read them, they do not understand them or understand the implications
of what they are agreeing to. However, establishing a (potential) legal right to use a customer’s data is
insufficient. The organisation must also act ethically and consider the reputational impacts of how it uses
data. For example, some users of Google’s free Gmail service have launched lawsuits to try to prevent
Google from analysing the content of their emails. Google’s terms and conditions give it the right to read
and analyse the content of all emails sent or received by a Gmail account. It appears to use this data mainly
to better target advertising.
Transparency around how an organisation uses data can be coupled with including that usage as part
of the value proposition offered to customers. In this way, the organisation can win customer support for
their data collection and use.
Pdf_Folio:278
Business
Data Results applications
sources and Data storage Data analysis
databases Data SCM
visualisation
Internal
data OLAP, CRM
Data marts Queries,
EIS, DDS
External Data Decision E-commerce
data warehouse support
Pdf_Folio:279
System request
Service level
agreement Planning Project plan
Support and
Analysis
maintenance Systems
development
Working Systems
system proposal
Implementation
Design
and testing
Systems
design
specification
Source: CPA Australia 2021.
Cost of poor-quality data = Lost business + Cost to prevent errors + Cost to correct errors
Pdf_Folio:280
What is the key advantage of iterative systems development approaches over linear approaches?
DATA MIGRATION
An organisation adopting a data strategy will almost certainly introduce new systems that have been
designed specifically for the organisation’s big data and advanced data analytics requirements. The
introduction of the new system will require the organisation’s existing data to be migrated from its legacy
systems to the new systems.
Data migration is a complex process. Successful data migration establishes a robust and high-quality
data set that will be the foundation for the organisation’s future data and analysis needs. On the other
hand, a poorly executed data migration program can compromise the quality of the organisation’s data
before its new strategy even begins.
Data migration will usually occur using a project methodology to ensure all required steps are planned
for and occur in a timely and cost-effective way. People within the organisation should be closely involved
in the planning for the migration to ensure disruption is minimised. A well-planned data strategy and
implementation plan will have used consultation to consider the impacts on the personnel and their needs
of the new system will have been integrated into the new system’s design.
People within the organisation will experience data migration directly as the systems and interfaces they
use to access and work with data will change. A change management program should include training and
support to help personnel realise the benefits offered by new systems.
Data migration is summarised in figure 4.9.
Test
migration
Data Final
cleaning migration
Maintenance
Security
Data
Data migration Validation New systems
mapping
Flexibility
Future proofing
Legacy Change
systems management
Consultation
The new system will use a different structure to existing systems. In some cases, it may be set up to
hold raw data so no details are lost, whereas in other cases there may be defined variables for each record,
though most big data systems can work with incomplete records where each record may include entries
only for some variables. It is important to ensure information from the old system is data mapped to the
new system.
Pdf_Folio:281
QUESTION 4.5
When data migration, from old to new systems, is taking place what can organisations do to improve
the quality of the data so that high-quality analyses can be produced?
CHANGE MANAGEMENT
An organisational culture represents shared underlying beliefs, norms, values, assumptions and expecta-
tions, which influence the typical patterns of behaviour and performance characterising an organisation.
The organisational culture strongly influences the way that things get done in an organisation. For an
organisation to create the most value from its data strategy, it needs to have a culture that values data and
analytics and relies on them for data-informed decision making.
In any major organisational transformation effort, achieving cultural change is probably the most
difficult. Change management, then, should be a major focus in the implementation of a data strategy.
Pdf_Folio:282
EXAMPLE 4.12
With the Data Strategy, we are building on our current capability and enhancing how we use data
now and into the future. We work hard to make the best use of our data, and we want to improve on
our current position, to be forward thinking, agile and proactive to deliver on our departmental vision.
The data strategy is summarised in figure 4.10, along with its alignment to other strategies.
Harness Share
Pdf_Folio:283
People IT
Plan IMT
Data strategy
Metadata
Project scoping Data asset design protocol Collecting and storing metadata
Data sharing
Data sharing and release policy Data sharing agreement Capitalise on sharing
protocol opportunities
Data hub
Stocktake existing assets Data hub Improve access
Promotion
Use existing forums Technology showcases Data showcases
Towards implementation: the roadmap
EXAMPLE 4.13
Validating Allowances
An organisation that provides weather allowances to its staff paid $90 to access current and historical
Bureau of Meteorology data for all the locations employees work in. That data was then used to build some
simple models that correlated allowance claims to the weather that actually existed in those locations
when those claims were made. This provided a layer of insight into those claims that would not have been
possible just a decade ago.
Small organisations often have an advantage over larger organisations when it comes to awareness of
what data exists in the organisation and awareness of the quality of that data. For example, banks and
insurance companies often have massive data volume and variety of data, but they often struggle to actually
understand how to access and use it. Smaller organisations often better understand the data they have and
how they can draw insights from it.
.......................................................................................................................................................................................
CONSIDER THIS
How would you describe and justify your data needs if you were contributing to formulation of a data strategy for
your organisation?
QUESTION 4.6
What types of assistance or advice could an accountant provide to improve data collection in
an organisation?
SUMMARY
Big data and analytics tools offer great potential to generate insights into almost every aspect of a business,
but businesses often fail to convert this potential into organisational performance. A data strategy is an
effective method to guide how the business will access, analyse and use data in a way that is aligned to
business objectives and so will maximise the ROI in data and analytics.
Pdf_Folio:285
KEY POINTS
4.1 Evaluate the effectiveness of integrating big data and data analytics to create efficiencies and
valuable insights in solving complex business problems.
• Big data and analytics provide almost endless potential to generate insights, but where resources
are constrained the investment should be guided by the information needed to create efficiencies
and generate the insights that are most valuable in the organisation’s decision-making.
• A data strategy guides the way data and analytics will be implemented into the organisation, with
a focus on creating efficiencies and valuable insights relevant to the organisation’s objectives.
4.3 Design the requirements for a robust data strategy including a plan for managing the
complexity of practical implementation.
• A data strategy should align with the business strategy and cover data requirements; analytics
requirements; implementation requirements including data governance, systems and change
management; communication; and interpretation and application.
• Various frameworks are available to focus a data strategy on the business’s information needs.
• Implementing a data strategy often requires significant changes to the organisation’s culture to
embed reliance on data for decision making, and to overcome data siloing and other effects of
organisational politics.
• A data strategy implementation plan should include detailed plans to develop or acquire the
necessary technology and to migrate data where necessary.
Pdf_Folio:286
BIG DATA
Big data refers to very large and complex data sets that cannot be stored or processed by traditional data
tools and methods. Big data can be generated by:
• businesses (e.g. consumer generated, financial statements, sales guidance)
• government agencies (e.g. fiscal data, economic conditions)
• financial markets (e.g. stock prices, bond yields)
• sensors (e.g. satellite imagery)
• IoT.
As we described in module 1, big data is often described in terms of the four Vs:
• volume — the large size of the data sets created by collecting and storing the data generated by devices,
transactions and interactions
• variety — the breadth and depth of data sets that arise from diverse sources
• velocity — the speed at which new data is generated and accumulated within data sets
• veracity — the variable quality of data sets given the diverse source of data collected.
The economics of data is based on the idea that value can be extracted through analytics. We should
therefore add value as a fifth ‘V’ to any description of big data. Companies that are able to use big data
astutely are able to derive better insights, improve decision-making, and potentially gain a significant
competitive advantage.
The enterprise needs to conduct the cost–benefit analysis of collecting and analysing data to determine
whether there will be sufficient ROI before undertaking a big data project initiative. Cost‐effective
mobilisation of massive‐scale data is one of the characteristics of big data. Ultimately, the value of big
data arises from the actions the business takes based on its analysis. In earlier modules, we have already
examined some of the applications that financial institutions are building based on big data to optimise
customer experience.
Conventional Analytics
Conventional data generation and capture is pre-determined and pre-designed. Specific data is required to
inform specific analyses and thus support certain types of decisions. Conventional statistical analysis often
involves using a sample of data to reach conclusions about the population from which the sample is drawn.
This process is known as statistical inference because it infers with varying degrees of confidence that
the patterns observed in the data (sample) are present in the wider population from which the data were
collected. The inference is based on a probability model and requires the sample to be selected according
to statistical principles.
Pdf_Folio:287
EXAMPLE 4.14
Implementation Challenges
The enterprise’s infrastructure, technology, and data management practices need to be transformed and
changed to accommodate the rapid shift in gathering data. The number of devices generating and
transmitting data such as sensors, tablets and smartphones continues to increase at a very fast pace,
and such information‐sharing options represent a tremendous challenge to enterprises that want to reap
the benefits of harnessing the data by using traditional data processing methods. Some of the challenges
in implementing big data include the following.
• The existing relational database approach does not work best when dealing with unstructured data that
needs to be processed faster at scale.
• The cost of implementing big data might be outside the budget for many enterprises, but special-
ist service providers are making big data and advanced analytics increasingly available even to
smaller businesses.
• The laws related to data privacy issues and compliance are not yet firmly established. Many enterprises
will maintain a conservative approach to big data adoption until there is clarity and confidence around
the intellectual property rights governing the data, ownership title, consequences and liability due to
data inaccuracy.
• Traditional data visualisation and analytics resources are not able to present the data and its insights in
a format so that it is simple and efficient for the users to realise value from big data.
• Existing information governance and management practices cannot cope with the new features of big
data. Adopting big data requires changes such as moving to the cloud and new security measures.
• Getting the right analytics skills and talent to manage and maintain big data and produce actionable
insights remains the biggest challenge for the enterprise. Accountants are accustomed to working with
structured data. One of the challenges for accountants is to extend their work to the use of semi-
structured and unstructured data.
• Many enterprises do not have the IT infrastructure to implement big data, which requires capital
investment. The maturity of the enterprise to deploy new technologies to support big data infrastructure
plays an important role in the success of the big data project.
• A shift to big data can only be effective if it is accompanied by a cultural shift to value and rely on
the insights analytics can provide. Transforming organisational culture can be a difficult and time-
consuming activity, but it is crucial to successful big data implementation.
Pdf_Folio:290
Developing Skills
Working with constantly changing topics, data sizes and data complexity forces deeper consideration of
analysis and aids development of additional skills. Within the finance function, analysts develop necessary
skills through practice, diligence, and dedication. The following approaches can help develop the skills
required to work with big data.
• Grasp unfamiliar data. Create a systematic approach to exploring the data in order to understand what
information it could reveal.
• Invest sufficient time. Speeding through the analysis will lead to mistakes and incorrect conclusions.
Allow enough time to thoroughly understand the data. Ensure the statements included in reports and
visualisations are supported by the data.
Pdf_Folio:291
Pdf_Folio:292
Data Sources
The next key consideration is the data source, which may be internal or external. Internal data is the data
that the business currently holds. This includes private or proprietary data that is collected and owned or
controlled by the business. In addition to the internal data an organisation already holds, the organisation
can capture data not currently captured or generate new data to meet the data needs that it identifies.
External data refers to the information that exists outside a business. External data are either public or
private. Public data are data that anyone can obtain by collecting it for free, paying a third party for it or
getting a third party to collect it. Private data would usually be sourced from another business or third-party
data supplier.
QUESTION 4.7
Explain why organisations may choose to anonymise customer data. Provide an example of a
business application where anonymising data is not viable.
Pdf_Folio:296
Explain why overfitting data leads to poor predictive performance from a statistical model.
DATA QUALITY
The quality of a supervised learning model depends on the accuracy of the classifications used in the raw
data. In some cases, the value of the target variable is known because it is an inherent component of the
data. Web logs will show whether a person clicked on a link or didn’t click. Bank records will show whether
a loan was paid on time or not. However, in other cases, the value of the known target must be supplied by
a human labelling process to accumulate enough data to train a model.
Email must be labelled as spam or legitimate, documents in legal discovery must be labelled as relevant
or irrelevant. In either case, the data mining algorithm can be led astray if classifications have been
inconsistently applied.
OUTLIERS
The greater the volume of data, the greater the chance of encountering erroneous values resulting from
measurement error, data-entry error, or the like. If the erroneous value is in the same range as the rest of
the data, it may be harmless. If it is well outside the range of the rest of the data (e.g. a misplaced decimal),
it may have a substantial effect on some data mining procedures.
The term ‘far away’ is deliberately left vague because what is or is not called an outlier is basically
an arbitrary decision. Analysts use rules of thumb such as ‘anything over three standard deviations away
from the mean is an outlier’, but no statistical rule can determine whether such an outlier is the result of
an error. In this sense, an outlier is not necessarily an invalid data point, it is just a distant one.
The purpose of identifying outliers is usually to call attention to values that need further review.
Examining the data may reveal an explanation for the outlier data. In the case of a misplaced decimal,
this is likely. Alternatively, there might be no obvious explanation but the value is clearly wrong; or the
value may be within the realm of possibility and should be left alone. All these are judgements best made by
someone with domain knowledge, that is, knowledge of the issue being considered. Statistical procedures
can do little beyond identifying the record as something that needs review.
Pdf_Folio:297
MISSING VALUES
Typically, some records will contain missing values. If the number of records with missing values is small,
those records might be omitted. However, if there is a large number of variables even a small proportion
of missing values can affect a lot of records. Even with only 30 variables, if only 5 per cent of the values
are missing (spread randomly and independently among cases and variables), almost 80 per cent of the
records would have to be omitted from the analysis, assuming all the variables are used in the model. (The
chance that a given record would escape having a missing value is 0.9530 = 0.215.)
An alternative to omitting records with missing values is to replace the missing value with an imputed
value, based on the other values for that variable across all records. For example, if among 30 variables,
household income is missing for a particular record, one option is to substitute the mean household income
across all records. Doing so does not, of course, add any information about how household income affects
the outcome variable. It merely allows the analysis to proceed without losing the information contained in
this record for the other 29 variables.
Some data sets contain variables that have a very large number of missing values. In other words, a
measurement is missing for a large number of records. In that case, dropping records with missing values
will lead to a large loss of data. Imputing the missing values might also be useless, as the imputations are
based on a small number of existing records. An alternative is to examine the importance of the predictor. If
it is not very crucial, it can be dropped. If it is important, perhaps a proxy variable with fewer missing values
can be used instead. When such a predictor is deemed central, the best solution is to invest in obtaining
the missing data from another source.
Significant time may be required to deal with missing data, as not all situations are susceptible to
automated solutions. In a messy data set, for example, a ‘0’ might mean two things: (1) the value is missing,
or (2) the value is actually zero. In the credit industry, a ‘0’ in the ‘past due’ variable might mean a customer
who is fully paid up, or a customer with no credit history at all — two very different situations. Human
judgement may be required for individual cases or to determine a special rule to deal with the situation.
Pdf_Folio:298
Pdf_Folio:299
Data Wrangling
Data wrangling is the process of putting data and data sets into a useable format, ensuring that missing,
inconsistent or anomalous values are detected and rectified to prevent corruption of the analytics results.
Research has shown that data wrangling can take up to 80 per cent of a data analyst’s time.
ML techniques can potentially be applied to data wrangling to:
• automate tasks such as detecting and removing outliers that skew the results
• remove duplicate records
• standardise value formats
• validate the accuracy of data
• generate new data from the existing clean data.
Real-Time Monitoring
Due to the volume and velocity of big data, analytic techniques such as business dashboards are used to
create KPIs and metrics that provide an overview of performance. These dashboards lack granular detail
that might be important to detect and alert on significant events.
Using AI, continuous, real-time monitoring of events can occur. For example, there may be a pricing
glitch for a short time that causes a product in an online store to be sold for $5 instead of $500. If
word quickly spread online and a short spike in sales occurred, this may be missed by a sales dashboard
showing summarised metrics for the store over the period of a day. AI can quickly detect and highlight
anomalous events such as these and alert stakeholders in a timely manner.
Automating Action
AI can automate actions based on the results of analytic processes. This could include generating
notifications or a report that highlights key insights to assist human decision making.
EXAMPLE 4.15
RISKS
The use of ML to automate and optimise analytics processes carries various risks that must be considered
when evaluating the suitability of the solution for any particular business problem.
The key risk is the generation of inaccurate or biased predictions that may cause financial loss or
reputation damage to an organisation. Issues around data quality and model transparency tend to be the
underlying causes of these risks.
Pdf_Folio:301
QUESTION 4.9
Assess the pros and cons of black-box artificial intelligence and machine learning models.
MACHINE LEARNING
Module 3 introduced basic ML concepts and we have touched on some more of these concepts and
applications earlier in this module. In essence, the ML process involves employing algorithms to
automatically identify patterns in data, then refining the algorithms with more and more data to improve
their accuracy. ML algorithms used in data analytics include:
• supervised learning
• unsupervised learning
• semi-supervised learning.
Supervised learning can be thought of as finding known patterns in unknown data. We must have data
that has been ‘labelled’. The labelled data are examples, where the desired answer or outcome of interest
(e.g. purchase or no purchase) is attached to the input data. This labelling is considered the supervision.
The labels have been derived from human recognition. Training data are the data from which the
classification or prediction algorithm ‘learns’, or is ‘trained’, about the relationship between predictor
variables and the outcome variable.
Once the algorithm has learned from the training data, it is then applied to another sample of data (the
validation data) where the outcome is known, to see how well it does in comparison to other models. If
many different models are being tested, it is prudent to save a third sample, which also includes known
outcomes (the test data) to use with the model finally selected to predict how well it will do. The model
can then be used to classify or predict the outcome of interest in new cases where the outcome is unknown.
Supervised learning is conceptualised in figure 4.11.
In predictive analytics, the aim is to build an analytical model predicting a target measure of interest
(Hastie, Tibshirani & Friedman 2011). The target is then typically used to steer the learning process during
an optimisation procedure. Two types of predictive analytics can be distinguished.
• Regression. In linear regression, the target variable is continuous, Common examples are predicting
customer lifetime value, sales, stock prices, or loss given default.
• Classification. In classification, the target is categorical. Common examples are predicting churn,
response, fraud and credit default.
Pdf_Folio:302
Model
learning data
Algorithm Verification
Data learning
outcome
Output
Source: CPA Australia 2021.
Unsupervised learning can be thought of as finding unknown patterns in known data. It does not use
labelled data. It attempts to find unknown structure or patterns in the data on its own. The process can look
for similarities or anomalies in the data. This approach is useful for finding features in the data that can be
used to perform categorisation of new data. Unsupervised learning holds perhaps the greatest promise for
AI. Unlike supervised learning, the unsupervised learning algorithm has no information about the data it
needs to process. It proceeds by grouping together (clustering) and gathering similar information. This is
a very effective technique when the analyst has no (or little) idea of what is contained in the information
(contrary to supervised learning, where the analyst has an idea of the expected result), which can reveal
something that would not naturally come to mind.
Supervised, unsupervised and semi-supervised learning approaches are used by several ML techniques.
Supervised and unsupervised methods are sometimes used in conjunction. For example, unsupervised
clustering methods are used to separate loan applicants into several risk-level groups. Then supervised
algorithms are applied separately to each risk-level group for predicting loan default propensity.
We will discuss supervised and unsupervised ML techniques in the following sections. We will also
cover two useful algorithms of ML: decision trees and support vector machines (SVMs).
Classification Regression
Basket 1
Basket 2
Basket 3
Basket 4
Pdf_Folio:305
Income
Low High
Medium High
risk risk
Class 2
Support
vector
Support
vector
Maximum margin
Class 1
Pdf_Folio:307
.......................................................................................................................................................................................
CONSIDER THIS
Can you think of applications for ML analytics that would provide valuable information to you that you cannot
currently obtain?
Pdf_Folio:308
EXAMPLE 4.17
Pdf_Folio:309
Very high
Country of domicile
High
Machine Medium
Main source of wealth
learning
classification
Low
Industry
Very low
Years as customer
Pdf_Folio:310
Factor 1
Factor 2
Factor 3
Factor 4
Factor 5
Factor 6
Factor 7
Factor 8
Pdf_Folio:311
Oversampling
Grow
Undersampling
Shrink
Villalobos and Silva (2017) tested both oversampling and undersampling when training the classification
algorithms during their research.
Choosing the Best Algorithm
Each ML method — for example, regression, support vector machines, decision trees — has its strengths
and weaknesses. Not all methods perform equally for a specific problem. The approach that has been
found to be the most successful is comparing methods for a given problem to select a suitable one
(Kovalerchuk & Vityaev 2005). Villalobos and Silva (2017) used the above approach, trying several different
ML algorithms and comparing the results.
Table 4.14 shows the results of the different ML algorithms used to classify suspicious transactions.
(Oversampling) (Undersampling)
% correct (Oversampling) % correct (Undersampling)
classification AUC classification AUC
Pdf_Folio:312
QUESTION 4.10
How could detecting fraudulent transactions be improved by combining digital technologies with
human judgement?
SUMMARY
The value of data arises from the actions the business takes based on its analysis. Analytics brings together
data from diverse, but relevant sources, and analyses it to generate insights that decision-makers can use
to solve business problems and achieve the business’s objectives.
Accounting and finance professionals need to develop sufficient skills in big data, analytics and data
mining to enable them to work with data scientists to plan and implement analytics technologies and
models that will produce valuable information for the organisation. They also need to be able to assess the
quality of data to be used in analytics and understand how the data and the techniques used may influence
the output. Data interpretation, discussed in the next part, must always been done with knowledge of the
context in which the analysis was performed.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
4.1 Evaluate the effectiveness of integrating big data and data analytics to create efficiencies and
valuable insights in solving complex business problems.
• The economics of data is based on the idea that value can be extracted through analytics.
• Companies that are able to use big data astutely are able to derive better insights, improve decision-
making, and potentially gain a significant competitive advantage.
• The enterprise needs to conduct the cost–benefit analysis of collecting and analysing data.
Ultimately, the value of big data arises from the actions the business takes based on its analysis.
4.2 Appraise the impact of big data and data analytics on the work of accounting and finance
professionals to determine the organisation’s resource needs.
• As businesses adopt big data, there will be increasing demand for professionals who can act as
interpreter between data scientist and decision maker.
• Finance professionals will require the skills to manage the complex data (both financial and non-
financial) to provide insights to the wider business.
Pdf_Folio:313
Pdf_Folio:314
Wisdom
Knowledge
Information
Data
The DIKW pyramid provides us with a conceptual model to show how data must be processed and
synthesised with existing knowledge to provide real value to decision makers.
Conceptually, analytics can be viewed as the means to transform data to information, and arguably
transforming the information most of the way to knowledge in the model of the DIKW pyramid.
Interpretation is what happens for us to arrive at knowledge. It is the synthesis of facts and ideas in the
human mind to form new ideas.
DOMAIN KNOWLEDGE
Domain knowledge is theoretical and practical knowledge that is specific to a business field, industry,
or discipline. It may be contrasted with general knowledge that can be applied to different domains. An
example would be a software developer (general knowledge) who has experience in the banking sector
and expertise in developing electronic payment systems infrastructure (domain knowledge).
It would seem desirable, or even crucial, for data scientists to possess domain-specific knowledge in
order to create useful and accurate analytics to address complex business problems. Although having
individuals with both data science and business domain expertise is the ideal, academics and practitioners
have accepted this is challenging and unrealistic to expect.
Data scientists typically do not have the domain knowledge, or a tendency to appreciate the business
problems they are expected to solve. Often it is easier for someone with existing domain knowledge and
expertise to develop their data literacy to a high enough level that they can act as a ‘data translator’ between
data scientists and decision makers. This role is important at all stages of the analytics process, including
the interpretation of data analytics results. It is the application of the domain-specific knowledge of the
analysts and decision-makers who consume the results that creates and leverages the insights.
.......................................................................................................................................................................................
CONSIDER THIS
Assume your organisation plans to invest in more analytics capabilities to help solve the complex business problems
it faces. Think of some specific business problems that would require your domain knowledge and thus your
involvement in designing and interpreting the analytics approach. What problems might arise if you (or others with
your domain knowledge) are not involved?
Do you have sufficient data literacy skills to collaborate effectively with data scientists? Identify the skills you need
to develop further.
QUESTION 4.11
Assess the potential implications of separating management of data from those who use the data.
Pdf_Folio:316
DATA VISUALISATION
Data visualisation plays two key roles in analytics:
• it aids interpretation of the data and analytics outputs during and after the analysis process
• it aids communication of findings and recommendations during the reporting process.
Below, we will briefly outline the role of visualisation in data interpretation. Its use in reporting will be
described in the final part of this module.
In addition to statistical results, such as probability values, analytical techniques often use data
visualisation to assist exploring the data during the analysis process. Charts, such as histograms, probability
distributions, boxplots, scatterplots, and decision trees, are some of the visualisations that are typically
produced by analytics tools.
An understanding of the different types of data visualisations is a key component of data literacy, as it
relates to the abilities to read and analyse data. Some of the simple, but useful, visualisation approaches to
aid interpretation include visualisations to show the following.
• Magnitude comparison. Comparing the size or other measures of data, of items within a data set
or between different data sets. The differences might be relative or absolute. The data points are a
snapshot, a picture at a moment in time or of an aggregation of time, such as over a year. Special
cases of comparison include deviation (the amount of variation from a fixed reference point), part-to-
whole composition relationships, or distributions (where there are areas of greater compared to lower
magnitude).
• Change over time. Presenting trends or how measurements of the data vary over a defined time period,
the differences can be relative or absolute.
• Correlation. Showing relationships or an association between different variables, data points or
data sets.
Confirmation Bias
A tendency to seek out information that confirms the beliefs you already have and ignore the information
that contradicts those beliefs
Example
An investor hears a rumour that a company is about to declare bankruptcy and considers selling their
shares. They tend to believe news that supports this and ignore news that contradicts it.
Anchoring
A tendency to stick to a psychological benchmark, which could be the first value, figure, solution, theory
or argument that comes to mind
Example
A seller of a property frames all price negotiations against the psychological benchmark of their purchase
price for the property.
Herd Behaviour
A tendency to follow the actions of a larger group (those actions could be rational or irrational) even though
most would not make the same choice individually
Example
Although analysis indicates an asset is above its intrinsic value, a trader buys into a speculative bubble
through fear of missing out on quick gains.
Optimism Bias
A tendency to overestimate the likelihood of positive events, thinking that you are at a lower risk of a
negative event than others
Example
Due to the positive outlook of the national economy, a business expands its operations into another state
despite analysis indicating significant financial risk.
Extrapolation Bias
A tendency to assume that recent experience or trends will continue into the future based on simplistic
analysis, although many variables are involved
Example
A company that grew its earnings by 6 per cent per year over four years projected their growth based on
these numbers far into the future, but the effect of other variables made this forecast inaccurate.
Availability Bias
A tendency to rely on the most recent or easy to recall information because it is readily available, rather
than collecting new data
Example
There was a market crash during the middle of a year, causing lingering negative memories of the event
on a massive scale. Even though the market recovered, and the annual return was still 12 per cent for that
year, a few years later the perception was that the market was flat or down in that year. This was due to
availability bias of more recent negative news than positive news.
Source: CPA Australia 2021.
CORRELATION
Correlation is the degree of relationship between two variables. A positive correlation exists when the
value of one variable moves in a particular direction and the value of the other variable moves in the same
direction. A negative correlation exists when the value of one variable moves in a particular direction and
the value of the other variable moves in the opposite direction.
The strength of correlation is measured by a value known as the correlation coefficient, r. This is also
known as Pearson’s correlation coefficient.
The correlation coefficient r has the following range:
• r = 1.0 indicates a perfect positive correlation between two variables
• r = –1.0 indicates a perfect negative correlation between two variables
• r = 0 indicates that no correlation exists between the variables, either positive or negative.
The correlation is said to be high or strong where r is closer to its extremes at plus or minus one. The
correlation is said to be low or weak if r is closer or equal to zero. Figure 4.22 plots data sets that exhibit
a variety of correlations.
Figure 4.23 describes how to calculate the correlation coefficient, r, for the variables x and y.
a × b (call it ‘ab’)
a squared (call it ‘a2’)
b squared (call it ‘b2’)
Pdf_Folio:319
EXAMPLE 4.18
The calculated correlation value is 0.916 rounded to three decimal places. Our result of r = 0.916
demonstrates a high positive correlation between professional experience and income for this particular
data set. This is also evident from the shape of the scatterplot of the data shown in figure 4.25.
Pdf_Folio:320
140 000
120 000
100 000
Income ($)
80 000
60 000
40 000
20 000
0
0 2 4 6 8 10 12 14
Experience (years)
Source: CPA Australia 2021.
CAUSATION
Causation is a cause-and-effect relationship between two variables. One variable is causing changes in
another. A correlation between two variables does not necessarily mean that a change in one is causing
the change in the other; there may be no causation relationship between them.
An example is shown in figure 4.26, examining the relationship between ice cream sales and shark
attacks in a fictional seaside region. Although there appears to be a high correlation between the two
variables, is it plausible that ice cream sales cause shark attacks?
30
25
20
15 Shark attacks
Ice cream sales
10
0
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May
Source: CPA Australia 2021.
Pdf_Folio:321
Proving Causation
Proving causation can be a time-consuming, complicated and expensive process involving a great deal
of research and experimentation, so an evaluation should be made as to whether proving causation is
important or not.
It might be significant if the causal relationship is integral to decision-making. But, with big data
predictive analytics and ML, it may be the case that all an organisation is interested in is the fact that
a correlation exists and they can predict an outcome (they may not have any interest in ‘why’ that
outcome occurs).
An instance where it might not be important to prove causation could be in a home insurance predictive
model, where a correlation between the average rainfall in a city and the average claim per dwelling per
year exists. Just knowing the correlation exists allows the company to predict risk, even if causality cannot
be proven.
An example where it might be important to prove causation could be an online retailer’s coupon discount
percentage correlating with sales volumes for a product range. Decision-makers would likely want to know
if there were other variables that led to the sales volume patterns in order to develop promotional strategies.
Numerous patterns exist that explain causal relationships between variables. The relationship between
causes and effects can vary in complexity. Figure 4.27 summarises patterns of causation that are commonly
used in business analytics.
A causal relationship may be non-existent. As with proving the existence of a causal relationship,
proving the non-existence of a causal relationship with 100 per cent certainty is logically impossible,
as there could always be numerous unknown or hidden confounding factors.
Proving causation involves isolating the causal relationships or factors through conducting experiments
and asserting that the results meet certain criteria.
Various approaches that can be used to assess the existence of causal relationships between variables
are described below.
Random Experiments
Random experiments, or randomised controlled trials, are often difficult to conduct and do not occur often
outside academic research. They rely on the researcher’s ability to control (randomise) the independent
variable or treatment, which is often not possible in business where variables can involve financial markets,
consumer behaviour and other complex, external factors.
Random experiments involve assigning a test subject randomly to either a treatment group (affected by
the hypothesised causal factor) or a control group (not affected by the hypothesised causal factor).
Where treatments can be controlled by the researcher, such as presenting different options or features
to visitors for an online e-commerce store, A/B testing can be used.
A/B Testing
In A/B testing, each experiment subject is randomly assigned to one of the variant groups to compare
which is more effective. Although it is called A/B testing, more than two variants can be used.
For example, a company may want to test a new design for the landing page of its website. It could do so
by conducting an A/B test. A website visitor may be randomly assigned to group A (the existing landing
page) or group B (the new landing page). The experiment could be conducted to determine whether the
new design results in more clicks (higher click-through rate) on the ‘contact us now’ button.
