Multinational Management
Multinational Management
Digitized by Google
Multinational
Management
David P. ~utenberg
Queen's University, Canada
~
Little, Brown and Company Boston Toronto
Digitized by Google
~u~ A~rfI
Ht>
(,;,(. ~
,~91
Rutenberg, David.
Multinational management.
Bibliography: p.
Includes index.
1. International business enterprises - Management.
I. Title
HD62.4.R87 658' .049 81-23641
ISBN 0-316-76365-9 AACR2
MV
Digitized by Google
........-........-c_=___
. ~---
/-"I..,y~
~".;2007b /
APR 1 9 1984
Contents
Introduction xi
Acknowledgments xv
Introduction 10
Experience Curves in a Product-Structured Corporation 12
Risk-Spreading in a Geographically Organized Corporation 18
Geocentric Balance in a Matrix Organization 23
Bibliography to Chapter 1 33
PART ONE
Multinational Financial Management 35
Digitized by Google
vi Content.
History of the Company 38
Foreign Exchange Risk and Hedging 41
Introduction 46
Polycentric View of Exchange Management 49
An Ethnocentric View of Exchange Management 56
The Geocentric View of Exchange Management 59
Conclusion 64
Questions from Other Viewpoints 66
Bibliography to Chapter 2 68
Ethnocentric - Introduction 81
Geocentric Model- A Generalized Network 82
Adjusted Transfer Prices A(pfj} 86
Fees and Royalties A(ffj} 90
Intersubsidiary Loans A (if: J+1} 92
Dividends A(d~} 95
Local Loans 98
The Optimal Capital Structure of a Subsidiary 99
Systems within Systems 100
Implementing the Model in a Polycentric Corporation 102
Conclusion 103
Questions from Other Viewpoints 104
Bibliography to Chapter 3 106
Digitized by Google
Content. vii
Introduction 122
Polycentric Process 122
Ethnocentric Process 125
Geocentric Process 131
Closing Decisions 136
Conclusions 137
Questions from Other Viewpoints 139
Bibliography to Chapter 4 141
PART TWO
Multinational Manufacturing 145
F I V E: Logistics 147
Introductory Note to the Case 147
Ascendant Electric of England Ltd. 148
Cost Review of Supplying Los Angeles Customer by Sea 148
Thinking about Airfreight 150
Negotiating Strategy 152
Introduction 153
First Phase: Product-Centered View of Logistics (Ethnocentric) 153
Second Phase: Subsidiary-Centered View of Logistics
(Polycentric) 163
Third Phase: Geocentric View of Logistics 164
Conclusions 170
Questions from Other Viewpoints 170
Bibliography to Chapter 5 172
Introduction 182
Ethnocentric Production·Focused Production Smoothing 182
Polycentric Nation·Focused Production Smoothing 184
Geocentric Production Smoothing 187
Digitized by Google
viii Content.
Conclusions 202
Questions from Other Viewpoints 202
Bibliography to Chapter 6 204
Introduction 220
How Governments Perceive Factories 220
Plant Location Decisions in a Geographically Decentralized
Corporation 223
Ethnocentric Plant Location to Minimize Cost 228
Geocentric Perspective on Plant Location 231
Questions from Other Viewpoints 236
Bibliography to Chapter 7 237
PART THREE
Multinational Marketing 241
Introduction 255
Ethnocentric SimUitude 255
Polycentric Production Smoothing 259
Geocentric Launch 263
Ethnocentric Appendix 267
Geocentric Appendix 270
Questions from Other Viewpoints 273
Bibliography to Chapter 8 275
Digitized by Google
Contents ix
NI N E: Pricing 277
Introductory Note to the Case 277
Tyler Abrasives Inc. 278
Background of Genag 279
Tyler's International Operations 280
The Second Meeting with Genag 282
The Selling Process 285
Sales Overseas 286
Further Developments with Genag 286
Summary 290
Questions 291
Introduction 293
Umbrella Pricing 294
Constraints on Price Policy 296
Data Required for Multinational Pricing 298
Three Pricing Schemes 299
Polycentric Pricing Scheme 300
Geocentric Pricing Scheme 302
Ethnocentric Pricing Scheme 303
The Three Pricing Schemes Discussed 305
Appendix 1. A Mathematical Formulation of the Pricing Model 306
Appendix 2. Mathematical Formulation of Geocentric Pricing
Scheme 309
Appendix 3. Mathematical Formulation of Ethnocentric Pricing
Scheme 310
Questions from Other Viewpoints 312
Bibliography to Chapter 9 314
Introduction 331
Polycentric View of Product Design 331
Ethnocentric View of Product Design 333
Geocentric View of Product Design 334
Conclusions 339
Appendix. Two Market Research Techniques 339
Digitized by Google
x Content.
Questions from Other Viewpoints 341
Bibliography to Chapter 10 343
PART FOUR
Multinational Executive Development 347
Introduction 358
Future Management Needs 358
Program Development 363
Bibliography to Chapter 11 372
Index 379
Digitized by Google
Introduction
xi
Digitized by Google
xii Introduction
ously, the marketing function may be run in a decentralized polycentric
manner. The three typical organizational patterns are described in Chap-
ter 1.
In this book, you will be working to understand the operating decisions
of an executive at the headquarters of a large multinational corporation,
with competent functional managers in each subsidiary. Your challenge is
to identify decisions that are worth centralizing. There is no point in
headquarters making decisions the subsidiary is able to make on its own;
unnecessary duplication is both wasteful and frustrating. Each chapter of
this book deals with a decision that might justify duplicated effort. The
first chapter deals with the organizational problems of making a global en-
terprise cohesive. Plow straight through Chapter 1. Don't linger on details,
but make a note to yourself to reread Chapter 1 when you finish the book.
The body of the text deals with decisions to be made in finance, produc-
tion, and marketing. Each chapter begins with a case. Working in a corpor-
ation you will be faced with caselike problems, but you will have no case
book to support you. As a student, run the risk of feeling befuddled: Read
the case and rough out a solution before continuing with the chapter. In
corporate terms, the information in the chapter corresponds to detailed
staff work that usually follows an executive's rough decision. Such staff
work shows how approximations could be refined and develops the
broader consequences of that rough decision. View each case as an
opportunity for you to practice making a rough decision, then read the
chapter to see how your decision could be extended and refined.
1. The Corporation, Not the Society. The focus of this book is the
corporation. Most governments are concerned about multinational
corporations and have legislated to control their operations. For more
thoughts on this extremely important issue, read Behrman (1970),
Blake and Walters (1976), and Bergsten, Horst and Moran (1978) [see
Bibliography to Chapter 1] .
2. Headquarters, Not Subsidiaries. This book views worldwide corpora-
tions, as personified by their headquarters. Someday researchers will
explore the delicate balance of life in a subsidiary, especially the
aspiration of each subsidiary manager to maximize autonomy, but we
are concerned only with headquarters.
Digitized by Google
Introduction xiii
Digitized by Google
xiv Introduction
Digitized by Google
-----;....---...-.-------::=-=:..:~ - - - - - -
Acknowledgments
xv
Digitized by Google
xvi Acknowledgments
MIT has been a valuable critic, and first gave voice to the distinction
between financial risk and the ability of the multinational to adjust its
physical assets.
Chapter 3 originated in my doctoral dissertation at Berkeley. When I
published some of the raw ideas, Representative Vanek of Ohio denounced
maneuvering liquid assets as "a cruel insult to the taxpayer who does his
share" (Congressional Record, Vol. 116, No. 117, July 13, 1970, pp.
H6640-4). I felt obliged to analyze the concepts more thoroughly.
I was fortunate to be a consultant to the corporate planning group of
Westinghouse Electric Corporation when Patrick Lynch, Ed Uber, and
Lou Greulich were struggling to conceptualize a corporate stance to help
the Westinghouse Nuclear Division negotiate its role with the government
of France. The government of France had other ideas, so the participants
learned a great deal. Ion Amariuta, now at Coopers-Lybrand, taught me
about joint ventures, and John Matthews and Joan Lang of IREX guided
me in Eastern Europe. Executives in many other multidivisional corpora-
tions contributed ideas that are woven into Chapter 4.
Chapter 5 probably originated when I worked on oil refinery planning at
Standard Oil of California. and included tankers in linear programs. I am
grateful to transportation managers at Gulf Oil, National Steel, and U.S.
Steel. Michael Potter of Massey-Ferguson and Donald Bridewesser. now an
independent consultant, sharpened my thinking.
Chapter 6 owes most to Vlf Peter Welam of Boston University's Brussels
campus, with whom I worked to develop many algorithms for production
smoothing. For this book, I wanted a very simple algorithm, and developed
it with Kenneth Fraser, now with the Canadian Department of National
Defence.
I evolved my understanding of plant location working with Ram Rao,
now at Purdue University. After we had completed one joint paper. we
proudly explained its ideas to Lee O'Nan of Alcoa, who explained the
problem of uncertainty due to rivalry, which led us to model preemptive
rivalry. Chapter 7 shows that heritage.
The theme of Chapter 8 is the coexistence of inadequate information
and rivalry when a new product is launched. I am extremely grateful to
W. W. Cooper. now of the University of Texas, from whom I learned to
model partial information. and to Roger Wets (Kentucky). Stanley Garstka
(Yale), and Morris de Groot (Carnegie-Mellon) with and from whom I
learned stochastic programming with recourse and the nature of optimal
adaptivity.
Chapter 9 originated at IBM World Trade when I worked for Emil
Schell. The chapter took form with the help of mathematical ideas from
Erwin Diewert of the University of British Columbia, tractor price data
from Norman McDonald of Corporation House Ltd., and experience curve
concepts from my students who joined the Boston Consulting Group.
Tim Shaftel. University of Arizona, and I wrote two papers which
Digitized by Google
Acknowledgments xvii
provided the skeleton for Chapter 10. As Manfred Padberg, now McKinsey
Germany, worked to apply the ideas of product design, the case study on
Black and Decker Canada Inc. took form. We are very grateful to Jack
Beckering for the many hours he spent educating us.
For Chapter 11, my intellectual debt is again to W. W. Cooper, University
of Texas at Austin. The U.S. Navy was forecasting that retirements would
lead to an acute shortage of civilian aircraft mechanics, just as their air-
craft were becoming more complex. Bill Cooper and his colleagues re-
sponded to this ONR challenge by modeling simple jobs as stepping stones
to more complex jobs. Simultaneously, I worked with executives who
were bemoaning their inability to expand abroad for lack of qualified
managers, so I saw the obvious analogy. My work with Richard Fleming of
Pittsburgh has enriched the human qualities of this chapter. Material
about executive development is so extensive yet nebulous as to deserve
an entire volume. Joan Huang, then a Queen's student, helped me prune
it to fit this book.
I sometimes think of this book as a stained glass window. The visions
of the whole have been shared with me by experienced international
executives. The stone structure and lead latticework is the rigorous theory
which I have drawn from my academic colleagues at Berkeley, Camegie-
Mellon, and Queen's. Individual pieces of colored glass have come from
the vivid examples of managers and students.
The richness of these streams of people has encouraged me to become a
good listener, and I am grateful to each in addition to those specifically
acknowledged. Conversely, I am responsible for any errors and misinter-
pretations that remain in my thinking and in this book.
Digitized by Google
Multinational Management
Digitized by Google
CHAPTER ONE
Digitized by Google
2 Multinational Management
Digitized by Google
Organizational Structures and Tensions 8
Historical Growth
Citibank opened its first international branch in London in
1902, then opened in five Asian countries that same year. It
moved to Buenos Aires in 1914, and opened six more Latin
American branches the following year. By 1939 it had 100
overseas offices. The branch network contracted sharply
during World War 11 but accelerated rapidly after the mid-
1950s. Between 1956 and 1965 the number of overseas
branches increased from 69 to 125. By 1973 Citibank was
operating in 95 countries. Overseas it had 242 branches, 320
affiliate offices, and 69 banking subsidiary and representative
offices. The Bank of America, its closest rival, was a poor
second with 103 branches in 44 countries.
International Organization
In 1973, the bulk of Citibank's international activities was
organized geographically, under the IBG (see Exhibit 1.1).
The group was led since its creation by G.A. Costanzo,
executive vice president. Five divisions, each headed by a
senior vice president, reported directly to him: (1) Asia and
Pacific, (2) Canada and Caribbean, (3) South America, (4)
Europe, and (5) South Asia, Middle East, and Africa. In each
country where the bank operated there was a senior officer
(commonly referred to as the senof) in charge of all Citibank
activities. Each senof reported to one of the division heads.
Traditionally, the senofs were the main line of Citibank's
international organization and were very strong guardians of
their domains. In some foreign countries, where the bank had
a major share of business, a senof may have controlled as
many as 5000 employees.
Within a country, the various business segments are orga-
nized basically along the same pattern as in the United States.
A corporate Banking Department in each country handled
the banking needs of both local and multinational industrial
corporations in that nation. Depending on the partiCUlar
country, the corporate banking unit might be broken down
further by industry and then by account manager.
Digitized by Google
.
INTERNATIONAL BANKING GROUP
1 I I I
Operations Personnel Planning, International
Administration Development and Monetary
and Credit Product Advisor
Management
I I I I I I I
DIVISION 1 DIVISION 2 DIVISION 3 DIVISION 4 DIVISION 5 DIVISION 6 Foreign World
o
cO"
Asia and Canada and South America Europe South Asia, International Exchange and Corporation
""
N"
(J) Pacific Caribbean Middle East, Banking International Division
Q.
and Africa Canter Money Market
cr
'< Division
c;
o
a-
( i)
EXHIBIT 1.1
International Banking Group. Note that in Divisions 1 to 5, each country
senof reports directly to a division head. (Source: company records.)
-----.....-..~.- ..:.. ----
Organizational Structures and Tensions 5
Digitized by Google
6 MulHnational Management
Internal Problems
Even when local versus global client preference was not an
issue, problems of internal communication existed between
the two groups in Citibank. If. a U.S. company wanted to
build a plant in Argentina, for example, the fmn's financial
officer might go to the bank's account manager in the CBG
for assurance of the necessary loan. To help his client, the
account manager would then make contact with the IBG.
"Among other grim realities," recounted one CBG manager,
"we didn't know who the bank officer in Argentina was.
We were a big bank there with hundreds of people, and so
you'd write to 'The Manager.' God knows whether he ever
got it." To avoid this, the CBG would present its case to the
IBG in New York and they would filter the request down to
the appropriate party in the field.
This process did not become cumbersome until the 1960s.
The personal network of relationships within the bank had
smoothed the flow; but with increases in the volume of busi·
ness and the number of people involved, the screening process
in New York broke down. In the mid·1960s de facto direct
Digitized by Google
~------------------~-~~~
Matrix Resolution
On August 6, 1973, William I. Spencer, President of First
National City Bank, sent a memorandum to all officers and
supervisors. The memo began:
Digitized by Google
8 MultiMtioMI Management
Questions
Executive Vice President Thomas C. Theobald assumed re-
sponsibility for the new World Corporation Group. January 1,
1974 was set as the date to officially transfer all accounts and
begin operating the WCG. During the coming months many
decisions would have to be made, including the following:
1. How would the business objectives and strategy of the
new group be articulated?
2. Which corporate customers would be transferred to the
WCG, what criteria would be used to make the selections,
and how many multinationals are headquartered in each
nation?
3. How should interactions between WCG and IBG be
structured and managed, with particular regard to (a)
financial, (b) human, and (c) physical resource alloca-
tions in all countries?
4. How many people would be needed, where would they
come from, and where would they be assigned?
5. What were the management information system require-
ments?
Digitized by Google
Organizational Structures and Tensions 9
Digitized by Google
Introduction
the 1970s multinational executives lived with the backlash from that view.
Although the public was awed by the vast resources of multinationals,
the executives knew that only governments can claim sovereignty. Attempts
to change government policy by bribery usually fail.
The problems of the 1970s revolved around the question of how a head-
quarters is organized to relate to its operations abroad. Such relationships
are affected primarily by the nature of the operations. There are technical
licensing agreements with autonomous partners, intimate cross-licensing
in which know-how is exchanged for equity (and the hope for more),
many kinds of joint ventures, and even management service contracts.
In many cases wholly owned subsidiaries are acquired but not fully
assimilated, and even those that the corporation itself has developed vary
in structure, for most grow rapidly and unevenly.
During the 1970s most corporate organizational structures were evolving.
The formal structures may have remained unchanged, but informal
practices evolved in three ways. The first organizational evolution was that
power relationships matured among headquarters, regional centers, and
national subsidiaries, especially in corporations containing merger frag-
ments. Because the most effective power may not be overt, the subsidiary-
headquarters interface may take many possible positions. Suppose, for
example, there has been a trend to centralization. The identical decision
may initially have been made locally, then made locally with a memo
to headquarters, made locally with merely the formality of a review at
headquarters, prepared locally for presentation at headquarters, made
at headquarters then sold to the subsidiary, or perhaps finally made at
headquarters with a memo to the subsidiary. During this evolution (or
an equivalent trend to decentralization), the formal organization chart
may have remained unchanged.
The second organizational evolution is that managers become more
familiar with how decisions should be made in their own corporation.
This is more than a question of familiarity with an office manual, for it
involves an executive's ability to categorize a decision in whichever of
the ways just mentioned is appropriate and to do so with confidence
and without anxiety. As an analogy consider a child playing in a yard;
a parent is nearby. If the yard is fenced, the child can have autonomy
inside that boundary; the parent need not monitor the child contin-
uously. Conversely, if the yard has no fence, the parent must oversee
constantly to assure that the child is safe. This second evolution refers to
establishing, agreeing upon, and testing mental fences.
The third organizational evolution is that managers around the world
now feel more legitimate in their roles. The subtle difference between a
federation and a confederation is that members of a confederation agree
on a goal. A m"ultinational corporation is a confederation with the econo-
mic goal of long-run profit. Managers pursue many other goals, but it is
easy to predict that the capital shortages forecasted for the 1980s will
Digitized by Google
12 Multinational Management
Digitized by Google
Organizational Structures and Tensions 13
EXHIBIT 1.2
Organization by product can lead to ethnocentrism.
Digitized by Google
14 Multinational Management
Digitized by Google
Organizational Structures and Tensions 15
Digitized by Google
16 Multinational Management
Digitized by Google
Organizational Structures and Tensions 17
Digitized by Google
18 Multinational Management
ization will be justified by the high cost of research and development and
thus the fixed cost of introducing another product. Few new products
will be introduced.
Chapter 11. International Executive Development. Planning guidelines
for managerial rotation will be explicit for Americans, who will be moved
among subsidiaries and then back to headquarters. The acute shortage of
local managerial talent will necessitate that nationals stay within their
subsidiaries, where their contribution is greatest.
Digitized by Google
Organizational Structures and Tensions 19
lCorporation Headquarters l
EXHIBIT 1.3
Organization by nation can lead to polycentrism.
Digitized by Google
20
buyer takes the advantage during negotiations, the price may be bid so
high or so low that lower total profits are achieved than if an enforcer
insisted on joint profit maximizing (Fouraker and Siegel, 1960).
Multinational manufacturing corporations emphasize the uniqueness of
their products, a uniqueness sustained by both patents and image building
in a milikf,t~ B¥~eeuse the buying may have no
and thu euhsidiary insuffjdf,nt ,:dtn«rmative outlets, like
monopoly by of a headquurtuee uuueufive.
in such difficult; it if h4Iee a
intrmgrator and enfueef'f~
decentralizatiun nn,cftjd a polycentlic the
corporation needs a headquarters with a staff of honest auditors, diligently
snuffing out personal cheating. It also needs a judiciary. Commercial trans-
actions always give rise to misunderstandings, and arms-length deals
between autonomous subsidiaries likewise occasionally go into dispute.
The headquarters executive hears evidence and renders verdicts in the
traditj,)ri t,ffrness and enterprising lenhuc head-
quarteec must inspiru have great ft~ills.
These discussed in ~~'FFft,U""
The the headquartecu b!~fjness
is to anticipate and m¥~binfftf
This calls for a deep and sympathetic understanding of each subsidiary,
joint venture, or licensee. The appropriate measure of whether head-
quarters achieves this is whether it can accurately anticipate how the su b-
sidiary will accept its guidance.
Exxon consists of 13 affiliated companies, most organized on geograph-
ical lffSO Middle Easf, Exxon USrt,
Each But bo± eUhan-
sions moves dfr,mtion of headqumteuf, which is
organief,h staff departmuntc, corporafn law,
and fulctinns. Jerome Ru~uw p. 156) of ertic-
ulated the delicate fabric of decentralization in a geographically decen-
tralized organization.
Digitized e
Organizational Structures and Tensions 21
[In the early 1960's Massey-Ferguson,) Ltd., a huge international farm equipment
manufacturer whose operations had been organized along regional lines, adopted
an organizational structure built around a series of largely self-contained market-
ing and manufacturing operations units ... [These are] centered on important
individual markets (the United Kingdom, France, Germany, the United States,
and Canada). Supplementing these units is an export marketing unit to cover sales
in parts of the world where [Massey-Ferguson] has no manufacturing operations.
Longer-range corporate strategy - determination of the basic world-wide product
line, decisions on major facilities, and changes in the logistic product flow from
production sources to markets - is set at corporate headquarters. But these
decisions are heavily influenced by operations unit judgments and recommenda-
tions. Each unit is responsible for determining the product lines best suited to its
local markets.
Digitized by Google
22 Multinational Management
prices will not be mandated, except for an equity rule that all purchasing
subsidiaries be charged the same f.o.b. price for the same item by a
particular manufacturing subsidiary.
Chapter 4. Multidivisional Expansion to a New Nation. The corporation
will rarely expand to a new nation except at the request of a licensee in
trouble or a floundering joint venture. Hence the product line that is
introduced will be the one that is requested, rather than the one best_
suited to the corporation's competitive position on a worldwide basis.
Chapter 5. Logistics. Details of shipping between subsidiaries are to be
arranged by those concerned.
Chapter 6. Production Smoothing. If factories in some nations are
excessively busy while others are not, regional officers in headquarters
may suggest transferring export orders to less busy subsidiaries. Unfor-
tunately, the lack of standard designs will limit this possibility.
Chapter 7. Multinational Plant Location. Marketing subsidiaries will
inform headquarters of every instance of government and customer
pressure for them to manufacture locally and will speak of factory and
market as one organic unit. Headquarters staff analysts, who might argue
for a rationalization of production, tend to find that data are unobtainable.
Chapter 8. New Product Launch. Each national subsidiary will introduce
new products when it wants to. There may be a casual sharing of informa-
tion such as "product XB155 is selling well," but there will be no com-
prehensive market research to clarify the underlying relevance of this
information to other subsidiaries.
Chapter 9. Pricing. No pricing scheme will be imposed on subsidiaries.
Hence national sales managers will constantly bicker about price; those in
high-priced nations will use incidents of product leakage to accuse managers
in low-priced nations of poaching customers.
Chapter 10. Product Design. Products will be designed to suit each
national subsidiary even though this means forgoing the mutual benefit
of experience curves. By tailoring design to its national requirements, the
subsidiary can have its products reclassified by customs inspectors to
lower rates of import duty, to meet local safety and engineering standards,
to use local grades and types of raw materials, and to cater to local cus-
tomer tastes.
Chapter 11. International Executive Development. Nationals will be
employed in each subsidiary to the greatest extent possible, and only a
few bilateral movements of managers will be planned. Nationals will be
sent to headquarters to learn technical skills needed in their homeland.
During their temporary absence, headquarters managers may fill their
positions, but this will be limited by the shortage of qualified head-
Digitized by Google
--------.. ~.-.
Digitized by Google
24 Multinational Management
Top
leadership
Matrix Matrix
manager manager
Two-boss
managers
EXHIBIT 1.4
Matrix roles (from Kramer, 1981).