Pdf_Folio:322
A B A B
A B
A B
Mediators
(A causes M, which then causes B.)
A M B
Observational Studies
When it is not possible to control the independent variable or randomly assign subjects to treatment groups,
an observational study can be conducted. In an observational study, the subjects are only observed
passively. There is no interference or attempt to influence outcomes by the researcher. The aim of the
study is to measure variables of interest and identify patterns or relationships between them.
For example, in research to study a correlation between a certain set of professional qualifications and
income, it would not be possible to randomly assign subjects or control the qualifications, so observational
study techniques would have to be used.
Because there are numerous characteristics or possible confounding variables that, like the main
variables of interest, are not controllable, they must be accounted for in the study. To try and isolate these
potential confounding variables, two common methods exist: matching and multiple regression.
Matching involves selecting subjects with similar characteristics other than the independent variable.
In the qualifications versus income example, the study might match pairs of subjects with the same gender,
residential location, occupation, ethnicity, years of professional experience and so on.
Of each pair selected, one would have the professional qualification and the other would not. This,
however, relies on selecting appropriate characteristics to match and it is not realistic to account for all
possibilities.
Multiple regression is used to account for the effect of confounding variables on the dependent variable
of interest. The effect of confounders must be modelled and taken into account in the outcome.
In the example, we could model how years of professional experience affects income and account for
the effect on each subject. This would be done in a similar manner for the other characteristics.
Pdf_Folio:323
QUESTION 4.12
Differentiate causation and correlation. Which can be demonstrated visually through a scatterplot?
Evaluate explanatory
ry
dec nicate anato
bu
outcome.
ke
exp
As ess q
sin
xp ecific on.
n.
Co icate
isio
lan
p
u
ato
mm
un
ry
mm
ti
Co
Int make
ta. -qu ct
and
ty
exp ret t deci
y
rel dent plan lect
De ato
da ood olle
an a ator
ali
erp
lan he re sion
cid ry ults
c
ex Col
t, g nd
e
ev ify
a
Analyse
s
explanatory
I
Conduct an analysis
.
Pdf_Folio:325
e
erv
bs
Or
ien
t
Ac
t
ide
Dec
The OODA loop is a military decision-making framework that has been adapted for use in the business
world. The OODA loop was developed by the United States Air Force to train its pilots to decide and act
more rapidly than their adversaries, giving them the advantage in air-to-air combat.
The emphasis on the speed of the cycle enables agility. When applied to business, the model provides an
organisation with a competitive advantage if it can execute the loop faster than its competitors. Execution
requires the transformation of data, to knowledge, to action in a timely manner (Bumblauskas et al. 2016).
.......................................................................................................................................................................................
CONSIDER THIS
Do you use a formal decision-making framework to ensure you approach decision-making consistently? If not, what
might be the risks of not using a framework?
ITERATIVE APPROACHES
Regardless of the specific data-informed decision-making framework used, an iterative approach should
be taken. Information that is valid now will not necessarily be valid in future, whether that is tomorrow
or in a year. It is important to monitor data inputs as well as outputs and make adjustments as necessary.
For example, changes in the internal or external environment may affect the relevance of KPIs over time
(Bumblauskas et al. 2016).
Another factor that requires the iterative approach, is that often the questions that we seek to answer
through analysis need to be refined over time. Renowned mathematician John Tukey in his paper ‘The
Future of Data Analysis’ noted that while precise answers to oversimplified problems can provide insight
and guidance, they are not necessarily the best answer to a particular data analysis problem (Tukey 1962).
Although we think we start with the right question, most often that isn’t the case. An iterative approach is
required, where we try different processes and along the way we form an idea of what the data can tell us,
but also what other questions we can ask and what questions we want answered (Peng 2019).
Pdf_Folio:326
EXAMPLE 4.19
Amazon
Amazon is a company that drives the use of data-informed decision-making within its organisation. The
company’s founding CEO, Jeff Bezos, classified decisions as ‘Type 1’, which are non-reversible decisions,
or ‘Type 2’, which are reversible decisions.
Type 1 decisions are given more resourcing than Type 2.
The company has a principle of commencing decision making once there is a ‘preponderance of
evidence’, which is about 70 per cent. This is because there can be a course correction later on.
.......................................................................................................................................................................................
CONSIDER THIS
What do you think about Amazon’s decision-making principles?
Once we establish that we have enough knowledge to inform a decision, we can form a conclusion. The
conclusions drawn from our analysis will feed into the next step, hopefully facilitating informed decisions
and driving business strategy forward.
We need to realise that these findings may not be understood or may even be ignored if they are
not presented effectively to the decision makers. Data analysts must become skilled in the art of data
storytelling and visualisation in order to communicate their findings with key stakeholders as effectively
as possible. Part E looks at this aspect in detail.
Decision Types
According to decision-making theory, not all decisions are weighted equally in importance. Some can
be considered routine and low impact, and at the opposite end of the spectrum, others are one-off and
potentially high impact. The latter types of decisions can influence hundreds or thousands of stakeholders’
lives, or affect the fate of the company itself. Table 4.15 illustrates the spectrum of decision weights.
A typical approach to programmed decision making involves establishing guidelines or policies that
allow managers to make streamlined, timely and consistent decisions. The routine nature of these decisions
provides accumulated data on past outcomes. The rules are developed based on those successful outcomes.
Examples may include disciplinary matters or technical support issues.
Pdf_Folio:327
Decision type
A non-programmed decision making scenario involves a novel situation, where no prior experience or
developed strategy exists. A unique solution must be developed, taking into account multiple variables and
possibly unknown information. An example of this scenario could be: what should the marketing strategy
for our new product be?
Many decision-making models take these concepts into account, balancing efficiency and collaboration
against decision quality and importance.
Data-informed decision making may be employed to guide decisions over the entire spectrum.
Decision-Making Strategies
We will briefly look at some steps to assist non-programmed decision making. All decisions come with
a certain degree of risk. This means that, when making decisions, the aim is to select the best possible
choice available, though it may not be the perfect choice. Hence, it is important that we have strategies and
processes to assist with evaluating and mitigating the risks of the decision.
The process guides the decision maker through the steps of setting objectives, exploring and prioritising
alternative choices, assessing the strengths and weaknesses of the top alternatives, and choosing the ‘best’
alternative. It assists the decision maker to come up with mitigations for foreseeable consequences of the
decision.
The detailed problem and risk analysis is designed to remove cognitive bias from the decision-
making process.
The decision-making strategies consists of four steps.
1. Evaluate the situation. Identify the concerns, outline priorities and plan involvement.
2. Analyse the problem. Describe the exact problem or issue by identifying and evaluating the causes.
3. Analyse the decision. Identify and evaluate alternatives by performing a risk analysis for each and then
make a final decision.
4. Analyse potential problem. Evaluate the final decision for risk and identify the contingencies and
preventive actions necessary to minimise that risk.
Once a decision has been made, communicated to the organisation and implemented, monitoring and
evaluation can take place. The data-informed decision-making model is iterative, creating a continuous
loop of decision-making, evaluation and refinement. Improvements can be made to chosen strategies as
new data comes in or as the organisation grows in analytical maturity. Our existing analysis questions may
also require adjustment, or the existing hypotheses may have to be re-written.
Pdf_Folio:328
QUESTION 4.13
Outline the digital skills professional accountants will need to work in a multi-disciplinary team
using big data to provide business insights to organisations.
SUMMARY
Data interpretation refers to the process of extracting insights from analytics outputs. The DIKW pyramid
is a useful way to conceptualise the difference between data, information, knowledge and wisdom and thus
how analytics takes us from data to decision-making.
Data interpretation requires domain knowledge. This refers to specialist knowledge in a specific field.
While data scientists will usually perform analytical procedures, they will usually not have sufficient
knowledge of the business problems or questions to be addressed. Thus, the accounting and finance
professional brings their domain knowledge to the process.
In interpreting data, it is important to remember key statistical and decision-making concepts such as
overcoming cognitive bias and not confusing correlation with causation.
The use of a decision-making framework helps create a structured and consistent approach to decision-
making. Each organisation should adopt a framework that suits its needs. One of the key decisions to be
made in formulating a framework is how much information a decision process needs.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
4.2 Appraise the impact of big data and data analytics on the work of accounting and finance
professionals to determine the organisation’s resource needs.
• Accounting and finance professionals have domain knowledge in some of the most important
aspects of business and thus should collaborate with data scientists and others in related analytics
work.
4.6 Analyse data analytics results to assess the ongoing suitability of models used and improve
business plans and processes.
• Part of data interpretation is to assess whether the insights gained are adequate to resolve the
business problem. Analytics is an iterative process — data and models should be refined to ensure
outputs meet the business’s needs.
Pdf_Folio:329
EXAMPLE 4.20
Anscombe’s Quartet
The four data sets in table 4.16 have the same variance in x, variance in y, mean of x, mean of y and linear
regression.
I II III IV
X Y X Y X Y X Y
Pdf_Folio:330
If we plot the linear regression for each data set, we obtain figure 4.30.
I II
12.5
10.0
7.5
5.0
III IV
y
12.5
10.0
7.5
5.0
5 10 15 5 10 15
x
Source: CPA Australia 2021.
Pdf_Folio:331
FIGURE 4.31 Data points and linear regression for the four data sets
I II
12.5
10.0
7.5
5.0
III IV
y
12.5
10.0
7.5
5.0
5 10 15 5 10 15
x
Source: CPA Australia 2021.
It was impossible to identify patterns in the data sets in example 4.20. These data sets had only 11 rows.
Consider big data sets with billions of rows of potentially unstructured data. Visualisations are a crucial
part of interpreting the outputs and will often be used during the analysis process to shape the analytical
procedures undertaken, both with conventional analytics and AI-enabled analytics.
QUESTION 4.14
STEP 1: STORYTELLING
Storytelling is a powerful communication technique because our brains are wired to receive, understand
and store information in that particular form. People remember stories, not statistics. Storytelling leverages
cognitive processes that result in a higher level of engagement and increased recall.
As organisations are adopting data-informed decision making, not all their members are able to
comprehend the analysis results or grasp the complexities of the techniques involved. There may be a
lack of confidence or even scepticism regarding the insights that are derived. This can create barriers to
developing a culture of data usage and data-driven transparency or may mean that members are not inspired
to act on the knowledge that analytics uncovers.
Storytelling using data can provoke conversations and encourage deeper analysis of the insights. It
transforms the results into a format that is accessible, allowing the storyteller to reach a wider audience of
both internal and external stakeholders. Weaving the data into a narrative helps the professional persuade
others to act on the insights.
Technique Applications
Pdf_Folio:335
130
Revenue
125
Indexed growth (%)
120
100
135
130
125
120
Indexed growth (%)
115
110
105
100
95
90
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Operating profit margin
Revenue
Source: CPA Australia 2021.
Comparing Magnitudes
A comparison of magnitude could be the difference between items within a single data set or the differences
between several sets of data. Many common charts types are useful for comparing magnitudes. There is
often no one correct choice, as there are multiple visuals that can work effectively.
Pdf_Folio:336
7000
6000
5000
4000
3000
2000
1000
0
Apples Oranges Pears Bananas Grapes
60
50
40
30
20
10
0
0–20 21–40 41–60 61–80 81–100
(continued)
Pdf_Folio:337
components within
each measure. 6 00 000
5 00 000
4 00 000
3 00 000
2 00 000
1 00 000
0
Central East South West
Furniture Office supplies Technology
40 000
20 000
0
North West South East
0
Units in stock Damaged Refurbished Saleable units
Pdf_Folio:338
5 000 20%
0 0%
rs
ry
es
r
re
ea
ea
e
ie
ag
ic
ag
ke
le
th
nc
pa
or
et
w
el
db
gg
O
ic
m
ra
ot
ss
er
Ap
w
St
an
Lu
os
ag
nd
Fo
Je
ce
C
Fr
U
Ac
Profit Cumulative total
(continued)
Pdf_Folio:339
Kansas
Pdf_Folio:340
Applications
Operations Software
Infra-structure
Networking Costs
Hardware
Support
Services
Voice Personnel
Operations
Operations
Administration
Development
Development Operations
respect to a certain
measure. M100 5346 1785 9707 5476 2664
Product model
A B C D E
Store
0 9737
10
–5
Sample 1 Sample 2 Sample 3
(continued)
Pdf_Folio:341
10 000
8 000
6 000
4 000
2 000
400
200
All of the charts in table 4.18 have useful applications. There are, however, some other visual displays
that are best avoided as they are confusing or otherwise difficult to interpret. These include pyramid charts,
circular charts and charts with three-dimensional effects.
Pdf_Folio:342
14 000
12 000
10 000
8 000
6 000
4 000
2 000
0
2011 2012 2013 2014 2015 2016
1.4
1.2
1.0
0.8
0.6
0.4
0.2
Jun 03 Jun 10 Jun 17
2160 Printing
1825 Software
Software 1504
1406 Couriers
Stationery 1285
Couriers 1254
Hardware 1152 1105 Stationery
956 Packaging
Packaging 804 820 Hardware
725 Postage
Postage 672
(continued)
Pdf_Folio:343
7
6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Time (months)
Fixed income Property Other Shares
Pdf_Folio:344
Design
Development
Review
8
8
8
8
8
8
8
8
8
8
1 8
1- 8
18
-0 01
-0 01
-0 01
-0 01
-0 01
-0 01
-0 01
-0 01
-0 01
-0 01
-0 01
-0 01
20
07 1-2
09 1-2
11 1-2
13 1-2
15 1-2
17 1-2
19 1-2
21 1-2
23 1-2
25 1-2
27 1-2
29 -2
-0
05
Source: CPA Australia 2021.
Pdf_Folio:345
60K
55K
50K
45K
Charges
40K
35K
30K
25K
20K
15K
10K
16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54
BMI
Source: CPA Australia 2021.
60 000
50 000
40 000
Total costs
30 000
20 000
10 000
0
0 10 000 20 000 30 000 40 000 50 000 60 000
Total revenue
While scatter plots and bubble charts are appropriate for visualising correlation between individual data
points, we need a different visual display for showing magnitude comparison or change-over-time with
more than two variables. This is because we need to show the ‘in-between’ or interpolated data. This can
involve 3D charts or 2D contour charts, but both are challenging to read.
Pdf_Folio:346
This section has covered a variety of charts that can be used to present data based on the aspects
of interest, for example, relative measurements or changes over time. Although visualisations can
communicate information effectively, there are certain considerations that can impact the delivery. These
considerations involve visual perception theory, which is explored in more detail in the following section.
The goal is to make the data-to-ink ratio as close to 1.0 as possible. To achieve this, we need to remove
unnecessary ink in the visualisations. If erasing something in the graphic does not change the information
conveyed, then it should be erased (Knaflic 2015). Details that are necessary but are not part of the main
message should be put in the background (e.g. using light grey, as opposed to black text and lines).
Gestalt Principles
Gestalt principles deal with the psychology of visual perception — how humans assess what they see
as a whole from many individual parts (such as shapes, colours or spacing). Gestalt principles help us to
understand how the audience is likely to perceive the elements of a visualisation. They help us maximise
the data-to-ink ratio, but avoid removing too much ink from the chart. The Gestalt principles relevant to
eliminating unnecessary detail in data visualisations are described in table 4.20.
Pdf_Folio:347
Increasing power Connection • We perceive visually connected objects as grouped. For example, we
to eliminate can show relationships between objects by visually connecting them, for
unnecessary detail example, using a straight line.
Similarity • We perceive objects with similar properties (e.g. colour, shape, size or
orientation) as a group.
EXAMPLE 4.21
Pdf_Folio:348
140
Indexed growth (%)
110
100
90
80
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
Pdf_Folio:349
140
Indexed growth (%)
130
After losing market share, we outsourced our R&D
operations to Lambda Group three years ago as a strategy
120 to drive innovation and competitive advantage.
110
100
90
80
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
Comparing the ‘before’ and ‘after’ charts, we can see how the data-to-ink ratio has been improved and
attention is easily drawn to the key parts of the story. This reduces the level of mental workload required
by the audience to understand the information contained in the visualisations, making our communication
more effective.
VISUALISATION TOOLS
Often visualisations are created in Excel, which enables various types of conventional charts. However,
when working with more complex data or real-time data, more sophisticated options are worth exploring.
Once the preserve of data scientists and business intelligence specialists, visualisation tools are being
made increasingly available to others in organisations. Figure 4.37 provides a brief overview of some of
the leading visualisation tools.
Microsoft Power BI
Power BI is an all-in-one business intelligence and analytics platform provided as-a-service or as a
desktop client, but it is particularly highly rated for its visualisation capabilities. Visualisations are created
directly from reports and can be shared with users throughout an organisation. As well as a large number
of inbuilt visualisation styles, new ones are constantly being created by the AppSource community or if
you want to get your hands dirty with coding, they can be created from scratch using the developer tools
Pdf_Folio:350
DASHBOARDS
Dashboards allow for easier comparisons between metrics, which facilitates stakeholder decision making.
Dashboards are a visualisation tool that combine multiple related visuals that update as the data changes.
Pdf_Folio:351
Region
Revenue (YTD) Sales category breakdown Revenue monthly trend Central
100% East
South
100K West
80%
% of total sales
Category
60% Sales Furniture
Office su..
40% 50K Technolo..
Net profit margin
20%
0%
February May August November
0K
Month of order date (2014) February April June August October
Month of order date (2014)
Top sub categories (units sold)
Sub-category Sales by region
2000 Region
1500 200K
Quantity
150K
1000
Sales
100K
500
50K
0 0K
Binders Paper Furnishings Art Phones Accessories Storage Central East South West
Working capital
Days sales outstanding Days inventory outstanding Days payable outstanding
0 1.5K 18k
16 16 16 16 16 16 16 16 16 16 16 16 7k 5k
1K
Cash conversion
20 20 r 20 r 20 20 20 l 20 20 20 t 20 20 20
n b a y n u g p c v c 16 16 16 16 16 16 16 16 16 16 16 16
Ja Fe M Ap Ma Ju J Au Se O No De
n
20 20 r 20 r 20 20 20 l 20 20 20 t 20 20 20
b a y n u g p c v c Net due < 30 days < 60 days
Ja Fe M Ap Ma Ju J Au Se O No De
Accounts receivable Accounts
< 90 days > 90 days
turnover payable turnover
Bullet Charts
A bullet chart is typically used to display a measured value against a target value. As such, they are often
used to display information about KPIs. The user can quickly see the actual versus target value of the KPI
and evaluate progress or compliance with an objective.
Revenue
$0k $10k $20k $30k $40k $50k $60k $70k $80k $90k $100k
Gauges
A gauge is similar to a bullet chart, as it also displays an actual value against a target value. The appearance
of a gauge can vary, from a simple vertical or horizontal bar to an arc (similar to a gauge on a physical
vehicle or machinery dashboard). The gauge may allow the values to have text labels or may just use
markings along an axis.
Return on assets
100%
30%
– 14%
0%
Progress Bar
A progress bar is normally used to indicate progress towards a certain goal by showing a percentage out
of 100 per cent. It is appropriate when the absolute value is not important. The progress bar is typically a
horizontal bar; however, it can be a vertical bar or arc.
60%
Directional Indicators
Directional indicators use a triangle or arrow, oriented with the point in an upward or downward direction
to indicate whether a metric has increased or decreased relative to a previous point in time (a day, week,
month, year etc.). Directional indicators are most often combined with a raw text value for the metric (the
green, upward facing indicator implying that clothing sales increased from the day before).
$11 726
Pdf_Folio:355
Goals Metrics
Strategic • Reduce average accounts receivable days • Average accounts receivable days
to 32, down from 41 last year • Turnover ratio
• Collection effectiveness index
The types of metrics and KPIs to be chosen, can be broken down to the following categories (Rasmussen
et al. 2009).
• Input. Input metrics are a measure of the resources that have been used to produce, develop or maintain
an output. The output could include delivering a service, manufacturing a product or conducting an
activity.
• Output. Output metrics are a measure of the quantity of something that was produced, provided,
conducted or delivered.
• Efficiency. Efficiency metrics are a measure of how much input was required to produce an output. This
is typically a ratio of the resources used to the amount produced. It can be a measure of cost efficiency
or productivity.
• Quality. Quality metrics are a measure of how well something has been produced or delivered. The
metric can be quantitative or qualitative. It can include concepts of satisfaction, timeliness, accuracy or
some other measure of meeting internal or external expectations or requirements.
• Outcome. Outcome metrics are a measure of progress towards an organisational goal, benefit to
customers, or the extent that an activity, product or service has impacted the target.
Pdf_Folio:356
EXAMPLE 4.22
Designing a Dashboard
We have been asked to create a dashboard for the Chief Financial Officer of a retail company.
The CFO has requested a simple strategic dashboard, with basic metrics to monitor one strategic
objective: to increase the company’s net profit by 10 per cent from last year.
The dashboard will be displayed on a large screen on the office wall, so we need to make the
visualisations easy to read, and avoid using small text.
Step 1: Metrics
After consulting the CFO, we have a list of metrics for a basic net profit analysis dashboard:
• Net profit (year to date)
• Revenue (year to date)
• Gross profit margin (against a target)
• Operating profit margin (against a target)
• Net profit margin (against a target)
• Net profit by region
• COGS (cost of goods sold) breakdown (materials, returns, production, shipping)
• Expenses breakdown (rent, utilities, salaries, travel, support services, insurance, marketing, other).
Step 2: Visualisations
Our process for selecting appropriate visualisations is as follows.
The CFO has told us that the KPIs they consider most important are net profit and revenue for the year
to date. They would just like the raw figure, and do not require the target to be shown. Rather than create
a bullet chart, we can use the raw values displayed in a large font, so they stand out.
Pdf_Folio:357
Net profit
$841 744
Revenue
$7 235 987
Step 3: Layout
We have been told the profitability indicators are the next most important metrics.
We decide to use bullet charts, so the CFO can see the profit metrics against their targets. As they are
related, we can group them together as a vertical stack. Due to the typical relative proportions, we can
place them in a descending order, forming a sort of funnel chart appearance.
We also make sure the choice of colours does not create too much contrast with the rest of the chart.
As the absolute values of the target marks are difficult to measure against the x-axis, we put small text
labels against them.
Profitability
Metric F
Gross profit margin 81.75
85
Operating profit margin 16.62
20
Net profit margin 11.63
14
0 10 20 30 40 50 60 70 80
Actual (%)
As each individual has their own preferences for visualisations, it is recommended to consult with
the user when making the various decisions. An alternative to the bullet charts could have been a set
of gauges.
Profit margins
0 82 85 0 17 20 0 12 14
After we consulted the CFO for their preferences, they indicated they wanted to use the bullet charts.
The last set of metrics required enable the CFO to have a quick view of the composition or part-to-whole
breakdown of net profit by geographic region, COGS and expenses.
For composition visualisation, our options include bar charts, pie charts, proportional area charts, tree
maps, Sankey diagrams, waterfall charts, pictorial charts or Pareto charts. Given the limited screen real
estate, we avoid most of these (as they require a large area to present the detail) and consider bar charts,
pie charts and proportional area charts.
Pie charts should be avoided where there are more than two or three categories and the proportions
are similar, which leaves us with either proportional area charts or bar charts. Bar charts lose some of the
part-to-whole impression that proportional area charts convey and can be difficult to label the categories
when restricted to a small space.
As a compromise, we have chosen a stacked bar chart with a wide column width, and assigned text
labels to each category. One disadvantage of using the text labels is that the smaller components will
not show their labels. This is a rule imposed by the software. The CFO was happy with just the top
categories showing.
Pdf_Folio:358
800K 1200K
4M
Europe 1000K
600K Materials
Salaries
Total ($) 3M
Total ($)
Total ($)
800K
400K 600K
2M Asia
Other 400K Production
1M 200K
Australia 200K
0M 0K North America 0K Shipping
Metric F
Gross profit margin 81.75
85
Revenue
Operating profit margin 16.62
20
Net profit margin 11.63
14
0 10 20 30 40 50 60 70 80
Actual (%)
800K
1200K
4M
Europe
1000K Materials
Salaries 600K
3M
800K
Total ($)
Total ($)
Total ($)
400K 600K
2M Asia
Other 400K Production
1M 200K
Australia
200K
North America Shipping
0M 0K 0K
.......................................................................................................................................................................................
CONSIDER THIS
Do you use a dashboard? If not, would one be helpful? What would you include and how would you present each
piece of information?
QUESTION 4.15
Briefly outline the main benefits and challenges in using data visualisation dashboards.
To conclude, we will look at an example that draws together all of the topics we have studied in
this module.
Pdf_Folio:359
Income Statements
for the year ended 30 June 2020
Group1,2,3 Bank2,3
Note $M $M $M $M $M
Interest income:
Interest expense 2.1 (11 552) (16 485) (15 807) (10 660) (15 434)
Operating expenses 2.4 (10 929) (10 928) (10 687) (10 745) (10 633)
Net profit after income tax 7 459 8 101 8 742 10 168 7 783
from continuing operations
Pdf_Folio:360
Balance Sheets
As at June 2020
Group1,2,3 Bank1,3
Note $M $M $M $M
Assets
Assert at fair value through Income 5.3 46 545 33 677 46 284 33 128
Statement
Investment securities:
Loans, bills discounted and other 3.1 771 547 755 173 – –
receivables
Total assets 1 014 060 976 502 963 747 910 851
Liabilities
Deposits and other public borrowings 4.1 701 999 636 040 631 301 573 851
Pdf_Folio:361
Liabilities at fair value through Income 4.2 4 397 8 520 3 888 7 961
Statement
Debt issues 4.3 142 503 164 022 113 323 131 094
Bills payable and other liabilities 7.2 13 188 10 068 11 866 8 687
Total liabilities 942 047 906 853 895 672 845 727
Shareholders’ Equity
Non-controlling interests 5 55 – –
The question is, will future disruption and new forms of competition impact on this, as shown by the
rapidly (by historical standards) growing number of authorised deposit-taking institutions in Australia that
are in the bank category?
Financial institutions, and in particular the large ADIs, have enjoyed a privileged position in the Australian
financial system. While they face high levels of regulatory oversight and compliance obligations, they are
also protected by these regimes as they also act as effective barriers to entry for competitors. This is
somewhat logical as the regulatory agencies have a primary purpose of protecting the efficacy of the
financial system and therefore the economy and prosperity of all in the community in so far as having
a stable and efficient financial system is a key ingredient for this. For participants in the Australian
financial system, they have also been largely rewarded with a stable financial system that has had few
significant institutional failures (notably including through the GFC), deposits have been protected and
thus consumers have confidently participated in the system.
One might argue, however, that there are storm clouds on the horizon. Central to this is the continued
lack of trust in these institutions. For example, the 2019 IPSOS Trust in Professions survey showed just
13 per cent of people believe bankers are trustworthy while 52 per cent believe they are untrustworthy
(see figure 4.43). This, together with increased competition, digitisation, changing trends in consumerism,
potentially lower transactions costs and effort for consumers make this an industry ripe for further
disruption.
Take a minute to reflect on this and your own interactions with your bank/financial institution. Have
you refinanced/switched in recent years to get a better deal? When was the last time you went into
a branch? What was your reaction to the 2017–19 Royal Commission into Misconduct in the Banking,
Superannuation and Financial Services Industry? How has the COVID-19 pandemic changed your views?
Pdf_Folio:362
Doctors 69% 8%
Scientists 62% 8%
Teachers 60% 9%
Polisters 9% 35%
• Character — a measure of creditworthiness based on indicators such as credit score and credit history
• Capacity — a measure of ability to pay back the loan based on income and expenses
• Capital — a measure of how much of your own money you are investing (e.g. a down payment)
• Collateral — a measure of the asset used to secure the loan
• Conditions — a market analysis of trends of the business’s industry
Source: CPA Australia 2021.
In order to obtain insights into the impact of change lending obligations, the financial institution would
apply credit scoring models to historical loan application data and determine the extent to which revised
parameters (an example simple data file is shown in figure 4.46) — including less verification of client
information by analysts — lead to potentially different credit decisions.
• Were any previously funded loans not funded in the new model?
• Were any previously approved applications that became non-performing loans not funded in the
new model?
• Are any not-funded applications recommended to be funded in the new model?
• Examine applications that the credit scoring model recommended for funding that the credit analysts
did not recommend — would this outcome change?
Pdf_Folio:364
Debit Payment
history Work
history
Age
Worth
Assets
Once these parameters are adjusted and applied to the historical data, the credit model can be re-run
to determine the updated recommendation with a simple infographic used to present the differences from
the original decision. This can then be matched against actual outcomes in terms of loan performance for
those loans that were approved. This may also reduce the level of human intervention in the process, as data
validation will be less onerous, particularly in relation to client expenses and serviceability calculations.
Indeed, simplifying these parameters could allow for data to be drawn directly from client financial records
in relation to easily identifiable transactions such as income and debt payments.
Once the revised data sets and credit scoring outcomes are in place, data scientists will apply econometric
analysis to predict outcomes and determine the impact of the revisions to the data and decision-making
process by back testing the existing loan portfolio. The more difficult component will be assessing loans
that were historically rejected that become accepted with the revised process. A match pairs (matching
these loan characteristics to loans that were approved) approach, could for example be used to predict
loan performance outcomes in these cases. Iteratively, parameters can be adjusted and tests repeated to
determine both the elasticity of the parameters to both funding decisions and subsequent loan performance.
This will then provide a picture of the resultant loan portfolio and the characteristics of it (volume of loans,
credit risk exposure, default risk profile, and implied capital adequacy and liquidity requirements).
Pdf_Folio:365
[ INPUT ] [ OUTPUT ]
Class 1
Class 2
Class 3
Class 4
[– 30)
[30 – 40)
[40 – 50)
[40 – 60) Risk score
[60 –)
[– 30K)
[30K – 40K)
[40K – 50K)
[50K – 60K)
[60K –)
Bias
Default risk
MEDIUM
HI
W GH
LO
Post
Pre
MAX
NO
Volume of lending
MEDIUM
HI
W GH
LO
Pre Post
MAX
NO
Historical risk
(based on existing risk
scorecards)
Source: CPA Australia 2021.
Pdf_Folio:366
SUMMARY
The way humans process information means visualisations are far more easily understood than text. Data
visualisation is the process of using visual representations to explore, interpret and communicate the
meaning of data and analytics outputs. The data visualisation process involves understanding the context of
the analysis, choosing the appropriate visual to use, applying cognitive principles to increase effectiveness
and being able to tell a story with data.
Various tools are available to help create appropriate visualisations. Whichever is chosen, fundamental
principles should be applied to each visualisation created to ensure it effectively communicates.
A storytelling approach is an effective way to communicate information and recommendations to stake-
holders. Storytelling engages the audience’s attention and memory more effectively than an information
that is not accompanied by a structured story narrative. A useful story structure involves establishing the
goal, setting the stage, developing the story and closing the story. The close will often involve a call to
action — the stage at which the potential value of data is realised through making data-informed decisions
and acting on them.
For real-time management, dashboards are a useful visualisation tool as they enable various real-time
metrics to be displayed in a single view. The dashboard can be designed to meet the user’s specific needs.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
4.4 Evaluate the effectiveness of various data visualisation tools in presenting complex data to
convey desired messages to a range of audiences.
• Visualisations should comply with cognitive principles to maximise their effectiveness.
• A range of software tools are available to produce visualisations from data. Each has its strengths
and weaknesses — and the choice of tool should be based on the individual’s visualisation needs.
• Dashboards combine multiple related visuals that update in response to data. They enable easy
comparison between metrics to facilitate stakeholder decision making.
• A well-designed dashboard provides actionable information related to business objectives to
decision makers. The content of the dashboard will depend on whether it is for strategic, tactical
or operational purposes.
Pdf_Folio:367
REVIEW
Organisations that adopt data-informed decision-making generally outperform organisations that place less
reliance on data. With data and processing power now more available than ever before, many organisations
are expanding the use of analytics in order to derive insights to address business problems and create
efficiencies.
Becoming a data-driven organisation requires data literacy throughout all levels of the organisation.
Many accounting and finance professionals are experienced data owners, analysts, information creators and
information providers, but the increasing potential of data mining, particularly AI and ML technologies,
has created the need for an expanded set of data skills. In particular, understanding the applications of
big data and ML-based analytics is essential to evaluating their role in the organisation, promoting their
appropriate use and taking advantage of the insights they can provide.
Accounting and finance professionals need to attain and encourage data literacy throughout the
organisation, ensure the organisation’s data strategy captures and analyses the data needed for business-
critical decisions, and be able to work with data scientists and others in the analytics process.
Data literacy and awareness of fundamental principles in data interpretation are required so that analytics
outputs can be applied to solving business problems. Once outputs are interpreted and recommendations
are formed, the professional needs to develop effective ways to communicate their insights and recom-
mendations to stakeholders. Storytelling is an effective approach and enables the use of data visualisations
that will help stakeholders understand the meaning of the data. For real-time analytics, dashboards can be
designed with a focus on the metrics most relevant to the user’s needs.
REFERENCES
Aaker, J 2014, ‘How to use stories to win over others’, Lean In, video, viewed accessed April 2020, https://leanin.org/education/
harnessing-the-power-of-stories
AASB (Australian Accounting Standards Board) 2016, ‘AASB Research Report No.2: Accounting judgments on terms of
likelihood in IFRS: Korea and Australia’, accessed May 2020, https://www.aasb.gov.au/admin/file/content102/c3/July_2016_
(KASB-AASB)%20Accounting%20Judgments%20on%20Terms%20of%20Likelihood%20in%20IFRHellips.pdf
Accenture 2020, ‘The human impact of data literacy: A leader’s guide to democratizing data, boosting productivity and
empowering the workforce’, accessed May 2020, https://www.accenture.com/_acnmedia/PDF-115/Accenture-Human-
Impact-Data-Literacy-Latest.pdf
Bank for International Settlements 2018, ‘Financial instability: can big data help connect the dots’, Remarks by Luiz Awazu
Pereira da Silva and Goetz von Peter, 29 November, https://www.bis.org/speeches/sp181203.pdf.
Bhargava, R & D’Ignazio, C 2015, ‘Designing tools and activities for data literacy learners’, Workshop on Data Literacy,
Webscience.
BI Survey, 2016, ‘Global survey on data-driven decision-making in businesses’.
Black, K 2019, Business analytics and statistics, John Wiley & Sons, Milton Qld, p. 9.
Brynjolfsson, E, Hitt, L, Heekyung, HK 2011, ‘Strength in numbers: How does data-driven decision making affect firm
performance?’, SSRN Working Paper 1819486.
Bumblauskas, D, Nold, H, Bumblauskas, P, Igou, A, 2017, ‘Big data analytics: transforming data to action’, Business Process
Management Journal, vol. 23, no. 3, pp. 703–20, Emerald Publishing Limited.
Carney, M 2013, ‘Flush with $20M from Peter Thiel, ZestFinance is measuring credit risk through non-traditional big data’,
Pando, accessed May 2020, https://pando.com/2013/07/31/flush-with-20m-from-peter-thiel-zestfinance-is-measuring-credit-
risk-through-non-traditional-big-data/
Commonwealth Bank 2020, 2020 annual report, accessed 27 April 2021, https://www.commbank.com.au/content/dam/
commbank/about-us/shareholders/pdfs/results/fy20/cba-2020-annual-report-print.pdf
Crozier, R 2019, ‘Suncorp uses AI to drill into insurance claim data’, IT News, accessed May 2020, https://www.itnews.com.au/
news/suncorp-uses-ai-to-drill-into-insurance-claim-data-532517
DalleMule, L & Davenport, TH 2017, ‘What’s your data strategy’, Harvard Business Review, May–June, https://hbr.org/2017/05/
whats-your-data-strategy
Dasgupta, S 2020, ‘Improving anti-money laundering compliance with dynamic customer risk profiling’, Risk Management,
12 February, accessed May 2020, http://www.rmmagazine.com/2020/02/12/improving-anti-money-laundering-compliance-
with-dynamic-customer-risk-profiling/
DESE (Department of Education, Skills and Employment) 2020, ‘Our data strategy’, Australian Government, https://www.
employment.gov.au/our-data-strategy
Ding, K, Hoogduin, L, Peng, X, Vasarhelyi, MA, Wang Y n.d., Clustering based peer selection with financial ratios’, Rutgers
University, accessed May 2020, http://raw.rutgers.edu/docs/wcars/40wcars/Presentations/KexingXuanYunsen.pdf.
Frické, M 2018, ‘Data-Information-Knowledge-Wisdom (DIKW) pyramid, framework, continuum’, in L Schintler L &
C McNeely (eds), Encyclopedia of Big Data. Springer, Cham.
Gartner 2020, ‘Creating a data strategy’, Joe Maguire, Gartner Research, 26 November, https://www.gartner.com/en/documents/
3975595/creating-a-data-strategy
Hanegan, K 2019, ‘Do you have what it takes to make data informed decisions’, Qlik, accessed April 2020.
Harle, P, Havas, A & Samandari, H 2016, ‘The future of bank risk management’, McKinsey & Company, 22 July, https://www.
mckinsey.com/business-functions/risk/our-insights/the-future-of-bank-risk-management
Hastie, T, Tibshirani, R & Friedman, J 2011, The elements of statistical learning: data mining, inference, and prediction,
2nd edn, Springer Series in Statistics, Springer, New York cited in Verbeke, W, Baesens, B & Bravo, C 2017, Profit driven
business analytics: a practitioner’s guide to transforming Big Data into added value, John Wiley & Sons, Hoboken, New
Jersey.
Hendry, J 2021, ‘CBA applies to become data recipient under CDR’, ITNews, 28 January, accessed March 2021, https://www.
itnews.com.au/news/cba-applies-to-become-data-recipient-under-cdr-560346
Hurley, M, Adebayo, J 2017, ‘Credit scoring in the era of big data’, Yale Journal of Law and Technology, vol. 18, no. 1, article 5,
pp. 149–215.
IBM 2020, ‘BondIT’, accessed July 2020, https://www.ibm.com/case-studies/bondit
InfoSys & CIMA 2014, ‘Using analytics to reduce days sales outstanding (DSO)’, InfoSys, CIMA, https://www.cimaglobal.com/
Documents/Thought_leadership_docs/Management%20and%20financial%20accounting/using-analytics-to-reduce-dso.pdf
Ipsos 2019, ’It’s a fact... scientists are the most trusted people in the world’, accessed 27 April 2021, https://www.ipsos.com/en-au/
its-fact-scientists-are-most-trusted-people-world
Keipper, S 2019, ‘How NLP and machine learning harnesses insights from unstructured data’, EY North America Financial
Services, accessed May 2020, https://www.ey.com/en_au/financial-services/how-nlp-and-machine-learning-harnesses-
insights-from-unstructured-data
Kennedy, K 2013, ‘Credit scoring using machine learning’, Doctoral thesis, Technological University Dublin.
Knaflic, CN 2015, Storytelling With Data, John Wiley & Sons Inc., New York.
Kotter, JP 1995, ‘Leading change: Why transformation efforts fail’, Harvard Business Review, vol. 73, no. 2, March–April, p. 61.
Kovalerchuk, B & Vityaev, E 2005, ‘Data mining for financial applications’, in Chapter 57, O Maimon & L Rokach (eds), Data
Mining and Knowledge Discovery Handbook, pp. 1204–41, Springer, Boston, MA.
KPMG 2017a, ‘Supply chain risk — a global perspective’, 21 March, https://home.kpmg/au/en/home/insights/2017/03/supply-
chain-risk-global-perspective.html
KPMG 2017b, ‘Ethical supply chains — investigation and risk prevention’, 21 March, https://home.kpmg/au/en/home/insights/
2017/03/ethical-supply-chains-investigation-risk-prevention.html
Ladley, J 2016, ‘Business alignment techniques for successful and sustainable analytics’, CIO, 13 May, https://www.cio.com/
article/3068642/business-alignment-techniques-for-successful-and-sustainable-analytics.html
Lampitt, A 2012, ‘Big data visualization: A big deal for eBay’, infoWorld, accessed May 2020, https://www.infoworld.com/
article/2616353/big-data-visualization--a-big-deal-for-ebay.html.
Marr, B 2015, Big data: using smart big data analytics and metrics to make better decisions and improve performance, John Wiley
& Sons, Hoboken, NJ, USA.
McKinsey, 2010, ‘The case for behavioral strategy’, McKinsey & Company, accessed April 2020, https://www.mckinsey.com/
business-functions/strategy-and-corporate-finance/our-insights/the-case-for-behavioral-strategy
Morrow, J 2018, ‘Developing a data literate workforce’, Qlik, accessed April 2020, https://www.qlik.com/us/-/media/files/training/
global-us/qlik-education-data-literacy-program-strategy-and-framework.pdf
Peng, R 2019, ‘Tukey, design thinking, and better questions’, Simply Stats, accessed April 2020, https://simplystatistics.org/2019/
04/17/tukey-design-thinking-and-better-questions/
Provost, F & Fawcett, T 2013, Data Science for Business: What You Need to Know about Data Mining and Data- Analytic
Thinking, O’Reilly Media Incorporated.
Rasmussen, NH, Bansal, M & Chen, CY 2009, Business dashboards: a visual catalog for design and deployment, John Wiley &
Sons Inc., New York.
Rowley, J 2007, ‘The wisdom hierarchy: representations of the DIKW hierarchy’, Journal of Information Science, April 2007,
vol. 33, no. 2, pp. 163–80.
SAS 2018, ‘The 5 essential components of a data strategy’, SAS, white paper, https://www.sas.com/content/dam/SAS/en_us/doc/
whitepaper1/5-essential-components-of-data-strategy-108109.pdf
Shmueli, G, Bruce, P & Patel, N 2016, Data mining for business analytics: concepts, techniques, and applications with xlminer,
3rd edn, John Wiley & Sons Inc., New York.
Smartsheet 2020, ‘Your quick-start guide to data-driven decision making’, accessed April 2020, https://www.smartsheet.com/
data-driven-decision-making-management.
St. Jeor, C 2019, ‘Reducing false positives for AML: Part II – building a model in SAS Viya’, Zencos, accessed May 2020,
https://www.zencos.com/blog/aml-analytics-how-to-predict-productive-alerts/
Tableau 2020a, ‘A data strategy framework: How to implement and scale for success’, Tableau, https://www.tableau.com/learn/
articles/data-strategy-framework
Tableau 2020b, ‘Meet Tableau Blueprint’, Tableau, https://www.tableau.com/learn/blueprint
Thomson, J 2018, ‘Leveraging data analytics in decision making: The new opportunity for finance’, Forbes, accessed April 2020,
https://www.forbes.com/sites/jeffthomson/2018/09/27/leveraging-data-analytics-in-decision-making-the-new-opportunity-
for-finance/?sh=2bf786d13acc.
Tufte, E 1983, The Visual Display of Quantitative Information, Graphics Press, Cheshire CT.
Tukey, J 1962, ‘The future of data analysis’, Annals of Mathematical Statistics, vol. 33, no. 1, pp. 1–67.
Vasal, A, Vohra, S, Payan, E, & Seedat, Y 2019, Closing the data value gap: How to become data-driven and pivot to the new’,
Accenture, accessed May 2020, https://www.accenture.com/_acnmedia/pdf-108/accenture-closing-data-value-gap-fixed.pdf.
Villalobos, MA & Silva, E 2017, ‘A statistical and machine learning model to detect money laundering: An application’, Anahuac
University.
White, A 2019, ‘Our top data and analytics predicts for 2019’, blog, Gartner Blog Network, 3 January, https://blogs.gartner.com/
andrew_white/2019/01/03/our-top-data-and-analytics-predicts-for-2019/
Wigglesworth, R 2018, ‘Spurious correlations are kryptonite of Wall St’s AI rush’, Financial Times, accessed June 2020,
https://www.ft.com/content/f14db820-26cd-11e8-b27e-cc62a39d57a0
Woodhead, B 2008, ‘Accenture wins BHP outsourcing deal’, Financial Review, 30 June, accessed March 2021, https://www.afr.
com/technology/accenture-wins-bhp-outsourcing-deal-20080630-jaoli.
RISK MANAGEMENT,
GOVERNANCE AND
REGULATION
LEARNING OBJECTIVES
PREVIEW
The digital transformation of accounting and finance, the technologies enabling it, and the integration
of data into most key business activities, are creating a new and rapidly changing risk environment.
Accounting and finance professionals need to be able to advise on and participate in strong risk governance
to ensure the organisation has robust processes and structures to identify and manage risks that threaten the
business’s ability to achieve its objectives. Likewise, compliance has become a more complex challenge in
the context of digital finance. Innovative financial technologies (FinTech) have challenged regulators who
seek to protect the interests of the community while not stifling innovation. Approaches to compliance
with regulation are themselves being disrupted by regulatory technologies (RegTech).
Part A of this module presents a comprehensive risk management process based on ISO 31000 Risk
management. Working through this process identifies, analyses and responds to risks that can threaten the
business’s ability to achieve its objectives. Adopting the best practices in ISO 31000 includes monitoring
and evaluation of both the risk environment and the performance of the organisation’s risk frameworks and
policies. This, along with awareness of emerging risk management innovations, provides the knowledge
and skills to ensure ongoing improvement in the organisation’s management of risk.
Part B introduces the concept of risk governance, which refers to how decisions are made and authority
is exercised in relation to the organisation’s management of risk. We describe good governance and the
ideal composition of a governance Board. We then describe a range of poor practices and emerging threats
to good practice in order to establish a comprehensive understanding of the requirements of good risk
governance. This part of the module concludes with a discussion of data governance, which sets the scene
for the in-depth discussion of data security and privacy obligations and risks in part C.
Part C introduces the industry standards and best practices for ensuring the security of data held and used
by an organisation. We explore cybersecurity and other threats to data, including the potential severity of
the consequences of a security breach. We conclude with a discussion of data ethics and the need to protect
data to ensure the privacy of individuals. Application of the material in parts A and B to the risks detailed
in part C enables the development of mitigation strategies for cybersecurity threats.
Pdf_Folio:371
Part D explains and applies a compliance framework to enable the creation of an action plan to achieve
compliance and manage potential compliance breaches. This part includes a discussion of emerging
compliance issues.
Part E examines how regulators have responded to FinTech by seeking to balance community protection
with an environment that enables innovation. It explores how organisations can comply with regulations
while developing or taking advantage of FinTech innovations.
Part F concludes the module with a discussion of RegTech — the application of technology to help an
organisation comply with its regulatory obligations.
Pdf_Folio:372
RISK GOVERNANCE
Risk management is a component of an organisation’s governance, risk management and compliance
(GRC) framework. Governance sets the overall management approach through which executives and
senior managers direct, control, and align the organisation to the business goals. Compliance is the way
the organisation complies with its stated internal and external risk requirements (Open Compliance and
Ethics Group 2020; CIO Australia 2017).
There are many tools and applications that can support the different components of GRC, but it is
the GRC framework itself that provides the structure and approach that management may use to achieve
their business goals. Within the framework are strategy, process, policy, technology and people. A unified
approach between these factors increases consistency and reduces duplication of remediation efforts. Risk
governance is discussed in detail in part B of the module.
Example 5.1 examines how the Australian banking sector’s risk governance failings led to increased
regulatory oversight.
EXAMPLE 5.1
Pdf_Folio:373
Board or executive management discussions are not just focused on execution questions such as ‘How
are we managing technology risks?’ The discussions and focus are also placed on how technology-related
risks endanger an organisation’s strategy.
For example, companies that utilise or interact with the Internet of Things (IoT), such as car and smart
device manufacturers and designers, know that their strategy to succeed requires a clear understanding of
cyber-related threats (e.g. hackers gaining access to networks via in-car navigation systems, or to smart
fridges that are linked wirelessly to the user/owner’s internet network). The organisation must keep track
of these risks and close loopholes that could be exploited wherever possible.
Control C
Control B
Control A
Event with
large potential Residual
exposure risk
Risk Owners
A risk owner is a person or group who is assigned the responsibility for a particular risk within the risk
management framework. An organisation will thus have many risk owners, each with responsibility for a
specific risk or set of risks. The risk owner is generally accountable for:
• the completion of risk analysis and the accuracy of associated data
• performing the risk assessment to rate the risk
• ensuring the risk is maintained in a ‘well-controlled’ state
• ensuring remediation tasks are implemented to rectify any identified issues.
.......................................................................................................................................................................................
CONSIDER THIS
Are you a risk owner in your organisation?
QUESTION 5.1
What role does an Audit and Risk Committee play in an organisation’s risk management strategy?
Pdf_Folio:376
QUESTION 5.2
Outline four responsibilities a ‘risk owner’ has and why an organisation will typically have many
risk owners.
ISO 31000
The key standards framework for risk management is the ISO 31000 framework. It includes principles
and guidelines on implementation, risk assessment techniques and a guide to risk management vocabulary.
This standard looks at risks from a strategic perspective. It aims to help an organisation develop a risk
management culture to increase the awareness of monitoring and managing risk (Hutchins 2019). The
standard includes a risk management framework and process. We will work through a process based on
ISO 31000 in section 5.3.
ISO 27001
The ISO 27001 standard relates specifically to information security management. ISO 27001 details the
specifications for an information security management system (ISMS). The standard looks at people,
process and technology as part of an ISMS. It provides the security controls that can be implemented to
mitigate risk.
ISO 27005
The ISO 27005 standard also relates specifically to information security risk management. This standard
sets guidelines for conducting an information security risk assessment with ISO 27001 and presents
information security risk management best practice. In contrast to ISO 31000, it considers best practices
in the context of information security, rather than the business as a whole.
QUESTION 5.3
Compare and contrast the risks covered by the ISO 27001 and ISO 31000 frameworks.
Pdf_Folio:377
Monitor
An organisation first needs to establish the scope, context and criteria of its risk management process.
This relates to:
• how the risk management process interacts with the organisation as a whole
• the internal and external environment and thus the sources of risk and the conditions in which the process
will be applied
• the criteria against which risks will be evaluated in terms of their impact on the business’s objectives.
The inclusion of communication and consultation in the process recognises that:
• risk management must be informed by consultation with each function across the organisation
• risk management processes and culture need to be communicated throughout the organisation so
that compliance and audits can be shared, creating a more effective risk management program (PwC
Australia 2019).
Recording and reporting involves the documentation and reporting of risk management processes and
outcomes to facilitate communication and decision making.
Monitoring and review encompasses all aspects of the risk management process to:
• analyse risk information including trends, updates and changes
• understand where risks have changed through both internal and external lenses
• ensure that controls to manage risk are designed and implemented effectively
• identify emerging risks.
We will discuss monitoring and review in more detail after we step through the assessment and
treatment stages.
RISK PROFILES
Organisations face different types of risk. An organisation’s risk will vary in nature, scale and complexity
depending on:
• the types of products or services they offer
• the diversity of their location and outsourcing of their tasks
• the volume and size of their transactions
• the balance of their retail and wholesale clients
• whether they give financial advice
• how many people are in the organisation.
Pdf_Folio:378
Information Assets
Information assets are essentially any valuable information that the organisation owns or controls. To
incorporate risks to information- and data-related assets into the risk profile, it is necessary to understand
the nature and value of these assets.
Information assets may be broadly categorised as follows.
• Data. Data is possibly the business’s largest asset. Depending on the nature of the business, data
could include patient health records, details of customer spending habits or personally identifiable
information (PII). Consider what the data is worth to the business; what it is worth to a competitor or
a cyber attacker; and what it would cost to replace.
• Intellectual property. A business might have intellectual property such as research, designs, copyright,
patents, trademarks or even trade secrets. Consider what these are worth to the business and what they
are worth to a cyber attacker or a foreign intelligence service.
• Funds. All businesses need funds to survive. Consider the impacts if a cyber attacker drained the
business’s bank accounts or found a way to siphon small amounts over a long period of time without
being detected.
• System availability. Other items of value within a business include hardware and software, including
web servers, file storage, general applications and how they contribute to the business’s systems being
available. Consider what it would cost the business in lost time and productivity if a cyber-attack disabled
part of the network. If the business provides online services, consider the impact on users who are unable
to access their accounts or services.
Organisations’ information asset listings are often spread across departments. It is good practice to
collate and incorporate these into a central inventory. The inventory must then be maintained to reflect
purchases, disposals, deletions, destruction or other changes to the assets. This process should involve
consultation with managers or team leads across the organisation as well as the use of IT tools to determine
how information assets are accessed and used, both internally and externally.
Like any business asset, the measures taken to protect an information asset are based on its value and
the costs of protecting it. It is not possible to protect everything— it would be too expensive — so the
business needs to prioritise its assets based on value and other relevant considerations. The value of an asset
should consider:
• original cost of acquisition
• cost of acquiring a replacement
• implementation or development costs to support a replacement
• current or expected cost of controls to maintain or protect the asset
• cost of impact to production and productivity if an asset is compromised.
Risk profile
asset/process Low Medium High
(continued)
Risk profile
asset/process Low Medium High
Economic Company is likely affected Moderate impact of interest Growth rate is significantly
by interest rates. rates. affected by seasonal
factors, inflation and
interest rates.
Social Social attitudes do not Social attitudes have Stakeholders and staff
greatly influence our some influence on market have high social attitudes
position in the market. position with variable and shared beliefs
shared beliefs. around health, work,
demographics.
Technological Consumers are not Technology changes are Consumer access and
influenced or likely to be unlikely to influence our use of technology has
affected by changes to business model. a significant impact on
technology. company objectives.
Emerging technologies
have, or are likely to,
influence our business
model.
Legal The business is regulated The business is regulated The business is sig-
by regional or national by regional or national nificantly regulated by
laws. Moderate or lesser laws. global laws concerning
obligations are imposed consumers, intellectual
due to company size. property, data protection
and tax.
Personal Consideration for Financial, minimal Must comply with Must comply
information data protection and PII, not required the Privacy Act with global data
privacy regulations to comply to the protection/privacy
Privacy Act 1988 obligations
(GDPR and other
requirements)
Pdf_Folio:380
Complexity in cybersecurity
Risk profile Compliance requirements
Pdf_Folio:381
Identify Risk
Business objectives are set and agreed upon during the planning cycle and are available in the organisation’s
long-term plans. The analysis of the operating environment enables the identification of the risks that
could impact on the delivery of business plans. The identification of risk can be done in many ways:
through clean sheet reviews, regular management reviews or ad hoc reviews when there is a change to the
business environment.
Once a risk event has been identified and articulated, an assessment of the risk is conducted to determine
whether further analysis is required, a risk owner is identified, and the risk is entered into a risk register.
Risk issues
Unauthorised activity/
Secure assess to
Business interruption
insider sabotage
database/server or
digital workplaces
Impact
Cause
Risk event
Cybersecurity attack
Generally, a bowtie is developed in a workshop with the relevant SMEs, risk owners and potential control
owners present.
When the risk event, causes and impacts have been agreed upon, the controls are identified. A critical
control is a control that significantly reduces the impact or likelihood of a risk. The number of critical
controls must be appropriate to the risk event and must play a key role in achieving the business objective.
The control documentation is developed by the control owners of each control to identify and describe
the components that must be in place and working effectively for the control to be effective. The control
documentation generally comprises the:
• design standard — what needs to be in place or provided for the critical control to be effective (based
on organisation or industry standards, established systems, processes, documents)
• operating standard — the activities or tasks that occur to ensure that the control operates as designed
• verification approach — the method and strategy to obtain assurance on the control health, using
various methods as evidence.
Risk Register
A risk register is a risk management tool to record details for identified risk, including risk ratings, nature
of the risk, owner, manager and mitigation measures.
A risk register is the key document used throughout the risk management process from identification
through reporting. When risks are identified, they can be added to the register, with supporting information
added as the risk assessment proceeds (e.g. risk treatment, improvement actions).
Pdf_Folio:382
1 Rare The event may only occur in exceptional Less than once in 15 years
circumstances.
2 Unlikely The event could occur at some time. At least once in 10 years
3 Possible The event will possibly occur at some time. At least once in three years
5 Almost certain The event will occur in most circumstances. More than three times per year
Health Near miss Minor injuries Lost time injury Lost time injury Death/fatality
Minor first requiring medical < 30 days > 30 days or permanent
aid injuries, treatment but not disability
not requiring hospitalisation
further medical
treatment
Financial Less than $50 001– $100 001– $500 001– More than
impact $50 000 $100 000 $500 000 $3 000 000 $3 000 000
(continued)
Pdf_Folio:383
This table can be used to assess any risk event against its impacts. A single risk will have multiple
impacts of different degrees; therefore, assessing each risk against each impact type will provide a
complete assessment. Digital risks will mostly have service interruption, financial, reputation and com-
pliance impacts, rather than health or safety impacts. Using a consistent table allows comparison of all
risks within the organisation.
This results in a final risk rating for each risk. Figure 5.5 is a risk rating chart.
1 2 3 4 5
Almost certain
5 10 15 20 25
5 Medium High High Extreme Extreme
Likely
4 8 12 16 20
Low Medium High High Extreme
4
Possible
3 6 9 12 15
Low Medium Medium High High
3
Unlikely
2 4 6 8 10
2 Negligible Low Medium Medium High
Rare
1 2 3 4 5
Negligible Negligible Low Low Medium
1
Likelihood
Impact
Reducing Risk
Risk treatment is the action taken to modify the identified risk to a level the organisation considers
acceptable.
The ways an organisation can treat risk are as follows.
• Transfer. An organisation may choose to transfer the risk to a third party, usually in the form of
insurance or outsourcing.
• Mitigate. An organisation may choose to mitigate a risk by taking action to reduce the likelihood of an
event occurring, but not avoiding the risk.
• Avoid. An organisation may choose to avoid the actions that would lead to the risk.
Alternatively, the organisation can choose to accept a risk. An organisation will need to take on
some risks, as without some risk it is difficult to achieve a profit. Choosing to accept a risk is called
risk retention.
There are thus four risk management options: transfer, mitigate, avoid and accept.
A risk treatment chart can help define accountability, response urgency, minimum treatment required
and the review frequency for the risks. An example organisation risk treatment chart/matrix is shown in
figure 5.6.
Pdf_Folio:385
Minimum
Risk level Accountability Response treatment Description Review
required
Immediate
action required
in consultation
with executive
leadership
Reject and
team to either
Extreme Board or CEO Urgent avoid, transfer Immediately
avoid the risk
or mitigate
entirely, transfer
it or to reduce
the risk to a low,
medium or
high rating.
Managers are
to be assigned
CEO or to these risks
executive Accept and and treatments
High Important to modify, Monthly
leadership mitigate
team reduce,
transfer or
eliminate the
risk is required.
Manage by
specific
Executive Operational controls, Monthly/
Medium Accept
manager/manager process monitoring or quarterly
response
procedures.
Manager/ Manage by
coordinator/ Capture in risk Quarterly/
Low Accept routine
team leader register annually
procedures.
Manage
Manager/
through
coordinator/team Refer to
Negligible Accept compliance Annually
leader/supervisor compliance
checks and
processes.
Pdf_Folio:386
Least
effective
Source: CPA Australia 2021.
Eliminate the hazard Remove this data entry step in the process.
It may not be practical to completely remove data entry
as this is a core part of the business. While elimination is
the best type of risk control, it is not always practical and
other options need to be considered.
Substitute the hazard with a safer alternative Replace human operators with automated bots.
This option might be reasonable, but it may not always
be practical due to technology, capital or productivity
constraints. Replacing the human operators might be
expensive and could have flow-on consequences that
would be more significant and costly.
Engineer the control to reduce the risk Implement a system of control validation to the data entry
to stop data corruption.
While this is another good idea, engineering controls can
be subject to failure if the solution created is bespoke,
and the design is not proved and tested in the industry.
Administrative control to reduce risk Establish policies and procedures, including assigning
access control to reduce exposure. Provide training to
staff that have access to the database.
This control would be very effective, but it is heavily
reliant on employee training and constantly enforcing
the procedures that have been put in place.
Key Controls
Key controls either alone, or in conjunction with other controls, significantly reduce the likelihood and/or
impact of risks. There may be more than one key control for a risk. Elements to be considered when
determining a key control should include the following.
• Is the key control the only barrier/layer of protection available for the occurrence of the event or the
impact exposure (e.g. a firewall barrier to stop external cyber threats)?
• Can the key control be used to prevent or mitigate multiple risk causes/impacts? Is this key control
independent from other controls?
Pdf_Folio:387
Effective There is little scope for improvement. • Processes (controls) operating as intended
and aligned to policies/procedures
• Subject to ongoing monitoring
• Reviewed and tested regularly
Adequate There is some scope for improvement. • Processes (controls) generally operating as
intended yet inadequacies exist
• Limited monitoring
• Reviewed and tested, but not regularly
Inadequate There is a need for improvement or action. • Processes (controls) not operating as
intended
• Processes (controls) do not exist or are not
being complied with
• Have not been reviewed or tested for
some time
Pdf_Folio:388
Differentiate risk evaluation and risk treatment. Provide an example of a typical task from each
process.
Organisation risk profile Reporting to Risk and Audit Committee an outline of Six-monthly
the organisation’s risks.
The risk management report to the Committee is
prepared half yearly and follows a consultation and
review process with risk owners, and endorsement by
accountable Vice President and CEO.
Risk input into two- and five- A risk review is conducted and incorporated into the Annually
year plan business planning and budgeting process.
Leadership team meetings Individual sessions with key business units across the Monthly
(VP and Head of Functions) organisation to ensure risk management and process Quarterly
health indicators reside on each agenda. Bi-annual
Planning session to review content for the Risk and
Audit Committee and plot a way forward. This includes
identification of new risks, analysis or review of
existing risks and control assessment, consideration of
any other inputs.
Pdf_Folio:389
SUMMARY
In this part, we have established the need for effective risk management and explored a comprehensive
risk management process based on best practice guidelines from ISO 31000. We noted that in the context
of data-related risks, the information security standards ISO 27001 and ISO 27005 are also important.
Effective risk management requires identification, analysis and evaluation of risks, development and
implementation of risk treatments and ongoing monitoring. This ensure the business’s risk profile matches
its risk appetite — how much risk it is willing to accept to achieve its objectives. Effective risk management
relies on strong governance, which will be discussed in part B.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
5.1 Assess risks and design mitigation strategies to ensure data security for organisational
outputs.