Digitized by Google
Organizational Structures and Tensions 25
Digitized by Google
26 Multinational Management
In Dow, Coming's matrix management there is a Business Board for each of the
company's ten [product] businesses. The only full time board member is the man·
ager of the business. His position in the organization is at once critical and
tenuous. It is critical because his direct responsibility is the profit yield generated
from the business he is charged with managing. It is tenuous because on paper he
does not have direct control of the resources needed to accomplish his task. His
operative body, his total resource is composed of representatives from the market-
ing, technical service and development, research, manufacturing, and economic
evaluation/control functions [see Exhibit 1.5]. These are the Business Board memo
bers. They report directly to their functional group heads (vice president of
marketing and distribution, director of technical service and development, and
so on).
Organizationally, there is a strong dotted-line relationship running from the board
member to the board manager. More important than organizational lines, however,
is the clear understanding of where the profit responsibility lies. [Exhibit 1.5]
illustrates the structure and communications pattern of a board. Its manager
reports directly to Dow Coming's top management. His primary task -to generate
Research
/ NPN_tall~ ~
Economic
M..ketlng
representative ----........
1 evaluator
Controller
Business
manager
/ \
Manufacturing Technical service and
representative development representative
EXHIBIT 1.1>
The business board.
Digitized by Google
Organizational Structures and Tensions 27
profits-is accomplished through the total and combined support of his board
members.
Periodically, however, basic decisions must be made that inevitably involve hard
choices among competing alternatives. '!be manifest content of planning discus-
sions concerns projects .nd programs; the latent content involves power and posi·
tion .... Planning is not synonymous with forecasting, not by any stretch of the
imagination. Rather it is a dynamic process by which inside and outside interests
arrive at a new balance of power-reflected in a new structure and new policy-
designed to establish the parameters of executive decision making for some period
of years, but not forever. (Sayles and Chandler,1971, p. 42)
Digitized by Google
28 Multinational Management
Digitized by Google
Organizational Structures and Tensions 29
tion regarding escalation to higher levels. While it is important to define the issues
to which the power of nonconcurrence applies, it is likewise important not to
define them too precisely. The more each circumstance is defined, the greater the
temptation to build up staffs to address them -and hence the tendency to create
a bureaucracy. The virtue of this system is that decisions can be made quickly
without dispute or constant referral to higher line or staff authority.
Digitized by Google
30 Multinational Management
visual and conceptual tools as well as oral and written directives he forces
other managers to make commitments and decisions by affecting their
trade-offs (e.g., schedule versus cost) and priorities. In shaping the re-
sponse of his ''team'' to his centralized beat, he may also seek to change
the resonance characteristics of the system. He can change the character
of the system's response with a shift in organizational relationships. For
example, he can insist that a group now report to Division B instead of
Division A, or consult earlier, or more often, or must now give its consent
before x, y, and z occur.
Textbooks say that managers plan, direct, and control. But Sayles and
Chandler say that integrators bargain, cajole, intervene, coach, confront,
and give orders. These six activities use information - prodigious quantities
of both raw data and adroit studies. Yet the ideal is not to use these six
activities but to have the personal discipline to avoid meddling, letting
subordinates solve problems themselves.
Integrators are searching for impersonal means of achieving visibility for
troubled areas. If a problem can be acknowledged simultaneously by all
those people who are affected, its very visibility creates pressures for its
solution. Sayles and Chandler (1971, p. 115) report that "managers in
other countries have shown a similarly high degree of interest in the use of
visibility-type pressures to pry out vital information. These become
especially appealing to adopt in any society in which the normal desire
to conceal failure is reinforced by heavy cultural stress on the importance
of maintaining one's honor at all costs." Behrman (1970, p. 68) corro-
borates this view that control means thwarting the temptation to hide
problems.
Digitized by Google
Organizational Structures and Tensions 31
and decisions to wrestle market share from specific rivals for long-term
reasons.
Usually the product management teams are suituated in the head-
quarters nation. Sometimes they are situated in other nations. When a
subsidiary has a product management team within it, it is said to have a
global product mandate.
Some global product mandates reflect natural comparative advantage.
For example, Canada generates more hydroelectricity than any other
nation except the Soviet Union, so engineers of Canadian General
Electric have become expert in the design and service of waterwheel
turbines. General Electric salesmen anywhere in the world who receive
inquiries about waterwheel turbines work with the Canadian team that
has the global mandate in waterwheels. As another example, northern
Italy produces excellent and inexpensive appliances, so NV Philips of
Holland moved its appliance management to Italy, from where it now
controls a world network of plants.
Other global product mandates are actually part of world mandates that
arose from government technical specifications. For example, standards
of television transmission in Europe are different from those in North
America. NV Philips turned over to its North American subsidiary the
development, design, and manufacture of television receivers appropriate
for North America.
A final kind of global product mandate is assigned to a subsidiary for
strategic reasons. Black and Decker power hand tools use the same motor
components in drills, sanders, jigsaws, and other tools. Economies of scale
are substantial in manufacturing these products, but the Canadian market
is so small that manufacturing costs exceeded U.S. manufacturing costs
plus Canadian import duty. Instead of shutting down the Canadian plant,
the headquarters executives assigned to Canada the worldwide sander
production. The resulting volume cut unit costs not only of sanders but of
drills and jigsaws also, with the result that all became viable.
Digitized by Google
32 Multinational Management
taxes paid to the world minus interest earned on liquid assets. After
analysis, these options will be implemented only when they are preferable
to normal banking transactions in each nation.
Chapter 4. Multidivisional Expansion to a New Nation. Diligent effort
will be made to identify important stakeholders in each new nation and
gauge their predispositions to the corporation. Product divisions will be
introduced into the nation in a carefully orchestrated sequence. As the
stakeholders' tolerance of the corporation increases, further divisions
which may be more profitable to the corporation can be introduced.
Chapter 5. Logistics. The total product flow on some routes will justify
the corporation's chartering its own vessels. These charters will be
managed in such a way as to pressure rate reductions (or nonescalation)
from liner companies on other routes.
Chapter 6. Production Smoothing. Demand for each item fluctuates in
each market. One attraction of being multinational is the portfolio effect
of a reduced fluctuation of total demand. This means switching markets
from one factory to another and reswitching them as demands change.
Chapter 7. Multinational Plant Location. A multinational corporation
has tremendous bargaining power and flexibility until it builds aplant.
Once built, the plant may be a hostage. This suggests that there will be
innumerable studies about possible plant locations, which therefore
should be computerized.
Chapter 8. New Product Launch. A nation can be used as a test market
for the world. This suggests grouping the nations to launch in stage 1,
stage 2, and so forth. Nations grouped for the first stage must be selected
to yield valuable and timely information. This can be used to cancel or
advance the product's launch in other national markets.
Chapter 9. Pricing. Prices for the line of goods are determined in one
benchmark country (e.g., the United States). Simultaneously an optimal
markup factor is calculated for each nation by which the benchmark
prices are multiplied to obtain local prices. This scheme is simple enough
to permit central coordination, yet it is also responsive to overall market
conditions in each nation.
Chap ter 10. Produc t Design. Marketing considerations other than price
lead to a proliferation of tailored designs as described in the section on
polycentric organization. Production considerations tend toward a single
design to be marketed throughout the world, as described in the ethno-
centric section. The geocentric resolution is to modularize the design by
subassemblies (and styling panels). Some modules should be standard
around the world; others should be tailored.
Chapter 11. International Executive Development. Multilateral moves
Digitized by Google
Organizational Structures and Tensions 33
and the use of third country nationals must be analyzed. Because the
number of foreigners a government will allow to hold executive positions
in a subsidiary may be restricted, visas should be allocated with care. The
emphasis on integration implies nurturing personnel who can tolerate the
role ambiguity such a position implies.
Each chapter of this book describes a difficult integration and each out-
lines a computer model. The computer is an impersonal demander of infor-
mation and an impersonal means of detecting exaggerations. The models
are merely tools that can be used by powerful executives as part of their
personal political processes, and thus every prospective executive should
be aware of their existence and their possible uses.
Bibliography to Chapter 1
Aiken, Michael and Gerald Hage, "Organizational Interdependence and Intraorganiza-
tional Structure," American Sociological Review, Vol. 33 (1968), pp. 919-929.
Behnnan, Jack N., Nati01Ul1 Interests and the Multi1Ultional Enterprise: Tensions
Among the North Atlantic Countries (Englewood aiffs, NJ: Prentice-HaII, 1970).
Bergsten, C. Fred, Thomas Horst and Theodore H. Moran, American Multi1Ultionals
and American Interests (WashingtoR: The Brookings Institution, 1978).
Blake, David H. and Robert S. Waiters, The Politics of Global Economic Relations
(Englewood Cliffs, NJ: Prentice-Hall, 1976).
Boston Consulting Group, Perspectives on Experience (Boston: Boston Consulting
Group, 1970).
Bower, Joseph L., Managing the Resource Allocation Process: A Study of Corporate
Planning and Investment (Cambridge: Division of Research, Harvard Business
School,1970).
Brooke, Michael Z. and Mark van Beusekom,Intemati01Ul1 Corporate Planning (London:
Pitman, 1979).
Channon, Derek F. and Michael Jalland, Multinati01Ul1 Strategic Planning (London:
Macmillan, 1979) .
./ C1ee, Gilbert H. and Wilbur M. Sachtjen, "Organizing a Worldwide Business," Harvard
Business Review, Vol. 42, No. 6 (1964), pp. 55-67.
Curhan, Joan P., William H. Davidson, and Rajan Suri, Tracing the Multinatio1Ulls: A
Source book on U.S.-based Enterprises (Cambridge, MA: Ballinger,1977).
Davis, Stanley M. and Paul R. Lawrence, Matrix (Reading, MA: Addison-Wesley, 1977).
Dundas, Kenneth, The Management of U.S. Subsidiaries in Canada, doctoral disserta-
tion, University of Western Ontario, 1979.
Etzioni, Amitai, Modem Organizations (Englewood Cliffs, NJ: Prentice-Hall, 1964).
Fouraker, Lawrence E. and Sidney Siege), Bargaining and Group Decision-Making:
Experiments in Bilateral Monopoly (New York: McGraw-Hill, 1960).
Galbraith, Jay R., Organization Design (Reading, MA: Addison-Wesley, 1977).
Goggin, William C., "How the Multidimensional Structure Works at Dow Coming,"
Harvard Business Review, Vol. 52, No. 1 (1974), pp. 54-65.
Digitized by Google
34 Multinational Management
Digitized by Google
PART ONE
Exchange Management in a
Multinational Corporation
Dozier Industries
37
Digitized by Google
38 Multinational Financial Management
Dozier Industries
Richard Rothschild, the chief financial officer of Dozier
Industries, returned to his office after the completion of his
meeting with two officers of Southeastern National Bank. He
had requested the meeting in order to discuss financial issues
related to Dozier's first major international sales contract,
which had been confinned that morning, January 13, 1976.
Initially, Rothschild had contacted Robert Leigh, a vice presi-
dent at the bank who had primary responsibilities for Dozier's
business with Southeastern National. Leigh had in turn sug-
gested that John Gunn of the bank's international division be
included in the meeting since Leigh felt that he lacked the
international expertise to answer all the questions Rothschild
might raise.
The meeting had focused on the exchange risk related to
the new sales contract. Dozier's bid of £425,000 for the
installation of an internal security system for a large manu-
facturing finn in the United Kingdom had been accepted. In
accordance with the contract, the British finn had transferred
by cable £42,500 (10 percent of the contract amount) as
deposit on the contract with the balance due at the time the
system was completed. Dozier's production vice president,
Mike Miles, had assured Rothschild that there would be no
difficulty in completing the project within the 90-day period
stipulated in the bid. As a result, Rothschild was planning on
receiving £382,500 on April 13, 1976.
Digitized by Google
-~--------~-------~------------
EXHIBIT 2.1
Dozier, Inc.
Sales and Income Summary
Year Ended Sales Net Income
December 31 ($000) ($000)
1963 314 28
1964 397 34
1965 521 43
1966 918 86
1967 2,127 179
1968 3,858 406
1969 5,726 587
1970 7,143 702
1971 9,068 857
1972 8,646 309
1973 5,471 (108)
1974 5,986 (16)
1975 6,427 82
Digitized by Google
EXHIBIT 2.2
Dozier, Inc.
o{DecemSer 1975
ASSETS
Cti00~nt assets
and setiUiititiii m ,286
l,titiounts rp'~~r1L!ti"mti
Inventories 1,113,533
Total current assets $2,120,566
titid equipili0ilit
,906
Less accumulated depreciation 1,316,702
Net plant $2,898,204
assets
bilitistment
Total assets $5,243,770
UABILITms CND
Current liabilities
Accounts payable $ 467,291
Notes payable - bank 326,400
butal curr4?ut HtiHilities 77:%,691
b,mg-term liaHilities
Notes payable $ 275,000
CZ$mkrkOn equitb
bUilimon st,,,,,,
Rt4?EU4?rves
Retained earnings
Total Equity $4,175,079
liabilW0rrz equity
------
EXHIBIT 2.3
Dozier, Inc.
Bid Preparation
Materials $414,250
Direct labor 208,410
Shipping 35,000
Direct overhead* 104,205
Allocation of indirect overhead $ 50,246
Total cost $812,111
Profit factor 48,726
Total $860,836
Digitized by Google
42 Multinational Financial Management
EXHIBIT 2.4
Dozier, Inc.
Interest and Exchange Rate Comparisons
January 14, 1976
United United
States Kingdom
Digitized by Google
Exchange Management in a Multinational Corporation 43
EXHIBIT 2.&
Digitized by Google
44 Multinational Financial Management
EXHIBIT 2.6
Digitized by Google
Exchange Management in a Multinational Corporation 45
Digitized by Google
Introduction
Management by objectives and the evaluation of managerial performance
both require that someone distinguish between outcomes over which a
manager has control and those over which he has not. In evaluating the
performance of a manager in another nation, should the headquarters or
the subsidiary take responsibility for exchange rate changes?
During the 1970s an American would say that the West German deutsche
mark had appreciated, and that the British pound had devalued. Suppose
a German-headquartered multinational kept U.S. $1000 in a bank account.
Because it keeps its balance sheet in deutsche marks, the German corpora-
tion would record a foreign exchange loss on this asset. On the other
hand, a British headquartered multinational, keeping a similar U.s. $1000
in a bank account, would record a foreign exchange gain on this asset. The
German headquarters would take a dim view of the losses incurred by their
American manager. The British headquarters would express delight at the
strong performance of their American asset.
Evaluation is simple as long as the manager of the American subsidiary
deals only in U .S. dollars. In a geographically organized corporation,
managers of autonomous national subsidiaries might deal only in their
local currency; if they did, their evaluation would be simple. Usually,
however, multinational corporations undertake a worldwide flow of sub-
assemblies, finished products, technology, and intersubsidiary loans. This
means that while managers of subsidiaries are evaluated on the basis of
their performance, its measurement usually involves several foreign
currencies.
Let us see which exchange rate U.S. corporations use to evaluate their
foreign subsidiaries. Pure trading cannot be planned and should be trans-
acted only if today's spot exchange rate is right. Five percent of U.S.
corporations both budget and evaluate performance at the spot rate
(Business International Money Report 1977); presumably these are the
trading corporations.
In 20 percent of U.S. corporations, the budget for foreign operations is
approved at the then current spot rate. At the end of the year, performance
is evaluated at the end of year spot exchange rate, which places the entire
foreign exchange burden on the subsidiary. This type of subsidiary-
management evaluation process is common in ethnocentric corporations.
Polycentric multinationals budget commitments using forecast exchange
rates, then evaluate the performance of subsidiary managers at year-end
spot rates. Business International reports that 50 percent of U.S. corpora-
tions operating abroad evaluate in this way. Although this evaluation pro-
cess still renders the subsidiary manager susceptible to exchange rate risk,
the foreign exchange burden he experiences is less than that of the subsidi-
ary manager of an ethnocentric corporation.
Fifteen percent of U.S. corporations budget commitments at a forecast
46
Digitized by Google
Exchange Management in a Multinational Corporation 47
rate and also evaluate subsidiary managers at this same forecast rate (even
if the forecast turns out to have been incorrect). Geocentric multinationals,
which tend to employ this scheme, view the subsidiary more as a cost
center than a fully autonomous profit center. The activities of the entire
subsidiary network are rationalized, and the subsidiary manager's evalua-
tion is based on pis contribution to the corporation as an aggregate system.
As such, the subsidiary manager is shielded from the effects of unforeseen
foreign exchange rate fluctuations. Nevertheless, the subsidiary manager
should be motivated to adjust his marketing and production decisions to
take advantage of whatever exchange rate occurs.
A forward contract occurs when two individuals agree that on a specified
future date they will exchange a stated amount of one currency for a
stated amount of another. Usually one of the individuals is a bank. A bank
active in forward markets keeps track of its exposure in each currency for
each day of maturity. By the rates it quotes, it attempts to offset its pur-
chases with its sales of each currency so as to control this exposure. As
we saw in the Dozier Industries case, it makes little difference whether the
corporation borrows abroad in order to repatriate now, or whether it
enters the forward market. Either method allows it to avoid the risk of
exchange rate fluctuations. In the commercial market one can obtain
forward rates for I, 2, 3,6, and 12 months for many currencies and even
longer contracts for a few currencies. As the Dozier case suggested, if a
corporation has an exposed asset abroad, it can today borrow in that
nation, in order to repatriate the proceeds immediately. This creates an
exposed liability to offset the exposed asset. In this way the corporation
can hedge for much longer than one year (Neave and Rutenberg,1981).
Evaluation of manager at year
end
Today's
spot ETHNOCENTRIC
rate
Exchange
rate used
for budget Today's
forecasting forecast
of the POLYCENTRIC GEOCENTRIC
year end
spot rate
EXHIBIT 2.7
The subsidiary's budget is set at the beginning of the year; performance
is evaluated at year end.
Digitized by Google
48 Multinational Financial Management
As marketing and production managers have to make long-term commit-
ments in several currencies, it seems sensible to insulate their marketing
or production decisions from the risk of exchange rate fluctuations.
Four psychological theories encourage such separation. First, expectancy
theory ( a model of perception) states that if effort leads to outcome, then
the effort will occur if the outcome is desirable. Conversely, a manager
will exert less effort if he perceives little link between his effort in marketing
or production and the outcome that was swamped by an exchange rate
change. Second, operant conditioning theory predicts that good efforts
that are not rewarded will be extinguished; to extrapolate, therefore, a
swing in exchange rates could reward and reinforce wrong production or
marketing decisions. Third, a general performance model states that
blocking path goals leads to frustration and stress, made manifest in pas-
sive withdrawal, active hostility, and depression. The corporation, facing
an unfavorable exchange rate shift, does not want to cause this behavior.
Finally, dissonance theory would be relevant if a conscientious manager
undeservedly gained from a favorable change in exchange rates. If the
manager had a low tolerance for this guilt, he might subconsciously sabo-
tage his own marketing or production efforts the following year to "even
the score."
For internal budget purposes, polycentric and geocentric corporations
need a forecast of exchange rates. They obtain the short-term rates of up
to one year from the data provided by the commercial market, but they
also have the data with which to forecast long-run exchange rates for some
currencies. The simplest conceptual model for forecasting is that of covered
interest arbitrage. Just as Dozier Industries could borrow three-month
Europounds, it could also have borrowed for a longer period, say five
years.
Let us examine a five-year .horizon. An investor has two options. The
first is to invest in a five-year V.S. security at the five-year rate (ru,r)' At
the end of five years, one dollar of investment would become
" (1 + ',)5
.------------------,I
I
I I
Today's I I Future spot
spot rateI Irate
I I
I I
I ,~
start 'us (1 + 'us)5 end
EXHIBIT 2.8
Covered interest arbitrage states that the two routes should be equally
profitable.
Digitized by Google
Exchange Management in a Multinational Corporation 49
Many money markets are efficient and this forecast is adequate. How-
ever, for some currencies other factors must be accounted for by the head-
quarters treasurer when forecasting rates. Most governments have, at some
time in the past, imposed restrictions on the flows of capital out of their
nation. A "blocked" currency is one with tight restrictions. Such restric-
tions allow the internal rates of interest to be lower than they otherwise
would be in a world eqUilibrium. Foreign investors fear that such restric-
tions might be imposed, thus trapping their capital. To compensate for
this perceived risk, the rf must be higher than in the absence of fear,
making the forecast future rate a bound.
The headquarters should forecast not just for a five-year maturity but
for all future years in all currencies. Today's spot rates and the bond yield
curves (Malkiel, 1970) for each national currency are all they need. Un-
fortunately, some countries are too unstable to have a bond market and
an analyst must use his or her best judgment when estimating rates.
This introduction has dealt with the service that the divisions should
expect from the headquarters treasury group: forecasting rates as well as
budgeting and evaluating based on these rates. Marketing and production
managers can get on with their work, with exchange rates taken into ac-
count. We will now study models that underlie the rate-setting process in
polycentric, ethnocentric, and geocentric corporations.
Digitized by Google
50 Multinationol Financial Management
mutual funds are internationally diversified. American investors achieve
their international diversification by buying shares in U.S.-based multi-
national corporations. Errunza and Senbek (1981) calculate the positive
relationship between international involvement and stock market premium.
Chen, Ricks, and Shawky (1977) found that the actual returns on the
stocks of multinational corporations are affected by a world market in-
dex and an exchange risk factor. Investors are compensated for bearing
both volatility and exchange risks. This implies that an investor forming a
portfolio consisting primarily of the stocks of multinationals will gain
from international diversification.
Fundamental to a polycentric view is the concept that many smart
people are trying to profit in exchange rates. If a corporation has no more
inside information than anyone else, it will sometimes gain, sometimes
lose, and on average break even. No abnormal systematic gain can be ex-
pected if the prices move as a random walk. Giddy and Dufey (1975)
showed that a random walk model adequately fits the spot movement of
the currencies of Canada, France, and the United Kingdom for 1972-1974
data. Using a longer data base, Levich (1979) showed that for the cur-
rencies of Switzerland and Italy a random walk model is a good descrip-
tion of the data; for Belgium, France, and Germany the random walk
model is a poor fit; and for Canada, the United Kingdom, the Nether-
lands, and Japan a random walk model does not fit, presumably because
of erratic government intervention. Nevertheless, in none of these cases
is the nonrandom pattern distinct enough or long term enough that
trying to benefit from it will cover the transaction costs.
The polycentric headquarters views each subsidiary as a capital invest-
ment, the value of which is determined by the discounted value of its
future repatriated revenue flows. The corporation lacks the organization
to reallocate its assets once invested, in response to the changing states of
the world economies. Only intermittent executive action will force re-
deployments. Although the headquarters decision to invest may have been
based on a risk-retUrn perception that was appropriate then, each national
subsidiary of a geographically organized corporation pushes its own
interest in new investment. Actually larger subsidiaries sometimes use
their greater political presence within the corporation to block detailed
studies that compare the financial risk of one nation with another.
The headquarters treasurer of a polycentric corporation tends to view
fluctuating exchange rates as random events occurring with shifts in the
world's economic and political structure. Such interest rate movements
can rarely be forecasted or systematically reacted to by the corporation.
Hence the polycentric attitude is to avoid wasting money on hedging
fees.
Nevertheless, in two circumstances the headquarters treasurer may
have the incentive to hedge. First, the treasurer could attain a higher
return after taxes if on occasion he hedged to renew a tax credit. Second,
Digitized by Google
Exchange Management in a Multinational Corporation 51
he may decide to bet against a government that has chosen to peg its
exchange rates.
If a corporation has paid more taxes abroad than it would pay at home
on its global income, the V.S. government not only receives zero tax but
actually grants the corporation a tax credit. Corporate tax credits in the
United States may be carried forward five years and back two years. A
corporation with unused five-year-old tax credits can either let them ex-
pire or find a way to "renew" them. Hedging provides a legitimate way.