• A risk management process based on ISO 31000 is best practice for the assessment and
treatment of risks.
• Data security practices are further informed by ISO 27001 and ISO 27005.
• A business continuity plan and disaster recovery plan, based on a business impact assessment,
are important aspects of ensuring the business can continue to achieve its objectives.
• A number of innovative technologies are available to help an organisation manage data-
related risks.
Pdf_Folio:392
GOOD GOVERNANCE
We introduced the concept of governance in section 1.1. The actions, processes and traditions (which are
put in place to support risk governance) focus on the identification, assessment, evaluation, communication
and management of risks.
Good governance creates appropriate systems and controls within the organisation to ensure a balance
between appropriate risk-based decisions and commercial decisions intended to develop, grow and evolve
the business.
Different organisations require different governance strategies. While the importance of good gover-
nance becomes more significant as an organisation grows, small organisations benefit greatly from good
governance as well.
Accountability
Authority
Decision making
Governance
Source: CPA Australia 2021.
Pdf_Folio:393
Authority
Authority is the ability or power (typically through authorisation) to direct, command and monitor
activities. Authority should have set limits and escalation protocols within an organisation.
Ensuring there is clear understanding and authority for responsibilities includes balancing:
• functional versus operational decision making
• global strategies versus regional strategies.
The organisation should establish appropriate authority for key governance functions such as audit, risk,
compliance and legal.
Key organisation questions that relate to authority include the following.
• Who has a voice in making decisions?
• Is the right level of authority set at the right level within the organisation?
• Do people with authority have the correct skills and support from the organisation?
Decision Making
It should be clear who is responsible for making decisions within the organisation. The decision-making
process should involve data gathering, option optimisation and clear criteria for making optimum decisions.
Internal and external stakeholders should be able to understand the organisation’s decision-making
process.
Key organisation questions that relate to decision making include the following.
• How are decisions made?
• Are decision makers clear on the processes involved in arriving at a decision?
• Are there delegated authorities for Board and management in the organisation?
.......................................................................................................................................................................................
CONSIDER THIS
Are accountability, authority and decision-making responsibilities clear in your organisation? If not, what problems
have arisen?
QUESTION 5.5
‘It is always easier to implement a governance framework with risk management earlier rather than
later, and ensure that it is embedded in every business activity.’ Discuss this statement.
Principle 1
Lay solid foundations for management and oversight.
Roles and responsibilities should be clearly defined, and delineated between the Board and manage-
ment. Performance should be reviewed regularly by the Board and management.
Principle 2
Structure the Board to be effective and add value.
To discharge its duty effectively and add value, the Board should be an appropriate size for the business
and comprise committed people with appropriate skillsets and experience for the organisation’s industry.
Principle 3
Instil a culture of acting lawfully, ethically and responsibly.
Create a positive culture from the top down, and continually reinforce the ideals of acting lawfully,
ethically and responsibly across the organisation.
Principle 4
Safeguard the integrity of corporate reports.
Put appropriate processes in place to make sure the integrity of corporate reports is maintained.
Principle 5
Make timely and balanced disclosure.
Make timely disclosures of any information that would cause a reasonable person to expect the share
price, or value of the company, to be significantly affected.
Principle 6
Respect the rights of security holders.
Provide shareholders with appropriate information and give them the ability to exercise their
shareholder rights.
Principle 7
Recognise and manage risk.
Establish a risk framework that is sound and ensure that an effectiveness review is completed
periodically.
Principle 8
Remunerate fairly and responsibly.
Compensate directors and executives with sufficient remuneration to ensure the organisation can attract
and retain high quality personnel that are also motivated to create value for shareholders while remaining
aligned to the organisation’s values and risk appetite.
Source: CPA Australia 2021.
Board
The Board is responsible for defining the risk appetite to meet the strategy of the organisation.
Their responsibilities include:
• defining the risks (nature and extent) allowed to meet the organisation’s strategic objective
• understanding and reviewing the appetite, framework and strategy for managing risk that will be
maintained by management
• monitoring management to ensure the risk management framework is sound and operates within the
risk appetite definition that has been established.
Data Owner
The data owner may be an executive or manager. Being a data owner is not usually their main role. A
data owner:
• delegates responsibility for data protection to the appropriate custodian(s)
• has the final corporate responsibility for data protection
• determines data classification levels and security requirements for data protection.
Data Custodian
The data custodian is generally a member of the IT team and will be delegated data maintenance tasks by
the data owner. For example, a security administrator may be responsible for enforcing access rights to
data. This would include:
• running backups and restores
• maintaining security controls
• defining user rights.
Pdf_Folio:397
Auditors
Internal and/or external auditors provide independent assurance that the organisation maintains appropriate
security objectives and compliance.
QUESTION 5.6
How can the CIO and CISO complement each other to ensure ongoing improvement in risk
management?
COMMUNICATION OF RISK
Communicating organisational risk management is a challenge when implementing and maintaining
good organisational governance. Poor communication leads to poor understanding, which can cause an
organisation’s risk strategy to become lost and/or become ineffective.
Some consequences of poor communication around risk governance include the following.
• Complacency can arise from a lack of understanding, caused by poor communication, and lead to over-
confidence and insufficient appreciation of risks.
• The organisation could become exposed to lawsuits due to impropriety, investigation by regulators, and
the reputational damage that could follow. Lawsuits, whether they are won or lost, result in unfavourable
outcomes that include lost time and money.
• Poor communication can make it unclear which risk management practices need to be implemented.
This can lead to mishandling of intellectual property and data, resulting in lost or stolen assets.
Pdf_Folio:398
EXAMPLE 5.2
Wirecard
A lack of governance contributed to Wirecard’s demise as follows.
• 1999. Wirecard is founded and begins offering online payment services for websites. Later, they expand
to offer electronic payment transaction services and risk management.
• 2005. Wirecard avoids the usual scrutiny required to list on the German stock exchange when it takes
over the listing of a defunct call centre. Wirecard acquires the online bank X.com and renames it
Wirecard Bank. With the purchase of the bank comes a bank licence and licensing for Visa and
Mastercard, meaning Wirecard can issue credit cards and handle money for merchants.
• 2011–14. Wirecard makes a series of obscure acquisitions after raising €500 million from shareholders.
Wirecard claims the rapid growth is organic and that they have also acquired superior technology.
• 2015. Questions are raised about inconsistencies in Wirecard’s accounting. Wirecard is accused
anonymously of money laundering. Wirecard denies it and the German financial regulator launches
an investigation into the accuser.
• 2017. Wirecard’s auditor reports a clean audit and share prices rise.
• 2018. A series of investigations sparked by whistle-blowers become public. Wirecard announces it will
sue the Financial Times and Singapore authorities who have launched a criminal investigation after
raiding Wirecard offices. Meanwhile, Wirecard’s auditor approves the 2018 accounts with only minor
qualifications.
• 2019. Under pressure from its own investors, after more whistle-blower stories are published, Wirecard
conducts a special audit, using new auditors.
What went wrong?
Wirecard seemed destined for failure near the end, but governance must have been lacking for things to
get as bad as they did.
Wirecard seemingly suffered from both questionable business dealings, which were potentially deli-
berate, and poor governance decisions in terms of authority, decision making and effective business
practices (particularly with accounting practices).
How could good governance have mitigated some of Wirecard’s problems?
• Strategy review. When Wirecard made a series of obscure acquisitions there was an opportunity for the
Board to question the organisational strategy.
• Accountability and transparency. Fundamental governance principles of recorded minutes at executive
meetings was lacking. Minutes not only promote accountability and transparency; they also facilitate
effective audit and regulatory reviews.
• Financial oversight. When questions were raised about inconsistencies in Wirecard’s accounting, the
organisation had the opportunity to revise oversight of their accounting practices.
Pdf_Folio:399
FIT-FOR-PURPOSE
A key factor in good governance for an organisation of any size is creating a risk appetite statement. An
organisation can have good risk management policies in place, but if there is weakness in risk culture, the
implementation of these policies can break down.
A risk appetite should define what types of risks are acceptable to pursue and thus what new products,
customers, and markets should be targeted. A risk appetite statement needs to be used to help embed
systematic risk management into every part of the business. This will make sure that the tolerances
and limits of risk, which have been determined in the risk appetite statement, can be used consistently
throughout the organisation to make informed decisions.
Risk tolerance is the amount of risk a business will assume in order to achieve its financial goals. This
should be aligned to the risk culture and should be considered when the risk profile is understood. All
businesses have a certain level of risk appetite; otherwise there would be no growth.
Risk tolerance is the tactical component of risk appetite and is a measure of the maximum amount of
risk the business can manage (Department of Finance 2020). The organisation will define what levels of
risk are acceptable for a specific risk or category of risk.
Considerations for managing risk include:
• what risks need to be escalated
• what re-allocation of resources are required
• what current controls are in place
• what potential new treatments would be required.
Tolerability is a measure of acceptance of the risk level remaining after taking account of the
effectiveness of controls and mitigating actions. Risk tolerability is typically included within the risk
management framework, with a relevant example of risk tolerability in order to provide clarity.
Pdf_Folio:400
An Agile Approach
Many technology companies and other smaller businesses now operate using Agile work practices. This
typically involves flattening the traditional hierarchies in organisations and creating smaller self-led teams
to tackle issues in two-week ‘sprints’. Due to the smaller team size and responsiveness required, Agile is
an approach which lends itself to smaller, more dynamic organisations rather than larger or more heavily
regulated businesses such as banks.
Agile can work in opposition to traditional corporate governance — especially at the project level where
Agile involves an iterative and incremental approach which is harder to monitor. For example, a team
working in rapid two-week sprints could miss compliance issues that the Board needs to address as there
could be two or more sprints between each Board meeting.
However, Agile and governance do not have to be mutually exclusive. Compliance checks can be worked
into the Agile cycle as ‘gates’ that teams must go through before advancing to the next stage of the project.
For example:
1. Checkpoint at review of the scope and schedule.
2. Checkpoint after preliminary design decisions.
3. Checkpoint before customer testing.
4. Checkpoint when product/service ready for launch.
5. Checkpoint for ‘lessons learned’ or post implementation reviews to build in learnings to future projects.
QUESTION 5.7
Assess the pros and cons of Agile work practices from a governance perspective.
The
organisation’s
values, strategy
and vision
Formal Informal
Demonstrate formal culture in the boardroom Demonstrate informal culture in the boardroom
through processes and an agenda. through board dynamics.
• Recruit members and directors who represent • Recruit a dynamic Board who have a range of
organisational values. relevant experience and diverse opinions.
• Evaluate the Board and governance policies • Align decision making to the organisation’s
against organisational strategy and vision. culture.
• Form agendas that prioritise the organisation’s • Encourage dynamic direction using critical
culture. thinking skills and emotional intelligence.
Develop formal culture within the organisation through Develop informal culture within the organisation
reporting to the Board. from top down through sensing and send the
• Report to the Board in line with organisational right message.
culture. • Provide the opportunity for staff to network with
• Develop reports, review processes and measure the Board and spread an organisational culture
performance in line with the organisational culture. organically.
• Feed compliance information to the Board through • Use walk-around management techniques to
formal channels. hear and acknowledge employees’ cultural
suggestions and hear concerns.
• Employees and management take the pulse of
other organisations’ cultures and make internal
suggestions for cultural improvement.
These approaches were largely reflected in the Banking Royal Commission findings (Allens 2019).
List the governance and control issues that organisations should consider when planning an
implementation of AI for workplace automation.
Step 3: Policies, procedures, Data governance establishes the policies, guidelines, procedures,
standards and oversight standards, roles, responsibilities and ownership of the data, within
the data management process.
Step 4: Management The policies and procedures established in the previous step
inform the data management tasks of creating, collecting, using,
maintaining, protecting, storing and controlling data.
Step 5: Data The data has now gone through a management process that is
aligned with the organisations goals, rules and procedures, and with
external compliance for how data needs to be managed.
Pdf_Folio:405
Ethics
Considering ethics, in the context of data, requires us to define what is morally acceptable in terms of
creating, collecting, using, maintaining, storing or controlling data. The consequences of failing to uphold
ethical standards can include a negative impact to an organisation’s reputation.
Data ethics principles generally include:
• informed consent
• transparency
• ownership and control
• fairness and equality
• preserving privacy.
Privacy
Privacy can be defined as freedom from being monitored or disturbed by others. In the case of data,
this applies to the individuals or organisations that the information relates to — an organisation has an
obligation to protect the privacy of individuals by protecting data that is identifiable with those individuals.
The consequences of failing to respect the privacy of individuals or organisations can include harm to an
organisation’s reputation, as well as legal and financial costs.
Data ethics and data privacy will be discussed in more detail in section 5.17.
Pdf_Folio:406
This can include intentional and accidental acts or omissions, which can arise both internally and
externally for an organisation.
The goal of cybersecurity is to protect an organisation’s operations or IT assets against cyberthreats.
Data Security
Data security is practically synonymous with cybersecurity. They both consist of the same infrastructure,
principles, threats, controls and mitigations. Technically, cybersecurity might include protecting against
a few non-data-related threats like making a smart car crash or an industrial control system malfunction.
However, the same security practices that help prevent these issues are also used in data security.
Like privacy, the consequences of failing to secure information and technology assets can include:
• damage to the organisation’s reputation
• legal sanctions
• other financial costs, such as compensation for those affected by a security breach.
The main objectives of cybersecurity and data security are confidentiality, integrity and availability of
data. Data security will be discussed in more detail in section 5.11.
Compliance
Compliance refers to obeying laws or regulations. In the context of data governance, this means complying
with laws or regulations related to privacy or data security. The consequences of failing to comply with
these rules can include the imposition of sanctions or financial penalties on an organisation. This is
discussed further in part D.
Pdf_Folio:407
Optimising
Continuous
Managed improvement
Measured, focus
Defined monitored
Organisation and controlled
Repeatable based,
Project based, proactive
Initial manageable
Unpredictable, but informal
reactive, poor
controls
A maturity model is a tool that is used to develop, assess, and refine an expansive program. As measuring
performance based on return on investment (ROI) or reduction of cost is inappropriate for data governance
programs, another method must be constructed to assess effectiveness.
SUMMARY
In this part, we have defined risk governance, examined the elements of risk governance and what standards
apply to good governance, described governance roles and responsibilities, and discussed data governance.
Risk governance requires the culture of the organisation to be shaped so that people are aware of and
engaged with the appropriate risk management processes of the organisation. A positive governance culture
arises from clear accountability, good risk monitoring and the willingness to learn from mistakes.
The future challenges for governance relate mainly to new technologies which fundamentally change
business processes and introduce a range of new risks, thus prompting a need for increased emphasis on
data governance. Governance is also influenced by changing expectations that organisations will address
corporate social responsibility.
Pdf_Folio:408
KEY POINTS
5.1 Assess risks and design mitigation strategies to ensure data security for organisational outputs
• Data governance aims to create strategies and implement policies and procedures that address
the ethical, privacy, security and compliance goals of the organisation.
5.3 Analyse the requirements for an effective risk governance framework to support the
development of risk management policies including security and privacy policies.
• An effective risk governance approach addresses accountability, authority and decision making.
• A culture of risk management should be created in the organisation, led from the top.
• Data governance aims to create strategies and implement policies and procedures that address
the ethical, privacy, security and compliance goals of the organisation.
5.4 Evaluate existing risk governance frameworks and risk management policies to ensure the
ongoing improvement of the policies and frameworks.
• The elements of good governance provide a foundation from which to evaluate risk management
performance.
Pdf_Folio:409
Other
Analog Digital
things than
information information
information
ICT security
Information security Cyber security
CPS 234 security
ISO/IEC 27001 NIST
standard – APRA
IoT-connected
Email devices
Cloud Employee
services PCs
Employee Employee
intranet intranet
Safe zone Employee Mobile Safe zone Email
PCs devices
Physical Physical
records records
Servers
Public Wi-Fi
Pdf_Folio:412
NIST
The NIST Cybersecurity Framework (NIST CSF) provides voluntary guidance around cybersecurity
risk. It is a regarded as the industry standard and best practice in cybersecurity. Utilising the NIST
cybersecurity framework, along with appropriate cybersecurity controls will complement and support an
effective ISMS.
The NIST CSF Core is a set of desired cybersecurity activities and outcomes, arranged using simple,
non-technical language to facilitate communication between stakeholders. The Core comprises three parts:
1. functions
2. categories
3. subcategories.
Functions are the backbone of the Core and are the five main elements of a successful cyber-
security strategy. The five functions included in the Core are identify, protect, detect, respond and
recover. Think of them as the basic incident management tasks. They apply not only to cybersecurity
risk management, but to risk management in general. Each function of the Core contains categories
used to identify specific tasks or challenges. The categories help define the cybersecurity objectives.
Table 5.9 shows the categories within each function in the NIST framework.
Function Category
Pdf_Folio:413
ISMS
An ISMS is a combination of processes that together help an organisation to manage its information
security by assessing risks and reducing them through proactive action. An ISMS will assist in protecting
information and will incorporate a number of controls and supporting activities within the framework.
The controls are a combination of people/process/technology controls which protect, detect, and respond
to information and security threats. Figure 5.15 describes some of the key characteristics of an ISMS.
Control optimisation
will be tailored to your Contiuous improvement
ISMS assists with
organisation, with processes will assist
optimum decision making
ISO/IEC 27001 including with maintaining a
for your organisation.
114 optional controls strong control environment.
to manage risk.
Pdf_Folio:414
5. Leadership
7. Support
8. Operations
9. Performance evaluation
10. Improvement
Project Mandate
Establishing a project mandate helps initiate the project prior to the project commencing. Project mandates
will help provide a defining document to ensure clarity of purpose, responsibilities, communication
methods, background, scope, interfaces and a business case.
Key information/cybersecurity considerations are as follows.
• The project mandate will include all of the information security and cybersecurity objectives, the
organisation’s current perceived maturity and existing initiatives.
• This will ensure each area of the organisation has clear communication and all interfaces defined.
• The ISMS implementation project needs to be communicated across the organisation, which will ensure
a shared understanding and a more effective risk management program.
ISMS Establishment
Following the completion of the steps 1–8, including the asset identification and risk assessment activities,
the outputs above will be incorporated in the policies and procedures that constitute the ISMS.
The requirements in ISO/IEC 27001, together with best practice guidelines in ISO 27002, will be key
sources of information to successfully implement an ISMS.
Key information/cybersecurity considerations are as follows.
• The ISMS framework documentation, using the ISO/IEC 27001 structure and principles, will help with
ensuring sufficient information for a best practice information security system.
• The ISMS documentation will include consideration of the ISO/IEC 27001 clauses (per ISO 270001
section above).
– 1, 2 and 3. Scope, normative references, and terms and definitions (provision of relevant context and
references)
– Context of the organisation (internal and external factors to consider in achieving objectives)
– Leadership (management support, including roles and responsibilities, and monitoring processes)
– Planning and risk management (development of risk management plans including risk treatment
plans)
– Support (right resources, right people, right infrastructure)
– Operations (execution of plans and processes)
– Performance evaluation (monitor, measure, analyse and evaluate the ISMS)
– Improvement (continuous improvement and action implementation)
PHYSICAL ASSETS
Physical asset controls focus on securing areas and equipment to deter, detect or prevent unauthorised or
unintentional access.
ISO/IEC 27001 provides guidance for controls to manage physical access to physical assets. Example
control types include:
• implementing perimeter security
• having entry controls
• designating secure areas
• limiting public access
• conducting equipment maintenance
Pdf_Folio:418
INFORMATION ASSETS
Security controls for information assets will need to ensure confidentiality (only the authorised persons
have the right to access information), integrity (only the authorised persons can change the information)
and availability (the information must be accessible to authorised persons whenever it is needed).
Example control types for information assets include:
• backups
• data encryption
• firewalls
• access control lists
• subnetting.
TYPES OF CONTROLS
Controls can be implemented by three primary areas of an organisation.
• Management security. Think about management security as the overarching design that provides
guidance for the rest of the organisation. They are the rules and procedures that need to be followed
to implement the security environment.
• Operational security. After the network has been installed, operational security refers to how effective
the controls are. Think of things like who has access, how access is authenticated, and the general
security of the network.
• Physical security. Physical controls protect the physical domain of an organisation. Think of things like
hardware, data, and protection from people – such as restricting access to buildings.
EXAMPLE 5.3
A Complex Attack
Hayes-Mueller Group LLC was a large corporation that owned hundreds of retail stores across North
America. The company used the services of many external contractors for transportation, infrastructure,
and maintenance for its physical stores around the continent.
The company created a web-portal that its contractors and business partners could log into to
submit contract bids, invoices, and integrate with Hayes-Mueller Group operational IT systems for
workflow efficiency.
Temio Security Industries was a vendor that provides physical security monitoring such as alarms and
closed-circuit television (CCTV) within stores. It had access to the company’s contractor web-portal.
Cyber attackers sent a phishing email to an employee from Temio Security Industries. The employee was
tricked into downloading and executing a malicious attachment that installed malware on their computer.
The antivirus software on the workstation failed to detect the malware.
The malware monitored activity on the compromised workstation and recorded keystrokes typed by
the user (a technique known as keylogging). During this period the employee logged into the Hayes-
Mueller Group web-portal using their username and password. This enabled the attackers to steal their
credentials. A short time later the attackers logged into the web-portal using the stolen credentials and
began looking around the portal application, gathering information about the application, the server, and
the company’s IT infrastructure.
Hayes-Mueller Group required two-factor authentication (2FA) for some vendors who accessed highly
confidential data. Vendors that were not in this group did not require 2FA, as there was a belief that
the low-security vendor access was segmented from the high security ones. This belief turned out to
be mistaken.
The vendor web-portal software (which was developed in-house) had several technical vulnerabilities
that the attackers found and exploited, allowing them to take control of the server. The web server was
directly connected to the Hayes-Mueller Group corporate network, which also hosted the company’s other
main servers, databases, and staff workstations.
With control of the server, the attackers were able to communicate with other computers on the
corporate network. They installed their own hacking software on the server and conducted network
scanning to gather information on the other connected computers and their technical vulnerabilities.
This allowed the attackers to move laterally within the corporate network by compromising other servers
through various technical exploitation techniques. Some of the technical exploitation was possible due to
out-of-date software on the machines.
Pdf_Folio:420
Temio
Security
Industries 7
1 Attackers
Attacker’s
2
external server
Vendor
web portal 6
3
Internal
corporate
network
5
POS
terminals
Key:
1. Compromise of Temio Security Industries credentials for
vendor web portal.
2. Exploitation and compromise of vendor web portal.
3. Lateral movement in Haynes-Mueller Group internal
corporate network.
4. Installation of malware on POS Terminals.
5. Moving stolen credit card data to a compromised
internal server.
6. Moving stolen card data to external servers.
7. Retrieval of stolen card data by attackers.
Source: CPA Australia 2021.
Pdf_Folio:421
Advantages
Advantages to outsourcing cybersecurity management to external professionals include the following.
• Supplement internal knowledge. Using the right tools, technology and approach, the organisation can
supplement its cyber security incident response through tools such as a secure monitoring system.
• Cost savings. Due to the distribution of security applications and facilities across the organisation,
significant fees can be incurred. Using a team of external experts through a subscription service can
cost a fraction of the amount that would be required to build internal team capability.
• Security expertise. IT security professionals for in-house teams can be difficult to identify, recruit, and
retain in today’s environment, and due to in-house ICT responsibilities can often struggle to find time
to keep up. External professionals, or consultants, sometimes have the additional capability (and the
incentive) to keep up with the latest security trends.
• 24/7 support. Outsourcing to security providers generally enables the organisation to provide real-time
cyber security reporting 24/7, 365 days a year. This is critical because of the uncertainty of cyber-attacks
and the unpredictability of cyber criminals.
• Compliance management. Cyber security plans that ensure the organisation adheres to specific regula-
tory measures can be difficult to implement. Regulations are constantly changing and if security plans
don’t change with them, there is a risk of being non-compliant. External professionals are sometimes
better placed to keep on top of changes in the industry, and ensure the company meets industry
regulations. The external professional will need to have experience in both compliance and completing
specific audits.
Pdf_Folio:422
Outsourcing Options
Outsourcing options include the following.
• Virtual CISO. Using a virtual chief information security officer (CISO) is a good option for organisations
that may not have the budget for a fulltime CISO, but want robust cybersecurity leadership for their
organisation.
• Consultants. If the organisation only wants to outsource some of its cybersecurity, it could bring in
cybersecurity professionals to assist with different components such as the cyber strategy.
• Cybersecurity-as-a-service (CSaaS). CSaaS is typically a subscription model where part of the organ-
isation’s cybersecurity management is outsourced (Stickman 2018; IT Governance 2021b). CSaaS
companies typically offer services such as:
– consultation
– anti-virus updates
– penetration testing
– cybersecurity health checks
– compliance with laws and industry standards
– business continuity planning
– network vulnerability scan
– training
– cybersecurity incident plans.
The most important part of cybersecurity management is the strategy. External consultants can help
develop a strategy, but the strategy and the strategy communication architecture should always be retained
and maintained inside the organisation.
.......................................................................................................................................................................................
CONSIDER THIS
What circumstances would challenge your company’s current cybersecurity strategy?
QUESTION 5.9
Explain the importance of the three key elements of a mature cybersecurity approach.
QUESTION 5.10
What are the key advantages and disadvantages of outsourcing cybersecurity management?
Pdf_Folio:423
EXAMPLE 5.4
.......................................................................................................................................................................................
CONSIDER THIS
With the benefit of hindsight, Target’s cybersecurity failings seem obvious. Are you aware of any cybersecurity failings
in your own organisation? Do people take basic cybersecurity measures such as using strong passwords and locking
screens seriously?
Pdf_Folio:424
EXAMPLE 5.5
Equifax
In 2017 government hackers gained unauthorised access to a website belonging to the US credit bureau
Equifax. The initial compromise of the website was achieved by the hackers exploiting an unpatched
security vulnerability in the software that was running the site.
Once the attackers gained access to the website, they were able to use the same vulnerability to access
other internal Equifax information systems. The intrusion allowed the hackers to access and steal the
company’s data over 76 days, until their activities were detected and stopped.
The stolen data included full names, dates of birth, home addresses, social security numbers and other
personal data for approximately 148 million US residents, 15 million UK residents and 20 000 Canadian
residents. Additionally, over 200 000 credit card numbers were also stolen from some of these individuals.
The ensuing investigation revealed that in addition to the outdated software, Equifax had poor security
practices including:
• not encrypting sensitive information
• a lack of network and database segmentation (isolation)
• insufficient security monitoring.
If these security controls had been better, the company may have thwarted the breach or detected
it earlier.
In July 2019 Equifax reached a settlement with the US Federal Trade Commission (FTC), Consumer
Financial Protection Bureau (CFPB) and US states and territories in relation to damages caused by
the breach. The settlement was intended to compensate individuals and require Equifax to undergo
organisational changes to avoid future data breaches. The settlement included victim compensation of
USD300 million, USD175 million to US states and territories involved, plus fines of USD100 million (Federal
Trade Commission 2019).
Informed Consent
Consent refers to the permission granted by the creator or owner of data, for it to be collected and used
by an entity for a particular purpose. Informed consent means consent given by a person armed with
sufficient knowledge of the risks and benefits that arise from granting that permission. This knowledge
should describe how their data is collected, stored, used, shared and retained. Ethics requires that consent
be respected by the data collector or user. An example of a policy implementation for this principle is to
maintain an audit trail of where data has come from and what consent has been given for its user.
Transparency
Transparency refers to ‘openness’ about the methods used for analysis and use of the data, both to
the creators and to other practitioners. When transparency allows other data professionals to understand
methods, scrutiny can help to ensure the highest quality practices. Transparency relates closely to informed
consent. It involves making people aware that their data has been collected; of how they disclosed it and
how it was generated or collected from them. It should also disclose how the data will be used and retained,
and communicate the risks and social consequences that arise from its collection and use.
Preserving Privacy
Privacy requires that safeguards are applied to ensure private or sensitive information is not shared
without the creator’s consent. Often legal compliance requirements for privacy have a low threshold, so an
organisation should strive to go above and beyond the minimum (Accenture 2016). We will now discuss
privacy in more depth.
DATA PRIVACY
Data privacy is concerned with the proper handling of personal data, principally consent, notice and
regulatory obligations. More specifically, practical data privacy concerns revolve around:
• how data is legally collected and stored
• whether, or how, data is used, disclosed or shared with third parties
• the integrity of personal information and the ability to correct errors to this integrity
• regulatory restrictions pertaining to the rights of individuals to be left alone from interference or to
access their personal information.
Pdf_Folio:426
Insecure
Privacy non-compliant Privacy compliant
Secure
Source: CPA Australia 2021.
Pdf_Folio:427
QUESTION 5.11
Differentiate cybersecurity ‘attack’ surface and prevention controls and discuss how they relate to
increased use of IoT technology.
QUESTION 5.12
List seven data rights that helps provide a good framework for ethical data practices.
SUMMARY
An awareness of the wide-ranging threats to an organisation’s data provides the digital finance professional
with an appreciation of the need for robust data risk management and cybersecurity practices.
NIST and ISO/IEC 27001 provide the guidance required to implement best practice cybersecurity using
an information security management system. An important of the cybersecurity strategy is the selection of
controls to protect physical and information assets. Controls will provide management security, operational
security and physical security. They will include preventative controls to stop cybersecurity incidents
happening, detective controls to identify when a cybersecurity breach is attempted, and corrective controls
to minimise damage and restore normal operations.
In addition to data security, organisations are ethically and legally obliged to ensure the privacy of
personal data that they hold. The EU’s GDPR regulation is the most stringent international data privacy
regulation and has become a data privacy benchmark for organisations.
The key points covered in this part, and the learning objective they align to, are as follows.
Pdf_Folio:429
5.1 Assess risks and design mitigation strategies to ensure data security for organisational
outputs.
• ISO/IEC 27001 includes a set of implementation steps to assess and treat data security risks.
• NIST includes an extensive set of controls.
5.3 Analyse the requirements for an effective risk governance framework to support the
development of risk management policies including security and privacy policies.
• NIST and ISO/IEC 27001 establish best practice for cybersecurity and information security
management systems.
• The types of security controls that should be implemented depend on whether the organisation is
seeking to protect physical assets or information assets, or both.
• A SOA maps controls to risks and is used in the implementation of an ISMS. The SOA links the risk
assessment and risk treatment stages of risk management.
5.4 Evaluate existing risk governance frameworks and risk management policies to ensure the
ongoing improvement of the policies and frameworks
• NIST and ISO/IEC 27001 establish best practice for cybersecurity and information security man-
agement systems. Ongoing monitoring and review are needed to analyse current risk information,
understand changes, monitor controls, and ensure emerging themes are identified.
Pdf_Folio:430
Leadership
The standard describes the need for commitment to the compliance management system by the senior man-
agement team. There should be a comprehensive compliance policy and this should outline organisational
roles, responsibilities and authorities.
FinTech companies that tend to focus on innovation, and potentially operate with a less mature
compliance culture, need to ensure there is an appropriate focus on senior management leadership.
Planning
The planning for compliance objectives should align with the compliance policy and outline what actions
should be taken to address compliance risks. The objectives need to be measurable and monitored.
Ensuring the right level of skill and appropriate resourcing in up-and-coming FinTech firms (that
sometimes can fail to adequately resource compliance) will help with compliance execution.
Support
Support for the compliance objective should ensure that adequate resources are provided, that staff receive
sufficient training and skills, and that all relevant stakeholders hold an awareness of the compliance
management system.
Consistent and sustained communication about the management system and its objectives should
be undertaken at governance, management, and employee level. Keeping documentation about the
management system is important to ensure that the policies of the organisation are consistent with
the system.
Operation
The operation element describes the establishment of controls and procedures, and discusses implications
for outsourced processes. The key implications in FinTech are monitoring:
• controls (ensuring the establishment of correct internal versus external roles and responsibilities)
• communication protocols (managing appropriate and timely communications, particularly around the
follow-up of issues)
• intellectual property (ensuring correct legal and commercial processes for protection of
IT/digital assets).
Performance Evaluation
The performance evaluation element describes the processes that compliance owners employ to ensure
effective internal controls are in place to monitor regulatory compliance. Supporting processes include the
monitoring, measurement, analysis and evaluation of the compliance management system, what audits are
undertaken, and the management review of the system.
Pdf_Folio:432
Improvement
Improvement describes corrective action that needs to be taken for non-conformity and non- compliance
in the management system and raises issues to be considered for continual improvement.
FinTechs, with their short and rapid innovation cycles, need to ensure improvement cycles are optimised
to maintain a competitive advantage.
Looking Forward
An Accenture survey (Shorten 2020) found that compliance professionals were struggling to keep pace
with the compliance changes in the financial sector. The survey also found that there were three things a
compliance function needed in order to meet today’s — and tomorrow’s — compliance challenges.
• New type of chief compliance officer (CCO). CCOs need to be proactive and lead from within. CCOs
are integral to all elements of ISO 19600:2014. The chief compliance role will typically be responsible
for monitoring the organisation’s compliance program, reporting back to the governance Board and
ensuring improvement of the compliance program. The CCO should be digitally fluent and have good
analytical skills.
• New roles and responsibilities. With changes to operating models, employee roles should be reviewed
to ensure they remain suitable. Responsibility should be assigned to best promote and encourage good
and effective risk management.
• New technology. New technology will be deployed on an industrial scale to identify risk earlier and
detect potential compliance issues. Technology can also drive automation of controls.
• The risk owner is responsible • Provides expertise, advice • Provides highest level of
for compliance management. and case input to 1st line of objectivity within organisation.
• Designs, maintains and defence on compliance and • Provides assurance on
performs effective day-to-day implementing compliance performance of 1st and
compliance control measures. control measures. 2nd lines of defence.
• Supervises correct execution • Sets compliance minimum • Performs audits, where
of compliance control requirements. necessary and requested
measures. • Stays abreast of compliance by the audit committee or
• Identifies compliance risks developments and alerts the Board.
in existing procedures by management to emerging • Typically internal audit
performing risk self- risks. performs these roles.
assessments in cooperation • Monitors adequancy and
with 2nd line of defence. effectiveness of compliance
• Typically, operations and management, controls
management perform and reports.
these roles. • Facilitates risk self-
assessments.
• Typically risk management
and other internal functions
(e.g. HSE, engineering
assurance) perform these roles.
FIGURE 5.20 Outline of KPMG’s compliance framework and recommended responses to compliance risks
Third Parties
Risk: Money laundering and terrorism financing risks and supplier fraud
KPMG’s suggested approach: Conduct KYC and know-your-supplier (KYS) programs. These programs
include applying customer identification procedures to all customers and suppliers. Both programs have
the goal of ensuring there is sufficient information to mitigate risks in third-party relationships that can
affect the organisation.
Corporate Ethics
Risk: Employee misbehaviour or a lack of ethical culture
KPMG’s suggested approach:
• Review the organisation’s code of conduct. Implement a targeted communication program.
• Run a risk management program for the organisation’s people. Provide compliance training.
HR, Competition and Data Protection
Risk: Non-compliance with regulation and compliance risks including competition laws, employee mobility,
overtime and bonuses
KPMG’s suggested approach:
• Evaluate existing programs to ensure they align with regulations.
• Set up a competition law program.
• Assess liabilities.
• Set up digital solutions.
• Run training and communication programs.
• Conduct a gap analysis of existing data retention and protection processes.
• Implement technology-based solutions.
Anti-Bribery, Corruption and Fraud
Risk: Anti-corruption legislation and fraud
KPMG’s suggested approach:
• Run an anti-bribery and corruption program.
• Implement a fraud prevention management system.
Compliance Organisation
Risk: Gaps in program design and the effectiveness of the compliance program
KPMG’s suggested approach:
• Conduct a compliance maturity assessment and a compliance program transformation.
• Set up an industry-specific best practice compliance management system.
Risk: Inadequate controls
KPMG’s suggested approach:
• Ensure there are adequate compliance policies and procedures.
• Integrate compliance controls within existing control systems.
• Create reporting lines for whistle-blowing.
Compliance Investigation
Risk: Incomplete or non-existent investigations
KPMG’s suggested approach: Conduct internal training on investigations.
Source: CPA Australia 2021.
Pdf_Folio:435
EXAMPLE 5.6
Pdf_Folio:436
Pdf_Folio:437
EXAMPLE 5.7
QUESTION 5.13
.......................................................................................................................................................................................
CONSIDER THIS
What circumstances might result in the risk of non-compliance in your organisation?
SUMMARY
In this part, we have provided an overview of compliance and regulations, and how these concepts intercon-
nect with governance and risk management. The compliance industry standard, ISO 19600:2014, provides
guidance for compliance best practice. Developing a compliance framework helps the organisation achieve
compliance. Compliance is a fast-changing area that requires ongoing awareness and review.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
5.2 Evaluate the relevant regulations and tools to provide advice to the organisation on achieving
regulatory compliance.
• ISO 19600:2014 provides guidance for best practice.
• A compliance framework provides a structure to guide compliance.
• Awareness of regulation is a key to achieving compliance. Numerous regulations exist, many of
which relate to AML/CTF and KYC procedures and data privacy.
• RegTech can help an organisation achieve compliance.
5.5 Design an effective action plan to deal with compliance and potential compliance breaches.
• A compliance framework provides a structure in which to ensure responsibilities and reporting.
Pdf_Folio:439
Technological Change
With so much business moving online, new privacy issues are created and with increased access to data
there are a lot of new opportunities for criminals. Biometrics are a major issue for FinTech regulators.
Pdf_Folio:440
Lack of Awareness
Consumers (including businesses) may not fully comprehend all of the risks associated with the FinTech
products and services being promoted to them. The implementation of new FinTech solutions may
also offer manufacturers and distributors an opportunity, whether inadvertent or deliberate, to mislead
consumers in a variety of ways, and to potentially expose them to fraud or significant production or
monetary loss. This begs the question of who is responsible for ensuring customers are fairly and properly
informed (whether it is the vendor, or if the customer must bear at least some onus for understanding what
they are signing up for).
In the event of misconduct, products and services with complex value chains tend to complicate the
responsibilities and accountabilities in the resolution and compensation process. As a result, customers
can become embroiled in lengthy and costly disputes. The paternalistic approach (outlined below) is a
form of response to the challenges associated with this issue.
Governance Risks
The senior management and Boards of firms may lack the crucial awareness and understanding of FinTech
and its inherent risks, which are needed to enable them to identify, control, measure and manage those risks
effectively when adopting new FinTech.
The accountabilities and responsibilities for FinTech and risk management may also be poorly commu-
nicated, or simply lack clarity.
Audit and risk committees, who have the appropriate skills and composition on technology, can provide
Boards and delegated executive committees with the right advice about:
• financial management and sustainability
• controls to prevent and detect fraud and errors
• external financial reporting requirements including audits, risk management, and mitigation
and compliance.
Legal Risk
New FinTech continues to raise challenging, yet-to-be-resolved legal questions, particularly where cross-
border operations must engage appropriately with national legal and regulatory frameworks.
In adopting FinTech, companies may find themselves necessarily changing how they operate — and at
times struggling to meet the conduct-of-business, market-dealing and anti-money laundering requirements.
Regulatory sandboxes allow FinTech companies to test and experiment with a range of regulatory and
compliance techniques free from costly or damaging legal consequence. Sandboxes will be discussed
in more detail later in this module. Some of the other legal risks to FinTech are outlined below in
greater detail.
Financial Reporting
Emerging technologies can substantially impact and affect the financial reporting within organisations.
As FinTech evolves, ensuring the key stakeholders for financial reporting (auditors, audit committees and
management) have a strong grasp of roles and responsibilities is key. As the use of emerging technology
increases, audit techniques are evolving to ensure that management’s internal controls over financial
reporting remains current and sufficient.
Self-Regulation
Many of those in the FinTech industry have taken a pre-emptive and cautious approach through self-
regulation such as ePayments Code and the Code of Banking Practice. While these are both voluntary and
optional codes, there is potential that they will form part of future regulatory focus. These self-regulatory
practices restrict the provision of certain services — even to financially capable customers who are willing
to take on risk, and even if it is legally permissible.
One example is the refusal of products or services to certain customers who a FinTech company may
deem to be unsuitable. A specific example is that the FinTech may determine who should be able to access
reduced interest balance transfers of credit card balances. While this service is of considerable benefit to
customers who are able to responsibly take advantage of this, offering this service may encourage some
customers to avoid confronting their mounting debt problem, and if allowed to engage with the service
this would increase their potential for indebtedness (Khoury 2017). These practices are often motivated
by the industry taking a pre-emptive approach to self-regulation to avoid the imposition of stringent and
complex regulation, which may impede the ever-changing FinTech landscape.
Example 5.8 revisits the example or Wirecard to explore whether stronger regulation could have
prevented its failings.
Pdf_Folio:442
We can expect FinTech-related risks to continue to evolve and increase over time as the adoption of
FinTech solutions, and the industry itself, grows, and regulators will have to work quickly to keep up.
ASIC
The Australian Securities and Investments Commission (ASIC) regulates corporations, markets,
financial services and consumer credit in Australia. ASIC oversees the financial system, promotes
informed participation, and makes information about Australian companies available to the public.
Laws and Licensing Requirements
ASIC administers the Corporations Act 2001, Insurance Contracts Act 1984, Australian Securities
and Investments Commission Act 2001 and National Consumer Credit Protection Act 2009 (National
Credit Act).
ASIC licensing requirements include the following.
• Australian financial services licence (AFSL) for financial services/products (e.g. non-cash payment
products and foreign exchange contracts). Companies with an AFSL must submit to ASIC financial
reports including profit and loss statements, balance sheets, note disclosures, an audit report, and an
auditor’s report.
• Australian credit licence for consumer lending.
Pdf_Folio:443
AUSTRAC
The Australian Transaction Reports and Analysis Centre (AUSTRAC) monitors the financial system
for potential criminal activity, including serious crime and organised crime. AUSTRAC collects financial
information to analyse for financial intelligence.
Companies must submit to AUSTRAC a compliance report that details how they met their AML and
CTF obligations.
Laws and Licensing Requirements
AUSTRAC administers the Anti-Money Laundering and Counter-Terrorism Financing Act 2006.
AUSTRAC licensing requirements include registration for AML/CTF compliance and lodgement of
suspicious matter reports.
RBA
The Reserve Bank of Australia (RBA) regulates and supervises the financial system, along with three
other bodies which make up the Council of Financial Regulators (CFR). RBA’s Payments System Board
is responsible for the safety and stability of the payments system.
Laws and Licensing Requirements
The RBA administers the Payment Systems (Regulation) Act 1998. Payment systems are arrangements
that allow consumers, businesses and other organisations to transfer funds usually held in an account
at a financial institution to one another. It includes the payment instruments —cash, cards, cheques and
electronic funds transfers which customers use to make payments.
Banking Activities
To conduct a banking business within Australia, an entity is required by APRA to be an ADI. As described
in s. 5(1) of the Banking Act, a banking business must both take money on deposit (other than as part-
payment for identified goods and services) and make advances of money, or carry out other financial
activities prescribed by the regulations. Receiving deposits and making advances, or at least the intention
to provide both of these services to customers, are required to set up a banking business (APRA n.d.). The
Banking Act allows only corporations to carry on banking business in Australia. This means APRA cannot
consider applications from associations, partnerships or unincorporated entities.
The Banking Regulations 1966 built on this definition, in certain circumstances, to allow for purchased
payment facilities, and credit card acquiring and issuing. In addition to practices such as retail lending
and deposit taking for individuals, banking activities also consist of acceptance and discounting of local
bills of exchange (such as promissory notes and bank cheques) as well as foreign exchange activities
(converting funds from foreign to local currency or accepting foreign deposits). Primarily, however, lending
and acceptance of funds are the core activities of a banking business.
Financial Activities
Any individual or entity conducting a financial services business in Australia must hold an AFSL to
permit the delivery of those services, unless an exemption applies. ASIC issues AFSLs and supervises the
conduct of its licence holders. Under the Corporations Act, financial services include making a market
for financial products, providing financial product advice, operating a registered managed investment
scheme (i.e. collective investment vehicles), providing traditional trustee company services, and providing
a custodial or depository service. Under the Corporations Act, a financial product is also a facility through
which a person makes a financial investment, manages financial risk or makes non-cash payments. For the
purposes of the Corporations Act, financial products would include securities, debentures, derivatives,
bonds, and so on.
Credit Activities
Any individual or entity conducting activities that relate to a form of credit contract or consumer lease to
which the National Credit Code applies, if that activity is also specified in the National Consumer Credit
Protection Act 2010 (Cth) must obtain an ACL. However, the National Credit Code applies only if the
debtor is a strata corporation or natural person.
Inside or Outside?
Regulation for FinTech organisations can be split into two informal categories.
• Internally facing. Internally facing regulation is concerned with how an organisation regulates itself,
such as risk management and governance. APRA prudential standards are examples of how FinTechs
must uphold standards that manage their liquidity and financial risk.
• Customer facing. Customer-facing regulations are laws designed to protect consumers. These regula-
tions are especially relevant for FinTechs because they can significantly impact on customer experience.
For example, customers must be appropriately identified based on the KYC system in order to comply
with AML laws. This means FinTechs must ask more questions and gather more information than other
technology companies that do not have to comply with such laws.
Human resourcing, and external advice, must be correctly set up to comply with regulatory requirements
and licensing requirements, which can sometimes be extensive.
GLOBAL REGULATION
Where an organisation does business affects how many regulations apply. With online financial products
the jurisdictional boundaries can be less defined.
Different countries have different regulations. While no regulation is truly global, some regulations have
far-reaching impact beyond national boundaries. There are some international regulations that Australian
providers need to know about.
GDPR
The GDPR is a regulation on data protection and privacy in the EU. The GDPR has extra-territorial effect
if an entity processes personal data of an EU data subject. So, if a FinTech business targets a client in
Europe from outside of Europe, the FinTech will be responsible for GDPR compliance. The penalties for
non-compliance are high. If personal data is handled in Europe, or personal data is processed for European
data subjects, compliance with the GDPR is required.
Pdf_Folio:446
CRS
The Organisation for Economic Co-operation and Development (OECD) developed the Common
Reporting Standard (CRS) in 2014. It is an information standard for the automatic exchange of
information (AEOI) regarding financial accounts on a global level, between tax authorities, whose purpose
is to combat tax evasion. It is very similar to the Foreign Account Tax Compliance Act 2010 (FATCA),
the original US regulations. Many FinTechs that operate payments or bank accounts often have to comply
with FATCA/CRS obligations globally.
.......................................................................................................................................................................................
CONSIDER THIS
To what extent is global regulation relevant to the activities of your organisation or your client’s organisation?
Australian financial services businesses must report financial transactions and suspicious matters to
AUSTRAC. AUSTRAC uses the reports to analyse the data and create intelligence reports. These reports
are shared to government agencies to uncover financial crimes. This reporting helps put responsibility on the
banks to find out who is benefiting from transactions. The onus is on them to monitor the transactions.
Example 5.9 examines what can happen when an organisation turns a blind eye to FinTech regulatory
requirements.
EXAMPLE 5.9
EXAMPLE 5.10
Self-Regulation: BNPL
Between exponential growth and consumer uptake in the FinTech sector lies the gap where regulators
struggle to catch up. To build consumer trust in their products and quiet their critics, FinTech companies
may need to self-regulate. This is particularly true in the case of Afterpay and other BNPL products and
services, including Klarna, Zip, OpenPay, Latitude, humm and Brighte. The Australian Financial Industry
Association (AFIA) BNPL Code of Practice came into effect on 1 March 2021, having been developed
in response to a Senate inquiry and ASIC’s calls for stronger protections due to increasing numbers of
consumers using BNPL products while BNPL companies continued to operate outside traditional credit
laws. The Consumers’ Federation of Australia submitted to AFIA that BNPL arrangements should be
controlled under the National Credit Act.
BNPL companies are contentious on two major fronts. These companies issue contracts that
prevent merchants passing on surcharges to customers, and they also feature a perceived lack of
consumer safeguards.
Contracts that Prevent Merchants Passing on Surcharges to Customers
In February 2020, following word that RBA intended to review its ‘no surcharge’ rule and possibly take
policy action, Afterpay argued in a submission to RBA that BNPL companies should have the right
to prevent retailers passing on their fees to customers. For instance, where a retailer may add a 0.6–
1.5 per cent surcharge for payment by debit or credit card at checkout or POS, thereby covering the
actual cost of the transaction for the retailer, Afterpay would continue to have the right to prevent the
retailer adding a similar surcharge each time a customer pays by Afterpay.
Afterpay’s fixed transaction fee of 30 cents plus a 3–7 per cent commission on the sales value translates
to a much higher cost to the retailer than that of a credit card company.
Afterpay’s argument is that BNPL represents an innovative, growing sector worthy of self-regulation,
and that it should neither constitute a payment system nor have regulated fees due to its small size,
uniqueness and value proposition to both consumers and merchants.
Although not all retailers necessarily want to surcharge — just as not all retailers allowing customers
to purchase by debit or credit card opt to pass on the debit or credit card surcharge now — the option
to do so would empower retailers to negotiate for fairer fees. Afterpay argues that the enforcement of
surcharges could do more than reduce revenue or damage business models — it could curtail innovation.
Yet by being able to prevent merchants from passing on their fees to consumers, BNPL companies gain
an unwarranted advantage by being able to market themselves as free to all customers who pay on time.
Lack of Consumer Safeguards
Another factor setting BNPL apart from other payment options is its window into the Gen Z and Millennial
demographics, often regarded as the most difficult to reach consumer market in the world. Yet the second
major criticism levelled at BNPL relates to its lack of consumer safeguards, largely in the form of credit
checks that would prevent vulnerable consumers from purchasing beyond their means.
While the Code of Practice includes commitments to ‘only provide our BNPL Products or Services
to customers aged 18 and over’, and to ensure that late fees are ‘fair, reasonable, and capped’, ASIC
criticised it as insufficiently vague and lacking in consumer protections, while consumer groups also raised
concerns. For instance, what constitutes a ‘fair, reasonable and capped’ late fee?
Pdf_Folio:449
.......................................................................................................................................................................................
CONSIDER THIS
What are the pros and cons of self-regulation from an industry perspective, a regulator’s perspective and a
customer perspective?
Marketing
ASIC ensures financial products (including credit products) are not sold in a misleading way to consumers.
ASIC’s Regulatory Guide 234 Advertising financial products and services (including credit): Good
practice guidance gives industry guidance on what can and cannot be said when selling financial products
(ASIC 2012).
The following advice is included in the guide.
• To say financial products are ‘free’ will be subject to scrutiny from ASIC.
• If an organisation or person is selling a financial product, they will need to make sure they are
appropriately licensed with an AFSL to permit them to give financial product advice.
Pdf_Folio:450
Onboarding
This involves collecting and verifying at least the full name and address, or date or birth or three years of
transaction history of the customer.
Generally, KYC information collection requires either face-to-face verification or, in the case of
FinTechs which almost exclusively operate online, electronic verification against independent and reliable
electronic databases.
The KYC process can be complicated, and the collection of the right information that may be verified
instantaneously using third party vendors is particularly important.
Transactions
A FinTech allowing transactions to be performed across borders will have to consider not only regulations
in Australia but also the jurisdiction that the payment is being made to. This includes ensuring services
are not provided to those that involve transactions that may be subject to sanctions, including United
Nations sanctions.
Sanctions include regulations on individual countries (six countries in the world have active United
States Sanctions as of 2020) or individual persons, such as terrorists, in some countries. In Australia,
DFAT offers a publicly available consolidated list of individuals and entities that have sanctions against
them (DFAT 2020).
FinTech businesses must take steps to ensure they do not breach sanctions laws by having appropriate
screening procedures in place.
Reporting of Issues
Each transaction is subject to transaction monitoring under AML/CTF regulations. This requires FinTech
providers to spot suspicious transactions and report them, within prescribed time limits, to AUSTRAC.
Disputes
Disputes arise from time to time with consumers. Generally, all AFSL holders will have to be members of
an external dispute resolution provider.
The Australian Financial Complaints Authority (AFCA) is the external dispute resolution scheme for
consumers who are unable to resolve complaints with member financial services organisations. When
members sign up to AFCA, they effectively allow AFCA to decide on disputes so that consumers do not
have to take FinTechs and other organisations to court for relatively small matters (AFCA 2020).
FinTechs must therefore have an internal dispute resolution process as well as a process for dealing with
external disputes raised to AFCA.
Pdf_Folio:451
Scalability
AML/CTF laws are likely to impact on a provider’s reporting requirements, with mandatory breach
reporting. Practical scalability may become an issue requiring RegTech to assist with moving from a
handful of customers to hundreds of thousands, or millions.
QUESTION 5.14
SUMMARY
FinTech providers operate in substantially different ways to conventional financial services providers
and institutions. Regulation is essential to balance the interests of consumers and the industry (both
new entrants and established players). Regulation focuses on consumer protection while enabling and
encouraging innovation and competition in the industry. Some parts of the FinTech industry have sought
to develop self-regulatory approaches to avoid the imposition of external regulation.
The increasingly global nature of finance requires many organisations to be aware of international
regulations, as these have effect when data and transactions cross national boundaries. As an evolving
area, it is crucial for FinTech and other financial industry organisations to be aware and prepare for
regulatory change.
Pdf_Folio:452
KEY POINTS
5.2 Evaluate the relevant regulations and tools to provide advice to the organisation on achieving
regulatory compliance.
• FinTech regulation is not yet well established in Australia. FinTechs have been provided with
regulatory sandboxes to enable them to innovate. However, more formal regulation is likely to
develop over time, and will seek to balance the interests of stakeholders.
• Some parts of the FinTech sector have begun to develop self-regulatory approaches.
• Transactions and data that cross borders or involve parties in other jurisdictions may be subject
to international laws. Compliance with GDPR will usually achieve compliance with most other
international laws relating to data protection and privacy.
Pdf_Folio:453
WHAT IS REGTECH?
RegTech is the use of technology to improve regulatory processes. For example, onerous compliance
obligations such as record keeping and reporting could be solved by using DLT, where transparent audit
trails of transactions could be provided to regulators on demand.
RegTech is developing in parallel with — and partly in response to — FinTech. It is becoming an
increasingly important facet of the financial services industry. There are two main drivers for RegTech:
• increased regulatory compliance
• technology innovations.
RegTech is used to improve regulatory processes including regulatory monitoring, compliance
and reporting.
Initially introduced in many organisations to improve internal processes (in order to reduce compliance
costs), RegTech became increasingly important as a way to monitor risk in order to comply with increasing
regulatory requirements following the global financial crisis (GFC). The GFC was triggered by the practice
of financial lending companies packaging risky loans into high credit rating securities (RBA n.d.). Many
of these lenders were not subject to the same regulations as banks and when the residential market in the
United States experienced a downturn, borrowers defaulted on their loans. In addition to the lax regulations,
there were cases of fraud including overstating of borrowers’ incomes and overstating the safety of
the securities.
The GFC prompted new regulations intended to prevent high-risk loans being packaged together and
offered as securities, and to prevent fraud. This included ringfencing investment and retail banking, plans
for recovery of loans, and resolutions, revisions and reforms to the more recent prudential standards by
APRA (RBA 2019).
Following these initial regulatory developments, jurisdictions around the world have continued post-
GFC to implement growing numbers of regulatory requirements of financial institutions, in order to
protect consumers. Regulators are also trying to align their capacity to effectively monitor and analyse
the increasingly digitised nature, and the sheer volumes of data generated by the post-GFC reporting
obligations of the financial market.
The other main driver of RegTech is technological innovation, both in terms of FinTech-related
technology advances which require technological oversight, and also in terms of the technology which
makes RegTech possible. With the advancement of technologies, RegTech has the potential to help build
a better financial system by reconceptualising finance and regulation.
Traditional paper-based, human-powered compliance processes are simply not capable of keeping up
with the depth and breadth of FinTech developments. Many of the same technologies that fuel FinTech
can also be employed to assist with compliance oversight.
Agility
To declutter big data
Analytics
To mine big data
Speed
To process and generate
reports quickly
RegTech holds great promise for financial institutions by easing compliance reporting, providing
insights into possible risks and helping them keep up to date with constantly changing financial
industry regulations.
RegTech can enable progressive organisations to design faster and more robust business processes to
manage their responsibilities around regulation and compliance. RegTech uses analytic tools to mine and
declutter intertwined big data sets, which enable large amounts of data to be processed and analysed in a
shorter period of time. Using ML algorithms, RegTech systems are also capable of reducing the number of
false positives, that human resources would ordinarily have to assess on a case-by-case basis. This enables
the organisation to make better use of its human resources — they can focus on investigating actual threats,
dealing with actual non-compliance and work on other value-adding business activities.
The benefits organisations can expect from implementing RegTech include the following.
• Increased revenue. RegTech may increase an organisation’s competitiveness and in turn, revenue, by
increasing customer satisfaction, especially in the onboarding of new customers.
• Reduced costs. Increased efficiency means that fewer people will be required to manage risk and
compliance. Therefore, this function will have reduced costs.
Pdf_Folio:455
EXAMPLE 5.11
Pdf_Folio:456
.......................................................................................................................................................................................
CONSIDER THIS
Does RegTech have application in your organisation or a client’s organisation? Would this likely be based on big data
analysis, blockchain, real-time monitoring or some other application of technology?
Pdf_Folio:457
Pdf_Folio:458
Pdf_Folio:459
EXAMPLE 5.12
India Stack
The India Stack is a platform established by the Indian Government to enable coordination and integration
of digital services. The approach to RegTech on the India Stack platform is intended to promote financial
services access. It involves four levels.
• Level 1: Biometric identification. A national system of biometric identification was created. Identity is
an important aspect of promoting financial sector access. For this, biometric identification cards were
created. More than 1.1. billion cards were issued over the first 6 years of the platform.
• Level 2: Establishment of bank accounts. This enabled the digitisation of transfers for national services
such as pension, health and other welfare. More than 300 million were created.
• Level 3: Common payment interface. The interface was created so individuals could make payments
through a common system supported by the Reserve Bank of India.
• Level 4: Electronic KYC initiative. This initiative enables individuals to keep digital copies of all the ID
documents they need to meet KYC requirements. It also means that individuals have control of which
financial institutions can access their data.
Pdf_Folio:460
THE FUTURE
Once the sector can overcome these obstacles, RegTech has the potential to be a transformative tool
that will revolutionise financial markets and regulations. A data-driven approach coupled with innovative
technology will enable the real-time monitoring of financial markets.
As markets are becoming more and more reliant on data, large corporations and technology providers
that have access to enormous quantities of data (such as Google, Alibaba and Apple), are more likely to be
able to accurately assess a customer’s credit risk and extend them credit (through a wide variety of services
and schemes), making them strong competitors to traditional financial institutions and banks.
As these non-traditional players enter the financial services marketplace, RegTech will need to continue
to evolve rapidly to provide regulatory oversight of these transformative organisations who primarily work
within the know-your-data paradigm, rather than the traditional KYC framework.
To conclude the module, example 5.13 brings together themes of risk, compliance and RegTech in
relation to a real-world business application.
EXAMPLE 5.13
Treasury, and many of the entities that made submissions, urged the need for caution before
recommending change. This is undeniably right.
Pdf_Folio:461
History shows, as Treasury submitted, that legislative simplification can be a long and difficult task.
Programs to simplify the law relating to income taxation and to reform corporate law have extended
over many years — well beyond the life of a single Parliament. And I do not doubt that simplifying
the law that relates to the financial services industry would be a large task. But there are two parts of
that task that can inform, and I consider should inform, what is done in response to this Report.
First, it is time to start reducing the number and the area of operation of special rules, exceptions
and carve outs. Reducing their number and their area of operation is itself a large step towards
simplification. Not only that, it leaves less room for ‘gaming’ the system by forcing events or
transactions into exceptional boxes not intended to contain them.
Second, it is time to draw explicit connections in the legislation between the particular rules that are
made and the fundamental norms to which those rules give effect. Drawing that connection will have
three consequences. It will explain to the regulated community (and the regulator) why the rule is
there and, at the same time, reinforce the importance of the relevant fundamental norm of conduct.
Not only that, drawing this explicit connection will put beyond doubt the purpose that the relevant
rule is intended to achieve. And, the further consequence will be to highlight the fact that exceptions
and carve outs like grandfathered commissions constitute a departure from applying the relevant
fundamental norm. Emphasising the fact of departure may assist in reducing both the number and
the extent of these qualifications.
Pdf_Folio:462
Pdf_Folio:463
Pdf_Folio:464
Outline the regulatory and compliance issues that FinTech companies should consider when
planning global operations.
SUMMARY
RegTech is the use of technology to improve regulatory processes including regulatory monitoring, com-
pliance and reporting. It is being driven by increased regulatory compliance and technology innovations,
such as cloud, blockchain, AI and big data.
The benefits of RegTech potentially include increased revenue, reduced costs, efficiency gains, reduced
risk and greater consistency in regulatory reporting and compliance requirements regulators are interested
in the use of RegTech to improve reporting and compliance as well as enhance customer experience.
RegTech is a new technology. The barriers to using it include the cost, lack of skills and lack of data, a
fear that it will lead to increasing compliance demands, risks associated with the new technology and the
organisations that supply it and the possibility of biases built into the technology.
Preparing for RegTech needs to look at not just the technology, but also the organisation’s processes and
how its people will be deployed into more value-adding activities.
The key points covered in this part, and the learning objective they align to, are as follows.
KEY POINTS
5.2 Evaluate the relevant regulations and tools to provide advice to the organisation on achieving
regulatory compliance.
• RegTech tools use technology for regulatory monitoring, compliance and reporting.
• The benefits of RegTech may include increased revenue, reduced costs, efficiency gains, reduced
risk and greater consistency in regulatory reporting and compliance requirements.
• Barriers to using RegTech may include cost, lack of skills, lack of data, the possibility it will lead to
increasing compliance demands, risks associated with the new technology and the organisations
that supply it, and the possibility of biases built into the technology.
• RegTech may be built in-house, built by a partner, purchased or accessed via a cloud
service provider.
5.5 Design an effective action plan to deal with compliance and potential compliance breaches.
• RegTech integrates technology into compliance solutions drawing on a range of approaches,
including advanced applications such as ML to automatically identify regulatory requirements, thus
facilitating compliance.