Suppose a U.S. corporation has a five-year-old foreign tax credit of
$480,000. Liechtenstein imposes no tax on income derived as a result of
foreign exchange fluctuations, so the V.S. corporation can have its Liech-
tenstein subsidiary hedge Italian lire by borrowing lire or by a forward
contract, so that the corporation will gain $1 million if the lira rises and
lose $1 million if the lira falls (numbers chosen to keep the arithmetic
simple).
Suppose there is 50 percent probability that the lira will rise and that
the corporation will gain $1 million. In that case, the corporation would
have had to pay an additional $480,000 V.S. tax (marginal tax rate of 48
percent). However, by applying its $480,000 excess foreign tax credit,
it can retain the entire $1 million. There is a 50 percent probability that
the lira will fall. In that case the corporation will lose $1 million, and the
unused old tax credit will expire at year end. However, the loss can be
carried forward five years. Though the old tax credit expired, a new one
lives onwards - until used to offset a profit.
The second occasion on which a headquarters treasurer may hedge is
when he can foresee that a pegged exchange rate might be changed. Not
all exchange rates fluctuate in a random fashion. Some governments
intervene in the exchange market to support their currency. This is termed
a "dirty float." The alert treasurer may decide to hedge when systematic
government intervention occurs, even though he can act only within
limits set by exchange restrictions of the foreign government.
"In many countries, even in those with convertible currencies, local
companies are only allowed to enter the forward market to cover specific
receivables and payables" (Chemical Bank, 1972, p. 14). For example,
V.S. dollars can be converted into French francs and vice versa -they are
both convertible. Anyone can easily convert V.S. dollars into Israeli
sheckels, but it requires the permission of the Israeli central bank to con-
vert sheckels to dollars in Israel. Conversion of sheckels in another coun-
try is viewed by Israel as a violation of Israeli law. The sheckel is not
convertible. Freedom of entry into the forward market or the bond mar-
ket for hedging purposes is rare when an exchange rate is pegged. Although
the world's currencies are commonly believed to be operating in a general
floating rate system, only a handful of them (admittedly among the most
important) float individually against all others. Most of the others are tied
to one of the key currencies or to a composite unit. The summary break-
Digitized by Google
62 Multinational Financial Management
down below from the International Monetary Fund Survey, February 20,
1978, illustrates the order within the apparently bewildering variety of the
world's currency map. Each year a few governments change the grouping
of their currency but the pattern remains.
Currencies Floating Independently. British pound, Canadian dollar,
Italian lira, Japanese yen, Philippine peso, Mexican peso, Spanish peseta,
Turkish lira, U.S. dollar.
Currencies Pegged on the U.S. Dollar. Bahamas dollar, Barbados dollar,
Bolivian peso, Botswanian pula, Burundian franc, Republic of China's
new Taiwan dollar, Costa Rican colon, Dominican peso, Ecuadoran sucre,
Egyptian pound, El Salvadorian colon, Ethiopian birr, Ghanian cedi,
Grenadan East Caribbean dollar, Guatemalan quetzal, Guyanan dollar,
Haitian gourde, Honduran lempira, Indonesian rupiah, Iraqi dinar, Jamaican
dollar, Korean won, kip of the Laos Peoples' Democratic Republic, Liberian
dollar, Libyan dinar, Maldivian rupee, Nepalese rupee, Nicaraguan cordoba,
Oman rial, Pakistini rupee, Panamanian balboa, Paraguayan guarani,
Romanian leu, Rwandan franc, Somalian shilling, South African rand,
Sudanese pound, Syrian Arabic Republic's pound, Thai baht, Trinidad
and Tobago dollar, United Emirates dirham, Venezuelan bolivar, Yemen
Arab Republic rial, Peoples' Democratic Republic of Yemen dinar.
Currencies Pegged on the British Pound. Bangladesh taka, Gambian dalas,
Irish pound, Seychelles rupee, the leone of Sierra Leone.
Currencies Pegged on the French Franc. The francs of Benin, Cameroon,
Central African Empire, Chad, Comoro, Peoples' Republic of the Congo,
Gabon, Ivory Coast, Madagascar, l\tlali, Niger, Senegal, Togo, Upper
Volta.
Currencies Pegged on the International Monetary Fund's Special Drawing
Rights (SDR) (historically the U.S. dollar but now a weighted average
currency): Burmese kyat, Guinean syli, Iranian rial, Jordanian dinar,
Kenyan shilling, Malawi kwacha, Mauritius rupee, Qatar riyal, the dobra
of Siio Tome and Pnncipe, Saudi Arabian riyal, Tanzanian shilling, Ugan-
dan shilling, South Vietnamese dong, the zaire of Zaire, the Zambian
kwacha.
Currencies in the European "Snake." Belgian franc, Danish krone, Luxem-
bourg franc, the Netherlands guilder, Norwegian krone, West German
deutsche mark. Each currency can snake between maximum margins
of ±2.25 percent with respect to a weighted basket of the other currencies.
Currency Composite Other Than SDR. Algerian dinar, Australian dollar,
Austrian shilling, Cyprus pound, Fiji dollar, Finnish markka, Indian rupee,
Kuwait dinar, Malaysian ringgit, Maltese pound, Mauritanian ouguiya,
Digitized by Google
Exchange Management in a Multinational Corporation 53
Digitized by Google
54 Multinational Financial Management
7r 0 = probability of no devaluation during period 0
1- 7r 0 = probability of a devaluation during period 0
dj = one of several estimates for the extent of devaluation [For
Dozier, d 1 , d 2 , d 3 , may be 10, 12, and 15 percent.]
Pj = the probability of devaluation to level dj, conditional on there
being a devaluation, such that Pj =1 (Possibly 75,20, and 5 per-
cent respectively, for Dozier.)
pjdj = the expected extent of devaluation assuming a devaluation will
occur (10.65% for Dozier.)
Pj dj Pj d j
.75 .10 .075
.20 .12 .024
.05 .15 .0075
.1065
Do =(l- 7r o)pj d j
Since Do > Co' Dozier should hedge period 0 now. Should period 1 be
hedged now? At the beginning of period 0 a hedge of any future period 1,
... , T can be looked at as a speculation in an option contract. That is,
quite apart from our interest in hedging period 1, we are also interested in
Digitized by Google
Exchange Management in a Multinational Corporation 55
Digitized by Google
66 Multinatio1llJ1 Financial Ma1llllIement
projects). However, he has great incentive to adapt his marketing and manu-
facturing decisions as the exchange rate changes.
This scheme is appropriate for subsidiaries whose exchange rate floats.
In such circumstances, the discrepancy between forecasted and actual ex-
change rates is likely to be small, and the subsidiary manager is pressured
to adapt rapidly. In some corporations, such as Alcan Aluminium Ltd.,
each subsidiary keeps its accounts in local currency and is allowed to hedge
if the subsidiary treasurer feels excessive risk. In theory, one subsidiary
could be buying and another selling futures in the same currency. In
practice subsidiary treasurers communicate, so they tend to agree as to
which currencies are weak. Note that even if each subsidiary hedged to
neutral in its local currency, commitments in the local currency would not
be hedged, so the corporation as a whole would likely have exposure.
In summary, the headquarters treasurer of a geographically organized
polycentric corporation will rarely hedge. Instead, he will passively absorb
the risks that the subsidiary marketing and production manager contracted.
Only in special circumstances, such as when old tax credits are renewed or
when, for political reasons, governments intervene in the currency market,
may the treasurer become involved with future contracts.
Digitized by Google
EXHIBIT 2.9
ASSETS
Cash on hand and demand and time deposits x
Marketable equity securities
Carried at cost x
Carried at current market price x
Accounts and notes receivable related
unearned discount x
Allowance for doubtful accounts and
notes receivable x
Inventories
Carried at cost x
Carried at current replacement price
or current selling price x
Carried at net realizable value X
Carried at contract price (produced
under fIXed price contracts) X
Prepaid insurance, advertising, and rent X
Refundable deposits X
Advances to unconsolidated subsidiaries X
Property, plant, and equipment X
Accumulated depreciation of property,
plant and equipment X
Cash surrender value of life insurance X
Patents, trademarks, licenses, and formulas X
GoodwUl X
Other intangible assets X
LIABILITIES
Accounts and notes payable and overdrafts X
Accrued expenses payable X
Accrued losses on firm purchase commitments X
Refundable deposits X
Deferred income X
Bonds payable or other long·term debt X
Unamortized premium or discount on bonds
or notes payable X
Convertible bonds payable X
Accrued pension obligations X
Obligations under warranties X
This table illustrates the exchanle rates to be used for translatinl balance sheet accounts under
the F ASB'. statement on translation accountlnl.
57
Digitized by Google
58 Multinational Financial Management
inflation, the net assets will be restated on a price level adjusted basis.
When the functional currency figures are translated into U .S. dollars at
current exchange rates any apparent gain or loss in the exposure in this
latest year will no longer be reported as part of the parent company's
income for this year as required for ethnocentric extensions, but rather
will be treated as a direct adjustment of the parent company's share-
holders equity.
Few treasurers wait until the year's end to determine corporate ex-
posure; they use pro forma balance sheets to forecast exposure, and they
hedge if it appears excessive. A treasurer who is anxious about foreign
exchange, as many in ethnocentric firms are, will do his best to bias the
exchange rate forecast he quotes to divisions. The treasurer adds a spread
between the buying and selling rates quoted for corporate use. He calls
the spread a "risk premium," and treasurers are known to deviously adopt
arguments of modem finance to justify the spread that induces operating
divisions to reduce their exposure. Such treasurers never think that some
foreign exposure might be desirable. (Beedles and Senchak, 1978; Errunza,
1977).
Transfer prices, budgets, and exchange rate information are biased by
both headquarters and the subsidiaries, since neither wishes to be burdened
with the risk of exchange rate fluctuations. The corporate atmosphere of
ethnocentric firms is not conducive to the development of a cooperative
and comprehensive exchange rate management policy.
Why do headquarters treasurers continue to hedge so ethnocentrically?
Their reasons are both psychological and bureaucratic. One treasurer told
me that his job is "like standing naked on a windy corner. Were you
covered? If not, why not?" After we had discussed the accuracy of vari-
ous exchange forecasts, another treasurer confided "I use Citibank so that
when we take a loss we can blame them." One rarely meets an interna-
tional treasurer in an ethnocentric corporation who feels that either his
international or his treasury superior understands his work.
Ethnocentric corporate treasurers pay substantial hedging fees to undo
diversification out of the "safety" of the home currency. The accumula-
tion of hedging fees can be surprisingly large. For example, even though
the British government formerly supported sterling futures, an internal
Ford Motor Company study showed that, through the years 1946 to 1969,
a policy of always hedging was appreciably more expensive than one of
never hedging but taking losses when devaluation occurred. The spread
between a buying and selling rate appears very low on a single 30-day
futures quotation, but the accumulation of these monthly fees from 1946
to 1969 was surprisingly large (Kohlhagan, 1975).
Worse than the expense of hedging is the fact that the very definition of
exposure is arbitrary. Exhibit 2.10 shows some of the different bases for
defining exposure. As can be seen, the same foreign subsidiary could be
perceived quite differently if its parent were American, Swedish, or Japan-
Digitized by Google
Exchange Management in a Multinational Corporation 59
EXHIBIT 2.10
ese. Even as the U.S. Financial Accounting Standards Board changes its
definition of exposure it will have to change it to another standard for
stewardship reasons. Consistent stewardship demands that there be a basis
for valuing all subsidiaries. However. managerial action to take advantage
of the rate change may not be congruent with the basis for valuation.
Attitudes toward exposure provide a clear test to distinguish between
an ethnocentric and a polycentric (or geocentric) treasurer. Suppose a
corporation has hedged a pegged currency, and there is a possibility of de·
valuation, although its probability is far from certain. Should the treasurer
sell that contract (unhedge) if someone offers more than its expected
worth? This frequently happens during the early stages of a run on a cur-
Digitized by Google
60 Multinational Financial Management
Digitized by Google
--- --
-
-~.
Options
Consider the problem faced by the corporation's marine division that
wants one new vessel and is negotiating to have it constructed in a Japanese,
or a Polish, or a Singapore shipyard. Negotiations are expected to last six
months before the division signs a purchase order with the most suitable
of these yards. If it is signed, there will be progress payments (at specified
dates) and then a final payment, when the vessel passes ocean tests, pay-
able in yen, zloty, or rupees. To chose from among these three shipyards,
the marine division must know what its U.S. dollar commitments will be
and hence should pay the treasurer at headquarters for an option on the
specified number of yen and again for an option on zloty and again on
rupiahs. The marine division can then negotiate with the three shipyards.
When there is an inadequate forward market, the options can flow to
the treasurer's department. One U.S. lumber company was negotiating to
purchase hardwood in Indonesia at a price specified in Indonesian rupiahs
or U.S. dollars, with the U.S. company being able to decide which to pay.
The hardwood buyers turned to their treasurer's department for advice on
how hard to negotiate for this clause. The most credible advice is for the
treasurer to quote a bonus to be paid the buyer to get such an option
(Merton, 1973).
Real Assets
The risky assets and liabilities of the multinational corporation are not
necessarily those that appear on the balance sheet, but are instead the
future streams of different currencies. Accurate exchange rate forecasts
are not possible, but the covariances between long-term exchange rate
changes can be forecast (Shapiro and Rutenberg, 1976; Goodman, 1979).
The marketing plans yield a forecasted stream of revenue. Each estab-
lished factory will require a stream of local currency for wages and domes-
tic purchases; there are interest payments, scheduled debt repayments,
and dividends. These flows must be forecasted on a few scenarios of ex-
change rates. A simulation of the cash flows of the corporation can then
be prepared. Scenarios of possible futures (including exchange rates and
restrictions) have to be written, so that each set of decision alternatives
can be evaluated for robustness.
An example should help make this clear. Consider a multinational cor-
poration, with an established market in Japan, about to develop an ore
body in western Canada. (The mineral will: be smelted in Canada, then
shipped to Japan.) To finance the mine and smelter should the treasurer
prefer to float bonds in Canada or Japan? Some treasurers would take a
Digitized by Google
62 Multinational Financial Mantlllement
balance sheet approach and argue that the assets in Canada should be off-
set by Canadian debt. Other treasurers would take a cash flow approach
and borrow in Japan. In their eyes the mine will require a continual flow
of Canadian dollars to pay for wages, maintenance, and local taxes, yet
the corporation has no sources of Canadian dollars. On the other hand, it
will have a continual stream of yen from the sales revenue, part of which
can be used for bond interest and principal repayment. A geocentric
executive would adopt this view and prefer to borrow yen. Specifically,
from a portfolio model a geocentric transfer should be able to calculate
the premium the corporation would be willing to pay to borrow yen.
Despite this premium he might nevertheless borrow Canadian dollars if
the government provided access to subsidized interest rates, or if the
treasurer could foresee a possibility of nationalization in which Canadian
bondholders would act as a local pressure group. Nevertheless, a geocentric
treasurer's basic consideration is future cash flows.
With estimates of the total cash flows, the treasurer can determine the
criteria for composing the corporate portfolio. There are usually hundreds
of different capital investments among which to select. The risk return
balance and the relationships between each investment and the existing
global portfolio must be examined to determine the asset mix whose risk
return level dominates all other combinations. At f"llSt glance it may ap-
pear that there is a bewildering array of possibilities. Most of the work of
a corporate planning group is to screen out the implausible alternatives;
only a few dozen decisions of financial importance are likely to emerge.
For example, many nations stipulate that the proceeds of a local bond is-
sue may be used only to build a factory in that nation. As another ex-
ample, Swiss bankers appear to stipulate (collectively) the maximum debt
they will float for each particular corporation. Even if a corporation
tries to envision global alternatives, it confronts so many constraints on
its viable financial options that their total is readily manageable. One such
constraint is the minimum expected return the corporation expects before
expanding into another market, on an efficient frontier between risk and
return. For a prespecified expected return, the combination of activities
which minimizes risk can be calculated by means of a quadratic program,
rather like a linear program which has a more complicated objective func-
tion to handle the covariance terms (Lietaer, 1971).
The geocentric treasurer utilizes exchange rate information received
from subsidiary managers as well as observation of imperfections in the
currency market - such as those caused by systematic government inter-
ference - to hedge selectively. Employing the decision rules established by
the hedging model mentioned in the section on the polycentric view, the
geocentric treasurer and managers occasionally participate in the futures
market. The treasurer may use information from hedging practices as in-
put into the portfolio of positive and negative cash flows in each currency.
A standard currency is required for operating a multinational accounting
system. If erratic government intervention into a national currency makes
Digitized by Google
Exchange Management in a Multinational Corporation 63
it unsuitable as a base currency, then this also means that the currency is
an unsuitable standard for multinational accounting, as appears to be the
case with the Israeli scheckel. This does not necessarily mean that the
physical headquarters has to be moved, but another currency may be
more suitable. For example, the Canadian-headquartered multinationals
Alean, Massey-Ferguson, and Moore now keep their accounts in U.S.
dollars.
In summary, the geocentric multinational will strive to maintain a bal-
anced portfolio. The concept of the geographically diversified portfolio
of the polycentric corporation is extended in the geocentric corporation
to allow for the redeployment of the corporation's marketing and produc-
tion facilities. The value of a risky asset to the individual subsidiary is
adjusted to match the treasurer's interpretation of the effects of economic
shifts and government policies on future income flows. The treasurer
makes this adjustment by biasing the subsidiary's forecasted exchange
rate. The portfolio's composition will change as the states of nature change.
Such a dynamic management practice as that pursued by the geocentric
corporation is necessary in the sometimes hostile environment in which
the corporation operates. Many nations where subsidiaries are located are
antagonistic toward capitalist corporations, particularly multinationals.
Because of the geocentric's rationalized production and marketing system,
it is particularly susceptible: A breakdown in one nation caused by
government policy or economic uncertainty may affect the entire network.
Simultaneously, the geocentric's ability to redeploy assets means that a
higher expected return and a lower variance of cash flow can be achieved
than that which is possible in a polycentric corporation.
The polycentric corporation fits on the mean-variance frontier. The
precise location may depend on assuming a fair amount of risk. Not 80
the ethnocentric corporation, whose costly financial constraints will also
E(r)
frontier
Standard deviation of
the profit stream
EXHIBIT 2.11
Risk-return space of the three corporate archetypes.
Digitized by Google
64 Multinational Financial Management
reduce its profitability below the frontier. The geocentric's ability to re-
deploy assets to adapt to the state of the world may (if it is managed
efficiently) allow it to be above the polycentric frontier of Exhibit 2.11.
Conclusion
The impact of exchange rate fluctuations on the multinational is twofold:
1. Short-term exchange rate fluctuations may be incorporated by head-
quarters into the year-end evaluation of a subsidiary manager's profit
performance. The subsidiary's earnings may be enhanced, or devalued,
according to random currency movements, the rate at which the sub-
sidiary's budget was forecasted, and the rate at which the subsidiary
is evaluated.
2. Many marketing and production decisions involve long-term commit-
ments by the organization, which causes the headquarter's treasury
office to be starkly aware of the effect devaluation may have on the
future income flows of those investments. Therefore, some multina-
tionals use exchange rate forecasts or introduce risk biases into their
decisions to commit resources.
If the morale of the subsidiary manager and the productivity of the
workers are to be sustained, operating managers must be shielded from the
effects of currency fluctuations. The degree to which the headquarters
treasurer insulates the subsidiary manager from the effects of exchange
rate fluctuations in evaluation procedures as well as the degree to which
he offers incentives for the manager to react to current conditions in a
manner that is most beneficial to the organization's total performance
vary with the organizational archetype of the corporation. The exchange
rate forecast and portfolio adjustment techniques employed by the trea-
surer and the cooperation that he receives from managers will greatly in-
fluence the long-term growth and profitability of the multinational.
The assets of the ethnocentric multinational are naively diversified;
capital investments are based on strictly adhered to return on investment
criteria. There is no mechanism by which to perceive assets as a coherent
whole. The ethnocentric treasurer stodgily operates according to world-
wide procedural guidelines. Subsidiary managers are evaluated in a manner
that exposes them to the variable movements of currencies. On the home
front, treasurers deplete corporate funds by hedging all foreign activities
in their attempt to avoid all risk of currency devaluation. Given that
hedging is the purchase of futures to protect against currency devalua-
tions, it is not an activity designed to maximize profits but rather a cost
incurred to avert a possible loss. Ethnocentric executives conduct their
foreign operations in a guarded and pessimistic manner. Income streams
Digitized by Google
-
-~-------
Digitized by Google
66 Multinational Financial Management
Digitized by Google
Exchange Management in a Multinational Corporation 67
Digitized by Google
Bibliography to Chapter 2
Adler, Michael and Bernard Dumas, "The Long Tenn Financial Decisions of the
Multinational Corpontion," in InterntJtio1JlJI Capital MtlTkets, edited by E.J. Elton
and M.J. Gruber (Amsterdam: North Holland, 1975), Chapter 9.
Beedles, William L. and Andrew J. Senchak, "Indirect International Diversification
Through U.S. Multinational Firms," Unpublished Working Paper 78-27, The
University of Texas, Austin, 1978.
Black, Fischer, "The Pricing of Commodity Contncts," JounuU of Fi1JlJnciGl Econo-
mics, Vol. 3 (1976), pp. 167-179.
Chemical Bank, Foreign Exc#uJnge ExpofU1'8 MtJ1UIIlf!ment, Multinational Cub Manage-
ment Group, Chemical Bank, New York,1972.
Chen, Andrew H., David A. Ricks, and Hany Shawky, "The Performance of Multi-
national Corpontions' Common Stocks and the Efficiency of International Capital
Markets," Unpublished Working Paper 77-74, CoUege of Administntive Science,
Ohio State University, 1977.
Errunza, Vibang R., "Gains from Portfolio Diversification into Less Developed Coun-
tries' Securities," JourntJl of InterntJtiolJlJl Bruinea Studies, Vol. 8, No. 2 (1977),
pp. 83-99.
Errunza, Vihang R. and Lemma W. Senbet, ''The Effects of International Opentions
on the Market Value of the Finn: Theory and Evidence," Journal of FiMnce,
Vol. 36, No. 2 (1981).
Folks, William R. and Stanley R. StaDaeU, "The Use of Discriminant Analysis in Fore-
casting Exchange Rate Movements," JounuU of InterntJtio1JlJI Bruinea Studies,
Vol. 6, No. 1 (1975), pp. 33-50.
Garstka, Stanley J. and David P. Rutenberg, "Computation in Discrete Stochastic
Programs with Recourse," Operatio".RuellTCh, Vol. 21, No. 1 (1973),pp.112-122.
Giddy, Ian H. and Gunter Dufey, "The Random Behavior of the Flexible Exchange
Rates: Implications for Forecasting," Journal of InterntJtio1JlJI Bruinea Studies,
Vol. 6, No. 1 (1975), pp. 1-32.
Goeltz, Richard K., "Managing Liquid Funds on an International Scope," Unpublished
Working Paper, Joseph E. Seapam and Sons, Inc., New York, 1971.
Goodman, Stephen H., "Foreip Exchange Rate Forecasting Techniques: Implications
for Business and Policy," JounuU of Finance, Vol. 34, No. 2, 1979, pp. 410-427.
Kohlhagan, Steven W., "Evidence on the Cost of Forward ~ver in a Floating System,"
Euromoney, September 1975, pp. 138-141.
Levich, Richard M., "Efficient Markets and International Financial Management,"
The InterntJti01JlJI Money Market: An Asseament of Forecasting Techniques and
Market Efficiency, Chap. 6 (J. A. I. Press, 1979).
Levy, Haim and Marshall Sarnat, "Devaluation Risk and the Portfolio Analysis of
International Investment," in InterntJtiolJlJl Capital Markets, edited by E.J. Elton
and M.J. Gruber (Amsterdam: North Holland, 1975), Chapter 5.2.