Pdf_Folio:465
REFERENCES
ACCC (Australian Competition and Consumer Commission) n.d., ‘Consumer data right (CDR)’, accessed August 2020, https://
www.accc.gov.au/focus-areas/consumer-data-right-cdr-0
Accenture 2016, ’Building digital trust: The role of data ethics in the digital age’, accessed 27 April 2021, https://www.accenture.
com/_acnmedia/PDF-22/Accenture-Data-Ethics-POV-WEB.pdf
Australian Institute of Cultural Directors 2016, ‘Types of directors’, accessed October 2020, https://aicd.companydirectors.com.
au/-/media/cd2/resources/director-resources/director-tools/pdf/05446-1-10-mem-director-t-bc-types-of-directors_a4_
web.ashx/
Allens 2019, ‘What this means for governance and directors’ duties. Better governance for all: It’s not just lessons for the banks’,
accessed October 2020, https://www.allens.com.au/campaigns/financial-services-royal-commission/what-this-means-for-govern
ance-and-directors-duties/
APRA (Australian Prudential Regulation Authority) 2019, ‘Transforming governance, culture, remuneration and accountability:
APRA’s approach’, accessed October 2020, https://www.apra.gov.au/sites/default/files/Transforming%20governance%2C%20c
ulture%2C%20remuneration%20and%20accountability%20-%20APRA%E2%80%99s%20approach.pdf
APRA (Australian Prudential Regulation Authority) 2020, ‘Understanding and managing the financial risks of climate change’,
accessed October 2020, https://www.apra.gov.au/understanding-and-managing-financial-risks-of-climate-change/
ASIC (Australian Securities and Investments Commission) 2012, ‘RG 234 Advertising financial products and services (including
credit): Good practice guidance’, accessed October 2020, https://asic.gov.au/regulatory-resources/find-a-document/regulatory-
guides/rg-234-advertising-financial-products-and-services-including-credit-good-practice-guidance/
ASIC (Australian Securities and Investments Commission) 2019a, ‘ASIC RegTech Initiative Series 2018-2019 Results and
Primary Observations’, accessed September 2020, https://asic.gov.au/for-business/innovation-hub/asic-and-RegTech/asic-
RegTech-initiative-series-2018-19/results-and-primary-observations/
Barlow, J 2017, ‘Compliance and risk management: Interrelated, but not the same’, Board Effect, 15 March, accessed August
2020, https://www.boardeffect.com/blog/compliance-risk-management-interrelated-not/
Bartlett, J 2019, ‘The £4bn OneCoin scam: how crypto-queen Dr Ruja Ignatova duped ordinary people out of billions — then went
missing’, accessed October 2020, https://www.thetimes.co.uk/article/the-4bn-onecoin-scam-how-crypto-queen-dr-ruja-
ignatova-duped-ordinary-people-out-of-billions-then-went-missing-trqpr52pq/
CapGemini 2018, ‘Leveraging the opportunities of open banking: The KYC platform’, 29 October, accessed August 2020,
https://www.capgemini.com/nl-nl/2018/10/leveraging-the-opportunities-of-open-banking-the-kyc-platform/
Chau, D 2020, ‘Westpac expects $900 million penalty for breaching money laundering laws’, 14 April, accessed September 2020,
https://www.abc.net.au/news/2020-04-14/westpac-1.4-billion-hit-to-first-half-earnings/12146360
CIO Australia 2017, ‘What is GRC and why do you need it?’, CIO, 11 July, accessed September 2020, https://www.cio.com/
article/3206607/what-is-grc-and-why-do-you-need-it.html
CISO 2016, ‘Understanding difference between Cyber Security & Information Security’ CISO Platform’, accessed October 2020,
https://www.cisoplatform.com/profiles/blogs/understanding-difference-between-cyber-security-information
CPA Australia 2019, ‘Regulatory burden — The impact of regulatory complexity’, accessed March 2021, https://www.cpaaustralia
.com.au/public-practice/toolkit/regulatory-burden
Pdf_Folio:466
Pdf_Folio:468
Pdf_Folio:469
GLOSSARY 471
business impact assessment (BIA) A process that establishes key objectives for disaster recovery.
business process management (BPM) A holistic discipline to optimise business activity flows in
support of enterprise goals, spanning application systems, human resources, customer relationships
and external business entities.
card-based payments Payments that involve a physical card, such as a credit or debit card, and a POS
terminal, where users need to insert or swipe their card.
cashless society A society in which the use of physical money would be eliminated.
causation The presence of a cause-and-effect relationship between two variables.
Central Bank Digital Currency (CBDC) scheme A (proposed) digital form of a country’s fiat
currency, which would be backed, regulated and administered by governments and work on either a
national or a supranational level.
centralised virtual currencies A virtual currency controlled by a single administrator (usually the
developer of the platform) who acts as the intermediary in transactions.
classification The predictive analytics process of determining what category or class a particular data
item should be labelled, which the target variable is categorical.
classifying ML ML that focuses on segmenting and clustering data in large data sets.
clustering An unsupervised learning technique that is used for automatic identification of natural
groupings of items in a data set.
cognisability (of money) See recognisability.
cognitive automation An approach to automation that creates new insights through data discovery that
combines internal and external data and then uses those insights in autonomous decision making.
cognitive bias When prejudice influences decisions or thinking, preventing clear, objective judgement.
cognitive ML ML focused on recognising and understanding text, speech, photos, audio and video in a
way that aims to mimic human thoughts.
collaborative filtering An unsupervised learning technique used in recommendation systems, which
aim to predict users’ interests and recommend items that would appeal to them.
collaborative networks Networks in which participants are interdependent, thus requiring a mutual
commitment to working with other members of the network.
collectively exhaustive A set of all possible events, such that an event outside that set cannot
possibly occur.
Common Reporting Standard (CRS) An OECD information standard for the automatic exchange of
information regarding financial accounts on a global level, between tax authorities, whose purpose is
to combat tax evasion.
competitive advantage The ability of a business to achieve better performance than its competitors and
defend that advantage over the long term.
compliance The process of ensuring adherence to all business rules, policies and guidelines that are
established internally or externally.
compliance framework A framework that sets out an organisation’s approach to managing its
compliance obligations and mitigating compliance risk, in order to achieve its compliance objectives.
conditional probability The likelihood of an event occurring based on the occurrence of another event.
confidentiality The quality of data and information that exists when the data is private and secure with
restricted view access.
consent The permission granted by the creator or owner of data, for it to be collected and used by an
entity for a particular purpose.
consortium blockchains A hybrid blockchain that combines the single highly trusted entity model of
private blockchains with the low-trust environment of public blockchains by giving pre-selected nodes
control of the consensus process.
continuous data Data with infinite possible values, corresponding to some continuous scale that does
not have gaps.
control documentation The design standard, operating standard and verification approach relating to
a control.
cooperative networks Networks in which participants share information and expertise, while remaining
independent; they interact only when necessary to harmonise efforts.
correlation The degree of relationship between two variables.
correlation coefficient, r A measure of the strength of a correlation.
corrective controls Controls implemented after a breach to minimise damage and restore normal
operation as soon as possible.
472 GLOSSARY
CPS 234 security standard An APRA standard that requires APRA-regulated entities to protect
information against security threats.
critical control A control that significantly reduces the impact or likelihood of a risk.
crowdlending See peer-to-peer lending.
cryptocurrencies A form of decentralised, fungible and convertible virtual currency secured using
strong cryptography, based on blockchain technology.
currency coins Cryptocurrency coins that have their own blockchain within which they function as
a currency.
cybersecurity The protection of data and ICT systems, and the response to attacks.
cybersecurity standards Published techniques or guidelines that guide the information security and
cybersecurity approach, and the processes needed to prevent or mitigate cyber threats.
dashboards A visualisation tool that combines multiple related visuals that update as the data changes.
data A fact or observation.
database management systems (DBMSs) Systems that integrate with data collection systems and
business applications, store the data in an organised way, and provide facilities for accessing and
managing that data.
data governance The oversight, authority, planning and development of policies, guidelines, procedures
and strategy for the access, protection and usage of data.
data-informed decision-making The practice of making decisions based on the actionable and verified
knowledge created through analytics.
data life cycle The existence of data from creation and initial storage to the time when it becomes
obsolete and is deleted.
data literacy The ability to read, work with, analyse and argue with data.
data literacy culture An organisational culture that embraces the use of data-informed decision-making
and fosters the growth of data literacy within an organisation.
data management The processes involved in creating, collecting, using, maintaining, protecting,
storing and controlling data.
data migration Moving data from one system to another system, usually involving improving the
quality of the data.
data mining The use of technologies that incorporate data science principles to produce information
from data.
data-oriented ecosystems The aspect of a digital ecosystem that provides organisations with data that is
relied on to understand customers and to make better decisions.
data ownership Who owns or is responsible for data and how they approach that responsibility.
data privacy Proper handling of personal data, principally in relation to consent, notice and regulatory
obligations.
data science A set of fundamental principles that guide the creation of knowledge from data.
data security How organisations protect their data, including technical safeguards that help ensure data
confidentiality, integrity and availability.
data strategy A strategy that comprehensively reflects the organisation’s data needs, aligned with
business strategy, and incorporating data requirements, analytics requirements, implementation,
communication, interpretation and application.
data-to-ink ratio A measure of how much of the content of a visualisation contains relevant
information.
data wrangling The process of putting data and data sets into a useable format, ensuring that
missing, inconsistent or anomalous values are detected and rectified to prevent corruption of the
analytics results.
decentralised finance (DeFi) The use of a distributed trust mechanism (such as blockchain) to eliminate
the need for central intermediaries in a finance system.
decentralised virtual currencies A virtual currency that uses cryptography to establish trust between
participants and thus eliminates the need for an intermediary.
decision tree Non-linear ML algorithms used for categorical or discrete values. They recursively divide
a training set of data until each division consists of examples from one class.
deep learning An ML approach in which artificial neural networks are used to mimic the complexity of
analysis performed in a human brain’s neural networks. Neural networks process data through a
layered structure of algorithms or decision points and thus a deep learning machine has a greater
capacity to evaluate multifaceted data than standard ML.
GLOSSARY 473
dependent event An event the probability of which is affected by previous events.
descriptive analytics Data analytics that summarise the results of past or current performance.
design standard In relation to a control, what needs to be in place or provided for the critical control to
be effective.
detective controls Controls designed to pick up when a cybersecurity breach is attempted or, if
successful, is underway so that a cybersecurity response can be made.
diagnostic analytics Data analytics that determine cause and effect.
digital currencies A currency that exists only as balances or records stored within databases.
digital ecosystem A dynamic integration of organisational departments, suppliers, tools, systems,
customers and external partners, brought together to increase data flow throughout an organisation and
drive business performance.
digital financial services (DFS) ecosystem The digitalisation of the financial industry; the digital
ecosystem comprised of financial services users, providers, infrastructure and governmental policies,
laws and regulations.
Digital Finance Cube A framework for exploring and analysing digital finance concepts through
interrelationships of digital finance business functions, digital finance technologies and technological
concepts, and digital finance institutions.
digital financing The use of alternative technology-mediated sources of financing.
digital literacy Possessing sufficient knowing about the applications, benefits, costs and limitations
of technologies to be able to assess their value to the organisation and support their effective
implementation.
digital mindset An attitude and set of behaviours that support the adoption of new technologies where
they can add value to business.
digital payment An exchange in which value is transferred electronically from one party to another.
digital roadmap A plan that guides an organisation’s digital transformation by identifying which digital
technology components are the best fit for their ecosystem.
digital transformation The process of moving analogue or traditional ‘non-digital’ technologies fully
into the digital space.
digital transformation strategy A strategy to adopt technologies in a viable, value-adding and
sustainable way to support business continuity and protect productivity and functionality.
digital trust A belief by consumers and other organisations that their digital practices are transparent,
honest, safe and reliable.
digital wallets Device-specific, platform-based, app-based or cryptocurrency based devices, software
or online services that a party can use to electronically transact with another party by means of
digital currency.
disaster recovery plan (DRP) A deliberate and documented approach that details how an organisation
can return to work after an unplanned outage or incident.
discrete data Data that has a distinct set of values, which can be counted and belong to a whole numbers
set (integers).
distributed applications (DApps) Applications that provide a web interface built over the top of one or
more underlying smart contracts.
divisibility (of money) The ability to be divided into small units.
domain knowledge Theoretical and practical knowledge that is specific to a business field, industry
or discipline.
donation-based crowdfunding A form of fundraising typically utilised by individuals or not-for-profits
to pool funding from the crowd in order to finance a project with a social purpose or cause, without the
expectation of having the funds returned.
durability (of money) The quality that money endures long enough to be utilised in any future
transactions.
edge computing The use of technology to move data storage and processing away from central servers
towards the end user, resulting in more efficient, reliable and faster data exchange.
electronic fund transfer (EFT) Payments that involve direct credit and debit transfers.
elementary event An event that consists of a set of one sole outcome.
e-money A regulated digital currency.
EMV A global standard for payments that deals with the authentication and security of chip-card
transactions.
474 GLOSSARY
enabling skills The broad and adaptable behavioural, data analysis and finance skills that employers are
seeking from employees to be effective in technology-disrupted workplaces.
Ethereum A blockchain designed for general purposes, including enabling the creation of
smart contracts.
Ethereum 2.0 The revised Ethereum platform that features various measures to improve
its performance.
event A set of one or more outcomes.
explanatory analysis In data visualisation, the use of analysis techniques to effectively communicate
findings to decision makers.
exploratory analysis In data visualisation, the broad and deep data analysis intended to find patterns
in data.
external data Data the organisation must obtain from outside, whether by purchasing access or setting
up systems to harvest the data.
financial instruments Monetary contracts that take place between parties.
financial product A facility through which a person makes a financial investment, manages financial
risk or makes non-cash payments. For the purposes of the Corporations Act 2001, financial products
would include securities, debentures, derivatives, bonds, and so on.
financial services Making a market for financial products, providing financial product advice, operating
a registered managed investment scheme, providing traditional trustee company services, or providing
a custodial or depository service.
financial services business A business offering financial services, required to hold an AFSL unless an
exemption applies.
FinTech 1.0 The first wave of the financial technology industry from 2010 to 2014, during which
financial technology businesses targeted areas that banks underserved or overlooked.
FinTech 2.0 The second wave of the financial technology industry, beginning in 2014, during which
financial technology businesses increasingly collaborate and cooperate with banks and financial
structures become increasingly open.
FinTech payments niche Companies that facilitate and/or support the transfer of money, particularly
non‐cash transactions.
FinTech solutions niche Companies that provide technology solutions to assist businesses and financial
institutions with financial services.
FinTech technology niche Companies that provide software and services to different financial services
subsectors.
functional strategy The organisational plan for each function, such as finance or manufacturing.
fungibility (of money) The uniform quality of an object that makes the object interchangeable and
indistinguishable, so that any one dollar is equal to any other dollar.
gap analysis A comparison of the current state and the desired state, to identify what must be done to
arrive at the desired state.
General Data Protection Regulation (GDPR) An EU regulation on data protection and privacy that
has extra-territorial effect if an entity processes the data of an EU data subject. It has become a
benchmark for organisations operating internationally.
Gestalt principles A set of principles that explain how people perceive the elements of a visualisation
and thus can be applied to create effective visualisations.
governance The process of setting and implementing clear business objectives, policies and standards
across the entire organisation and ensuring they are properly executed.
governance, risk management and compliance (GRC) framework A framework of strategy, process,
policy, technology and people that provides the structure and approach that management may use to
achieve their business goals.
hybrid blockchains A blockchain that combines the concepts of private and public blockchains in order
to benefit from key aspects of both concepts.
hybrid payments Payments that combine multiple payments technologies.
hybrid RPA A combination of unattended and attended RPA.
ICT infrastructure ecosystems The hardware and software services that capture, store and organise
data in a digital ecosystem.
impact The consequence of a risk event either in financial terms or in the context of other factors that
affect both business and society.
GLOSSARY 475
independent event An event the probability of which is unaffected by previous events.
Industry 4.0 The use of Network Age technologies in industry; in particular the combination of
analytics, artificial intelligence and automation technologies.
information The output of data analytics that has meaning and purpose relevant to an individual’s or
organisation’s needs.
information assets Any valuable information that the organisation owns or controls.
information security See data security.
information security management system (ISMS) A comprehensive system of policies, procedures
and additional controls involving people, processes, and technology to mitigate risk and thus protect
the confidentiality, integrity and availability of the organisation’s information.
informed consent Consent given by a person armed with sufficient knowledge of the risks and benefits
that arise from granting that permission.
initial coin offering (ICO) The cryptocurrency equivalent of an initial public offering.
initial DEX offering (IDO) An experimental fundraising mechanism within the crypto-based funding
space, in which a decentralised cryptocurrency exchange replaces the role of the single exchange in
an IEO.
initial exchange offering (IEO) A variation on the ICO, in which a cryptocurrency exchange acts as an
intermediary, and conducts and underwrites the sale.
innovation A change in the way technology is used.
integrity The quality of data and information that exists when the data is reliable, correct and restricts
edit access.
intelligent process automation A combination of RPA and AI that extends RPA capabilities to handle
unstructured data and complex processes.
internal data Data the organisation has already captured and stored through its existing processes.
interval data Similar to ordinal data, but with the additional property that the difference between any
two data values is measurable.
invoice crowdlending See invoice trading.
invoice factoring A contractual agreement with a third-party provider, which is responsible for
managing the entire sales ledger.
invoice finance A collective term for invoice factoring and invoice trading.
invoice trading Financing in which a business sells individual outstanding invoices through electronic
platforms in order to obtain working capital.
ISO 19600:2014 The ISO standard on compliance management systems.
ISO 27001 The ISO information security management standard.
ISO 27005 The ISO information security risk assessment standard.
ISO 31000 The ISO risk management standard.
key goal indicators (KGIs) Metrics that indicate the extent of achievement of key goals.
knowledge Created by integrating information with prior experience, skills and opinions.
large-value payment systems Payment systems that process large volumes of high-value critical
payments, principally interbank/inter-financial institutional transactions.
likelihood The chance (or probability) of a risk eventuating.
machine learning (ML) A type of AI in which the machine is able to learn from data to develop its own
algorithms and thus make predictions.
marketplace lending See peer-to-peer lending.
matching In an observational study, selecting subjects with similar characteristics other than the
independent variable.
measure of value The function (of money) of measuring the market value of goods, services and any
other transaction.
medium of exchange The function (of money) of intermediating the exchange of goods and services.
miners Nodes in a blockchain network that specialise in adding new data to the blockchain.
mitigate In relation to risk treatments, taking action to reduce the likelihood of an event occurring, but
not avoiding the risk.
money An item or a verifiable record that is accepted for payments and repayment within a specific
socio-economic context.
money supply The total of all the available financial instruments within an economy.
multiple regression In an observational study, a statistical technique used to account for the effect of
confounding variables on the dependent variable of interest.
476 GLOSSARY
mutually exclusive The situation in which a given set of outcomes or events cannot occur at the same
time. Outcomes are always mutually exclusive, whereas events may or may not be mutually exclusive.
natural language processing (NLP) A type of AI that applies computational linguistic techniques to
text to enable a machine to communicate with a human through natural language.
negative correlation A correlation in which when the value of one variable moves in a particular
direction, the value of the other variable moves in the opposite direction.
neobank An emerging bank start-up.
neobanking The sector created by emerging bank start-ups.
Network Age The current period of time, characterised by interconnectivity, enabled by the internet,
smart devices, mobile communications, and the rapidly diminishing cost of technology.
network effects Positive returns to scale; as network participant numbers grow, the value of
participating also grows.
New Payment Platform (NPP) An open-access, industry-wide infrastructure that enables near-real time
data-rich payments between accounts held in different institutions.
NFC payments The use of near-field contactless technology and a mobile wallet to enable payments
between smartphones and payment readers.
NIST Cybersecurity Framework (NIST CSF) A cybersecurity framework that serves as an industry
standard and represents best practice around cybersecurity risk.
nominal data Data that can be named, labelled, or categorised, but cannot be ordered.
non-programmed decision making Decision making in a unique situation where there is no prior
experience or pre-developed strategy to apply.
Notifiable Data Breach (NDB) Scheme An obligation under the Privacy Act that stipulates that
affected individuals, as well as the Data Commissioner, must be notified by way of a prepared
statement if a breach affecting personal data occurs.
objective probability A probability value calculated empirically (based on historical data or
measurements) or a priori (based on formal logic, reasoning and the known characteristics of events).
observational study Passive observation of the subjects to measure variables of interest and identify
patterns or relationships between them.
open accounting The sharing of accounting data with relevant stakeholders to increase its value as
a resource.
open banking An open marketplace for banking services, created through connectivity enabled by data
sharing, apps, APIs and analytics.
operating standard In relation to a control, the activities or tasks that occur to ensure that the control
operates as designed.
operational strategy How the business achieves its objectives through resources and processes.
optimising ML ML that focuses on applying algorithms to find the most efficient and effective outcome
for a process.
oracle A live-feed of data from external systems that are continuously monitored, treated as a source of
truth and are used to trigger execution of a smart contract when the defined conditions are met.
ordinal data Data with named categories that have a ranking system, and can be ordered.
organisational behaviour The shared goals, beliefs, values, expectations and the environment in which
the members of that organisation work.
organisational strategy The coordinated use of resources to achieve a set of organisational goals.
overfitting The situation that arises in AI in which algorithms become excessively complex and specific
in order to explain all variations that exist in the data, and thus do not necessarily translate to effective
analysis of other data.
outcome A single possible result of an experiment.
outlier A data point that is far away from the rest of the data.
partition A collection of outcome sets where the union of those outcome sets is collectively exhaustive
and each of the outcome sets is mutually exclusive.
paternalistic approach A regulatory approach that shifts the responsibility from the customer to the
industry, to ensure that organisations take steps to look after their customers, as opposed to requiring
the customer to understand complex disclosure documents and policies.
payment A mutually agreed transfer of funds, assets or goods, or services, in exchange for another form
of funds, assets or goods, or services.
Payment Card Industry Data Security Standard (PCI DSS) A guide to maintaining a secure
environment for all organisations that accept, process, store or transmit credit card information.
GLOSSARY 477
payment gateways Services that enable digital payments by transmitting customer transaction
information to the business’s merchant account, where funds are held until the transaction is cleared.
Once the transaction is settled, the bank releases the funds to the business’s bank account.
payment service provider (PSP) A third-party payment processing company that provides merchants
with everything they need to access digital payments, including usually a merchant account and a
payment gateway.
peer-to-peer invoice finance See invoice trading.
peer-to-peer (P2P) lending The model of lending in which individual lenders provide loans to
individual borrowers such as consumers or businesses, often via an electronic platform.
permissioned blockchains Blockchains that have a control layer that grants access to accepted users.
permissioned ledger A distributed ledger with a defined set of owners.
permissionless blockchains Decentralised public blockchains that have no central authority.
permissionless ledger A distributed ledger without a specific owner.
personal data See personally identifiable information.
personally identifiable information (PII) Information that can be used to identify an individual, either
directly, such as name, or in combination with other data.
PESTLE An external analysis based on examination of political, economic, social, technological, legal
and environmental factors.
physical assets Infrastructure, devices and other hard assets used in an organisation’s data and
information system.
portability (of money) The ability to handle and deal in money in small quantities that can be
exchanged or transferred in value.
portfolio lending See balance sheet lending.
positive correlation A correlation in which when the value of one variable moves in a particular
direction, the value of the other variable moves in the same direction.
pre-attentive attributes The aspects of a visualisation that people notice first.
predictive analytics Data analytics that forecast or estimate future performance.
predictive ML ML that characterises and analyses historical and recent data to predict an accurate
future outcome.
prescriptive analytics Data analytics that determine a course of action.
preventative controls Controls intended to stop cybersecurity incidents happening in the first place.
privacy coins Altcoins that facilitate fully anonymous monetary or information transactions
between users.
private blockchains A blockchain in which participation is by invitation and is either validated by the
initiator of the network or by predefined rules.
private permissioned blockchain A permissioned blockchain in which the control layer governs
transactions performed by the participants of the network.
probability The chance or likelihood that a particular event will occur.
programmed decision-making The use of guidelines or policies that allow managers to make
streamlined, timely and consistent decisions.
proof of stake A consensus mechanism based on collaborating nodes used by some blockchains.
proof of work A consensus mechanism based on competing nodes used by some blockchains.
property-based crowdfunding An entrepreneur or a property developer (seeker of funding) pooling
together funds from investors (providers of funding who are not ordinarily large-scale investors in
these cases) to pay crowdfunding mortgages, which are secured on specific properties, or to finance
property development.
public blockchain A blockchain in which every user is allowed to interact with the network by
submitting transactions and hence adding entries to the ledger.
public permissioned blockchain A permissioned blockchain in which the control layer does not
interfere with the blockchain decentralisation or authority.
qualitative analysis The use of critical thinking and subjective interpretation of unstructured
qualitative data.
quantitative analysis The use of mathematical and statistical techniques to analyse numerical data.
qualitative data Data distinguished by non-numeric characteristics.
quantitative data Numerical data distinguished by numeric characters that represent counts or
measurements.
478 GLOSSARY
random experiments An approach to proving causation by assigning a test subject randomly to either a
treatment group (affected by the hypothesised causal factor) or a control group (not affected by the
hypothesised causal factor) and then conducting trials.
ratio data Interval data that includes a natural zero starting point.
real estate crowdfunding See property-based crowdfunding.
recognisability (of money) The quality that the value of money is easy to identify so all parties to a
transaction can verify and accept the authenticity and value of the money.
regression The predictive analytics process of determining a target variable which is continuous.
Regression identifies which variables have an impact on an outcome of interest.
regulatory technology (RegTech) The use of technology to improve regulatory processes including
regulatory monitoring, compliance and reporting.
Reserve Bank of Australia (RBA) Australia’s central bank and co-regulator of the financial system.
restricted authorised deposit-taking institution (ADI) licence An APRA licence for neobanks.
retail payment systems Payment systems that settle obligations involved in the routine purchasing and
selling of goods and services.
reward-based crowdfunding Entrepreneurs incentivise financial donations in return for a service
or product.
risk The effect of uncertainty on objectives.
risk appetite The types of risks that are acceptable to the organisation to pursue, thus determining what
new products, customers, and markets should be targeted.
risk appetite statement A documented record of the organisation’s risk appetite, used to inform risk
management throughout the organisation.
risk criteria The criteria against which all risks are measured and evaluated, primarily in the risk
analysis and risk treatment steps of the risk management process.
risk management The process of mitigating and managing business-related risks and uncertainty
through establishing control mechanisms throughout the organisational environment, and the
processes required to meet the business objectives and governance requirements.
risk management function The business function with responsibility for the delivery of efficient and
effective risk management.
risk owner A person or group who is assigned the responsibility for a particular risk within the risk
management framework.
risk profile A written description of the organisation’s risks.
risk register A risk management tool to record details for identified risk, including risk ratings, nature
of the risk, owner, manager, and mitigation measures.
risk retention The decision to accept a risk.
risk tolerance The amount of risk a business will assume in order to achieve its financial goals.
risk treatment The action taken to modify the identified risk to a level the organisation considers
acceptable.
risk triggers Defined events, changes or analysis outcomes that may affect a risk or indicate a potential
change in a risk and thus require re-evaluation of the risk management approach.
robotic process automation (RPA) Software robots that automate simple-to-medium complexity tasks
and processes by applying a set of rules known as an algorithm.
scarcity (of money) The supply of money in circulation cannot be unlimited.
security controls Controls focused on securing information assets and related physical assets in order to
protect the organisation’s data.
security token offering (STO) Similar to an ICO, but ownership of the token is tracked against a
real-world asset, such as shares, and thus STOs are perceived as security offerings.
security tokens Tokens that represent a stake in a real-world asset.
semi-structured data Data that does not conform to a tabular format like structured data, but does
contain tags, markers or metadata that facilitate processing.
semi-supervised learning An approach to ML in which the machine is provided with some labelled
data but mainly unlabelled data. The labelled data is used to partially train the algorithm, then some of
the unlabelled data is fed to the partially trained machine.
sensitive company data Information relating to a company’s trade secrets, acquisition plans, financial
data, and supplier and customer information.
GLOSSARY 479
sensitive personal data Information relating to a person’s racial or ethnic origin, religious or political
beliefs, health, social relationships, employment, financial situation, sexual orientation, gender, and
so on.
smart contracts Transactions that self-execute based on a set of programmed conditions, usually hosted
on the Ethereum blockchain.
software ecosystems The aspect of a digital ecosystem created by the interaction of businesses via a
common technological platform to produce software and services.
spurious association When there is a correlation between two variables, but no causation exists.
stability (of money) The ability to sustain value over time so as to be used for transactions and avoid
transaction costs.
stablecoin A cryptocurrency coin pegged to a ‘real-world’ asset or benchmark in order to achieve a
relatively stable value.
standard of deferred payment The function (of money) of operating as a standard of deferred payment.
statement of applicability (SOA) A document that sets out the controls chosen in response to the risks
identified. It provides a link between the risk assessment and risk treatment stages of risk management
and is a core element of ISO/IEC 27001.
statistical inference Using a sample of data to reach conclusions about the population from which the
sample is drawn.
statistics The discipline that concerns the collection, organisation, analysis, interpretation and
presentation of data.
store of value The function (of money) as an asset bearing a relatively stable sustaining value.
strategic management The process of formulating and implementing strategies to accomplish
long-term goals and sustain competitive advantage.
strong AI Any AI instance that can exhibit natural intelligence (does not yet exist).
structured data Data that adheres to a pre-defined model, which makes it straightforward to analyse. In
a database, it conforms to a tabular format with a clear relationship between the rows and columns.
subjective probability A probability value derived from intuition and experience.
supervised learning An approach to ML in which the machine is provided with input data and the
corresponding output data so it can learn the algorithm itself and, thereby, generate valid output data
for new input data.
support vector machine (SVM) A type of algorithm used in ML for both classification and regression.
systems development life cycle (SDLC) A multiple-stage approach to developing high-quality systems
from planning and analysis through support and maintenance.
technology Any process by which a company transforms information and data, human labour or
economic capital, into products or services of greater value.
tokens Representations on a blockchain network of a real-world asset or entitlement to products or
services associated with the token.
total probability A calculation of the probability of an event from a set of conditional probabilities.
transfer In relation to risk treatments, shifting a risk to a third party, usually in the form of insurance or
outsourcing.
transient advantage The ability of a business to temporarily achieve better performance than its
competitors, but the inability to sustain the advantage, as competitors can quickly copy or improve on
the source of the advantage.
transparency Openness about the methods used for analysis and use of the data, both to the creators and
to other practitioners.
transparent AI An approach to AI that enables humans to understand how the machine reached
its conclusions.
unattended RPA RPA in which the bot interacts directly with computer systems to perform a business
process from start to finish without human involvement.
unconditional probability A probability calculation that ignores information regarding any historical or
external events.
unit of account See measure of value.
unstructured data Data that does not conform to a predefined structure and is not organised in a
predictable manner.
unsupervised learning An approach to ML in which the machine is provided with a complex set of
input data with no corresponding output data; the machine identifies the relationships in the data to
provide valuable insights for decision makers.
480 GLOSSARY
utility tokens Tokens for accessing dedicated products and/or services associated with the token.
value driver Anything that can be done to increase value.
verification approach In relation to a control, the method and strategy to obtain assurance on the
control health, utilising various methods as evidence.
virtual currency An unregulated digital currency.
weak AI Any AI instance that can perform a single task effectively.
wisdom The ability to use knowledge to make decisions and act on them.
GLOSSARY 481
SUGGESTED ANSWERS
MODULE 1
QUESTION 1.1
Changes span both the digital ecosystem (hardware) and applications (software).
The digital ecosystem has changed with the introduction and development of the internet, mobile
communications and smart devices.
Things are changing with the use of automation, analytics, artificial intelligence and machine learning.
New applications are possible because of faster computers and networked access to huge quantities of data.
Where the digital ecosystem and new applications overlap, organisations can use new technologies
to re-shape the finance function, specifically to provide personalisation of services and better business
intelligence.
These technologies are changing the tasks finance teams are working on because data is increasingly
an asset they need to protect and ensure systems incorporate the necessary controls to protect the data’s
integrity. This covers both financial and non-financial data.
Accountants should no longer be performing tasks that can be automated; their role is no longer about
the collection or calculation of financial data, but interpretation of both financial and operational data to
form strategic insights. For this they need to develop a digital mindset (including digital literacy).
QUESTION 1.2
Digital payments (also known as electronic payments or e-payments) refer to any payments that are
initiated, processed or received electronically. Our ability to make digital payments depends on a suitable
digital ecosystem, comprising suitable internet connections, mobile communications and smart devices.
Digital payments are arguably the oldest form of FinTech. Well-established technologies such as
EFTPOS, credit cards and direct payments through online banking are now common practice. We also
have new forms of digital payments such as digital wallets on smart devices.
As fewer people carry cash, offering digital payments has become a fundamental requirement of doing
business. This has accelerated in the pandemic environment, which created the need to avoid physical
contact while exchanging money.
Using digital payments also provides better integration with data collection and analytics systems. This,
in turn, provides organisations with comparative advantages from their business intelligence systems.
QUESTION 1.3
Structure innovations focus on organising company resources in unique ways that create value and can
be difficult for other companies to replicate — for example, changing to a flat management structure to
shorten communication lines from workers to senior management. This is usually done to help identify
ways in which the organisation can cut costs or improve efficiency.
Process innovations involve the activities and operations that produce a business’s primary offer-
ings. During 2020 many organisations were forced to develop online products and services; although
few changes were unique they were necessary to allow the organisation to continue operating in a
pandemic environment.
QUESTION 1.4
Service Channel Brand Customer engagement
• Make a product easier • This is the ways a • Ensure that customers • Uses insights into
to try, use and enjoy. company’s offerings recognise, remember customer aspirations
• Reveal features and connect with and prefer a to develop meaningful
functionality customers customers and users. company’s offerings connections between
might otherwise over those of them and the
overlook. competitors. business.
(continued)
Pdf_Folio:483
• Service innovations are • Both physical stores • Uses communications, • This is increasingly
increasingly delivered and online ecommerce advertising, service taking place in the
through electronic are channels. interactions, channel social media space.
interfaces, remote environments, and
communications and employee and
automated technologies. business partner
conduct.
QUESTION 1.5
FinTech 1.0 describes the first wave of FinTech developments. FinTech 1.0 start-ups fell into three
categories:
1. those that focused on payment processing
2. those that generated new lending models using peer-to-peer structures
3. those that helped banks through personal financial management and risk modelling tools.
FinTech 2.0 is characterised by:
• increasing collaboration and cooperation between banks and the FinTech companies
• a trend towards open financial structures.
In the context of FinTech 2.0, banks remain the cornerstone on which economies are built and are likely
to continue to play a central role, with much of their innovation driven by integrating FinTech companies’
technologies instead of competing with them.
Digital disruption no longer threatens traditional banks’ existence but is increasingly influencing how
banks do business. Banks must comply with open banking regulations, which will make it easier for
customers to change banks or selectively choose different institutions for specific financial services, but
banks still retain the advantages of their economies of scale and economies of scope.
QUESTION 1.6
When considering failure, it is important to separate causes from symptoms. With digital implementation,
the failure of projects is still possible even when they are running on time and within budget. Projects
can still fail because they do not provide the outcome the end user expects. This is a symptom of failure
and, although it is a problem, to reduce the risk of it occurring again we need to consider the cause of
this failure.
Inadequate resources, overly aggressive timelines, underestimated costs, overlooked requirements,
unanticipated complications, poor governance and human mistakes such as bad code can all lead
to project failure.
The main cause of failure in digital implementation is a lack of clearly defined and/or achievable
milestones and objectives to measure progress. The second most common cause is poor communi-
cation, followed by even a lack of communication by senior management, employee resistance and
insufficient funding.
QUESTION 1.7
Adoption of Airwallex services would remove the need for an organisation’s accountants to perform a
range of basic bookkeeping tasks and foreign currency translations. Integrated invoicing, payroll and
payments would automate transaction processing and accounting data collection. Forex translation across
all of those accounts would remove the need for separate calculations. The fees charged on forex
transactions would also be much lower than those charged by traditional providers.
The trade-off for all these benefits to the accountant, is that they would now need to provide oversight
and advice on integrating Airwallex into the organisation’s existing systems. This would include identify-
ing all the benefits and risks of the technology and ensuring sound controls were in place to prevent abuse of
the system. Once this is successfully implemented, however, accountants can focus on more value-adding
work such as interpreting data and identifying strategic insights.
Pdf_Folio:484
QUESTION 1.9
The CFO is central to organisational change in allocating resources, providing access to financial data and
applying performance metrics across the organisation. Although accountants have traditionally been the
main data analysts in organisations, they need to be proactive in expanding their role beyond the collection
and management of financial data.
Management accountants are better placed to be involved in digital transformation implementation as
they traditionally work on performance evaluation and are more familiar with handling non-financial
information. They can provide leadership and establish clear measurable goals by which the success or
failure can be determined.
The risk, for the management accountant, is that although they already use some non-financial data, they
may make the mistake of thinking that is sufficient when the digital transformation will produce a great
deal more data.
Overall, there is a potential problem as the CFO has the most say in whether the transformation project
is funded and goes ahead, while the management accountant has the most to gain from involvement in the
project and use of the new data generated. Better alignment between influence and benefits would increase
the chance of the project being successful.
QUESTION 1.10
Accounting and finance professionals have four broad roles they can adopt in supporting digital
transformation.
1. Connector. They work on redesigning operating models around the digital transformation, ensure
connections are made between front- and back-office roles, and try to make user experience as seamless
as possible.
2. Digital transformation leader. They work on reducing friction between existing staff and departments
as the digital transformation changes the workplace, they ensure that new data generated is supplied to
all who need access, and assist staff in making the most use of that data in the creation of actionable
business intelligence.
3. Insights generator. They move from reporting on past events to using the rich data sources to create
predictive insights.
4. New talent nurturer. They may develop expertise in data science or use their collaboration and
communication skills to build teams who can make full use of the data.
Pdf_Folio:485
QUESTION 2.2
Advantages Disadvantages/Challenges
• Reduced crime as there is no physical money • As currency becomes data, cybersecurity becomes
to steal a threat
• Reduced money laundering as cashless • Privacy is compromised as cashless transactions
transactions are traceable and the parties are usually are traceable and the parties are usually identifiable
identifiable • Reliance on availability and scalability of technology
• More cost-effective, as there is no need to print, to enable transactions
handle or store physical money • Reliance on interoperability of systems (including
• Foreign exchange transactions will be seamless as establishment of standards)
they will not need to involve handling of cash • Fees charged by payment service providers,
particularly if one or a few become dominant
• Potential exclusion of those with limited access to
technology or skill with technology
While it may appear that the disadvantages outweigh the advantages, note that most of the disadvantages
are technical and in time should become much less of an issue as technical solutions are developed. The use
of contactless payments has been encouraged, and in some cases required, in the pandemic environment
as businesses try to minimise all physical contact between staff and customers.
QUESTION 2.3
Pros of going cashless for small retail businesses are as follows.
1. Save time and money. Reduce the time spent on cash management.
2. Increase checkout efficiency. Speed can attract more customers to make a purchase.
3. Decrease risk. There is less cash at the register.
4. Improve accounting. Track every transaction, lead to more accurate accounting.
Cons of going cashless for small retail businesses are as follows.
1. Credit card fees. May be a burden on their books since their profit margin may not be very high.
2. Loss of customers. Some customers either don’t know how to use digital payments or simply prefer
using cash.
3. Costly. The required technology might be expensive for a small retail business.
A small retail business needs to consider whether they really want to take the lead in making this change,
especially if society is not yet ready.
It would be easier and more cost-effective to wait for large retail to go cashless as they can drive social
changes more efficiently.
A counter-argument is that a small business may want to make a point of going cashless as a way of
signalling how they are forward thinking or concerned for their customers due to a pandemic, and want to
reduce physical contact and the risk of disease transfer through physical currency.
Pdf_Folio:486
QUESTION 2.5
Smart contracts are transactions that self-execute based on a set of programmed conditions.
Two of the main advantages of smart contracts are speed and reduced risk. Traditional reliance on
physical documents leads to delays, inefficiencies and increased exposure to errors and fraud. For example,
for insurance, use of smart contracts can result in cost savings through automation and reduced processing
costs in claims handling.
Broadly speaking the potential advantages in the use of smart contracts in financial services are in three
key areas: risk reduction, cost savings and enhanced efficiencies.
The limitation of smart contracts is that they only work in situations where events have clearly occurred,
real contracts are often subject to disagreements about whether the contractual terms have been satisfied
and require independent judgement or arbitration.
QUESTION 2.6
As permissionless ledgers are immutable, errors such as a financial transaction to an incorrect account
cannot be revoked, modified or cancelled. This is fundamentally different to standard accounting practices
where incorrect transactions are easily reversed. With blockchain assets like cryptocurrencies there is no
way to force a reversal and there is no central authority to appeal to and request a reversal.
The closest to reversing a transaction on the blockchain is to create a side-chain from a point before the
offending transaction. Although very rare this has occurred in special cases, and even then some users of
the ledger decided to keep using the original chain, which is not deleted, because they did not believe the
justification for creating the new side-ledger was strong enough.
As long as the data stored is correct then the inability to change it is valuable, but when there are
problems with the data, the immutability removes any ability to correct it short of having every copy of the
blockchain deleted.
QUESTION 2.7
There is a misperception that cryptocurrencies are a black box, which makes them popular for illegal
activity and tax avoidance. However, most cryptocurrencies use public blockchains, so transaction histories
can still be traced. While there is certainly no KYC within the blockchain, whenever there is a transfer
between traditional currency and cryptocurrency there is a possibility of linking the KYC of a participating
bank or cryptocurrency exchange to the corresponding blockchain address. This depends on the bank or
exchange being KYC compliant and willing to work with government authorities on tracking activity
within the blockchain.
Cryptocurrency miners based in countries with active law enforcement tracking of cryptocurrencies have
even been reported as only including transactions in a mined block if connected to known, legitimate, users.
So even though many accounts remain anonymous it will become harder for those accounts to transact or
to transfer the funds into traditional currencies.
However, some cryptocurrencies are under development where it will not be possible to trace a chain of
transactions. With these any form of KYC will be difficult or impossible to impose.
Pdf_Folio:487
QUESTION 2.9
Decentralised finance (DeFi), removes intermediaries and enables both parties in a transaction to exercise
direct control over the source and use of funds.
Advantages are an increase in market efficiency, lower costs and privacy in transactions. Disadvantages
include the risk of cyber-attack, no real protection against harm, loss of funds or other malicious attacks.
DeFi remains in the early stages of development and adoption. This also means that the regulatory
response to DeFi is yet to be declared. One of the advantages of centralised financial systems is that
it provides clear targets for regulators. If DeFi lowers the barriers to entry in financial services, then
we should expect regulatory action to ensure new financial service providers offer reasonable protection
to customers.
QUESTION 2.10
Virtual currencies provide online payments, usually restricted to a specific application or platform.
Virtual currencies are common in online games and are used in reward schemes such as airline frequent
flyer points.
Cryptocurrencies are a much more complex form of virtual currency where distributed ledger technology
and cryptographic processes are used to control the creation and ownership of cryptocurrency tokens.
As no central authority can increase the number of cryptocurrency tokens beyond the rate set when the
blockchain was launched, inflation is limited. In contrast, if an airline decided to revalue its frequent flyer
points, or to cancel the system, there is little that the owners of those points could do to stop it.
QUESTION 2.11
The main reasons for a move away from traditional sources of finance (e.g. banks, stock market,
venture capital) to new alternative sources (e.g. equity crowdfunding, peer-to-peer lending, crypto) are
a combination of new enabling technology and regulatory changes.
The main enabling technology has been the development of more advanced internet-based systems
that allow two-way interaction between customer and platform. This allows financial firms to interact
with clients online and removes the barrier to entry of needing a branch network to contact and interact
with clients.
The technology would not have made anywhere as much of a difference if financial regulations had not
been relaxed to allow new financial business models to develop.
QUESTION 2.12
Traditional invoice finance models used factoring, where the financier took over collection of the selling
company’s accounts receivable. This is a model with significant economies of scale and specialisation, so
selling companies needed to be large and preferably have large individual accounts to sell.
With the advances in the internet, cloud-based accounting and platform financial models, it became
possible to apply crowdfunding principles to invoice finance and eliminate most of the economies of scale.
Risk was managed, either by spreading it over many investors (a form of crowdfunding) or providing
the financier with detailed access to the cloud-based accounting records. Costs were reduced by allowing
the selling firm to retain cash collection responsibilities.
Pdf_Folio:488
QUESTION 2.14
An initial coin offering (ICO) is the sale of crypto tokens to investors. ICOs enable firms to raise substantial
funds quickly and cost-effectively at a very early stage of the business’s development. Unlike other forms
of raising funds early in a firm’s life, these tokens can be utility or access tokens, so their sale does not
involve the sale of equity or taking on debt. They do not dilute ownership or control over the business.
Nevertheless, the tokens can have investment characteristics for the purchaser as many, but not all, of the
coins issued are tradable on cryptocurrency exchanges; therefore, it is possible for the investor to realise
gains through any increase in the token value.
Obstacles to using an ICO mainly relate to the fact that this is a new and largely unregulated market, so
many investors will not yet feel comfortable participating in the market or have the technological know-
how to store and trade tokens.
The relatively unregulated nature of the ICO, and crypto markets in general, also means that many
fraudulent ICOs have taken place with tokens quickly losing all value.
QUESTION 2.15
RegTech is more than just a subset of FinTech companies working on regulation problems. RegTech
promotes a fundamentally different approach to financial regulation, one based on a data-driven and data-
led mindset in the key areas of regulatory monitoring, reporting and compliance.
Instead of seeing FinTech companies as operating on the border of existing financial regulation, at times
in an apparently unregulated environment because regulation has yet to catch up with new developments
in FinTech. Changing to a data-driven regulatory system will make regulation central to all FinTech
companies. This is because data generation is a fundamental characteristic of all FinTech companies and
regulation just becomes another way in which the data can be used.
MODULE 3
QUESTION 3.1
Robotic process automation (RPA) is based on a series of predetermined steps (an algorithm). The
requirement for the steps to be predefined limits the scope of RPA applications, as does its reliance on
structured data.
Artificial intelligence (AI) mimics human thought through the use of algorithms developed by machine
and deep learning. Therefore, AI can be used to enhance RPA and extend RPA’s uses or the extent
of the process to which it can be applied. For example, AI can be used to examine unstructured data
(speech, handwritten or free text) and identify the structured data required by an RPA application. In this
application the addition of AI to RPA removes the need for manual data entry and therefore speeds up the
process. However, before this can happen there will need to be a time investment to develop the machine
learning algorithms.
Pdf_Folio:489
QUESTION 3.3
A balanced scorecard is a strategic and operational assessment tool that, in the case of automation, can be
used to assess both the applicability of automation and, once implemented, its performance. The balanced
score card approach uses a range of both financial and non-financial measures to assess performance
against organisational goals. For example, the organisational goal may be to improve customer satisfaction
and automation may have been selected as a means to achieve this. With a balanced scorecard approach,
measures will be selected to assess whether automation has contributed to this goal.
A small number of financial and non-financial measures (across categories including customer advocacy,
productivity, people, flexibility, scalability and controls) and their targets will be devised. Generally, targets
will be based on pre-automation performance. Assuming customer complaints are a proxy for customer
satisfaction, measures could include:
• cost of the customer complaints department
• number of complaints received
• average time taken to resolve a customer complaint
• ‘likes’ received by the customer complaints department
• average wait times for complaints calls to be answered
• staff turnover in the complaints department
• percentage of complaints that are escalated to a supervisor.
By comparing actual performance across the measures with the targets, the benefits of the automation
can be assessed.
With a change to any system, it is important to make sure that the benefits outweigh the costs and not
just in the short term or in the immediate area of the business where the automation was implemented.
Some costs and benefits may take time to appear, they may appear in another part of the business or for
stakeholders not originally considered. A balanced scorecard needs to be devised with these things in mind
so that a true and complete assessment of the automation’s contribution to the organisation is made.
QUESTION 3.4
Unattended RPA performs a business process from start to finish without human involvement while
attended RPA runs on the user’s computer and is triggered by the user as required.
The two main factors to consider when choosing between the two include:
• nature of the task
• type of data being processed.
Unattended RPA can be used for tasks that are triggered be events such as input of data, input from
another bot or a time-based schedule. Tasks that are ad hoc, require human activation or human intervention
are suited to attended RPA. Unattended RPA is suitable for tasks that are highly structured and reasonably
straightforward, while attended RPA is suitable for tasks that are complex or highly varied.
Unattended RPA requires high-quality structured data. Issues with missing or non-compliant data may
cause a bot to pause the process while it waits for input from a human.
Pdf_Folio:490
QUESTION 3.6
Both weak and strong AI are designed to replicate human decision making, but they differ in the range of
situations where they can do so.
Weak AI can manage a single simple task effectively and can mimic human behaviour within a
narrow context.
Strong AI displays natural intelligence — the ability to process and respond to information as a human
would respond. This remains a rare form of AI.
Due to the rarity of strong AI, most examples of AI found in Australian businesses are weak AI.
QUESTION 3.7
When adopting AI technology, organisations are likely to face the following risks:
• poor decisions being made based on the outputs of flawed models, including investment, recruitment,
lending and forecasting decisions
• legal repercussions from poor data security from discriminatory practices that result from flawed models
• inability to substantiate the output from black-box models
• reputational damage from any of the previously listed risks.
QUESTION 3.8
The key parts of an effective AI strategy are data, technologies and algorithms, infrastructure, organisation
and skills. Successful AI strategies are derived from and supportive of an organisation’s wider business
strategy. Therefore, an AI strategy should not be developed in isolation but rather enable and support the
wider business goals through the use of technology.
Each company’s AI strategy will be unique to its needs. The following process will ensure an AI strategy
that is fit for purpose is created and AI applications that add value are developed.
QUESTION 3.9
Supervised machine learning and unsupervised machine learning differ in the types of models they
generate and the data they use to generate the models.
If labelled data is available in sufficient quantity supervised machine learning can be used to develop
classification or regression models. Classification models result in applications that can classify images for
medical diagnoses, classify loan applications into accept or reject categories and decide if an email is spam
or not. Regression models result in applications that produce numeric predictions, forecast temperature,
estimating life expectancy and forecast financial market activity. In all of these applications the models
are capable ‘converting’ input data into expected or predicted results. The conversion process or model
is created by exposing the machine generating the model to labelled data – that is, data that consists of
historical inputs and their associated results or outcomes. For example, the demographic and financial data
of a loan applicant and whether they repaid their loan or not. The data is split into two sets that mirror the
two phases of model development. The first set is the training set, which the machine learns from to create
the model. The second set is the test set, which is used to confirm the accuracy of the model before it is
put into use with new input data, from which necessary predictions are made.
If unlabelled data is available, unsupervised machine learning can be used to develop clustering and
other types of models that explore and find patterns or associations within the data. Clustering models
result in applications that segment a customer base, make recommendations based on previous purchases
or to target marketing. Associative models can be used to reduce the breadth and size of datasets so
that only the most salient of data is retained. This will help to minimise data collection, storage and
processing costs and facilitate the creation of less complex data visualisations. Models are created by
exposing the machine creating the model to unlabelled data – that is, data that has not been tagged with its
classifications, properties or identifying characteristics. The machine then works with this data to identify
the properties and characteristics of the data to uncover patterns or associations within the data that allow
models to be created.
QUESTION 3.10
The primary advantage AI systems have over human analysts is the ability to process large volumes of
data quickly without getting tired or bored by the repetitive work. Note that human analysts have an
advantage over AI systems in that they are better at creative thinking. The additional factor that gives AI
systems an overall advantage over their human counterparts in financial analysis is their ability to consider
a much wider range of possible models. In some respects, this can be seen as AI systems compensation
for their inability to use creative thinking by instead quickly reviewing all possible models irrespective of
whether there is any logic to the model at all. All they are looking for is a model that works, they are not
constrained by unconscious bias or preconceived notions of how the world of finance works.
Pdf_Folio:492
QUESTION 3.12
Most data traditionally managed by accountants, such as quantitative accounting data, can be organised
as a table, and are classified as structured data. In contrast, accountants have made little use of qualitative
data that is unstructured. Examples of unstructured data include social media activity, surveillance imagery
and text. Because it is so vast and appears in unusual formats, this type of data is difficult to analyse using
traditional methods.
Machine learning applies algorithms to data, processing it to allow the practitioner to make sense of large
amounts of information, of various forms, procured from multiple sources. For instance, an accountant
could gain insights into the success of a new product line by monitoring social media activity or satellite
footage of delivery trucks leaving the distribution warehouse. Alternatively, accountants could analyse the
text found in the corporate press releases or interviews with corporate executives to discover potential
discrepancies with accounting information.
QUESTION 3.13
The outputs produced by RPA are the result of a clearly documented, pre-determined process, so the
organisational user knows how the output is generated and can easily review the supporting documentation
to explain it.
AI and ML, on the other hand, are often black-box solutions where the internal process that took the
input data to produce an output recommendation is a complete mystery. Often the underlying AI or ML
programs are based on methods developed externally to the organisation, so there are few, if any, within
the organisation that can explain the process, even in broad terms.
However, it does not need to be like this. Some modern AI and ML systems are being developed to also
provide information on the model used to produce the output, or information about the internal processes
that resulted in the final model. Nevertheless, more transparent systems are still subject to limitations
when the model is being updated dynamically every time new data is introduced as the model and/or the
processes used to select the model are frequently changing.
QUESTION 3.14
Corporate governance is the process of setting and implementing clear business objectives, policies and
standards across the organisation and ensuring they are properly executed.
Increased automation of business processes increases governance risks mainly because it reduces or
removes human involvement, meaning that awareness of procedures not working as intended is limited.
There is a risk that senior management will assume that governance tasks are being performed as expected,
without full oversight.
The FinTech subfield RegTech, or regulatory technology, is an area where automation can improve
the management of governance risks. Automation can benefit risk compliance, by ensuring processes are
followed reliably and consistently, provided the automation process is designed well (including suitable
control mechanisms), set up correctly and maintained properly.
Pdf_Folio:493
MODULE 4
QUESTION 4.1
The key driver of a data-literate organisational culture is that decisions are made based on insights gained
from interpreting the output of analytics. For this to be achieved it is not necessary for all staff to become
statisticians; however, it is important that the majority understand enough basic maths and statistics to
have an appreciation of how the organisation is generating and using data. It is also important for all staff
using data to be competent communicators of the data-driven outputs of any analysis, so they can make
data-driven decisions and bring others onboard.
For example, a data-driven organisation needs to have a culture where staff provided with a dashboard
showing key operational or financial metrics are able to drill down into the detailed data relevant to
their role. They must have both the access and knowledge to use that information to improve decision
making. The management also needs to ensure such relevant knowledge, skills, experience and actions are
encouraged across the whole organisation.
QUESTION 4.2
Most organisations are far from reaching their potential due to a combination of technical and
human issues.
On the technical side, the use of AI and ML remains a complicated and specialised field, where the
majority of AI and ML analytics projects are run by data and technology experts and do not scale up to
serve the whole organisation.
On the human side, organisations need people with the ability to understand and leverage analytics
insights, who can push and convince others to follow the direction suggested by the data. This requires
an organisation-wide commitment to data-informed decision making, which in turn requires people to
develop data literacy skills.
To realise the full potential of data analytics in business we need to understand that it is not just about
what the technology can do, nor what the technology can do with the data available, but what the technology
and available data can do for the organisation and to what extent the people in the organisation can accept
and interpret data and then form insights.
QUESTION 4.3
Structured data has a standard format, is usually numerical or simple text, and is usually stored in a table.
Unstructured data has no standard format. Common forms are long-form text (full sentences and
paragraphs), audio recording or image data.
Pdf_Folio:494
QUESTION 4.4
Due to the complexity of systems development, it is generally understood that a structured approach is
required. Development principles are often borrowed from project management practices. However, the
detail of the best structure for systems development is open to debate.
Early systems development models proposed a linear approach, working through a series of discrete
steps in a specific sequence. Under a linear approach any failure in the resulting system was considered to
be due to one of the steps being incorrectly completed and not a failure of the linear approach itself.
Problems with the linear approach were that it often produced systems that did not meet user
requirements and were delivered late.
The modern system development life cycle (SDLC) is an iterative process. This assumes that a single
pass through any step will not produce a good outcome. At each stage new information is developed
relevant to previous stages, so it is necessary to go back and modify the earlier stages in light of the new
information learned. There is a clear understanding that no amount of front-end planning will anticipate
every aspect of the eventual system.
Curiously the old linear model and modern iterative processes both have the same steps: planning,
analysis, design, implementation and testing, support and maintenance. They just have a different
understanding of how these relate to each other.
QUESTION 4.5
Test
migration
Data Final
cleaning migration
Maintenance
Security
Data
Data migration Validation New systems
mapping
Flexibility
Future proofing
Legacy Change
systems management
Consultation
Pdf_Folio:495
QUESTION 4.6
Accountants are very experienced in dealing with financial data and ensuring its quality, they need to bring
those skills to the management of non-financial data in an organisation.
Data, if used well, is a very important and valuable asset to the business. Just as accountants install
controls and perform audits to protect physical assets, they need to help develop controls and checks to
maintain the integrity of all data.
Organisations should use a data strategy framework to identify what they need to know, and then
prioritise obtaining the corresponding data. Accountants can help by, for example, performing a cost–
benefit analysis as it can be a useful approach to decide what further data should be collected and what
should not.
Ultimately the organisation needs to prioritise its investment in data and analysis with respect to how
the data will contribute to the organisation’s strategic objectives.
QUESTION 4.7
The two main reasons that an organisation may decide not to collect and store personal data, anonymising
it instead are as follows.
• Any customer data that includes personal identifiers will be subject to privacy legislation in the country
it is held. This imposes both financial costs and significant responsibility for maintaining the security
of that data.
• Collecting personal data requires collecting informed consent and allowing people to access their data.
Again, the costs of doing this may not justify the benefits of using personal data.
Organisations cannot, however, avoid collecting personal data when they intend to provide a customised
service, such as new product recommendations, which must be matched to a specific customer’s profile to
be meaningful.
QUESTION 4.8
Overfitting refers to the practice of fitting a complex mathematical model developed from an existing
set of data to a new set of data (e.g. a cubic model instead of a simple linear model). While either
could validly be used to describe the relationship between the variables, the complex model minimises
the statistical ‘error’ (the difference between the actual data used to produce the model and the values
predicted by the model). However, these ‘errors’ are not necessarily true errors in the model, they could
be due to measurement error or the seemingly random effects of an unobserved variable.
When the overfitted model is applied to new input data to generate new outputs, the outputs are poor
estimates of the true values. Therefore, overfitting produces a model that is less likely to represent the true
underlying relationship in the new data.
In summary, overfitted models do a good job of fitting the original data used to create the model, but
because the model is usually wrong it does a very poor job of fitting any new data applied to the model.
QUESTION 4.9
Black-box approaches are those where we cannot see the process that takes data inputs and creates an
output or recommendation. Hidden workings of the AI or ML model mean that any biases introduced by
the data are unclear and the rationale for the output cannot be examined and validated. This creates the
risk that biased or flawed data may produce analytics results that will lead to incorrect or unfair decisions.
Pdf_Folio:496
QUESTION 4.10
Supervised learning can be thought of as finding known patterns in unknown data. We must have data
that has been ‘labelled’. This labelling is considered the supervision. The labels have been derived from
human recognition. Training data are the data from which the classification or prediction algorithm
‘learns’, or is ‘trained’, about the relationship between predictor variables and the outcome variable.
When training an AI- or ML-based system to detect fraudulent activity, you first need a dataset of prior
fraud. This must be correctly coded so only genuine cases of fraud are classified as fraud, and vice versa.
However, data analytics is rarely perfect, so entrusting the AI or ML model without any human oversight is
not recommended.
Specifically, humans should be involved in checking data quality, ensuring classifications are applied
consistently, that data outliers are checked and managed consistently and any missing values are checked.
Human intervention in checking source data is critical to ensuring an AI or ML model works correctly.
QUESTION 4.11
The concept of data ownership refers to who owns or is responsible for the data. How they approach that
responsibility is important.
Data ownership problems exist where there are no clear policies defining responsibility and account-
ability for managing an organisation’s data. Such problems are exacerbated when those who manage the
data are not the same as those who use it. When those responsible for data management have not ‘bought
in’ to the value of data to the organisation they cannot be expected to allocate the necessary resources
to ensure quality data is collected and securely stored. Inconsistent data formats of various departments
create an additional set of problems as organisations try to combine individual applications into integrated
enterprise systems.
Technical teams have little control over data quality and do not rely on the analysis it supports while
business users do have control and are the users of analytics. Business users should therefore have
responsibility for data quality.
QUESTION 4.12
Correlation refers to the degree of relationship between two variables and is a pattern observable in the
data. For example, if ice-cream consumption increases in summer months then we should see a regular
pattern in measurements of temperature and ice-cream consumption volumes. This pattern could be
visualised by looking at numeric data in a table where one variable has been sorted by size and the
corresponding variable also shows some size ordering. Causation is a cause-and-effect relationship
between two variables and it often starts with an observed correlation. Causation takes it further by adding
an explanation as to why those variables move in an observed pattern. This is much harder than correlation
because a convincing cause needs to remove the possibility of alternative explanations. For example, is
the increase in ice-cream consumption in summer due to a belief that ice-cream will help us cool down on
a hot day or because of a habit followed since childhood? Usually we need to conduct more research to
narrow down to the most likely cause.
Only correlation could be demonstrated visually by a scatterplot because a scatterplot is a type of data
display that shows the relationship between two variables, which is the pattern in the data. A scatterplot
cannot show the cause of the pattern.