Lietaer, Bernard, Fi1JlJncial Manogement of Foreign Exc#uJnge: An OperatiolJlJl Tech-
nique to Reduce Risk (Cambrldle: MIT Press, 1971).
Uoyd, W.P., "International Portfolio Diversification of Real Assets: An Inquiry,"
JourntJl of Bruiness Research, April 1975. Vol. 3, pp. 111-120.
Malkiel, Burton G., The Term Structure of Interest Rates: Theory, Empirical Eui-
dence, and Appliclltio". (Morristown, NJ: General Learning Press, 1970).
68
Digitized by Google
Exchange Management in a Multinational Corporation 69
Merton, Robert C., "The Theory of Rational Option Pricing," Bell JouT1Ull of Econo-
mies and Manogement Scilmce, Vol. 4 (Spring 1973), pp. 141-183.
Mueller, Gerhardt G., International Accounting (New York: Macmlllan,1967).
Neave, Edwin and David P. Rutenberg, "When to Hedge Successively," Unpublished
Working Paper, Queen's University, 1981.
Roll, Richard and Bruno Solnik, "A Pure Foreign Exchange Asset Pricing Model,"
Journal of International Economies, Vol. 7 (1977), pp. 161-179.
Senchack, Andrew J. and Laura T. Starks, "International Diversification Through
Listed Foreign Securities," Unpublished Working Paper 78-28, The University of
Texas, Austin, 1978.
Shapiro, Alan H. and David P. Rutenberg, "Managing Exchange Risks in a Floating
World," FinancitJI Management, Vol. 5, No. 2 (1976), pp. 48-58.
Shapiro, Alan H. and David P. Rutenberg, "When to Hedge Against Devaluation,"
Management Science, Vol. 20, No. 12 (1974), pp. 1514-1630; reprinted in Inter-
national Capital Markets, edited by E.J. Elton and M.J. Gruber (Amsterdam,
North Holland, 1975), Chapter 6.1.
Digitized by Google
APTER
Paisley S.A.
Paisley S.A.
Marc Chemenceau, assistant to Vice President of Finance
Jean Lefevres of the Paisley company, was told to analyze
an aspect of the company's financial policies. He was a recent
MBA graduate of INSEAD the distinguished Graduate
School of Business at Fontainebleau, where he had specialized
in finance and international business. His vice president felt
that Paisley's financial resources were no longer being utilized
as effectively as they might be.
Paisley S.A. * was a French company that manufactured
and marketed commercial fragrances, dyes, and chemicals
worldwide. During 1981 foreign subsidiaries had substantially
increased their accounts payable to the parent company; in
1982 they would want additional large increases. These in-
creased demands on the parent company's financial resources
coincided with an increase in its own needs for working
capital and a drop in 1981 profits of the parent company.
These internal pressures promised that 1982 would be a
difficult year. Because of the high inflation rate at the time,
the newly elected Socialist government had limited the
amount of credit available to the economy; short- and long-
term interest rates were increased, and foreign payments
were rationed.
Paisley et Cie. had been established in 1935 by Mr. A. Begin
to commercialize synthetic perfumes he had patented.
Established perfume houses wouldn't touch his ersatz per-
fumes, so he manufactured dyes and then revived his syn-
thetic perfumes as commercial air fresheners. An air freshener
order from the Paris Metro finally launched the company. In
1960 his son, Robert Begin, assumed the presidency. He
ThiII cue was prepared by Lutz Holtield under the IUpervUion of Profeuor David
Rutenbera.
·S.A. (Societe Anonyme) means a stock corporation slmllar to an incorporated
U.S. corporation.
Digitized by Google
72 Multinational Financial Management
Digitized by Google
------
.---~---.. -----~
The Organization
The operating holding company, or parent company, had
three product divisions: fragrance, dyes, and miscellaneous
chemicals. Each of these divisions was then organized on a
functional basis.
The French subsidiaries reported directly to the top man-
agement of Paisley but also communicated widely with the
corresponding product divisions of the operating holding
company or with the R&D division. Their financial auton-
omy varied; some had an independent borrowing policy and
others depended on the parent holding company.
All the Paisley foreign subsidiaries were owned by Genmarc
holding company, incorporated in France. Direct exports
were channeled through this company. In organizational
terms, the product divisions of Paisley (or exporting French
subsidiaries) had direct contact with the foreign subsidiaries.
To have better access to certain markets Paisley had created
several wholly owned sales subsidiaries. Some, like the
subsidiary in Brazil, had set up limited manufacturing or
packaging operations because of tariff problems, patent
necessity, or governmental pressures. Genmarc had originally
been responsible for these subsidiaries. The changing nature
of the foreign operation from arm's-length export to limited
manufacturing had made it necessary to establish additional
links with the product divisions that sent goods to the foreign
subsidiaries. Thus Paisley product divisions usually bypassed
Digitized by Google
74 Multinational Financi41 Mantl6ement
Genmarc in their communications with the individual man-
agers of the foreign subsidiaries. Genmarc received dividends
and royalties from the subsidiaries and also could collect a
markup on export orders.
Financial Policies
The large integrated foreign subsidiaries functioned as if
financially independent of the parent company. Their high
profitability and relatively low dividend to Paisley allowed
them to finance their expansion from retained earnings.
When external financing was necessary they dealt with local
banks with which they had long developed relationships.
This was especially true of the U.K. subsidiary, owned by
Paisley since 1960.
The large subsidiaries reported to the parent company, but
their reports were treated casually. Their forecasts were
accepted as long as profitability and sales were growing.
Clearly, no one at Genmarc or Paisley was particularly com-
petent in international finance and accounting. No one sys-
tematically considered the impact on the subsidiaries of
different financial policies and accounting principles, of local
environmental variables, or of currency fluctuations.
The smaller subsidiaries were much more dependent on
Paisley. They financed their growth by increasing their
Genmarc accounts receivable. When additional financing was
required they asked Paisley for longer credit tenns or for a
fonnal loan. Usually Genmarc and the Paisley product divi-
sions approved their requests.
Conflicts had arisen occasionally about devaluation losses.
Subsidiaries with large accounts payable to Paisley denom-
inated in French francs were severely hit when a devaluation
in the local currency suddenly increased the local currency
value of these accounts. As a result, some subsidiaries had
started to borrow in local currency while others had the
burden of devaluation split on a 50:50 basis with Paisley.
By the end of 1981 both Vice President Lefevres and Marc
Chemenceau felt increasingly uneasy about the financial
needs for the coming year. The foreign subsidiaries' increase
in accounts receivable was imposing a heavy financial strain
on Paisley. The working capital requirements had also in-
creased because of larger inventories as well as an increase
in domestic accounts receivable. This was expected to con-
tinue well into 1982. Furthennore, price increases for
Digitized by Google
- - - - -- - -------"----:--"
Digitized by Google
""I
a.. EXHIBIT 3.1
Paisley S.ill"
mmmmmmnmmmmmmmmmmmmmm MMMMMMlllllUlmlmmmmmmmmll mmmmmmmmmmmmmmmmmmmmm mmmmmmmmmmmnnMmnmmmmll
FinanCii'::ll Il::il,ill~I::I, ,I'?,)reign subsidial"'iil!l:,!i ;1' ,9,!';I;I' (,11;1,1,;1 thousands)
1!,:I,I,llrchase Royalties
Extent 111,1" Finished Paid to
Country of Dividends Products finished Paisley
of Owner- Profit To From or row for
Incorporation ship (%) Sales (Loss) Paisley Paisley S.A. Materiah R&D
II!hl.pOrts
l'llations
(i)
120,240
EXHIBIT 3.2
Paisley S.A.
Balance Sheets of Selected Subsidiaries (FF Millions)
1982
1981 (foreclUlt)
UNITED KINGDOM
Assets
Fixed assets 27,810 29,520
Accumulated depreciation 8,640 9,090
Net fixed assets 19,170 20,430
Inventory 5,940 6,120
Accounts receivable
Other Paisley subsidiaries 5,040 4,230
Others 6,570 6,750
Cash 1,890 1,080
Total assets 38,610 38,610
Liabilities
Common stock 6,300 6,300
Retained earnings 18,990 21,060
Long-term debt 6,750 6,390
Short-term debt 4,410 2,700
Accounts payable 2,160 2,160
Total liabilities 38,610 38,610
BRAZIL
Assets
Fixed assets 159 218
Accumulated depreciation 17 22
Net fixed assets 142 196
Inventory 304 493
Accounts receivable 855 1,174
Cash 29 43
Total assets 1,330 1,906
Liabilities
Common stock 72 72
Retained earnings 261 ll6
Long-term debt
Paisley 145 247
Other 72 0
Short-term debt 99 130
Accounts payable
Paisley 638 1,268
Other 43 73
Total liabilities 1,330 1,906
77
Digitized by Google
'78 MultiratJtionai Fi1umcilll MtJMgement
EXHIBIT 3.3
Paisley S.A. Subsidiaries
Sources and Uses of Funds
Anticipated for the year 1982, (FF thousands)
UNITED KINGDOM
Sources of funds
Accumulated depreciation 450
Accounts receivable from others 810
Casha 810
Retained earnings 2,070
Total sources 4,140
Uses of funds
Fixed assets 1,710
Inventory 180
Accounts receivable subsidiaries 180
Long-tenn debt 360
Short-tenn debt 1,710
Total uses of funds 4,140
BRAZIL
Sources of funds
Accumulated depreciation 5
Long-tenn debt from Paisley 102
Short-tenn borrowings 31
Accounts payable to Paisley 630
Accounts payable to others 30
Total sources of funds 798
Uses of funds
Fixed assets 59
Inventory 189
Accounts receivable 319
Cash 14
Retained earnings 145
Repay long-term debt to others 72
Total uses of funds 798
GThe U.K. treumer has sta~d in another conted that the minlmum cub _ t he
could opera~ with Is FF 600 milUon • .o a further FF 880 milUon could be available.
Digitized by Google
---=.---.~-----------------
EXHIBIT 3.4
Digitized by Google
EXHIBIT 3.5
EXHIBIT 3.6
France Brazil 10 16
United Kingdom 10 6
Nontreaty 25 25
Brazil France 10 16
United Kingdom 25 26
Nontreaty 26 26
United Kingdom Brazil 34 0
France 10 0
Nontreaty 34 0
80
Digitized by Google
Ethnocentric - Introduction
This chapter is concerned with moving funds between subsidiaries. The
purpose of such moves is to get liquidity to each subsidiary when it needs
it, in a way that minimizes total taxes paid by the corporation.
Corporations differ in the guidelines they impose on subsidiaries for
handling liquid assets, but three patterns predominate, each corresponding
to one of the archetypes of headquarters-subsidiary relationships discussed
in Chapter 1. Some headquarters feel uneasy about subsidiary managers
holding liquid assets from year to year and insist on their declaring divi-
dends to headquarters. This assures control of the foreign managers by the
ethnocentric headquarters, although at a high cost in taxes. Geographi-
cally organized corporations tend to view each subsidiary as totally in-
dependent (which it legally is) and have little involvement in the manage-
ment of liquid assets; this policy has the fewest negative behavioral
consequences, although it results in a high level of total short-term assets.
Between these two extremes, the geocentric corporation moves liquidity
to subsidiaries when needed by manipulating transfer prices, managerial
fees and royalties, dividends, and intersubsidiary loans in order to maxi-
mize interest received on liquid assets minus total taxes paid on a world-
wide basis. As Robbins and Stobaugh (1973) noted:
Make no mistake about it-even though these large enterprises seemingly are de·
centralized, their worldwide network of units and intercompany links are in fact
controlled indirectly by their rule books, or "bibles" from headquarters. The "bi·
bles" are in the form of standard procedures, sometimes consisting of several vol·
umes, that specify such items as the limits of local borrowing, standard terms of
payment of intercompany accounts, and standard rules for management fees.
81
Digitized by Google
Geocentric Model- A Generalized Network
The subject of this chapter is the optimal use of tax havens, bilateral tax
treaties, nonuniform treatments of income received from abroad, and
national differences in income tax rates, import duties, and border taxes.
This is a partial analysis for tactical planning because it takes as given the
planned operations of each national subsidiary, and hence whether the
subsidiary is to be a net source or recipient of funds in each year.
Suppose there are N subsidiaries around the world, and the analysis
will be carried out over T years. A real problem deals with all N sub-
sidiaries; for simplicity, we deal with only three subsidiaries: i, i, and
k. The analysis begins by forecasting the net cash surplus/deficit for each
subsidiary for each period bf, and by setting the safe levels below which
bank deposits Kf·t + 1 and other liquid asset holdings Gf·t + 1 must not
fall. From there, a linear program is developed as a guide to the optimal
maneuvering of liquid assets between the subsidiaries. The linear program
can be solved on a computer.
For each mlijor product line or industry in a nation, a planner can plot
the industry attractiveness (growth rate is the usual proxy) against the
cash generation capability (related to market share) of the product line.
Planners have generally adopted the Boston Consulting Group names for
the four quadrants: cash cows, stars, dogs, and seedlings. In a growing
industry an enormous (but estimable) amount of cash is needed to gain
market share (to transform a seedling into a star). In a low-growth in-
dustry, an enormous cash flow can be generated for a few years by curtail-
ing research, exploiting workers, and raising prices (which transforms a
cash cow into a dog). Exhibit 3.7 is a tool used in many corporations to
conceptualize cash flow over several years. This specific example is taken
from the 1976 annual report of the Black and Decker Manufacturing
Company.
Even within the strategic deployment of each product line in each
nation, hard work is required to forecast the cash inflow requirements
or cash outflow. Inflation of input prices must be forecast. Revenue must
be estimated from pricing plans and market expansion intentions. The
most probable rates of taxation must be forecast to yield after-tax income
plus depreciation.
It is far more revealing to forecast pro forma cash flows and interest
income in nominal, rather than what economists call ''real,'' terms.
Deflating these values with anticipated inflation obscures the possible
conduits that can channel required liquidity to where it is needed at the
most advantageous exchange rates. The forecast exchange rates from
Chapter 2 are used in each year of the program. (Some nations employ
different exchange rates for different kinds of transactions, and during
the 1980s more nations will probably impose currency controls.)
82
Digitized by Google
ManeulJering Liquid Asset. 83
Unit Relative
Business Company growth market
unit buslne.. rate share
(%) (%)
1. European Power Tools 38 14-15 3.2-4.0
2. U.S. Consumer Tools 20 7-8 2.0
3. McCulloch 14 10-11 0.6-0.7
4. U.S Professional Tools 10 5-61.25-1.75
5. Pacific International 10 26 0.3-0.4
6. Canada 6 10 4
7. Australia 2 4.5 5
5
20%
o
Stars Seedlings
~
I.,.
.::
c
~
-; 10
i!
•
2
Cash
cows
Dogs
Digitized by Google
84 Multinational Financial Marurgement
b' + 1 b' + 2
k k
EXHIBIT 3.8
Inflows and outflows for subsidiaries i, i, and k.
b:= capital generated across all product lines-planned expenditures
art + 1 = amount of liquid assets carried by subsidiary i in the one year
period from t to t + 1, employed in the most profitable manner
K:· t + 1 = amount of cash held by subsidiary i (usually in a dollar account
in a U.S. bank) that can satisfy compensating balance requirements
tage of the higher interest rates obtainable on longer term loans. However,
to simplify exposition, we use only one-year arcs.
A feasible solution to the problem of maneuvering liquid assets is one
that gets funds to those subsidiaries that need them. Exhibit 3.9 pre-
sents the problem of maneuvering liquid assets in a multinational com-
pany as a network problem. Our objective is to minimize taxes paid to
the world minus interest received. The problem sounds very complicated.
Fortunately, all of these complexities can be handled within the problem
itself. Along the way from 0 to T we have to meet demands for liquidity.
In year T there will be some excess liquidity (assuming that the problem
is feasible). The more excess liquidity, the better our accomplishment
Digitized by Google
Maneuvering Liquid Assets 85
(less was lost to taxes, and more has been gained from interest). Hence
the objective function of this linear program is to maximize the sum
of liquid assets remaining after time T. These liquid assets are the three
arrows flowing off the right-hand side of Exhibit 3.8.
For any given pattern of intersubsidiary ownerships, there are at least
four ways to transmit liquid assets from one national subsidiary to another.
These control variables are:
1. Pj; Adjust transfer prices on shipments between subsidiaries i and j.
2. f;; Charge legal and managerial fees and royalties on technical knowl-
edge supplied to subsidiary j by subsidiary i.
3. lfl + 1 Make new intersubsidiary loans and repay old ones with interest.
Delayed invoicing is a loan.
4. d j; Remit dividends from subsidiary j to its legal stockholder sub-
sidiary i.
A linear program has constraints. Each node of the network represents
a separate constraint. For node i in year t the flows of funds in minus
d' + 1
1I
b' + 1
1
G'" + 1 /' + 1
J 1I
/' + 1
ki
b' + 1
k
PERIOD t PERIOD t + 1
EXHIBIT 3.9
The complete network.
Digitized by Google
86 Multinational Financial Management
the flows out must equal bf. In each year, N compensating balances must
be specified, as must N -1 dividend flows, N(N -1) transfer price flows,
and N(N -1) loan flows. The number of managerial fee flows will be either
N(N -1) or N -1, depending on whether fees and royalties have to follow
the ownership pattern. Thus, at most, T(aN'l - N - 1) flows will have to
be calculated. There are direct and indirect costs associated with each
transfer, which decrease the liquidity of the corporation (even government
irritation must be paid for with legal fees and ''political contributions").
These are referred to as "factors of attenuation." The model aids to
minimize these costs. The problem is a generalized network (also called
weighted distribution, or machine loading problem). Dantzig (1963,
Chapter 21) presents an excellent discussion of the generalized network
problem. Extremely fast computer algorithms have been developed by
Balachandran and Thompson (1975).
Because of the power relationships within a corporation, it is prudent
to begin to implement control over liquid assets by modeling variables
over which headquarters already has control (dividends and fees). Once a
dividend and fee model has been proven in use, headquarters can proceed
to model intersubsidiary loans and adjusted transfer prices and to imple-
ment the models in a prudent and consistent manner. We now consider
each of the four transfer methods in detail. Although the power, control,
and political implications of adjusted transfer prices are onerous, the data
analysis is the most simple, hence we begin with this method. Conversely,
to implement changes in the dividend policy is quite easy, but to analyze
it is difficult; hence we will analyze dividends last.
Digitized by Google
ManeulJering Liquid Assets 87
Exchange
rate
CD ~
i7Export
I~m.
tax Import
Income
tax
credit
tariff tariff
(usually none)
EXHIBIT 3.10
The factor of attenuation on transfer price manipulation A(pjtj ).
Digitized by Google
88 Multinational Financial Manogement
EXHIBIT 3.11
Maneuvering lI88ets by means of adjusting transfer prices. (A = total cost
of adjusting transfer prices on shipments from j to i; B = cash flows through
the arcs; C = marginal cost of adjusting transfer prices.)
income increases in j (zero tax) but decreases in i (75 percent tax). This
results in lower taxes paid to the world. Therefore the cost of flowing
funds through the Pij arc A(Pfj) is decreased by increasing the transfer
price when the marginal cost is negative.
If each subsidiary is a semiautonomous profit center, there is a possibility
that cooperation among subsidiaries could give way to competition. A
bilateral monopoly relationship is possible between i and j. When one
manufacturing plant (monopolist) is the only source for one marketing
outlet (monopsonist), the dynamics of transfer pricing from the manu-
facturer to the marketer frequently escalate so high that the marketer
ends up charging his customers a price higher than that which a joint
monopolist would charge. In other words, the consumers suffer, and the
manufacturer and marketer divide a profit pie that is unnecessarily small.
An imposed transfer price will work as long as producing-to-demand
remains in the self-interest of subsidiary i. There are two internal indirect
costs to flow through a Pii arc.
One indirect cost lies ID the profit forgone by subsidiary i as it tailors
its promotional efforts to its cost of goods, which is partly the manipulated
transfer price. (There is a temptation to do this in a polycentric corpora-
tion when each subsidiary is an independent entity and when it is eval-
uated on its return on investment.) Clearly the costs of acrimony over
negotiated and imposed transfer prices are greatly reduced if unit sales and
production quotas are set centrally.
The other external indirect costs arise from government inquiries. If the
price is too low, both governments will intervene. The tax authorities of
country j will see tax revenues forgone (in the United States, I.R.S.,
Section 482; see the 1967 Court of Claims case against Eli Lilly and
Company). The import tariff commission of country i will see dumping. If
the transfer price is too high, income tax will be forgone by country i.
These external indirect costs have to be paid in the time and effort at
investigatory hearings, additional advertising to counter ill-will, and actual
settlements.
Digitized by Google
------:::= - _. --.----.--.--.-------------.. -
Several governments have tried to define a fair transfer price. The U.S.
Treasury Department regulations specify that an arm's-length price be
used. The Treasury Department lists four methods to be used in sequence:
1. Comparable Uncontrolled Price Method. Monitor the prices of sales
between unrelated corporations and use those. Differences in the
quantity sold, quality, terms, use of trademarks or brand names, time
of sale, level of the market, and geography of the market may be
grounds for claiming that the sale is not comparable. Gradation adjust·
ments can easily be made for freight and insurance but cannot accu-
rately be made for trademarks.
2. Resale Price Method. The transfer price equals the resale price minus a
standard markup. There is considerable leeway in determining a
standard markup.
3. Cost Plus Method. No definition of full cost is given, nor is there a
unique formula for prorating shared costs over joint products. The
markup over cost allows room for maneuvering.
4. Another Appropriate Method. The Treasury regulations explicitly
state that while a new market is being established it is legitimate to
charge a lower transfer price: "It is specifically provided in the regula·
tions that goods may be sold, for a period, at a price which is below
the full cost of manufacture in order to establish or maintain a market."
The North American concern with transfer prices, in the corporate
income tax context, implies a concern that foreign companies bear their
"fair share" of the tax. In Europe, revenue is raised by a value added tax,
similar to a North American sales tax but applied on all intermediate
transactions, not just on sales to the final consumer. Mathewson and
Quirin (1978) note that the tax philosophies of the European Economic
Community (EEC) countries differ from those of North America in an
important respect:
The basic thrust of Europe in tax legislation, at least in recent years, has been to
define a tax base which is coextensive with the domestic population. This is most
clearly seen in the application of the value added tax (V AT) in which export sales
are exempt while VAT paid on all domestic purchases may be claimed as a credit
against tax payable on domestic content of export sales from VAT. While the
Europeans are willing to have foreigners contribute to domestic tax revenues via
corporate income taxes on exports, one suspects that faced with a choice between
the tax revenues and the exports, they would expect to find that the constraints
imposed by E.E.C. members on transfer pricing involving exports from the E.E.C.
to be of minimal significance.
Digitized by Google
Multinational
t
Algeria
RRRIBIT 3.12
ChRining invoices RfLEk€t flows.
summation
of fees
EXHIBIT 3.18
Summation of fees sequenced by their transfer cost.
Digitized by Google
92 Multinational Financial MarllJlIement
The costs indicated in Exhibit 3.13 will be incorporated into the linear
programming generalized network model. Once the network problem has
been run, we can study the period-to-period variations in the fee flows. We
can then adjust the fees and find the allocation methods that produce
the flows that best fit those required by the model. We then rerun the
program to fit the defmition.
Intersubsidiary LoansA(lf· J+ 1)
The most prevalent form ofintersubsidiary loan is a speed-up or delay in in-
voicing shipments between a pair of subsidiaries. This may be done
because normal terms of payment differ between the two nations, because
of tax savings in some cases, or to bypass restrictions on remittances. In
addition, actual shipping schedules can be accelerated in one direction and
decelerated in the other so that one subsidiary carries inventory for both.
Most amounts due from affIliates on open account are generally re-
quired to bear interest after six months, unless zero interest can be shown
to be a regular trade practice. Even within such bounds, the leading and
lagging of the invoices on imports and exports can create a substantial
loan. Where desirable, the company can institute a program of netting,
whereby all intercorporate payables and receivables flow only to the net
creditors of the corporate group. By netting, a company can:
1. Reduce its exchange purchases.
2. Provide a more flexible means of controlling its overall liquidity
picture.
3. Allow for a more timely analysis of the company's foreign exchange
exposure.
Exhibit 3.14 prepared by Business International provides a consolidation
of the limits granted by major countries.
Unless there are specific laws to the contrary, one subsidiary has the
right to grant a formal loan to a subsidiary in another nation and to be
repaid with interest. Years ago, some corporations abused this right and
had subsidiaries make zero-interest loans with no specified maturity. Such
loans look so similar to dividends (but are not taxed) that tax authorities
are now on the alert for "deemed dividends." Hence a loan must have a
reasonable interest rate and a reasonable repayment date. Where there is
leeway as to reasonableness of interest rate, the treasurer can select either
the minimum or the maximum to minimize total taxes (we discuss taxes
later in this section).
In determining the actual amount of funds a subsidiary will need to
borrow each year, it is not sufficient to look only at the net cash deficiency
of that year alone.
Digitized by Google
EXHIBIT 3.14
Australia 180 days Not allowed 180 days Not allowed Permission
required
Belgium 180 days 90 days 180 days 90 days Permission
required but
readily
available
Brazil Not allowed Allowed· 180 days Not Not permitted
no limit allowed G
Canada Allowed· Allowed· Allowed· Allowed· Permitted
no limit no limit no limit no limit
France 180 daysb Allowed· 90 daysC Not Permission
no limit allowed d required but
difficult
Germany Allowed· Allowed· Allowed· Allowed· Permitted
no limit no limit no limit no limit
Italy 90 days 360 days 360 days 30 days Not permittede
Japan 180 days' 180 days' 120 days 120 days Not permitted6
Mexico Allowed· Allowed· Allowed· Allowed· Permitted
no limit no limit no limit no limit
Netherlands Allowed· Allowed· Allowed· Allowed Permitted
no limit no limit no limit no limit
Spain 90 days 180 days 180 daysh Not Permission
allowed' required but
difficult
South 180 days Allowed· Allowed· Not allowed Not permitted
Africa no limit no limit except with
special per·
mission
Sweden 180 days Allowed· 180 days Not Permission
no limit allowed} required but
readily
available
United 180 days Allowed· Allowed· Not allowed Permitted
Kingdom no limit no limit
United Allowed· Allowed· Allowed· Allowed· Permitted
States no limit no limit no limit no limit
-Based on information from local bankinl sources.
tJExcept with special guthorization of exchange department of central bank for a maximum of 25%
of value of imports; but usually depends on ~rms of agreement; cexcept for raw materials, which
have a maximum of 180-360 days' lal time; except for 30% down payment on imported capital
goods; eltalian exchange office of Foreign Trade Ministry can authorize exceptions. Offsetting
debits and credits can be effected by banks only when two different companies are involved,
except for oil companles; 'but exporters can legally make use of 360 days for both leads and lags
J:f exports less than $500,000; 6except for so-called "invisible:' items, such as trade-related expenses;
includes 9o-day grace period and lonler lap are permitted; 'ex<:ept with very special permission or
when 25% down payment is required for placement of order;Jexcept if over Skr50,OOO if prepay·
ment is in accordance with normal commercial practice.
93
Digitized by Google
94 Multinational Financial MantJlfement
An analogy will help clarify this statement. Suppose your parents loan
you money to attend college and charge 10 percent interest. On gradua-
tion, principal plus interest totals $10,000. Intending to study for an
MBA, you estimate that your need for additional liquidity will be $5,000
the first year and $6,000 the second year. Finally, suppose that your
parents insist on one-year loans but grant new loans immediately. How
big will each loan be?
O. Now, at year 0, you need to repay the $10,000 debt and obtain
$5,000 more. So borrow $15,000.
1. At the end of year 1 you will need to repay the capital of $15,000,
interest on it of $1,500, and obtain new liquidity of $6,000. So
borrow $22,500.
2. On graduation at the end of year 2 you will need to repay the capital
of $22,500 and its interest of $2,250. Assuming that you have a job
after graduation you will not need new liquidity, but because you have
not yet accumulated funds to payoff your debt, you need to borrow
$24,750.
In this example your needs for liquidity flows of $5,000 and $6,000 had
to be reconstituted into loans of $15,000, $22,500, and $24,750. After
the maneuvering network is solved, a separate calculation will reconstitute
loans from the liquidity flows calculated in the maneuvering network.
Unfortunately the interest on intersubsidiary loans gives rise to taxes,
which reduce liquidity.
Suppose in year 1 subsidiary j is due to receive interest income from
subsidiary i. Nation i may impose a withholding tax on interest paid
abroad. Nation j may impose a withholding tax on interest received from
abroad. Both withholdings may be reduced by a tax treaty between nations
i and j. If subsidiary j pays taxes (it won't if it experiences a loss or
consumes unused tax credits), its revenue authorities mayor may not tax
interest income from abroad. Usually tax will be charged, but only partial
credit (counting the time value of cash) will be allowed for the withholding
taxes.
After the loans have been reconstituted from liquidity flows, and the
tax implications detailed, we will have to deduct these taxes from liquidity
and recompUte the optimum. If there are several ways to reconstitute
loans, select the one that minimizes global taxes. The maneuvering liquid
assets problem is intended to be recalculated every quarter, or at least
every year. The patterns of intersubsidiary loans (which determine tax
rates) appear relatively stable, so repeated iterations are rarely needed.
Thus far we have talked about intersubsidiary loans as though the ex-
change rate were forecast to remain unchanged. Suppose A is to receive a
loan and pay TA interest if denominated in A's currency, or TB if in B's
currency. When the loan is made, the exchange rate is one unit of currency
A = e~ currency B. From Chapter 1 we forecast that when the loan is
Digitized by Google
-_ _--
........... c:-_-=-_-::":
~"""'_=_ ___:::-:--_--:-_"--'_ _ _ __ _ __ _
repaid the exchange rate will be one unit of currency A = e~r 1 currency
B, with currency A expected to devalue so that etAB > e)J . If the loan
is denominated in the borrowing currency A, then B will show a foreign
exchange loss of e~B - e~+B 1 per dollar loaned to A, and an interest income
of e~; 1 r A rather than e~B r A • Subsidiary B will pay fewer taxes, and
saving will be (e~B - etA; 1) (1 + r A) (tax h+ 1). If the loan is denominated
in currency B, then A will have to pay back the appreciation of both
capital and interest. Subsidiary A's tax savings will be (e iB - elB+ 1)
(1 + rB) (tax i + 1). Thus the choice of denominating the intersubsidiary
loan in currency A or B depends on the tax rates and bounds on the allow-
able rate of interest. The choice does not depend on the extent of devalua-
tion (though it would be reversed if a revaluation were forecast).
Dividends A (d jj )
If there are N wholly owned subsidiaries, the intersubsidiary ownership
pattern is fixed, and N - 1 dividend arcs are known. The problem is to
estimate the cost function A(djl ) for each arc. The cost function (if non-
linear) and the restrictions on dividend flow depend on the tax base. It is
presumed that aspects of the tax base internal to each country have been
optimized so that in this formulation we need consider only intersubsid-
iary allocations.
First, the tax laws on dividends for three nations are discussed, starting
with the uncomplicated tax laws of Liechtenstein. Though they will not
be examined here, local taxes imposed by states, counties, and cantons are
substantial enough that they should be included in the cost functions. Tax
treaties between pairs of nations should also be studied, especially when a
company operates with a foreign corporate charter, a branch, or a more
unusual legal entity.
When planning a base company (a tax haven or financial clearing house)
in the popular tax haven countries of Surinam, Panama, Liechtenstein,
Switzerland, Cayman Islands, and Barbados, usually the most applicable
laws are those of the corporate headquarters, the V.S. Revenue Acts of
1962 and 1976 for a V.S. corporation. The combination of these two
Revenue Acts is so complicated that several chartered accountancy firms
have prepared computer simulations of the tax position of their client
corporations. The complications center around the "tainted income." We
examine one example, Subpart F income.
Digitized by Google
96 Multinational Financial Management
~
Liechtenstein
. Gennany
i Belgium
EXHIBIT 3.15
Maneuvering assets through dividend payments.
Digitized by Google
- - - - - = : . = . -_. ~ -:: ..- .-=-
...".. ..--=--:~--~ .. - - -
Digitized by Google
98 Multinational Financial MantJllement
48
43.2
35~~----------------------+---------~35
Foreign tax
o 63 100
Dividend payout rate
EXHIBIT 3.16
Total taxes as a function of dividend payout rate.
other U.S. holdings. Other nations have similar institutions. Such legal
entities of every nation must be analyzed, for they can reduce taxes paid
to the world. A corporation without such entities should consider ere-
atingthem.
Local Loans
In each nation all possible sources of funds should be studied, including
the assets that would have to be pledged as security, the amount that
could be borrowed, the interest rate (augmented by the opportunity cost
of restrictive covenents that bankers impose). Fixed assets such as factories
and inventories in a nation are valuable, primarily to local bankers in that
nation. Movable assets such as ships can be collateral in almost any nation,
so the solution procedure is to guess where they will best be mortgaged
and run the problem.
A local loan can be formulated very neatly within the network (Crum,
1974). In one year the bank makes money available. A year later that
same amount of money plus interest will be siphoned out of the subsid-
iary's liquidity, as seen in Exhibit 3.17.
Foreign banks commonly grant loans on an overdraft basis, so that the
borrowing company pays interest only on the amount of the granted loan
it has in use. In contrast, a U.S. bank grants a line of credit for which it
charges a commitment fee. The corporation can borrow up to the line of
credit; the borrowing company pays interest on the entire amount of the
granted loan (whether or not drawn down from the bank, though often a
lower rate is paid on the amount not withdrawn) and furthermore must
keep specified compensating balances on deposit in the bank. As the com-
pany has to keep transaction and precautionary cash, the compensating
balance requirement is not as onerous as it would be in a deterministic
world, though it should be reflected in a slight upward adjustment in U.S.
Digitized by Google
Maneuvering Liquid Assets 99
b'I
EXHIBIT 3.17
Local loan makes liquidity available in year t, but its repayment with
interest reduces liquidity in year t + 1.
N
:t Kt. t + 1 ;> compensating balance requirement in year t
i=1
Digitized by Google
100 Multinational Financial Management
taxes to the world are paid when the subsidiary's capitalization is mostly
debt. Unfortunately, the interest payments on debt are a persistent, non-
discretionary flow of funds. They are difficult to manipulate. Whenever
the tax advantage of debt is low, headquarters will want to fund its subsid-
iary with equity in order to have the maneuverability of dividends. This
theory suggests that it would be inappropriate to survey for average
debt/equity ratios, because an average of a high and a low is not meaning-
ful. Actually, most situations are less clear cut and the sensible approach
is to formulate this maneuvering model and simulate different capitaliza-
tion structures for particular subsidiaries.
Digitized by Google
-------=------- -- --- - - ----
the dual variable. These data provide a very clear guide for making short-
and medium-term loans in that nation, even revealing the term structure
for medium-term loans. This interest rate can be used as the cost of capital
for investment decisions.
Some banks are themselves multinational. In one nation the bank may
ration credit because of a high demand for loans within the permitted
interest rate. In another part of the world the bank may be short of de-
posits as it strives to expand. In exchange for depositing money in one
part of the bank, the multinational company can negotiate loan priority in
another part of the bank. The rationale behind this is that most currencies
are blocked and it may be easier for a company to move money than it
is for a foreign bank. The merit of such induced-swap transactions can be
evaluated with this model.
Raising Long-Term Capital. Long-term capital can be raised in most na-
tions, usually after negotiations with the host government and promises
that local investments will result. To support such negotiations, world-
wide manufacturing models of the type discussed in Chapter 7 must be
supplemented by worldwide financial schemes. Similarly, one can evalu-
ate the effects of mergers and divestments.
Restructuring the Pattern of Legal Ownership. If the company has N
subsidiaries, there are (N - 1)2 different patterns of intersubsidiary own-
ership possible even if each subsidiary is constrained to have just one legal
parent. If one removes the restriction that a subsidiary have one legal
parent the tax data required are forbiddingly more complicated and vague
than for cases of complete ownership. In practice the model presented in
this chapter should stimulate new patterns of intersubsidiary ownership
by eliminating some of the drudgery. A subcalculation, about which tax
lawyers are well informed, is the benefit of operating in nation i with a
corporate charter granted in nation j - operating in Ireland with a Bermu-
da charter is one possibility.
Inserting High-Profit Products. The Swiss pharmaceutical companies
prefer to produce their high-profit products in the nations where they in-
cur their largest cash drains to research and dividends. This practice, a
vivid example of Vernon's (1966) product cycle theory, reduces the need
to flow funds between nations, though at a high production and trans-
portation cost. Ideally, one would like to simulate the effect of inserting
a high-profit product into any subsidiary technically sophisticated enough
to manufacture it.
Competitive Analysis. When one corporation plans to overtake another,
it usually undertakes a thorough competitive analysis. The competitor's
annual report is usually a consolidation of real worldwide operations
Digitized by Google
102 Multinational Financial Management
which the first corporation would like to disaggregate and understand. A
corporation that has already created a maneuvering model of itself may
consider creating a similar model of its competitor. Although most of the
input information is limited to estimates, two kinds of publicly available
information have been commercially compiled.
First, the financial statements of most subsidiaries are available. Many
governments require publication of the financial statements of each
business incorporated and operating in their nation. With diligent work
one can collect the subsidiary financial statements of a competitor, and
with diligence one can assemble these into a financial model. This work
has already been done for almost 100 multinational corporations, in the
Source of Earnings Report of the Buttles Corporation (Plandome, New
York 11030).
Second, the dividend conduits are available from organization charts.
Multinational corporations have frequently evolved numerous subsidiaries
and branches in every nation. Verlag Hoppenstedt (D6100 Darmstadt,
West Germany) provides a compilation of the percentage ownership be-
tween every legal entity, its capitalization, and line of business.. U .S.
Organization Chart Service Inc. (La Jolla, California 92038), provides
organization charts showing not only titles and relationships but also
names in over 575 U.S. high-technology corporations.
Digitized by Google
- - - - - - = = ._-...-----.-._---.---
be a task force to set up the data base. This will be the most time-consuming
aspect of model implementation. This task force will, of necessity, include
tax lawyers, accountants, and financial analysts, as well as ,ystems analysts.
During the early stages of implementation it will be sensible to solve
the problem with a familiar linear programming code. Most corporate
O.R. analysts are familiar with linear programs, but few have used gener-
alized networks (an ordinary network does not include factors of attenua-
tion). Each flow is an activity, and each node is a constraint. Once a
"small" problem has been tested and used month after month, there will
be a push to add more subsidiaries.
Unfortunately, computer costs will become substantial if every legal
entity in every nation is modeled. Only at this point is it sensible to
switch to a generalized network. The power of the Balachandran-Thompson
algorithm is that its systematic parametric programming of the cost co-
efficients allows analysts to impose constraints that bind several flows. A
network formulation is attractive not only because it can be visualized,
but also because it can be solved faster than a general linear program. A
simple maneuvering liquid assets model can be inexpensively developed. A
full model representing 100 subsidiaries may cost $500,000 and require
one to three years of task work.
Conclusion
Digitized by Google
Questions from Other Viewpoints
Digitized by Google
Maneuvering Liquid Asset. 105
Digitized by Google
106 Multi1UJtio1UJl Fi1UJ1lcial Mtmagement
16. Societal Emotional Descriptive National. Your nation is identified
as a tax haven in the advisory newsletters written for multinational
treasurers. You are Minister of Finance and your cabinet colleagues
are always pressuring for expansionary programs that will require
more tax revenue. Lately, the Minister of Education suggested ex-
panding the national university. In your past encounters with him, he
has never had any interest in economics. Describe the nonfinancial
benefits your nation gains from being a tax haven. Draft a concise
memo.
Bibliography to Chapter 3
Balachandran, Venkataraman and Gerald L. Thompson, "An Operator Theory of Para·
metric Programming for the Generalized Transportation Problem," NaVGl Research
Logistics Quarterly, Vot. 22, No. 1 (1975). pp. 79-125.
Bums, Jane 0., ''Transfer Pricing Decisions in U.S. Multinational Corporations," Jour-
nal of InterntJtioMI Business Studies, Vot. 11, No. 2 (1980), pp. 23-39.
Crum, Roy L., "Cash Allocation in the Multinational Finn: A Constrained Generalized
Network Approach," Doctoral Dissertation, University of Texas at Austin, August
1974.
Dantzig, George B., Linear Programming and Extensions (Prince ton : Princeton Univer·
sity Press, 1963).
Eiteman, David K. and Arthur I. Stonehill, Muitinatio1UJ1 BUsiness Fi1UJnce, 2nd ed.
(Reading, MA: Addison-Wesley, 1979), esp. Chapter 11.
Eli LiUy & Company v. The United States, 372 F. 2d 990, 178 Court of Claims
[No. 293-61, Decided February 17,1967], pp. 666-733.
Fowler, D.J., "Transfer Prices and Profit Maximization in Multinational Enterprise
Operations," JourntJI of International Business Studies, Vol. 9, No. 3 (1978),
pp. 9-26.
Mathewson, J. Frank and G. David Quirin, "Economics of Fiscal Transfer Pricing in
Multinational Corporations," A Study Prepared for the Ontario Economic Coun-
cil, Toronto 1978.
Merville, L.J. and J. Petty, "Transfer Pricing for the Multinational Finn," Accounting
Review, Vol. 53, No. 4 (1978), pp. 935-951.
"Multinational Corporations and Income Allocation Under Section 482 of the Internal
Revenue Code," Harvard Law Review, Vol. 89, 1976, pp. 1202-1238.
Reier, Sharon, "IBM's Science of Simplification," Institutional Investor, Vol. 14,
No. 11 (1980) pp. 219-220.
Robbins, Sidney M. and Robert B. Stobaugh, "Multinational Companies: Growth of
the Financial Function," Financial Executive, Vot. 41, July 1973. pp. 24-31.
Rodriguez, Rita M. and E. Eugene Carter,Inter1UJtional Financial Management (Engle-
wood Cliffs, N.J.: Prentice-Hall, 1976); esp. Chapter 8.
Rutenberg, David P., "Maneuvering Liquid Assets in a Multinational Company: For-
mUlation and Detenninistic Solution Procedures," Management Science, Vol. 16,
No. 10 (1970). pp. B671-B684.
Digitized by Google
ManeulJering Liquid Assets 107
Seghers, Paul D., How to Do Business Abroad at Least Tax Cost, Englewood Cliffs,
N.J.: Prentiee·HalI, 1964.
Sbapiro, Alan C., "Payments Netting in International Cash Management," Journal of
International Business Studies, Vol. 9, No. 2 (1978), pp. 51-58.
Sbulman, James, "Wben the Price is Wrong- By Design," ColumbiIJ Journal of World
Business. Vol. 2 No. 3, May-June 1967, pp. 69-76.
Vemon, Raymond, "International Investment and International Trade in the Product
Cycle," Quarterly Journal of Economics, Vol. 30, May 1966, pp. 24-35.
Wentz, Roy A., "Corporate Transfers of Intangible Abroad," Tax ExecutilJe, April
1967, pp. 142-159.
Digitized by Google
CHAPTER FOUR
Multinational Expansion
to a New Nation
108
Digitized by Google
..,.--,--,.. .
~---==-=-==-=--=--.: _. -
Digitized by Google
110 Multinational Financial Management
most recent and most important were the radial tire in 1948
and the assymmetric tire for high-speed driving in 1965.
From the start Michelin had been a family company. One
of its most obvious characteristics was secretiveness. For
example, to maintain security during the erection of its Nova
Scotia plants, Michelin had brought its own construction
crews from France and had paid the Nova Scotian Federation
of Labour a large amount-$250,000 by one report-to
compensate for not hiring local workers.
The secrecy surrounding Michelin had not obscured its
fmancial success, however. Although reported fIgUres were
inadequate for complete analysis, Michelin's operating margin
appeared to be about 13 percent. In addition, the company
had a record of steadily rising profits, unlike the four U.S.
giants - Goodyear, Firestone, Uniroyal, and Goodrich - and
also unlike its chief European rival, Dunlop-Pirelli. By 1969
Michelin's tire sales exceeded U.S.$l billion in value. In
Europe alone, where it ranked first in tire sales, it employed
more than 70,000 persons. The Michelin group of companies
was totally self-financing. In 1969, its intemal sources of
funds, including posttax profits and reserves, amounted to
more than U.S.$120 million. These figures did not include
Michelin holdings in the Citroen automobile and Kleber-
Columbes companies, concems independently managed and
registered on the Paris Stock Exchange.
In its strategy, Michelin was known to stress quality control
and market planning but was believed to give top priority to
technical research, analysis, and development. Thus it was
credited with "radializing" most of the world's tire markets
outside the United States, while U.S. interest, too, was
clearly growing. In August 1971, U.S. market trends were
assessed as follows:
Digitized by Google
Multinational Expansion to a New Nation 111
Digitized by Google
112 Multinational FinanciDI Management
The Agreements
Before completing their negotiations, Michelin and Nova
Scotia had arranged for financial support from provincial,
local, and national governments, as seen in Exhibit 4.1.
Besides quantifiable loans and grants totaling about $85
million, Michelin would get three other kinds of assistance,
the value of which could not be precisely measured. These
three would come from the Federal Government and would
be additional to the tabulated Federal grants, which were
payable under the Area Development Incentives Act (ADIA).
Also provided under the ADIA was an accelerated de-
preciation, against income taxes payable, of the capital cost
of land and buildings. The company could write off the
building costs at 20 percent per year and the machinery at
50 percent per year, profits permitting.
Moreover, Michelin also would be allowed to import much
of its manufacturing machinery duty-free, under tariff item
Digitized by Google
Multinational Expansion to a New Nation 113
EXHIBIT 4.1
GAaumes 100 percent asseament of actual .22.5 mWion cost in Granton and .62.2 mWion in
Brid.ewater. Iporea adjustments for present value and imputed interest on tax saviD...
bFederal cub Irants were payable under the Area Development Incentives Act of 1964. Amounts
depended on the value of investment made and the number of jobs created. Aresa eUlible to re-
ceive the Irants were deailDated on the bula of employment statistics, non-fum-family income,
and income distribution.
Digitized by Google
114 Multirtlltiortlll Financial Management
Digitized by Google
Multinational Expansion to a New Nation 115
Industry Reactions
So far as the rest of Canada's tire industry was concerned,
early news about Michelin caused little stir. For one thing,
the company was concentrating on truck tires, a relatively
small segment of the tire market. For another, competition
had heard with pleasure of delays in Michelin deliveries and
start-up troubles at the plants. Thus little serious thought
had been given to the long-term implications of Michelin's
arrival.
This complacency was to disappear in January 1970, when
a Federal Government official informed the president of the
Rubber Association of Canada (RAC) of everything the
government was doing for Michelin. After listening in
"stunned silence," RAC's President W.V. Turner asked some
questions to ensure that he understood completely and then
reported back to the industry. Company managements were
said to be shocked: They saw the Michelin arrangements not
only as a source of increased competition but also as a betrayal
of the other tire companies by the government.