Pdf_Folio:497
QUESTION 4.14
Exploratory visualisations are used early in the data analysis process, and exploratory analysis is about
finding patterns through breadth and depth of analysis — for example, to determine which types of model
are likely to best fit the data. An analyst is likely to briefly use a range of visualisations before moving to
more formal analysis of the relationships between variables. These can be quickly produced and are often
‘rough and ready’ as there is no intention to keep them or let people outside the analysis team see them.
Explanatory visualisations are used on completion of the data analysis. They are used to more effectively
explain and communicate the results of the analysis to decision makers. As the data analysis results will
be seen by a range of users, it is important for them to be professionally presented using visualisation.
QUESTION 4.15
Benefits Challenges
• Dashboards can quickly communicate key • Important details might be omitted because of the
information. limited space in the dashboards.
• They are interactive. • Might not be easy to apply because they require the
• They are customisable (can be designed to meet matching the data to specific users’ needs.
the user’s specific needs). • May not effectively facilitate the user’s decision
• Allow live data feeds. making because not all users can interpret the data.
• Data overload could be avoided.
MODULE 5
QUESTION 5.1
An Audit and Risk Committee provides governance (oversight) over identified risks. It receives reports on
risk management within the organisation to monitor these activities.
As a governance committee it would be responsible for creating and reviewing risk management policies.
Policies should be reviewed on a regular basis to ensure they remain fit for purpose.
QUESTION 5.2
Risk ownership involves allocating each risk to a named group or individual to manage. For this reason,
an organisation will have multiple risk owners, each responsible for a risk or a set of risks.
As risks generally cannot be eliminated, risk management involves keeping track of the risk and taking
steps to ensure it does not become a serious problem.
Specifically risk owners should:
• analyse the risk and check all associated data
• rate the risk regularly
• take steps to keep the risk at reasonable levels
• fix identified issues arising from the risk.
Pdf_Folio:498
• Has a narrow framework, specific to information • Has a broad framework for risk management.
security risk management. • Takes a strategic perspective on risk management.
• Provides specifications for an information security • Provides guidelines for implementation and risk
management system. assessment.
• Considers the role of people, processes and • Aimed at helping organisations develop a risk
technology in risk management. management culture.
ISO 31000 is the key standards framework for risk management. Notably it is a framework including
principles and guidelines. In contrast, ISO 27001 is the specific for information security management and
sets out specifications and controls for this area.
QUESTION 5.4
Risk evaluation Risk treatment
• Involves comparing measured risk estimates against • Involves selecting the appropriate response to a
pre-determined risk standards. given risk.
• Should be performed periodically. • Four main treatments:
• Should produce an easily understood outcome. - Accept the risk.
• Must be backed by evidence. - Transfer the risk to a third party.
Example: A financial institution may estimate the - Mitigate the risk.
probability of losing $10 million due to the failure - Avoid the risk.
of a client business. It may then rate that loss by its • Treatment decisions need to be informed by risk
severity and use that information to review its lending evaluation measures.
practices. Example: Insurance is a common example of risk
transfer.
QUESTION 5.5
One of the key reasons a crisis can create problems is its unpredictability, not just about when it will
occur but also the nature of the crisis. To most people COVID-19 was unpredictable so planning for it
was impossible. Similarly, natural disasters are often things that occur somewhere else, until they happen
around you. Anticipating every possible thing that could go wrong and preparing for it just is not feasible.
Risk governance is instead about identifying broad types of risk and putting plans in place that provide
a framework for dealing with the problems the risk event creates. For example, both natural disasters and
pandemics can cut off access to a business’ physical premises. So, having access to business records online
would support business continuity. But a policy that ensures the records are online cannot be easily created
and adopted at the time the crisis arises or in the immediate aftermath.
QUESTION 5.6
The chief information security officer (CISO) is responsible for developing and implementing an
organisation’s information security policy. The CISO will be involved in the organisation’s risk
management strategy.
The chief information officer (CIO) is responsible for IT-related planning, budgeting and performance
evaluation. The CIO’s decisions should be based on an effective risk management program.
The CISO should take the primary responsibility of forming the organisation’s information security
plans and policy. The CIO then takes responsibility for making resources available to implement the plans.
Both should work on regularly reviewing the effectiveness of plans and their implementation, and advise
the other on where improvements are possible.
Pdf_Folio:499
Pros Cons
• Can incorporate ‘gates’ between stages to meet • Compliance issues can be missed due to the short
governance requirements. timeframes of Agile work practices.
• Suits non-traditional hierarchies — for example, • Does not suit traditional hierarchies.
small self-led teams. • Does not provide the longer time period required for
governance to develop.
QUESTION 5.8
Key governance and control issues with automation include:
• risks relating to cybersecurity
• social licence to operate
• workforce/change management
• human resources needs
• project management
• fit with company culture
• integration with supply chain
• data privacy.
QUESTION 5.9
Three key elements of a mature approach to cybersecurity are:
• consistency
• adaptability
• accountability.
Consistency is important because an inconsistent approach will create weak points in the attack surface,
which hackers will identify and exploit to gain access to the system. For example, an organisation’s
computer-use policy must require strong passwords.
Adaptability is important because the risk environment is constantly changing. This should also be linked
to consistency to ensure all parts of an organisation can adapt together, or inconsistencies will develop.
Accountability is important to ensure everybody follows the required procedures because they know they
are individually responsible for their actions. A lack of accountability will lead to a lack of consistency in
applying cybersecurity protocols.
QUESTION 5.10
Advantages
• Supplement internal knowledge. By using the right outsourced tools, technology and approach, an
organisation can supplement its internal knowledge.
• Cost savings. Using a team of external experts through a subscription service can cost a fraction of the
amount that would be required to build internal team capability.
• Security expertise. IT security professionals for in-house teams can be difficult to identify, recruit and
retain. It can also be difficult to keep up to date with security trends and developments.
Pdf_Folio:500
Disadvantages
• Sensitive data. Using and trusting outside parties with sensitive data or intellectual property can
potentially open the organisation to additional risks.
• Additional software/hardware requirements. External providers may mean the organisation must use
specific security software or they may suggest additional hardware requirements as part of the terms
and conditions of service. This can result in additional costs and potentially slow down ICT systems.
• Service-level agreements. Ensuring the service-level agreements and support matches the organisation’s
specific requirements can sometimes be difficult.
• Lack of control. Sacrificing and relinquishing key cybersecurity responsibilities to eternal providers
requires monitoring.
QUESTION 5.11
The attack surface relates to the points within a business that are exposed to any form of exploitation.
Networks and servers are highly interconnected so once the surface is breached access to extensive data
resources is gained by the attacker.
Prevention controls include firewalls, anti-virus software, encryption, risk analysis, job rotation and
account lock outs.
Prevention controls need to be deployed at every point on the attack surface to be effective.
With the development of the internet of things and edge computing, the number of devices on each
network has grown significantly. More connected devices means there are more potential access points for
attackers to exploit and results in the need for more prevention controls.
QUESTION 5.12
The General Data Protection Regulation (GDPR) provides a good framework for ethical data practices.
• The right to be informed. The company should have systems in place to inform individuals when
information is collected.
• The right of access. The company should have systems in place to allow individuals to request and
access all personal information the about them.
• The right to rectification. The company should have systems in place to correct information if provided
with evidence that it is incorrect.
• The right to erasure. The company should have systems in place to remove information if requested,
this should cover all backups as well as the main database.
• The right to restrict processing. The company should have systems in place to prevent data from being
used if requested by the individual who is the source of the data.
• The right to data portability. The company should have systems in place to convert data into commonly
used machine-readable formats.
• The right to object. The company should have systems in place to prevent data from being used for
marketing or profiling if requested by the individual who is the source of the data.
QUESTION 5.13
Having a specialist compliance officer ensures that there is at least one person focussed on compliance
issues. It is also more likely that this person will be in a position to maintain some level of independence
and objectivity as this is the sole focus of their role. Their KPIs will all be compliance related so there is
no danger of having to sacrifice compliance-related tasks in favour of other tasks.
Giving responsibility for compliance to someone with a broader job description creates the risk that that
individual will allocate insufficient time to compliance, especially as there is no-one with a greater role in
compliance to tell them otherwise.
A specialist compliance officer is also more likely to stay up-to-date with current best practices
in compliance.
Pdf_Folio:501
QUESTION 5.14
Regulatory sandboxes allow FinTech companies to test and experiment with new products and services
free from costly or damaging legal consequences. The goal is to give FinTech companies a chance to
develop and take a minimally viable product to market knowing it has value for users.
The main downside for FinTech companies is that as long as they stay in the sandbox they are not able
to fully test their products; nor will they know the full extent of the cost of their operations.
The main benefit for regulators is that it allows them to maintain oversight and knowledge of FinTech
innovations, giving them awareness of potential regulatory requirements so that they can plan for and
implement them before they are needed.
The sandbox approach protects consumers without hindering innovation, allowing new players in the
FinTech industry to test in a safe environment while giving regulators advance insights into potential
changes required in the regulatory environment.
QUESTION 5.15
The main issue for FinTech companies looking at international expansion are the regulatory and compli-
ance requirements they encounter (and must comply with) in each new jurisdiction.
They will also need to ensure compliance with rules imposed by multinational bodies such as the
United Nations.
There may also be additional compliance requirements in their home country, such as dealing with
AML/CTF rules if they were not previously dealing with international clients.
However, if they are already compliant in a heavily regulated jurisdiction like Australia, they should
meet many of the compliance requirements in other countries.
Pdf_Folio:502
INDEX
A/B testing 322 bias 200 automation 3, 143, 158–70
accept 385 sensitivity 199 balanced scorecard to evaluate
access to devices 31 usability 199 166–7
account reconciliation 180 volume 199 basics of 158–62
accountability 394 in different industries 197 combining technologies 162
accounting/auditing 111 agriculture 198 key automation technologies
Afterpay 95 aviation 198 158–61
AGI see artificial general intelligence finance 197 building automation strategy 168–9
agility, responsiveness and crisis healthcare 197 data 168
management 284–5 marketing 198 infrastructure 169
AI see artificial intelligence recruitment 198 organisation 169
Airwallet 97 skills 169
retail 198
Airwallex 44 technologies and algorithms
enablers 195
Airwallex increasing integration 26–7 168–9
ethical considerations 231–2, 404
Airwallex payment solution 8–9 and employment 232
in finance and business 196–7
algorithm 159 evolving 162
managing risks and challenges when
alternative finance governance, risk and compliance
implementing
definition 125 challenges 227–30
appropriate data 200
importance 126–7 of governance, risk and compliance
regulating 127 costs 201
228–9
tools, taxonomy of 125 data preparation 200
planning for success 167–70
traditional vs alternative lender employee communication 201
purposes of using 158
funding process 126 skills 201
technologies 159
alternative funding 130–4 methodology for AI strategy 202–4
artificial intelligence 160–1
investment-based crowdfunding preparing for an automated future
machine learning 161
131–3 233–4
robotic process automation
non-investment-based crowdfunding risks 198–9
159–60
133–4 to individual 198–9
technology use cases 226–7
alternative lending 127–30 to organisation 199
value creation using 226–7
balance sheet lending 128–9 to society 199 value in finance 233
invoice finance 129–30 strategy and business strategy 202–4 value of 163–7
peer-to-peer lending 127–9 types of 195–6 autonomous systems 232
alternative viewpoints 318 artificial moral agents 232 availability 397
Amazon digital ecosystem 24–5 artificial narrow intelligence (ANI) availability bias 318
analyse 247 195 avoid 385
analytical workforce 289 assess and strategise 289 awareness 318
analytics 3, 64, 244–7 association rule mining 304, 305
requirements 265 attack surface 411
anchoring 318 balance sheet lending 128
attended RPA 172
ANI see artificial narrow intelligence vs peer-to-peer lending 128–9
audit control processing 181
anonymising data 293–4 balanced scorecard to evaluate
auditors 398 automation 166–7
Anscombe’s quartet 330–2 augmented reality (AR) 53, 143
Ant Technologies 145 bank sector compliance 436–7
Australia, FinTech regulation 445–6 bank, regulations 458–9
Anti-Money Laundering and
APRA 444 banking business 445
Counter-Terrorism Financing Act
ASIC 443 banking consumer archetypes 142
2006 (AML/CTF Act) 446
AUSTRAC 444 Banking Executive Accountability
API see application programming
banking activities 445 Regime (BEAR) 398
interface
credit activities 445 banking, future of 142–7
API platforms 143
Apple Pay’s digital payments ecosystem RBA 444 banking-as-a-platform 146
33–4 regulation 445–6 banking-as-a-service 146
application programming interface (API) regulation and law 444 banks’ strategic response to FinTechs
16, 47–8 Australian credit licence 443 145
appropriate data 200 Australian Prudential Regulation customer preferences 142
AR see augmented reality Authority (APRA) 444 enablers of change
argue 247 Australian Securities and Investments regulatory framework 143–4
artificial general intelligence (AGI) 196 Commission (ASIC) 443 technologies 143
artificial intelligence (AI) 3, 160, Australian Stock Exchange (ASX) 395 neobanks in Australia 146
193–205 Australian Transaction Reports and open banking regime 147
basics of 193–6 Analysis Centre (AUSTRAC) 444 operational effectiveness and
benefits of 196–8 authorised deposit-taking institution efficiencies 144–5
brief history of 194–5 (ADI) 444 trust 146
data challenges 199–200 authority 394 banking-as-a-platform (BaaP) 146
availability 199 automating action 300 banking-as-a-service (BaaS) 146
INDEX 503
“BMIndex_PrintPDF” — 2021/9/15 — 13:30 — page 504 — #2
Bayes’ Theorem 255–6 causal relationships, criteria for performance evaluation 432–3
behaviour prediction 165 assessing 324 planning 432
bias 200, 317 causation 321–4 and risk management 213
in decision systems 232 causal relationships, criteria for support 432
biased analysis 294 assessing 324 compliance framework 433
biased data 294 common patterns of 323 development 433–6
bidirectional virtual currency 81–3 proving 322–3 effective framework 436
big data 2, 64, 244, 287–8 CBA ML strategy 218–9 KPMG 435–6
big data analytics 288 Central Bank Digital Currency (CBDC) looking forward 433
business using 288–9 scheme 79 conditional probability 254
conventional analytics 287–8 centralised databases 100 confidentiality 397
implementation 289–90 centralised virtual currencies 80 confirmation bias 318
challenges 290 change management 282–4 consensus algorithm 100
initial steps 289–90 Kotter’s eight-step framework for consent 426
in risk management 278 successful 282 consistency 324
skills required by finance changes over time 342–5 consolidation 289
professionals 290–2 channel innovations 13 consortium blockchains 106
core skills for digital future 291 chatbots and voice-based assistants Consumer Data Right (CDR)
developing skills 291–2 164–5 scheme 144
new skills for digital era 291 checkout-less stores 40 content categorisation 164
big data analytics 10 chief financial officers 61 context of data visualisation
Bigstone 45 chief information officer (CIO) 397 identifying the audience of
biometric data 271 chief information security officer presentation 334
bitcoin 8, 83–4 (CISO) 397 present the communication 335
BitDegree 136 classification 302–4 present to the audience 334
blockchain 8, 10, 103, 143, 233 classifying ML 212–3 contextual extraction 164
consensus mechanisms 103–4 closed loop systems 30 continuous data 253
in financial services 110–13 closed virtual currency 80–1 contract analysis 213
cloud computing 47, 64 control documentation 382
permissions and ownership 105–6
cloud-based payment systems 89–90 conventional vs smart contract 108–9
solution to double-spending problem
clustering 304 convertible virtual currency 81–3
104–5
Coca-Cola 178–9 cooperative networks 20
Board 396
coding 317 Corporations Act 2001 443–5
borrower and investor perspectives 139
cognisability 75 corrective controls 419
bots 159–60
cognitive automation 162 correlation 319–21
Bradford Hill Criteria 324
cognitive bias 317–9 correlation coefficient 319
brand innovations 13
common types of 318 cost reduction 289
breaches, cryptocurrency 122
strategies to overcome 318–9 costs 201
bubble chart 345
cognitive ML 212 costs of misclassification 299
business activity data 289
cognitive principles to visualisations COVID-19 229–30
business case for RPA 177–9
eliminate unnecessary detail 347 and alternative lending 126
business continuity plans (BCP) 388–9
coins 113–5 Boards 403
business functions
collaboration 318 cashless society response 79
human resources 183
collaborative filtering 305, 306 contactless payments 94
marketing and sales 183
collaborative networks 20 CPS 234 security standard 412
supply chain management 183
collectively exhaustive 254 created data 274
business impact assessment (BIA) 388
combatting financial fraud 181 credit model relationships 365
business operations data 289 commodity currency 77 critical control 382
business process management (BPM) commodity-backed currency 77 cross-border payments 40
172–3 Common Reporting Standard crowdfunded equity 46–7
business processes 184–5 (CRS) 447 crowdfunding 6–7
cybersecurity 185 Commonwealth Bank of Australia crowdlending 127
IT processing 184 (CBA) 447 crypto derivatives 118–19
product and service marketing 185 communication 265 crypto wallet 92
sales order processing 184 communication networks 29 crypto-based funding
business strategy 265 competition 32 initial coin offerings 135–6
business, benefits of digital payments and risk 272–3 initial DEX offerings 137
for 96 competitive advantage 5, 289 initial exchange offerings 136–7
buy-now-pay-later (BNPL) services compiled data 274 security token offerings 137–8
40, 95–6 compliance 213, 227, 309, 373 cryptocurrency 7, 78, 83–6, 448
regulatory response 448–50 action 436–8 applications 117–9
context of the organisation 432 benefits of 121–2
captured data 274 improvement 433 coins 113–15
card scheme rules 451–2 leadership 432 considerations and future outlook
card-based payments 88, 90–1 moving forward 123–4
cash flow data, finance 270 reimagining resources 439 defined 113
cash, and e-money decoupling 80 staying current 438–9 and finance professionals 121
cashless society 79–80, 97 operation 432 fraud risk 122–3
504 INDEX
“BMIndex_PrintPDF” — 2021/9/15 — 13:30 — page 505 — #3
INDEX 505
“BMIndex_PrintPDF” — 2021/9/15 — 13:30 — page 506 — #4
506 INDEX
“BMIndex_PrintPDF” — 2021/9/15 — 13:30 — page 507 — #5
distributed ledgers 73, 101–3 financial markets 45–8 future finance talent 233
divisibility 76 APIs, cloud and open accounting future-ready finance function 4
document summarisation 164 47–8
domain knowledge 316 neobanking 48 gap analysis 225
donation-based crowdfunding 134 P2P lending 45–6
gap in digital skills 31
DoS (denial of service) 411 SME finance 46–7
gender gap 31
driving innovation 201–2 financial product 445
general AI 196
durability 75 financial return 133
General Data Protection Regulation
financial review preparation 179
(GDPR) 428–9, 446–7
e-money 78–9 financial services 445
general employees and digital
advantages and disadvantages 78 business 445
transformation 62–3
cashless society 79–80 governance risks 442
Gestalt principles 347–50
Central Bank Digital Currency 79 legal risk 442–3
global environment
decoupling e-money and cash 80 technology and operational
risks 441–2 card scheme rules 451–2
hardware-based products 79
FinTech 141 customer journey
software-based products 79
banks’ strategic response to 145 disputes 451
earnings before interest and taxes (EBIT)
23 in business 36–40 marketing 450–1
EBIT see earnings before interest and current solutions in business 37 onboarding 451
taxes current use by accounting and finance reporting of issues 451
edge computing 48–9 professionals 40–3 transactions 451
electronic fund transfer (EFT) 88, 91 decentralised finance 49 scalability 452
electronic invoicing 6 edge computing 48–9 global financial crisis (GFC) 454
electronic money see e-money emerging 48–9 global regulation
elementary event 254 innovation incentives in financial Common Reporting Standard
emerging digital currency platforms 16 services sector 42–3 447
emotion analytics 165 basic necessity 42 Financial Action Task Force 446
employee communication 201 equity 43 Anti-Money Laundering and
employee morale 187 evolving preferences 43 Counter-Terrorism Financing Act
enabling skills 3 legacy systems with hope of 444, 446
enhanced productivity 289 innovation 42 General Data Protection Regulation
environmental changes 299 profitability 42 (GDPR) 446–7
equity 43 regulation 42 good governance 393
equity-based property crowdfunding innovation opportunities 49–50 principles of 395–6
132 innovation roadblocks 49 risk management strategy 394–5
ethereum 8, 107 innovation thinking in 43–8 governance
ethereum 2.0 107 financial markets 45–8 accountability 394
Europay, Mastercard, Visa (EMV) 89 payments 44–5 auditors 398
event 254 wealth management 43–4 authority 394
experimental data 274 innovations in retail 40–2
automation and AI 404–5
explanatory analysis 332–3 scope of 36–7
Board 396
exploratory analysis 330–2 strategic issues in 39–40
chief information officer 397
exploratory and explanatory vs traditional financial services
chief information security officer
visualisations 330–3 37–9
397
external data 273 FinTech 1.0 16
culture
extracting knowledge from data FinTech 2.0 16, 38
315–16 fit-for-purpose 400–1
FinTech companies 11
DIKW pyramid 315–16 FinTech mix 38–9 organisational culture 401–2
extrapolation bias 318 FinTech payments niche 37 positive governance culture
FinTech regulation 400
consumers and investors 440–1 data custodian 397
facial recognition 163
financial services firms 441–3 data owner 397
failure to locate responders 299
failures, cryptocurrency 122 firms 452 elements of 393–4
fiat currency 77–8 global regulation 446–7 future threats 403–5
fictional currencies see closed virtual key regulators in Australia 443–4 increased accountability 398
currency lack of awareness 441 IT security practitioners 397–8
fiduciary currency 78 paternalistic approach 441 lacking 399–400
finance 269–70 technological change 440–1 roles and responsibilities
finance innovation lab 4 too much information 441 396–8
Financial Accountability Regime FinTech solutions niche 37 security awareness trainers 398
(FAR) 398 FinTech technology niche 37 senior management 396–7
Financial Action Task Force Five Cs of credit 364 systems owners 397–8
(FATF) 446 fraud data 272 technology impact 403–4
financial inclusion 15 fraud risk 122 governance, risk management and
financial information analysis frictionless payments 98 compliance (GRC) framework
180–1 functional strategy 223 373
financial instruments 75 funds 379 government
financial literacy 98 fungibility 75 RPA in 185
INDEX 507
“BMIndex_PrintPDF” — 2021/9/15 — 13:30 — page 508 — #6
508 INDEX
“BMIndex_PrintPDF” — 2021/9/15 — 13:30 — page 509 — #7
management security 419 open virtual currencies see convertible planning for success 167–9
manipulation of behaviour 231 virtual currency platform/online-based 92
marketing and sales 183 open-loop systems 30 plausibility 324
Marketlend 45 operating standard 382 PledgeMe 47
marketplace lending operational dashboard 352–4 point-of-sale (POS) 40–1
127–8 operational efficiency 57, 289 poor data quality 301
matching 323 operational strategy 223 poor/inappropriate risk governance
measure of value/unit of account 75 operations 270–1 communication of risk 398–9
medium of exchange 75 opportunity assessment 53 governance is lacking 399–400
miners 103 optimising ML 212 portability 75
missing values 298 optimism bias 318 portal queries 180
mission statement 268 oracle 108 portfolio management 214
mitigate 385 ordinal data 253 POS see point-of-sale
ML see machine learning organisation positive correlation 319
mobile payments 8, 30, 44 blending mainstream and alternative pre-attentive attributes 347–50
mobilisation 54 financing 139 prediction tasks 163–6
modelling outcomes 365–6 borrower and investor perspectives predictive analytics 165–6, 245
money 139 predictive ML 207, 212
and currency 75–8 financing needs 138–9 predictors, importance of 311
functions of 75 funding sources 139 prescriptive analytics 245, 288
properties of 75–6 impact on current finance system prescriptive security 143
vs currency 76 140–1 preventative controls 419
money laundering organisational behaviour 249 privacy 406
detecting 309–13 organisational implementation model Privacy Act 1988 230, 380, 426, 428,
phases of 309 188 429, 437, 458
money supply 75 organisational politics and culture privacy and surveillance 231
MoneyPlace 45 277–8 private blockchains 105–6
MUFG Coin 117 organisational strategy 51–2, 374 private permissioned blockchain 105
multiple regression 323 strategic management 52 probability 253–61
mutually exclusive 254 environments and goals 52 in decision making 256–8
processes 52 in practice 258–61
narrow AI 195 outcome 254 in risk analysis 258–9
National Australia Bank 373–4 outliers 297–8 stage weighted vs individual weighted
natural language analytics 300–1 oversampling 311–3 260–1
natural language processing (NLP) 160 rare events 298–9 terms and concepts 254–6
negative correlation 319 process innovations 12–13
neobanking 48 process selection 179, 187, 193
neobanks 146 P2P lending see peer-to-peer (P2P) processing power 195
Network Age 2 lending product and service marketing 185
network effects 2 P2P payments see peer-to-peer payments product crowdfunding 46–7
network innovations 12 P2P technology 10 product innovation 58
New Payments Platform (NPP) 89, 448 partition 254, 295 product performance innovations 13
NFC payments 89 paternalistic approach 441 product system innovations 13
NFC see near-field communication payment 86 profit-sharing crowdfunding 133
NIST Cybersecurity Framework (NIST Payment Card Industry Data Security profitability 42
CSF) 413–14 Standard (PCI DSS) 413 programmed decision-making 327
NLP see natural language processing payment gateways 88 programming tools and languages
nominal data 253 payment service provider (PSP) 88 291
non-investment-based crowdfunding payments 44–5 project mandate 416
133–4 payments ecosystem 32–4 proof of concept (POC) for RPA 190
normalising and rescaling data 298 payments systems 28 proof of stake 103–4
not-for-profits PayPal 8, 88, 93–4, 437 proof of work 103
RPA in 185–6 peer-to-peer (P2P) payments 8 property-based crowdfunding
Notifiable Data Breach (NDB) Scheme peer-to-peer invoice finance 129 132–3
429 peer-to-peer (P2P) lending 45–6, protect assets 374
127–8 provoked data 274
objective probability 254 vs balance sheet lending 128–9 public blockchain 105
observational study 323 permissioned blockchains 105 public permissioned blockchain 105
onboarding 159–60, 451, 456–7 permissioned ledger 101
ongoing evaluation and improvement permissionless blockchains 105 qualitative analysis 317
55 permissionless ledgers 101 qualitative data 252–3
observe-orient-decide-act (OODA) loop personal data 427 quality 58
326 personalisation 97–8 quality of work 31–2
opacity of AI systems 231 PESTLE (political, economic, social, quantitative analysis 317
open accounting 48 technological, legal and quantitative data 252–3
open banking 16, 40, 144, 274–5, 439, environmental) 379–80 quick response (QR) code payments
448 physical assets 418–19 92, 93
open marketplaces 16 placement 309 quick wins — efficiency 190
INDEX 509
“BMIndex_PrintPDF” — 2021/9/15 — 13:30 — page 510 — #8
random experiments 322 ISO 27005 377 scatter plot 345, 346
RateSetter 45 ISO 31000 377 SDLC see systems development life
ratio data 253 minimise adverse impacts 374 cycle
raw values 335–6 objectives 374 search engine data 272
read 247 risk evaluation and risk treatment Second Life 82
readiness to adopt 195 384–9 security awareness trainers 398
real estate crowdfunding 132 risk identification 382 security controls 418
real-time monitoring 300 risk monitoring and action categories of 419
recognisability 75 389–91 information assets 419
recommendation engines 165 risk profiles 378–81 physical assets 418–19
regression 302, 304 responsibilities 376 statement of applicability 419–21
RegTech see regulatory technology risk governance 373–4 types of controls 419
regulatory compliance 180 value chain 375 security token offering (STO) 137–8
regulatory framework 143–4 risk monitoring and action semi-structured data 158, 252
regulatory response 448–50 audit 390 semi-supervised learning 303
regulatory risk 122 control validation 390 Senate Select Committee 457
regulatory sandboxes 144 ISO 31000 monitoring 390 senior management 396–7
regulatory technology (RegTech) key risk indicators 390 sensitive data 427
143–4, 454 risk triggers 390–1 service innovations 13
ASIC perspective 458 risk owners 376 singularity and superintelligence
Association membership 462–3 risk profiles 232
benefits of 455–6 analysis for 379–81 skills 201
future 461–4 information assets 379 smart contract technology
industry barriers to implementation risk levels 381 advantages and limitations of 110
458 risk rating 384 distributed applications 107–8
matching 456–7 risk register 382 ethereum 107
mitigate non-compliant marketing risk risk retention 385 oracle 108
461–4 risk tolerance 400 vs conventional contract 108–9
RAAS 460 risk treatment 385 smart contracts 107
Senate Select Committee 457 risk triggers 390–1 smart devices 143
strategies for 459 robo-advice 49, 214 smart regulation and compliance
top-down implementation 457–8 robo-advisors 9, 49, 214 143–4
trust and regulation 460–1 robotic process automation (RPA) SMART strategy board 267
remittances 111 159, 171 SME finance 46–7
representative currency 78 balanced scorecard 177–9 invoice financing 46
Reserve Bank of Australia (RBA) 444 basics of 171–6 product crowdfunding and
resources 272 benefits 176–86 crowdfunded equity 46–7
retail industry 118 building an effective strategy 188–92 Snowball Effect 47
retail payment systems 86–8 in business 182–6 social media data 272, 289
reward-based crowdfunding 133–4 business processes 184–5 social networks 10, 15
RFID (radio-frequency identification) industries 185–6 soft skills 61
sensors 271 business process management 172–3 software ecosystems 20–1
risk and organisational strategy 374 capability — effectiveness 191–2 specificity 324
risk analysis categories 173 speech-to-text and text-to-speech
impact 383–4 characteristics 171 conversion 164
likelihood 383 creating business case for 177–9 spreadsheet correlation function 320–1
risk rating 384 developing a solution 189 stability 75
risk appetite 375, 400 experimenting with 189 stablecoin 96, 116–17
risk appetite statement 400 in finance function 179–82 stage weighted vs individual weighted
risk communication management 399 implementation 190–2 probabilities 260–1
risk criteria 383 maturity model 190 stakeholder involvement 289
risk evaluation 384–9 risks and challenges for using 186–7 standard journal entries 180
business continuity plans 388–9 data security 186 standard of deferred payment 75
disaster recovery plans 388–9 employee morale 187 statement of applicability (SOA)
reducing risk 385–6 process selection 187 419–21
treat risk, controls 386–8 system change 187 statistical inference 287
risk governance 373–4 technical issues 186–7 statistical methods 291
good governance 393 success factors 188–9 statistics 252
risk identification unattended, attended and stock market data, finance 270
causes, impacts and controls 382 hybrid 172 store of value 75
identify risk 382 vs traditional automation 174 storytelling 333–4
risk register 382 RPA see robotic process automation strategic dashboards 352
risk management 227, 373 strategic management 52
best practice 377 sales and inventory reconciliation environments and goals 52
framework and process 375 181–2 processes 52
function 376 sales optimisation 161 strength of association 324
future innovation 391–2 sales order processing 184 strong AI 196
ISO 27001 377 scarcity 76 structure innovations 12
510 INDEX
“BMIndex_PrintPDF” — 2021/9/15 — 13:30 — page 511 — #9
INDEX 511