Digitized by Google
116 Multinational Financial Management
Industry-Government Relations
As one company executive put it, by 1970 the Canadian tire
industry was "sick of playing 20 questions with the govern-
ment." Such an attitude grew from historical events. For
example, the government had encouraged the tiremakers to
work together for maximum efficiency during World War 11,
but in 1952, by bringing a successful price-fixing suit against
the companies, it had effectively frightened them away from
any future efforts at cooperation. Similarly, the government
had encouraged the tiremakers to produce a full line of tires
in Canada by setting tariffs at a high 17.5 percent, but during
the 1960s consideration had twice been given to U.S.-
Canadian free trade in tires. That is, two proposals had been
advanced to include tires under the U.S.-Canada Automotive
Trade Agreement, which in effect had brought the U.S.-
Canadian free trade in motor vehicles-provided U.S. makers
would produce an agreed proportion of their output in
Canada. This agreement had permitted longer production runs
on both sides of the border. It had also increased Canadian
output: From under 50,000 in 1964, unit shipments of
passenger cars had risen to 1.13 million in 1971, with Canada
exporting about 428,000 more units to the United States
than vice versa.
Besides the two proposals to put tires under the U.S. - Can-
ada Auto Pact, in 1969 another disturbing proposal emerged
in the course of discussions between the tire companies and
officials in the Ministry of Finance. According to one execu-
tive who was present, officials had formed the opinion that
the Canadian tire industry was poorly managed, lacking in a
sense of direction, and fraught with bickering. The govern-
ment was concerned about industry efficiency, and at one
point the ministry officials had threatened to make the
industry more efficient simply by lowering tariffs on all
imports.
Digitized by Google _ _
-----==------ ----- -- - - ----
Industry-Governmen t Negotiations
From the outset, Canada's tire industry spokesmen maintained
that they had no objection to the government's loans and
grants to Michelin -these being forms of assistance that were
available to any company. Rather, the bone of contention
was the remission-of-duty agreement. To the industry it
seemed that this agreement meant that the government was
not only bankrolling Michelin but was also creating an effec-
tive rationalization for the company-one such as they had
been seeking, but one from which they would not benefit.
Based on this analysis, the industry's first hope was to get
the government to back off from its duty-free concessions. It
soon became apparent, however, that the government was
fully committed.
The tire manufacturers then started meeting together in
RAC to see if they could agree on what the government
could do for them to compensate for the Michelin deal. This
effort, too, soon ran into problems. According to RAC's
President Turner, one reason was that all the companies had
replaced their presidents in 1969 and 1970, and the new men
did not have as good a relationship as the old guard. In addi-
tion, the companies had very different strengths and weak-
Digitized by Google
118 Multinational FinanciIJl Management
Digitized by Google
Multinational Expansion to a New Nation 119
Digitized by Google
120 Multi1llltional FinancitJl Management
Remedy Sought
The remedy sought by the V.S. companies was a so-called
countervailing duty. Vnder V.S. law, the V.S. Treasury must
grant the industry's request should investigations show that
the foreign government had set up a subsidy program to
Digitized by Google
Multinational Expansion to a New Nation 121
Digitized by Google
Introduction
Risks are incurred when a multinational corporation expands to a new
nation. Since most corporations are multidivisional, headquarters must
choose which division to urge to move first into the new nation. Few
headquarters executives are conscious of making such a choice. This
chapter considers the selection of the most suitable division. The chapter
is arranged into polycentric, ethnocentric, and geocentric sections; analyti-
cal models are presented in each.
Polycentric Process
Most polycentric corporations became multinational by a process of
creeping incrementalism. Expansion to a new nation may have begun
with an unsolicited export order. Slowly the export market expanded,
with rarely any attention from headquarters. Headquarters attention was
usually first drawn by a capital appropriation request for a warehouse in
the new nation. Such requests are usually presented as part of ongoing
operations, and are approved on the basis of return on investment (ROI)
calculations, based on the conservative assumption of cost savings. There
will rarely be a political forecast in the capital appropriation. The next
step may occur when the local management wishes to begin manufactur-
ing. In some developing nations, the government threatens to withhold
import permits unless the corporation performs some assembly opera-
tions. The largest, though intangible, asset of the corporation in a nation
is the present value of the expected stream of marketing revenue. Fre-
quently, however, it seems to be the real asset (e.g., the warehouse) that
becomes the hostage in executives' thinking. The weakness of a poly-
centric organization is that once it has expanded to a nation there is
only a weak organizational mechanism available to redeploy its assets
if this becomes necessary. Hence the initial decision must be sound.
Notwithstanding this concern, the essence of a polycentric corpora-
tion is that its executives perceive risk as though the corporation were a
geographically diversified mutual fund. Modem stock market theory sees
risk as having a price; financial risk can be bought and sold. A polycentric
corporation feels that it should diversify broadly and thus pool risk.
In headquarters, the contact officer manning a national desk may
experience ambiguity as to his role, as described in Chapter 1. His impor-
tance increases as corporate involvement within "his" nation is expanded
from importing to systematic sales, to licensing, to a joint venture, and
then to a wholly owned subsidiary. To the extent that self-interest in his
career colors his business judgment, he will favor the subsidiary's incre-
mental expansions and will have little motivation to perform adequate
122
Digitized by Google
-----==-==-=--=--- - - -- -- -----
Ask a busy U.S. international executive how he copes with the world information
explosion. If he is honest, he will point to a wastepaper basket fdled with unread
documents and explain that there just isn't enough time to get through everything,
and besides, most of the material in that wastepaper basket is worthless. (Keegan,
1968)
A Facilitating Methodology
Task forces are used when deciding which divisions should enter which
nation. The attention that a task force leader gives to "getting everyone
on board" is an indication of the importance of establishing enough
familiarity with and attendant knowledge of the alternatives. "N onmetric
scaling" could be used by this leader to bring some structure to vague
perceptions of the decision problem. This methodology helps pinpoint
Digitized by Google
124 Multi1UJtional Financial Management
important considerations and steers the company toward wiser investment
decisions.
Step 1 provides input needed to produce a non metric scaling map of the
perceptual space being used by each participant. From eight divisions
there are 28 feasible pairs of divisions. Twenty-eight cards, each bearing
the names of a pair of divisions, are given to each task force member. He
must sort the cards so that the top ranked card shows those two divi-
sions with the most similar priority for corporate funds in the nation
(they should be closest to tying for the same priority regardless of whether
the actual priority is high or low). Participants find it easy to decide
which divisions were most similar and which were least similar; inter-
mediate comparisons take longer. Each participant's deck of cards is run
through the computer program for nonmetric scaling; the computer prints
a two-dimensional depiction (as in Exhibit 4.2) of how the participant
configures the eight divisions' priority for funds. After the data are
plotted, each paper is trimmed into a circle to emphasize that the dimen-
sions (the meaning of the groupings) are unknown.
In step 2, the leader of the task force studies the pile of paper circles,
one by each participant, and groups them into a few somewhat similar
clusters. (There are mathematical techniques for clustering that could
be used.)
Step 3 is to articulate a rationale for the clusters.
Step 4 is to gather data on the disputed divisions and prepare exhibits
that highlight differences between these divisions in a Delphi-like pro-
cedure (Mason, 1969).
Step 5 is to reconvene the task force and concentrate attention on the
disputed divisions in an effort to clarify the true nature of the corpora-
X
Division 3
X
Division 2
X
Division 1
X
Division 8
X X
Division 7
X
EXHIBIT 4.2
One participant's nonmetric scaling of divisions.
Digitized by Google
----
------.----. -- .. -------_.- ---.--
tion's strategy for the new nation. Following the spirit of Mason (1969),
disagreements about the priority of various divisions will require the
executives to articulate unique features of this corporation's strategy for
the new nation.
In summary, in a polycentric organization there is no structure in which
to perform rigorous analyses of commitment investments in a new nation.
The methodology outlined in this section was designed to facilitate the
interaction of a task force whose leader lacks definitive power. Once the
task force is in reasonable agreement, it proves relatively easy to select a
division to enter each nation.
Ethnocentric Process
To an ethnocentric executive, risk is to be avoided. Many real capital
budgeting decisions are made de facto when a manager immediately
rejects (screens out) a project as being unworthy of detailed examina-
tion. The ethnocentric top executive is haunted by the fear that once a
nation gets mentioned on the agenda of things to investigate, a task force
will be created. The executive knows from experience that such task
forces are unstoppable. The task force will be beguiled by the natives;
seeing themselves (or their protegees) as managers of the new national
venture, they will usually report favorably. Thus the only way to suppress
such risk-seeking profligacy is to control agenda creation and snuff out
any mention of new nations.
An ethnocentric executive who feels uncomfortable about the foreigness
of wholly owned subsidiaries of which he might lose control will feel even
more anxious about joint ventures abroad in which he must share control
with a foreigner. In order to understand the attitude of U.S. executives
toward joint ventures and co-production agreements, Amariuta, Ruten-
berg, and Staelin (1979) conducted a mail questionnaire survey of the
vice presidents (International) of U.S. corporations. Executives were asked
how many joint venture proposals their corporations had rejected in the
last three years and at what level. It was emphasized that the joint ven-
tures could be anywhere in the world. The astonishing result was that
most proposals were rejected a priori, with no analysis whatsoever.
Working from the right side of Exhibit 4.3 we see that formal capital
appropriation analyses are used to help eliminate only the final 5 percent
of all starting proposals. Yet the evaluation tools of finance deal mainly
with such analysis. The second disturbing result is that so few proposals
are eliminated by creating alternatives against which to compare them.
To verify this, executives were asked to think about their most recent
international joint venture and recall how many alternative host countries
they had considered for this project. Over 77 percent had considered only
the nation chosen. Of the 23 percent who had bothered to compare
alternatives, each considered an average of three nations_ Fewer than a
Digitized by Google
126 MultilUltiolUll FilUlncilll Management
1.0
Accepted
joint
venture
9.47
Proposals
0.11
Reject after Reject after
7.88 considering formal analysis
Reject a priori alternatives
EXHIBIT 4.8
Severity of screening.
Digitized by Google
--==-==- ._-- -- .--,-:-:---:-:---........----
Knowledge
of
Eastern Europe
Perception of Perception of
riskiness of Inconvenience of
doing business doing business
in Eastern Europe in Eastern Europe
Required
rate of return
before investing
In a jOint venture
in Eastern Europe
EXHIBIT 4.4
Paths between knowledge and threshold rate of return.
able executives saw more risk but were oblivious to the expensive work of
negotiating and sustaining relationships in Eastern Europe.) Thus one
danger of an ethnocentric attitude is that such executives do not realize
that they have an enormous problem in coming to understand each
nation.
Some theory will put this survey in perspective. Think of the corpora-
tion exposed to a stream of potential deals. Whereas a domestic executive
can make commitments rather intuitively, theory would lead one to
expect an international executive to be more systematic - the territory is
less familiar, the range of opportunities is greater, and the number of
proposed deals is much greater.
Because time taken to screen is expensive, we would expect executives
to group proposed deals into:
1. Reject immediately.
2. Investigate further.
3. Accept immediately.
Digitized by Google
128 Multinational Financial MaruJgement
The Checklist
Several corporations have more than 100 product divisions, so many that
even the corporate planning group has difficulty keeping track of the
strategy of each. An important corporate planning task is to identify those
divisions that should be encouraged to move faster to a worldwide posi-
tion and those divisions that should be restrained from leaving the United
States. The viewpoint is rather ethnocentric. The planners presuppose that
attractive opportunities exist somewhere abroad and that no foreign
government resistance will be met. In this section we examine a checklist
used in the headquarters of a U.S. corporation to rank divisions as to their
readiness to go abroad.
Checklists are used more extensively in corporations than business
school teaching would indicate. They reduce the cost and pain of an
investigation. Such lists remind subordinates of factors to be considered
and they screen huge problems down to manageable size. The following
checklist gives 15 factors to be considered. The first two factors eliminate
a division. The next three factors limit the division's suitability. Eight
factors positively propel the division. If a division scores in the final two
factors, corporate planning begins an immediate analysis of its expansion
abroad. For each division, each factor could be checked.
1. Current Domestic Divisional Problems Preclude Diluting Management
Attention (Eliminating) Like the boy who ran away to sea to avoid
facing unresolved problems at home, some managers of troubled divisions
may want to escape by going international. If the origin of their problems
lies in their own managerial ineffectiveness, the division should stay
domestic. One possible retort could be that such a constraint is ethno-
centric, that the domestic market for this product is cut-throat, and that if
the corporation thought globally, it would expand abroad while shutting
down the domestic operations. Such conceptual flexibility is viable only
when the corporation has experience in performance evaluation and pre-
diction based on years of experience in many nations. Until then, current
domestic problems eliminate the division from the list of candidates.
Digitized by Google
- -~---:'~~---- .. -- -- -
Digitized by Google
180 MultinationtJl Financial Management
that a strong U.S. market position also means a large share of the world-
wide market.
10. Real Market Growth Greater Than 5 Percent (Positive) For the
division's product, worldwide market growth data can be adjusted for
inflation. The quality of worldwide data is usually terrible, but it is good
enough to determine whether real growth centers around 5 percent or
around 10 percent. Growth is important if entrenched competitors are to
accommodate themselves to a new entrant. A growth rate must be used to
calculate the present value of a stream of future income.
11. Emerging Markets Abroad (Positve) Based on a belief in experience
curves, the corporation desires a large share of the markets it enters. This
is easier to achieve when the market is newly emerging and no competitors
are entrenched.
12. Real U.S. Market Growth Greater Than 10 Percent (Positive) The
industry containing this division is growing very fast in the United States
and presumably it would grow fast abroad. If the corporation injects more
capital into the division now, it can dominate the market worldwide. The
real problem, however, is that managing such growth absorbs so much
executive energy that overseas growth seems superfluous.
13. Accepted U.S. Industry Leadership Position (Positive) There is
usually a leader in an oligopoly. If the division is the leader and wants to
remain the leader, it should expand abroad before its competitors do, or
it may forfeit leadership.
14. U.S. Marketing Position Threatened by Competitor's Sourcing Com-
ponents or Products Abroad (Compelling) Another company can try
to blackmail our corporation into not moving abroad, whenever our cor-
poration is a customer of one division of a company and a supplier to a
second division of the same company. If the second division plans to move
abroad, the company may threaten to find another supplier. When such
blackmail is threatened, headquarters (not the divisions) has a compelling
reason to analyze the costs and benefits of moving abroad.
15. Markets Currently Served from the United States Are Threatened by
Increasing Import Restrictions (Compelling) In a developing nation the
threat of tariffs is familiar. In a developed nation, a national corporation
may be selected by the government planners as the chosen instrument to
supply this particular technology. If our corporation builds a manufacturing
plant, it may be able to negotiate to avoid foreclosure from this national
market.
Digitized by Google
------:== _. ------.--_.-----""'"'"
Geocentric Process
At the negative extreme, the xenophobia within certain countries may lead to
either a complete bar to foreign investment or to the expUlsion of foreign com-
panies, with confiscation of assets. At the positive extreme, there exists a real
accommodation so that the foreign company becomes almost indistinguishable
from national companies.
The majority of nation-states are somewhere in the middle range of this spec-
trum. This basic fact colors any consideration of multinational corporate per-
formance. Foreign ownership implies differences in values and objectives as well
as language and customs. These foreign birds have strange and different colors
when they first migrate to new lands. Without any natural camouflage they may
become fair game for the local hunters. Time and experience means a natural adap-
tation or ultimate extinction. (Rosow, 1974, p. 147)
Digitized by Google
132 MultinatioTUJl Financial Management
though each nation had only one business climate. Actually few societies
are so monolithic as to have only one climate. Societies are pluralistic in
nature, with several stakeholders - individuals or groups with enough
power to block or facilitate a project. It is extremely important that some-
one in the corporation learn to differentiate the several stakeholders,
"the local hunters" in Rosow's quotation, and to rank each stakeholder's
preferences for one division over another. Behnnan (1969, p. 114) presents
a general list of possible stakeholders, to which I have added comments
and examples.
1. Private customers' acceptance of the corporation's product per-
fonnance.
2. Customers' (especially government agencies') perceptions of the
foreignness of the corporation. Most countries support local industry,
particularly if locally owned. However, in Romania, for example, the
novelty of dealing with a foreign corporation may be an assist.
3. National government goodwill. This may be subdivided into the Min-
istry of International Trade and Industry, the Ministry of Finance, and
any other government agency, if there are differences in their views.
4. Municipal government perceptions of the corporation as being either
a good or an unstable employer. In India, it may be necessary to also
consider the state government, because of its power to delay licenses.
5. The current and potential eagerness of local managers to adopt cor-
porate goals as their own. Many U .S. corporations still fear that their
Japanese managers are more loyal to Japan than they are to the cor-
poration. The degree to which this division's products are in the
national interest may ameliorate any conflict of loyalties.
6. Workers' and trade unions' perceptions of the prospect for adequate
wages. This is generally an asset, as multinationals have a reputation for
paying higher than average wages.
7. Other executives' opinions (pro or con) about the corporation.
8. Public opinion. In Hungary, the public is probably eager for contact
with the West. An investment in some supposedly capitalist nations,
however, may be viewed by that country's populace as exploitation by
the multinational company.
9. Corporate headquarter's perception of the likelihood that the division's
operation in this nation will quickly be a net cash producer and will
grow steadily. An investment in Romania might also be seen as a foot-
hold for further expansion in Eastern Europe.
10. The financial community's (including local banks) relationship with
the corporation.
An international executive should prepare a similar list, tailoring it to the
social structure of the particular nation under consideration. The challenge
is to have these stakeholders in mind when deciding which division to
move first.
Digitized by Google
---==~ -- ~.-=----:-.-:-.--:-:".=-::::'~------.
One-Stage Selection
One-stage selection implies that a division (or several divisions) will be
selected, moved into the nation, and left there to operate. The one-stage
selection has three steps. They are most conveniently performed on a
computer, with the questions posed on the computer screen (Saaty, 1977).
Step la determines the preferences of each stakeholder and thus de-
termines the direction in which each stakeholder is likely to use its force.
Consider each pair of divisions, and from the vantage point of the stake-
holder express his preference, choosing one of the following categories
for each pair:
Division rating Score in cell ij
Division i is unquestionably superior to division j 5
Division i is superior to division j 4
Division i is much better than division j 3
Division i is somewhat better than division j 2
Division i is equal to division j 1
Division i is somewhat less good than division j %
Division i is much less good than division j %
Division i is inferior to division j %
Division i is unquestionably inferior to division j Ifs
If there are d divisions there are (d -1) d/2 pairs of divisions. The com-
puter can arrange the preference data into a square matrix of divisions
(the diagonal cells are all one) and can calculate the eigenvector of the
matrix. The eigenvector is merely a scale of preference, on a line 1 through
10, as in Exhibit 4.5. This vector is d X 1 where there are d divisions.
Step 1 b brings together the preferences of all the stakeholders. There is
Digitized by Google
134 Multinational Financial Management
Matrix Eigenvector
Divisions: g h j k
g 1 5 4 yields 1 10
h 1/5 1 g j k h
1/4 1
j 1
k 1
EXHIBIT 4.11
Depiction of matrix and eigenvector output of the computer.
Digitized by Google
- - = = = - ---------------------
Digitized by Google
136 Multinational Financial Management
Closing Decisions
When a corporation wants to close down one of its product divisions in a
nation, the corporation should expect pressure to be exerted on its other
divisions. Major customers and suppliers of the remaining divisions, fearing
that this closing will be the first of many, will cool their relationships,
Digitized by Google
---==-=-==::":-::- - ...---..-.---.------
..----~
Conclusions
All multinational corporations are inevitably faced with the decision of
whether to enter a new nation. Sometimes the decision will be an easy
one; more often, as in the case of Hungary, that decision will be com-
plicated. A corporation's decision to expand its activities to a new nation
is based on a multitude of factors; some act as constraints while others are
stimulants. The purpose of this chapter is to develop procedures that
will help a firm's management scrutinize its own decisions concerning
international expansion. Attention has been focused on the divisional
managers, because most large corporations - multinational or not;
European, American, or Japanese - are organized by product division
Digitized by Google
138 Multinational Financial Management
lines, and the divisional manager should be a key participant in such a
decision; he alone can make it work.
The organization theories of Bamard (1938), March and Simon (1964),
and Cyert and March (1963) emphasize search, especially search in the
vicinity of existing practice. When Aharoni (1966) investigated how
corporations decided to study and evaluate a nation, he found evidence of
the notion that search is engrossing and that members of a task force
appointed to evaluate a particular nation almost always favored that
nation. What Aharoni only hinted at was that a task force usually consists
of all the people in the corporation familiar with the nation, thus preclud-
ing any informed opposition.
Designing corporate policy includes articulating the objectives of the
corporation, assessing its environment, and identifying corporate strengths
and vulnerabilities. Environmental assessment can become a deeply felt
commitment to ride with an environment. Searching a new environment
seems to involve personal risk: "Get your feet wet" and "Get burned a
few times" are common directives.
Strategic moves defmitely arouse strong emotions. In such a situation it
is difficult to find an appropriate problem analysis structure. Further-
more, if too much structure is imposed, interesting opportunities will
surely be excluded. If too little structure is imposed, the magnitude of
their task will engender feelings of anxiety that will overwhelm most
analysts. (Their usual coping behavior is to ignore all but one small corner
of a problem.) This chapter offers a sequence of analytical tools that pro-
vide some structure to the agonizing work of trying to help make strategy.
The models of this chapter are about as complicated as one can com-
fortably manipulate analytically. Reality is appreciably more complex.
For example, the model is for one point in time and yet attitudes change
(and at different rates). As Truitt (1970) noted, "In assessing risk ... it
behooves the foreign investor [and students of international business] to
bear in mind that the host government will be asking, 'Yes, but what have
you done for me lately?'"
Moreover, we've assumed that the stakeholders act in isolation, whereas
in some societies particular coalitions of stakeholders are possible. One
stakeholder without power to block a project may nevertheless work to
become the catalyst around which a blocking coalition will coalesce.
Finally, the company's power to influence such coalitions has not yet
been explored. Clearly, the corporation is bringing benefit to some sectors
of the nation. Hence, with a well-planned program of public and govern-
ment relations, it can become a force in its own right. The most perceptive
model of corporate action may be the early days of the Tennessee Valley
Authority. Actually (and ironically) the TV A was a government-owned
enterprise whose early existence was threatened by private enterprise
stakeholders. The chairman of the TV A, David Lilienthal, mastered a
process of slowly coopting stakeholders, then systematically expanding
TVA activities (Selznick, 19(9).
Digitized by Google
... o.M"-"-"'-_ •• -- oi_-.·-.-
Digitized by Google
140 Multinational Financial Management
Digitized by Google
Multinational Expansion to a New Nation 141
15. Societal Emotional Descriptive Global. If there were to be economic
warfare between national governments, in what way would multi-
national corporations be viewed? Specifically, how could a subsidiary
abroad become a hostage, influencing the actions of the headquarters
government? In what way could a subsidiary abroad be a means of
expressing the wishes of the headquarters government?
16. Societal Emotional Descriptive Subsidiary. The problem of this
chapter is only a partial statement. There are many corporations
that may be induced to compete to enter a nation. Select one of the
stakeholders, and from that viewpoint write a one-page briefing
memo on how to play corporations of different nationalities off
against one another.
Bibliography to Chapter 4
Aguilar, Francis J., Scanning the Business Environment (New York: Macmillan, 1967).
Aharoni, Yair, The Foreign Investment Decision Process (Cambridge: Harvard Uni-
versity Press, 1966).
Amariuta, Ion, David Rutenberg, and Richard Staelin, "How American Executives
Disagree about the Risks of Investing in Eastern Europe," Academy of Manage-
ment Journal, Vol. 22, No. 1 (1979), pp. 138-157.
Barnard, Chester I., The Functions of the Executive (Cambridge: Harvard University
Press, 1938).
Behrman, Jack N., "Some Patterns in the Rise of the Multinational Enterprise,"
Research Paper 18, University of North Carolina, 1969.
Bower, Joseph L., Managing the Resource Allocation Process: A Study of Corporate
PlIJnning and Investment (Cambridge: Harvard University Press, 1971).
Brada, Joseph C., "Markets, Property Rights, and the Economics of Joint Ventures in
Socialist Countries," Working Paper 76-64, Graduate School of Business Adminis-
tration, New York University, 1976.
Bradley, David G. "Managing against Expropriation," Harvard Business Review, Vol.
55, No. 4 (1977), pp. 75-83.
Charnes, Abraham, Frederick Glover, and Darwin Klingman, "The Lower Bounded and
Partial Upper Bounded Distribution Model," Naval Research Logistics Quarterly,
Vol. 18, No. 2 (1971), pp. 277-281.
Cyert, Richard M. and James G. March, A Behavioral Theory of the Firm (Englewood
Cliffs, NJ: Prentice-HaII, 1963).
Fouraker, Lawrence E. and John M. Stopford, "Organizational Structure and the
Multinational Strategy," Administrative Science Quarterly, June 1968, pp. 47-64.
Goodwin, Richard N., "Letter from Peru," The New Yorker, Vol. 45 (May 17,1969),
pp. 41-109.
Keegan, Warren J., "Global Intelligence: A Framework for Action," Worldwide P and
I Planning, July-August 1968, p. 48.
March, James G. and Herbert A. Simon, Organizations (New York: Wiley, 1964).
Marer, Paul and Joseph Miller, "U.S. Participation in East-West Industrial Coopera-
tion," Journal of International Business Studies, Vol. 8, No. 2 (1977), pp. 17-29.
Digitized by Google
142 Multinational Financial Management
Digitized by Google
Summary of Multinational Financial Management
The three finance chapters you have just read flow into one another to
build successively. Although the topics were carefully selected ;.0 typify a
short-, medium-, and long-term problem, these are merely a small sample
of the variety of real finance problems in a multinational corporation.
Consider a typical textbook in corporation finance. For each chapter
think how that activity could be done in a multinational corporation. The
guiding rule in most corporations is to decentralize an activity if possible,
subject to two caveats. First, the procedure for performing the activity
is often very centralized (for example, anyone can concoct a capital
budget proposal, but headquarters specifies the format of the write-up,
the mathematics of cash flow summarization, and the review procedure).
The second caveat is that some decisions can be better made centrally. In
essence, the justification for the expensive headquarters of a multinational
company is to identify and make these profit-creating decisions. Chapters
2, 3, and 4 identify three such decisions.
The essence of the foreign exchange chapter is to say that exchange rate
risk should not interfere with sensible business decisions. This does not
mean that the risk is to be ignored; managers must adapt to whatever the
rate is, while simultaneously providing guidelines for long-term planning-
a tricky balancing act.
Maneuvering liquid assets provides liquidity where it is wanted, but the
main motivation is that tax jurisdictions differ. Whereas a domestic firm
is stuck with one tax environment, a multinational manager can exercise
some discretion. This presents an awkward dilemma. Many people take a
lifetime to gain a thorough understanding of the tax system of one nation.
It follows obviously that a multinational executive cannot understand the
tax systems of 50 or 100 nations. Faced with this challenge some managers
give up and leave taxes to the tax specialists. Their decision analyses may
be precise - but they will be precisely wrong. The essence of maneuvering
liquid assets is that it is better to approximate taxes, to decide on the
managerially desired action, and to have the tax specialists check it to
revise the approximation. A normally operating company need take
interest in only a few corners of the tax codes of the world.
Chapter 4 does not take risk as an exogenous given; some of the risk is
what a corporation brings on itself by its actions and inactions. The
essence of the chapter is that any corporation is in coalition with other
powerful forces in a society. When a corporation suddenly grows or con-
tracts or quits, it disturbs the eqUilibrium of these forces, and they may
resist. National executives can manage these balances of force in an
intuitive manner.
Multinational managers might better supplement their political flair
with pedantic checklists and procedures. First, as a foreigner, the multi-
143
Digitized by Google
144 Multinational Financial Management
national executive lacks an intuitive feel for the situation. Second, the
stakeholders lack sympathy for the problems of a foreign corporation.
Third, the multinational's resources mean that its rate of expansion
or withdrawal can be much more sudden than the natural rise or fall of
indigenous companies. Finally, whereas an indigenous company has little
choice but to accept the political reality of its nation, a multinational can
choose where it wants to locate - and to choose well takes hard study.
Digitized by Google
PART TWO
Multinational Manufacturing
Digitized by
-==-="":-=~_ ___ . __c-=.=-.=--_c:-::.==:-----
CHAPTER FIVE
Logistics
147
Digitized by Google
Ascendant Electric of England Ltd.
John Larson, Director of Physical Distribution at Ascendant
Electric of England, felt squeezed between a rock and a hard
place. During the late 1970s the value of the British pound
had risen because of North Sea oil and the government's tight
monetary policy, so the cost of Ascendant's products had
risen in terms of most world currencies. Simultaneously,
from Ascendant's sales manager in America, its most impor-
tant export market after Saudi Arabia, came request after
request that the company lower its prices or it would con-
tinue to lose market share to Japanese competitors. In June
1980 Larson was asked by Ascendant's managing director to
do whatever he could to reduce the logistics costs of shipping
the electrical generators from Ascendant's plant in Rugby,
England, to the V.S. West Coast.
The generators were transported by ship. Larson, while
visiting London, found himself in a conversation about
airfreight and had learned that Boeing 747 airfreighters are
fully laden on their flights from California to Britain, but
have excess space on their way back, for which airlines
charge lower backhaul rates. Larson wondered whether it
would be economical to airfreight Ascendant's generators
to California, and what negotiating strategy he could use.
148
Digitized by Google
- - - - :::c_:--::
__ __:-:-_ _
~ _
Logistics 149
he had to figure the cost himself:
Packing cost $8.10 per generator
U.K. inland freight $7.03 per generator
U.K. shipping tenninal cost $3.02 per generator
Vessels of the U.K. to U.S. West Coast Conference loaded
containers at the Port of Felixstowe on the English coast, and
passed through the Panama Canal before heading north to
Los Angeles. The conference rate for electrical generators
was $163/1000 kilograms, so the remaining out-of-pocket
costs were:
Ocean freight cost $31.78 per generator
Los Angeles tenninal charges $ 2.87 per generator
Out-of-pocket shipping payment $52.80 per generator
In addition to this out-of-pocket cost, Larson decided to es-
timate the opportunity cost of inventory in transit. Within As-
cendant, this was commonly called the "pipeline inventory."
At the Rugby works generators were being assembled so rapidly
that the once-weekly shipment to Los Angeles was only a
fraction of one day's production. Hence, no additional
inventory was ever accumulated at the Rugby plant, to be
eannarked for Los Angeles.
The conference steamship company was using fast 24-knot
vessels, and on first investigation it appeared that the inven-
tory transit time via the Panama Canal was only four weeks.
Subsequent analysis, however, revealed that the actual transit
time was longer. Palletizing 64 generators and packing a
container is not in itself time-consuming, but each job had
to take its place in the queue of work. Once packed, the con-
tainers were held awaiting transportation to Felixstowe. The
ships usually kept schedule, and in Los Angeles the con-
tainers were offloaded rapidly. But then they sat on the
dock, sometimes for days, until the customer took delivery.
The customer had to schedule the assembly of diesel
generator sets into his other manufacturing work, and had
fallen into the practice of taking delivery only as he needed
the generators, thus deferring the payment of customs duties
until he actually needed the goods.
Ten days after the container is landed portside the cus-
tomer is supposed to be charged demurrage at $40 per day.
However, the shipping companies did not charge demurrage
to Ascendant for fear that they would lose its business.
Nevertheless, there had never in the past been a need to
negotiate the tenns of sale around the question of precisely
when the customer took delivery. Larson wanted to make
sure that if the goods arrived in Los Angeles early, Ascendant
Digitized by Google
150 Multinational Manufacturi"ll
Digitized by Google
___ • I
-~-
----c=::-=-::::-:-::"_c:-.~_:'.____ . - - - - - - - - - . - - - .--
Logistics 151
Cost
- - . , . . . - - - - - - Cost per 100 kilograms
'-----------::P-:'-Iv-ot------Weight
point
EXHIBIT ~.1
Digitized by Google
152 Multinational Manufacturing
Digitized by Google
Introduction
Ask the manufacturing manager of a subsidiary to identify his most im-
portant problem, and he will likely respond by describing union relations
and worker morale. These are his problems, not the problems of head-
quarters. From a headquarters viewpoint hisjob is to enhance productivity
and meet schedules. He must notify headquarters when he is going to run
the risk of a strike, when he has problems meeting quality, or when any
other phenomena affect the global production network. If a subsidiary
production manager is having trouble coping, the rest of the network
can accommodate his erratic output, but it cannot solve his problems.
The three chapters of this section deal with the global production net-
work as a system. The chapters build on one another, with logistics
coming first. In a multinational corporation, the global logistics depart-
ment is never large and is rarely powerful. The importance of logistics
lies in the fact that if it is poorly managed the interaction between manu-
facturing and marketing will sour. The perfect logistics department is
staffed by unsung heroes.
The three sections of this chapter are sequenced to accord with the
Gruber, Mehta, and Vernon (1967) international product life cycle.
During the first phase the newly developed product is exported from
the home market. During the second phase the rate of product innovation
slows (the product becomes more mature) and a network of plants can be
built around the world. In the third and final phase for a long-established
product, the inefficient plants (usually those where the product was first
built) may be closed to rationalize production.
153
Digitized by Google
154 Multinational Manufacturing
8. + Import duty
9. + Repackaging if necessary
10. + Inspection at the sales outlet
11. + Delivered inventory
12. - Export incentives
13. - Duty drawbacks
14. + Quotas
15. + Customer uneasiness about products imported from this nation
Between most pairs of ports the flow is so small that a corporation often
signs a contract with a freight forwarder to take care of all the paperwork
and to consolidate the shipments of several corporations into one container.
An ocean container is a large box (8 X 8 X 40 feet or 8 X 8 X 20 feet on
routes with more dense cargo) which provides good protection from both
ocean spray and pilferage. Nevertheless, even if a freight forwarder issues
one bill for all his services, the cost elements still exist. Therefore, the
discussion will proceed as if the corporation handled its own paperwork,
its own consolidation, and as if it is shipping on a regular schedule, say
monthly.
Packing
Most cargo is packed into containers to reduce the cost of breakage,
spoilage, or theft; most shipping charges are based on volume, not weight,
unless the shipment is very dense. One of the advantages of standard
container sizes is that it has become worthwhile to analyze how to package
items so that you can squeeze the largest number of units into the con-
tainer. Nevertheless, to pack 2,000 cubic feet of cargo into an 8 X 8 X 40
foot container is considered to be excellent packing.
Digitized by Google
-~
,
--====-== .--. -~-----------
-
Logistics 155
Transportation Charge
Liner conferences are price cartels. They originated when steam replaced
sail and the overabundance of fast tonnage led to drastic price cutting
in efforts to gain or retain customers. In 1875 shipowners on the London-
Calcutta route agreed on rates for certain cargos so that they could all
show a profit. Liner cartel pricing was and is encouraged by governments.
The U.S. Justice Department exempts conferences from antitrust laws,
and in many nations government-owned liners are conference members.
There are now more than 300 conferences. All have rules that provide
for unified action to prevent outside competition, uniform rates for the
various classes of cargo, and limits on internal competition.
In the past, when an outside shipowner challenged a conference with
lower rates, he faced the menace of a "fighting ship." Such a ship, sub-
Digitized by Google
156 Multinational Manufacturing
Particular shippers may be given special concessions, which are nonnally secret,
either directly in the fonn of reduced rates or indirectly by classifying particular
cargo items into a lower rate class than that prescribed i~ the tariff; for example,
virtually identical cargoes from different shippelS may be classified differently on
the basis of slight differences in packaging or in technical description. (Liner Con-
ference System, 1968, p. 45)
Digitized by Google
Logistics 157
Import Duty
Each item in the product line has a customs classification in each nation
and hence a rate of import duty. For example, an item entering the
Digitized by Google
158 Multinational Manufacturing
Digitized by Google
Logistics 159
Digitized by Google
160 Multinational Manufacturing
Inventory
The appropriate inventory models to use in the subsidiary are those that
depict a stochastic lead time (a probability distribution between order
entry and receipt of the goods). The amount of inventory held on average
can decrease if:
1. The subsidiary sales are predictable and controllable.
2. Replenishment shipments are frequent.
3. The lead time is predictable.
4. Accurate packing is persistently expected.
5. There are several alternate factories producing the item, lest produc-
tion at one be disrupted.
6. Airfreight could be used to expedite emergency shipments.
7. There are several possible suppliers within the nation lest import per-
mits be suspended.
8. The subsidiary has enough managers who can spend time managing
inventory.
Discretionary inventory is kept at both the factory and the subsidiary.
Most inventory should be kept at the subsidiary. However, three offsetting
factors are the interest cost on paying logistics charges earlier, the possi-
bility that the corporation may suddenly quit this nation, and the possibil-
ity of having to reexport if a sister subsidiary has stocked out.
Digitized by Google
Logistics
the present values of these two cash flows (discounted using dual variables
as discussed in Chapter 3) is the dollar export incentive to be deducted
from this logistics cost calculation.
Drawbacks lJiiduction)
foreign submi23kiffibliki23 may be built kiXport order.
think of that wzz:ve l1beZ23ad for assembly
United States. ixclude electronic
in Taiwnx lJffierican Motoxz: hiwing its enginz123
assembled in Mexico.) The government charges duty only on the value
that was added abroad. The U.S. Customs Service interpretation of Item
807.00 states:
Each nation makes and administers its own rules on duty drawbacks. Such
drawback bnducted from cost (after
nnYffi:ed by the conh f'neordkeeping and and discoun tieg
duty drawbacb the time lag duty was paid
hime the drawbeed
Digitized e
162 Multinational Manufacturing
Conclusion
The sum of these 15 costs divided by the quantity shipped equals the
logistics cost from the factory to the sales subsidiary.
If the cost of capital is high, and if the corporation is comparing ship-
ment from a nearby factory with shipment from a cheaper factory on
the other side of the world, it may be worthwhile to compute the costs
more precisely by including the time value of money. Thus there are two
more considerations: the flow of actual cash outflows incurred in making
the shipment and how long the journey lasts. With large, infrequent
shipments, the interest accumulates on the inventory on the ill"St day
that it is sold. We would like to compute the total unit cost of the
item.
First, on a time line as in Exhibit 5.2, mark all the actual expenditures
incurred for this shipment. This pinpoints the costs relating to each ship-
ment of inventory. Assume that the sales subsidiary keeps its inventory
on a first-in, first-out basis. Work backwards through time to when the
shipment cleared customs (mark on the time line when the duty actually
had to be paid), and back to the time this shipment was manufactured. All
those cash outflows on the time line, future valued to the day of first
use, constitute the delivered cost of this shipment. Present value these
costs to one datum point; the day of first use is most convenient.
In the geocentric section it will be seen that a company might consider
chartering its own ships. If it does, a ship is so huge and the route between
subsidiaries will have so many legs that the time value of money should be
computed.
Ocean Inland
Pack freight freight Usage period
Store Duty store IIIIIIII
f~:':~ I'"' "'' '
i i" " " " " , • Time
EXHIBIT 5.2
Cash outlays to present value.
Digitized by Google
Second Phase: Subsidiary.Centered View of Logistics (Polycentric)
During the second phase of a product life cycle, plants are constructed
abroad. The marketing manager of each nation vies for a plant. In his
striving, he is motivated to underplay logistics.
Let us stand in the shoes of a marketing manager who has spent the best
part of his life building a loyal market he and his company can be proud
of. Wholesalers and retailers know the company sales force and the
standards it represents. Both company and customer resist temptations to
cheat since they know the company is committed to stay. Let us pursue
a medical image. A sales organization depends on products for its life
support system. Only if the supply is adequate can the sales organization
live. If the product supply fails, there is no way in which the company
can avoid a depletion of goodwill through its entire channel of distribu·
tion. Furthermore, interruption of supply gives hostile competitors the
opportunity to increase market share by signing up distributors.
One may reasonably predict that the sales manager abroad would prefer
to have a factory in the safety of his territory. To this end he will con-
vince himself and attempt to convince headquarters that his customers
feel vulnerable without a local source of supply. All governments encourage
local manufacture (import substitution), although the apparent serious-
ness of their urgings must be gauged with care. We might expect the sales
manager to pass on to headquarters full reports of such urging. We might
even suspect him of dropping involuntary clues to the government agency
to state their requirements in the strongest diplomatic language. We might
further expect the sales manager to magnify the significance of corporate
errors. Of the thousands of containers he may receive in a year there will
inevitably be errors in packing and in the delivery schedule. Instead of
coping with these, we may expect him to exaggerate some of the difficulties.
It is said that every senior manager at one time or another has had to
put his job on the line and threaten to resign unless his request is met. We
may expect that the sales manager will time his threats to coincide with
occasions when the corporation depends on his profit contribution. In
the face of such pressures and in the absence of a strong analytical
staff to assert cost analyses to the contrary, a polycentric corporation
will have a proliferation of small local plants and hence international
logistics will not be a major corporate concern. Those few items that are
shipped between nations will be those on which no major sales commit-
ment has been established and for which even a sales manager cannot
deny the economies of centralized manufacture.
In a polycentric corporation, the logistics manager is under constant
scrutiny. His every mistake is enlarged out of proportion by those who wish
he did not exist. New plants are being constructed and so logistics flow
patterns are being changed continuously. Worse yet, any delay in the
168
Digitized by Google
164 Multinational Manufacturing
In the final phase of the product life cycle, inefficient plants may be
closed. Production will be rationalized. If several product divisions are
shipping between two ports, it is advantageous to coordinate their ship-
ments to keep a steady flow of full containers. Specifically, the corpora-
tion may be able to bargain with the carriers for a price cut or justify a
staff study of cheaper means of shipping.
Planners in geocentric corporations foresee a substantial increase in
intercorporate shipments of subassemblies and products. For decades
governments of developing nations have required that multinationals
build a local factory to produce part of their product line as the price
of admission to their national market. Miniature plants still abound, but
cost pressure is forcing corporations to build for economies of scale. The
emerging outcome is a network of single-product plants, one per nation,
each exporting most of its production while importing the rest of the
product line from the network members. More intercorporate shipments
imply a busier and more important corporate traffic department. Are
traffic departments ready for the task ahead? Unfortunately not: Some
tend to see logistics not as decision making but as paperwork.
Most corporations draw a sharp distinction between sales and market-
ing. Some salesmen have the wrong personalities for marketing. A similar
distinction should be drawn between freight specialists such as the fonner
rate clerks and logistics planners. A geocentric corporation needs a logis-
tics planning staff. The staff has two tasks: to reduce costs between each
pair of ports and to record these costs (current and projected) as a set of
computer-based transportation tableaux (one per major product category)
that link all plants to all markets.
Digitized by Google
Logistics 165
Digitized by Google
166 Multir&4ti0r&41 Manufacturing
The telephone is always-literally always at hand. In the car, the bathtub, at the
exclusive restaurant with the permanently reserved table, the phone is the ship-
Digitized by Google
Logistics 167
owner's lifeline .... "All I want is five minutes advance notice the first time the
Japanese steel industry falls 5% short of its goal. Just so I can fix my fleet long
term before the bottom drops out of the market." ... This is a world in which a
telephone commitment is a bond, a different, hard, calculating world. (Forbes,
August 1970, p. 20.)
Digitized by Google
168 Multinational Manufacturing
Digitized by Google
PLANTS MARKETS
rs,
, S;,;
Supplie;,; J MaiFz;,;?
dem;,;;,;;,;
.,.".bIOls. requir;,;m;,;;,;z;,;
• • •
• • •
• • •
m Xm1 Xm2 Xmn Sm
Demand: D1 D2 D3 ••• Dn
EXHIBI'3:'
Each F'zutwork can be uomputer as a
tableaR4,
Fz69
Conclusions
The logistics function in a multinational corporation is rarely large, nor is
transportation cost more than a small percentage of product cost. Neverthe-
less, neither of these observations should beguile one into underestimating
the usefulness of a well-managed logistics department. The profitability
of some multinationals depends on their ability to separate marketing
from manufacturing. To do so with calm assurance necessitates a reliable
and responsive logistics system.
It is more common to see the logistics function scattered over several
rival jUrisdictions within the corporation, staffed by former rate clerks
who have never been motivated to see the whole logistics cost. The
unfortunate consequences are that too many small plants are built,
sensible export markets are never launched, and periodic stock-outs
allow rivals to gain a larger market share. The payoffs in logistics manage-
ment are as asymmetric as they are in ambulance care. Improve a good
ambulance service, and the gain is hardly noticeable. But many ambulance
services are poor, and what a damnable waste that is.
Digitized by Google
Logistics 171
Digitized by Google
172 Multinational Manufacturing
Bibliography to Chapter 5
Ball, Robert, "Volkswagen Gets a Much Needed Tune Up," Fortune, March 1972.
Chorley, Lord and O.C. Giles, Shipping Law, 7th ed. (London: Pitman, 1980).
Cizaskas, Albert C., "French Exporters Are Backed by Diversified Credit Program,"
IMF Survey, June 7, 1976, p. 162.
CufOey, C.F.H., Ocean Freights and Chartering (London: Staples Press, 1962).
Devanney, John W., Ill, V.M. Livanos and R.J. Stewart, Conference RIlte-Making and
the West Coast of South America, Technical Report 72-1, Commodity Trans-
portation and Economic Development Laboratory, Massachusetts Institute of
Technology, 1972.
Goss, Richard, "Shipping Conferences," Journal of Transport Economics and Policy,
Vol. 5 (May 1971), pp. 173-183.
Gruber, William H., Dileep Mehta and Raymond Vernon, "The R & D Factor in Inter-
national Trade and International Investment of United States Industries," Joul7UJl
of Political Economy, Vol. 74,No.l (1967),pp. 207-215.
Hill, C.J.S., An Introduction to the Law of Carriage of Goods by Sea (London: Stan-
ford Marine Press, 1974).
Jones, Roger M., "Imagination in Bulk for the Seventies," Jones, Bardelmeir, Clements
and Co. Ltd., Nassau, Bahamas, 1971.
Leach, Rodney, "International Freight Transport Enters a New Era," McKinsey
Quarterly. Vol. 14 (Winter 1969), pp. 3-47.
Manca, Plinio, International Maritime Law, European Transport Law, 19,Justikiestraat,
Antwerp, Belgium, 1971 (especially "Affreightment" in Volume 11).
Metaxas, Bas N., The Economics of Tramp Shipping (New York: Oxford University
Press, 1971).
Officer, Lawrence H., "Monopoly and Monopolistic Competition in the International
Transportation Industry," Western Economic Journal, Vol. 9, No. 2 (1971),
pp. 134-156.
Digitized by Google
Logistics 173
Terpstra, Vem,lntemational Marketing (New York: Holt, Rinehart and Winston, Inc.,
1972).
"The Billionaire Sealords," Forbes, Vot. 116, August 1, 1970, pp. 20-23.
The Liner Conference System, Report by the Secretariat of the Conference on Trade
and Development (UNCTAD), United Nations, New York, TD/B/C.4/38 (1968).
The Liner Conference System, Report by the Secretariat of the Conference on Trade
and Development (UNCTAD), United Nations, New York TD/B/C.4/62 Rev. 1
(1970) (Sales No. E70.70IID.9).
Williamson, OUver, Markets and Hierarchies (Englewood Cliffs, NJ: Prentice-Hall,
1975).
Digitized by Google
CHAPTER SIX
Production Smoothing
174
Digitized by Google
Production Smoothing 175
Digitized by Google
176 Multinational Manufacturing
A Matter of Economy
This vice president, as well as other officials connected with
American multinationals, would do well to take a close look
at the experience of Ford, which decided on integration in
1967. At the time, the company's rationale for so doing
seemed sound indeed: Integration would avoid unnecessarily
duplicating the amount-some $100 million -that it costs to
engineer and produce a new auto model. "There's no sense
spending that a dozen times or more for each country in
Europe, when you need spend it only once," explains Gordon
Guthrie, general sales manager of Ford's German subsidiary.
So Ford decided to produce just one European line in place
of the completely different cars that used to be turned out
by its British and German plants. And the single line began
to reduce costs in another way, since the company began to
buy parts in bigger volumes - meaning lower prices - from
its outside suppliers.
Along with its integration of production, Ford also moved
to shift its marketing emphases from countries like Britain,
where auto sales were languishing in the 1960s due to govern-
ment credit controls, to countries like France and Italy,
where sales were growing. "In countries where markets were
static, we had some of our strongest and most imaginative
management teams; in other countries where the opportunities
for expansion were greater, we had much smaller resources,"
Stanley Gillen, then chairman of the integrated operation
(called Ford of Europe), told a company management meet-
ing in 1971.
Digitized by Google
Production Smoothing 177
Digitized by Google
178 Multinational Manufacturi1l/l
Digitized by Google
Production Smoothing 179
created ill-will. And there have been bitter fights over pricing,
with Ford of Europe pushing higher prices to maximize
profits at the same time that some operating companies were
trying to price their cars lower to be more competitive in
the market.
In fact, Ford of Europe, which was supposed to improve
communications, has instead created some bottlenecks of its
own. "Everything, including hiring a single worker, has to go
to Warley (Ford of Europe headquarters) for approval, and
it can take forever," says Heinz Allrup, who represents
workers at Ford of Germany plants. He cites the case of a
foreman who died six months ago. "We just got permission
to fill his job," Mr. Allrup says.
Another example: One day last year a Ford of Europe man
walked into a German plant and announced that a department
employing over 100 men would be closed. A worker's repre-
sentative got on the phone to Hans-Adolf Barthelmeh, then
chairman of Ford of Germany, and asked what was going
on. It was the first time Mr. Barthelmeh had heard of the
closing. (The shutdown was later rescinded.)
"General Animosity"
Hans-Adolf Barthelmeh is no longer with Ford of Europe,
nor are several other executives who chafed at the integrated
operation's various communications gaps and arbitrary trans-
fers. But integration has also caused disaffection at nonexecu-
tive levels. "There is a general animosity toward Ford of
Europe," says Mr. Allrup, the German workers'representative.
The reasons for this animosity are multifaceted. German
workers, for example, are angry over layoffs stemming from
British strikes. (Ford's Cologne plant recently said it would
layoff 4,450 workers in the weeks ahead, with one of the
cited reasons being the British three-day workweek.) British
workers, on the other hand, are fearful of losing jobs to
Germany.
But Ford of Europe's problems go beyond its work force.
Take, for example, the case of its marketing plans. The
German subsidiary was to have fed cars into France and
Italy; however, Ford officials say they have found it diffi-
cult to sell in France, since they are competing against
government-owned Renault, which doesn't need to make
any profit and can thus cut prices to the bone. And Ford
says it's also difficult to sell in Italy, where Fiat, which
specializes in the tiny cars preferred by Italians, holds a
Digitized by Google
180 Multinational Manufacturing
Questions
1. It would appear that the launch of the Taunus in Germany
should have been delayed one year. What explicit business
plan might have been followed to launch it in 1972?
2. One gets the impression that Ford bit off more than it
could digest in 1969 with this single-stage integration.
Layout a two-stage integration; state the activities you
would put in stage one and how long it would last.
3. In the years since this article was written, Ford has sys-
tematically resolved most of these problems and is now
accruing the benefits of European rationalization. Assume
that the confusion of reorganization would take three
Digitized by Google
--- - - ---=---------_. . . . . . .
Digitized by Google
Introduction
The buffer that stands between marketing decisions and production deci-
sions is a good logistics network and well-positioned inventories. Marketing
and production executives depend on one another; only the logistics/in-
ventory system separates them. Not only must their interface function
efficiently, but it also must have the resiliency to absorb market mistakes
and factory errors.
In a multinational corporation the logistics/inventory system is complex.
A factory does not know with certainty where its products are destined
to be used. A national marketing manager cannot be sure where his next
shipment will come from. A telex from a marketing manager may be
understood by him but incomprehensible when received at a factory in
another nation. And worst of all are the messages that the recipient
thinks he understands when he doesn't. Given the infrequency of ship-
ments, a marketing manager may wait more than a month before his
orders arrive, the containel'l are opened, and the error is realized. If
factories transship subassemblies, then a strike in one factory will strain,
and can impair, the worldwide system.
In each nation, production-smoothing problems and opportunities
are different. A unique feature of a multinational corporation is that
headquarters can alter the flows of products between nations to smooth
production. The subject of this chapter is the reallocation of markets
to factories to achieve optimal production smoothing for a multiproduct,
multifactory corporation. We will follow the three stages of an international
product life cycle. In the first phase, perception of the problem is dom-
inated by technical considerations so that a product focus should be
expected even if carried to ethnocentric extremes. Later the corporation
will build plants in many nations, each of which will be operated some-
what independently. Finally there will be enough confidence and cost
pressure to encourage geocentric rationalization.
182
Digitized by Google
Production Smoothing 183
will commit a pennanent sales force and well-nurtured brand name and
those to which it considers export not a pennanent commitment.
Some products are extremely difficult to manufacture. New petro-
chemicals, phannaceuticals, and integrated circuits pose incredible produc-
tion engineering problems. After the production process has been explored
in a pilot plant, the design engineers scale up to a commercial plant.
Unfortunately the first plant rarely works as designed. A carefully de-
signed sequence of test runs is needed to identify all problems, and
subsequent debottlenecking takes anywhere from a few months to several
years. After systematic test runs have been perfonned, the results can be
analyzed to give the design engineers more accurate parameters for their
next plant design. This means that the design of the early plants will
unavoidably become obsolete. In most plants there are substantial
economies of scale. Hence the planning dilemma is how large to dare
design the early plants, knowing that obsolescence will cut short their
economic life. In tenns of a trajectory of total plant capacity through
time, there is a critical rate of capacity growth. A multinational that
builds faster than the critical rate will accumulate obsolete large plants.
A diffusion of new and innovative products - products that customers
are learning to use-is underway. Nobody should expect new product
diffusion to be predictable: Any demand forecast is little better than a
guess. If the product is good, sales will rise, but erratically. If sales rise so
rapidly they exceed plant capacity, stockouts will occur. Stockouts during
the launch of a product have two consequences. First, stockouts will slow
the subsequent diffusion by raising in the minds .of potential customers
doubts about whether to take the trouble to adapt. Second, stockouts dur-
ing the launch phase of a product invite competitors who can get a free ride
atop the accumulated goodwill of the innovator's advertising and mis-
sionary sales effort. In summary, capacity should not be built faster than
the critical rate of growth. Actual demand should be built up more slowly
because sales forecasts are inaccurate and stockouts are expensive.
In this situation, the great advantage of a multinational corporation is
that it maintains a reserve of markets in which it has not yet launched the
product. When a giant new plant proves itself on stream, the corporation
can utilize any excess capacity by selecting appropriately sized nations
in which to launch the product. To tailor demand to fit the capacity avail-
able, the most suitable national markets are small but well developed,
and hence the product can rapidly achieve its sales potential. After each
major plant proves itself, the corporation can select a few more markets in
which to use whatever capacity is available. One suitable decision display
is a graph showing the ultimate sales potential of the product on one axis
and the percentage penetration of the market within one year on the
other (Exhibit 6.1). Each nation is a point on this graph. Confidence
bands drawn around the estimates of sales and penetration transfonn the
point estimate into a rectangle of plus and minus one standard deviation.
Digitized by Google
184 Multinational Manufacturing
100~-----------------------------
Percent
market U.S.
penetration West Coast
In one •
year
•
Japan
nce
•
o Sales potential
units/year
EXHIBIT 8.1
Sales potential versus market penetration indicates which markets are
small and predictable enough to fill a plant.
Digitized by Google
Production Smoothing 185
Inventory
Now consider an extra unit of demand arising during the peak period. The
is alreadd pmducing r;;;;bacity, unit will to be '~."'~"·"U'I:l.n
a time (:arried in iF:m4:,ntoryF:?eeded.
incremental cosf will be fhe inventory carrying cost hach to the produc-
tion period, plus what then was the cost of manufacture. If an earlier
period has a lower cost of manufacture, it may be economical to produce
the carryinh eost is Invenbfhh r'arry-
Overtime
Inventory cnrts nlay be hldr]r4"d by overtime~ A miniInmn rost
bfh;f;;;;icehn;tween
oveihin~,r is with wages k%wer .. ~ ..•.~•.~.,.~.
caused by worker fatigue. The cost and capacity of overtime also depends
on the plant's relationship with its union. If the wage level of workers is
already high, workers may be unwilling to work much overtime. If the
nnmhany labor and could On c%ther
some maintfbn hnod relationchip which can
ask the employees to work longer hours at peak demand, repaying the
workers with favors such as extra time off during slack periods. Overtime
is usually less of a problem if there is local unemployment, wages are
un:WCl~ece:C1alil~enlrrlC nnlations based on problen:l
of confmntation.
Layoffs
IthDugh are an lklI1LtlIl*"diate to pmPndion sm%x%thing,
dtendant p:mhlems it is r"~"',,,,,,h
nn;ait worh;;;;rc retirinh, fireh, puittinh rnduce the c'cmk-
force. Retirement need not be automatic at 65; workers over the formal
186 Multinational Manufacturing
Digitized by Google
Production Smoothing 187
Maintenance
Operating managers tend to underinvest in maintenance; thus common
corporate practice is to set up a maintenance fund for each machine.
Every unit of output pays a bounty to the fund. The maintenance fund
must be spent by the end of the fiscal year or else the budget authorization
expires. This practice assures that maintenance is accomplished, yet the
practice is highly dysfunctional from a production-smoothing viewpoint.
During a boom year it makes little sense to shut down the plant for an
overhaul simply to prevent the money from reverting to headquarters.
Conversely, during a slack year the fund is too small for a major over-
haul, even though production will not be missed and the quality and
alacrity of contracted help is superior. It is therefore imperative to lengthen
the cycle of each maintenance fund from one year to a full business cycle,
so that maintenance is scheduled for slack periods. To do this requires
study of the cost of delaying maintenance, in essence an analysis of how
near to flat the cost curve is in the region of the optimum maintenance
schedule.
Marketing
Usually sales departments receive goods at the standard accounting cost.
However, the current actual cost of each item fluctuates through time.
Were sales to receive goods at their fluctuating actual cost or its forecast,
the sales department would have an incentive to delay or advance the pro-
motion of sales. This would ease production-smoothing peaks, even
though it complicates the analysis.
Mathematical Techniques
Most real production planning models are large linear programs that embody
the cost of having workers switch from one manufactured item to another.
A mathematical model of production smoothing requires as input a forecast
of each month's required shipment of each item. The model computes
the lowest cost of achieving this output. Furthermore, its calculations
generate the incremental cost of producing each item in each period. This
incremental cost is the cost of augmenting the quantity demanded of one
item in one month leaving all other demands unchanged; all inventories,
overtime, maintenance schedules, and so on, are adjusted. The incremental
costs will be needed in geocentric production smoothing.
Digitized by Google
188 Multinational Manufacturing
markets and use recently built excess production capacity. Product divi-
sions existing within the corporation ensure organizational control and
coordination. The blend of product and national management provides
the matrix framework in which geocentric production smoothing can
occur. If several nations have at least one product in common (Exhibit
6.2), geocentric smoothing can allow for flexibility in both marketing.
and production of all products. For common products, markets can be
reallocated from busy to slack plants. This reallocation in turn eases the
multi-item production-smoothing problem within each plant.
Some sales territories are on a border between plants. If a marketing
manager's sales territory is between two or more plants, the marketing
manager may be able to act entrepreneurially. Assuming the item is
identical from each factory, he can buy from the factory that will deliver
more cheaply. Since production planning cannot begin until demand has
been estimated, it might appear that the border marketing manager's
entrepreneurship would complicate production smoothing. However, the
border marketing manager and the headquarters production planner have
a strong basis for agreement. From a corporate viewpoint, plants that are
busy should withdraw from border territories, leaving those sales to plants
with excess capacity. If a plant is operating near capacity, overtime will
be required for incremental production, raising the per unit real cost. Each
plant can deliver to the border at its incremental production cost plus the
logistics cost analyzed in Chapter 5. This sum of logistics plus incremental
production cost provides a clear guide for redistributing markets to plants.
Requiring that some items be of standard design to allow smoothing
does not mean that production methods and plant layouts must be
standardized. Quite the contrary. In sports, different positions on the
EXHIBIT 6.2
An item produced in two or more plants can be smoothed simultaneously.
Digitized by Google
Production Smoothi1llJ 189
same team all share the same "goal"; it is their methods of accomplishing
that goal that differ. Similarly, the corporate ability to smooth production
is enhanced if there are four distinct types of factories: lowest unit cost,
flexible multi-item, seasonal, and stockpile sites. Producing an item in
several of these factories simultaneously allows great flexibility. The bulk
of worldwide demand for a long-life cycle product can be produced by
the lowest unit cost factory, while the peaks and valleys of the demand
curve are satisfied by seasonal plants (Exhibit 6.3).
National labor forces differ significantly and it may be sensible to match
the labor force to the type of factory. The four types of factories are
operated in very different styles. The selection of plant managers, the
design of incentive systems, and the production process itself all depend
on the type of plant necessary and the workers available.
Total
Low unit cost
cost factory
Flexible factory
EXHIBIT 6.3
A comparison of low unit cost factory and flexible factory.
Digitized by Google
190 Multinational Manufacturing
Lowest unit cost plants generally take one of two extreme forms:
1. Highly automated, perhaps with government subsidized capital.
2. Labor intensive, where wages are very low.
The design of a capital-intensive plant is primarily an engineering job.
The equipment must be correctly situated, installed, and adjusted to run
at a coordinated speed, usually at full capacity. A cut in production would
raise the unit cost of overhead and capitalization. Highly automated plants
are located near their primary market, usually in an industrial nation.
Since the investment is long term, political and economic stability must be
considered. Surprisingly, local wages and union practices do not play a
vital role in plant location because the amount of labor required is small;
however, the availability of trained mechanics and craftsmen is crucial. A
well-established source of materials and manufacturing infrastructure is
also important.
The United States is perhaps the most popular country in which to
locate a capital-intensive factory. The domestic market is large yet easily
accessible through a sophisticated distribution network. Although unions
are powerful and wages are high, highly skilled mechanics and engineers
are readily available. Labor relations are improving, with the number of in-
dustrial disputes leveling off and the number of lost working days drop-
ping. The United States is politically and economically stable, with the
risk of nationalization minimal. Many multinational corporations' head-
quarters are in the United States, shortening lines of communications
between them and their large production facilities, thereby ensuring ade-
quate supervision of major investments.
For a labor-intensive low unit cost factory, the level of wages is a crucial
determinant of plant location. High labor intensity normally dictates a
location in an underdeveloped nation with very low daily wages. The
levels of industrialization and employment as well as the trend of social
reform within the country are important, because historically, wage level
rises as a country becomes more developed. The relatively low capital
investment required is at risk only if the payback period is long; usually
it is short.
Once the criterion of low wages is met, other characteristics of the labor
force must be examined. The strengths and practices of local unions are
important. For example, a low unit cost factory could work well within
a stable, well-defined relationship with a trade union. On the other hand,
if a union were too powerful, wages and productivity might be managed
to the detriment of the company.
A labor-intensive lowest unit cost factory poses different challenges
in design and management. Labor management in a labor-intensive
assembly line factory places a great emphasis on worker productivity.
Incentive systems and steady employment reward those who produce.
Hence, for motivational reasons, it is not prudent to carry a large inven-
Digitized by Google
Production Smoothing 191
tory of finished items at the plant. Because most jobs will be long term
and routine, the company has to invest in worker training. To minimize
the risk of losing a trained employee, turnover and layoffs must be avoided.
A normal production load of one shift allows flexibility to work systematic
overtime when extra output is needed for production smoothing.
As one example among many, the Philippines appears to be an attractive
location for a labor-intensive low unit cost factory. Although nearly 50
percent of the working population is employed in primary industries, the
manufacturing sector accounts for over 11 percent of workers and is grow-
ing. Philippine workers are mostly literate and so can adapt to manu-
facturing work. Productivity per employee has been rising rapidly. Wages
in the Philippines averaged $50 per month in 1975 and are rising
moderately. Unions exist and have the right to collective bargaining, but
recent changes in the government have made both strikes and lockouts
illegal. Compulsory arbitration is provided for, if no agreement can be
reached. Layoffs require severance pay and permission of the government
but are not impossible. Overtime is not difficult to arrange. These factors
combine to create a favorable situation for a labor-intensive low unit cost
factory that is a stable employer. In the Philippines martial law was
declared in 1972. The political situation within the country and through-
out the area has worsened since. A large capital investment in this area
might be unwise, but a low capital cost factory that would take advantage
of the low wages would have a short enough payback period that the
corporation could ignore long-run uncertainties.
To build a low-cost factory, the corporation has to make a substantial
investment - whether in automated machinery or in worker training.
When the factory is asked to phase out a declining product and phase in
a new one, the investment becomes vulnerable because managers and
workers of a low-cost factory have had little need to develop transitional
skills. Headquarters may have to send patient and sensitive experts to help
each transition.
Flexible Factory
Visualize a job shop. A flexible factory is a multi-item operation, with
flexibility of material flow, general purpose equipment (perhaps numer-
ically controlled), and a completely literate work force managed with an
emphasis on job enrichment. The challenge of managing a flexible factory
is that there cannot be a tight cost control system. For example, a standard
costing system provides a very reasonable ceiling above which cost should
not rise. But the attempt to minimize accounting cost is incompatible
with the need to change schedules and even interrupt production runs-
after all, the flexible factory was built to ensure tight time control.
Labor management in a flexible factory also has to be run without tight
control. Each worker has to be creative and imaginative in developing
Digitized by Google
192 Multinational Manufacturi1llJ
transitional skills. Cultural values and social status in many nations accom-
pany the individual's creative pride in knowledge of a craft. Participative
management can flourish in this type of atmosphere. Since a great deal of
autonomy must be granted, it is essential that each worker know his own
and his fellow group members' productivity. Worker selection and pro-
bationary review are crucial to keep slackers from contaminating the work
atmosphere. This requires that the foremen and superintendents have the
integrity and self-discipline to dismiss workers for poor performance. The
corollary is that although there may be dismissal for cause, there should
be no layoffs once workers have passed probation, except under the most
adverse circumstances.
All new items are produced in a flexible factory. All unforeseen diffi-
culties or problems are referred here. Although not all items would be
produced in a flexible factory, the factory has the capability to begin
production of any item within a short start-up time. Such standby capability
has its cost. The benefit is that the corporation can then dare to build low
unit cost factories in some of the impoverished and unstable nations of
the world. Then if a risky plant is lost anywhere in the world the com-
bination of the output from the flexible factory, stockpiles, and global
overtime can meet customer needs until a replacement plant can be rushed
into production.
A flexible factory requires a variety of skills and high degree of training;
thus one can best operate in a country with an established manufacturing
base and a well-educated labor force. The manufacturing base can provide
purchased components and material, and the educated labor force will
ensure adaptable and competent workers. For example, Ireland and
Japan both provide ideal environments for flexible factories, but for
different reasons.
The Republic of Ireland has a moderate but rapidly expanding manu-
facturing base, and it is within the European Economic Community
tariff barrier. Although the Irish domestic market is small, bulk transpor-
tation is readily available to the United States and Europe. Irish workers
have been ready to learn, and pockets of skilled labor throughout the
country have been developed through apprenticeship and training pro-
grams. Over 20 percent of the labor force is in the manufacturing sector,
with the rest scattered between primary and service industries. The
Irish government provides capital grants and loans for the building of
factories and the training of employees. The government imposes a
system of levy grants whereby a company receives a refund for training
workers. Since training is a major expense in setting up a flexible factory,
Ireland provides an attractive opportunity.
Union membership is entirely voluntary, and over 50 percent of the
Irish workforce is unionized. In turn, 95 percent of Irish unions belong
to the Irish Congress of Trade Unions (ICTU). In the early 1970s, unions
Digitized by Google
Production Smoothing 193
Seasonal Factory
Because it hurts to layoff workers, managers are tempted to delay action
by biasing upward the sales forecast. Since such delay results in an even
deeper layoff, it seems psychologically sensible that the layoff decision
should not be made in the factory itself. The scapegoat has to come from
outside - from a foreign headquarters, for example.
When several factories produce the same item, some can avoid a layoff
if others layoff more frequently and/or cut deeper. Can there be a
layoff factory designed for intermittent production? It would be socially
more acceptable if this were a seasonal factory. Whenever the corporation
has some items with a seasonal demand, it is worth evaluating a seasonal
plant. Such a plant is designed to be operated only a few months of
the year, frequently using rented equipment and old tools, dies, and
assembly jigs. The location criterion is to find a community accustomed
to seasonal work, whose pattern would be complemented by this plant.
In poor years the plant can be operated for only a month or so (though
to maintain a work force it is crucial to open the plant annually). In
Digitized by Google
Multinational
good years the plant can be run with large overtime for several months or
up until the community's next season. The personnel policies must accord
with this pattern, with overtime and layoffs commonplace.
Communities such as fishing villages, which are accustomed to seasonal
work, are easy to find. Finding a country in which seasonal labor is a
nntinnal phenomennn more difficult~ (::)untry is Finlnnd:
))h):re both men participate force, with
IInwever, due to and the produd
strong seasonll in the demn:nh
labor, especinUU r:(:rtdrrn Finland. rnpply also fluntn~
ates, dropping on average 10 percent during the winter. Labor disputes
are on par with other European countries, but workers, especially in the
north, are in a weak bargaining position. Because seasonal work is both
socially and economically accepted, unions are less likely to press inflexible
contracts.
The Finnish an'I1:~~r::::1:::~~:: like many other tries to repln(:n
t::~:T:1:1: horary or seasm::ni n:ith more nnni::pations. Seasnnll
however, opportunity to i:rzl00:i:n out the peaks
of I1n,"1YI,n::::::::::::~::f is in the inte)ertr the governmoHt
nOi:n:o:rili work when is at its
IIherefore, although Hi:ork may not by a
,","'''lA&::::
Stockpile Factory
A careful study of the income tax and property tax laws of every nation
reveal that SOilie subsidize that a subsidieilih
have a low holding cost. in Denmark
rOl)1:~1Ll component iT:wentory carryjnh percent of
(:orporation's cost This pleasant zili~T:'£:::es about beCH:1Hl
tax authoritie'£:: deduction of Ud of the
inventory value in computing taxable income for any year. Given a Danish
tax rate of 37 percent, the savings is 11.1 percent of the inventory value.
The allowance deducted in one year must be added back to current
taxable income in the following year. In other words, part of the corporate
tax is deferred one year. In Sweden the incentives are even greater. With a
,£:::,:nitu-down of 60 ::Hnwed and a hO percent, the
carrying to only 64 the corporatioH
of capital. Othnr tHH01rnments provido n:ili=ohousing and suh'£::i~
loans againrt Negotiation:r possible oecn1:1:r:r;:
governments employment:
Tax incentives are only one part of the cost of carrying inventory. The
multinational has some of its own capital tied up in inventory. As dis-
cussed in Chapter 3, a corporation's cost of capital may be different in
-----------..............======~.- =,--"----
different nations but tax costs restrain the corporation from moving the
funds. These funds are nevertheless available for inventory. Exhibit 6.4
shows that if inventory is to be held longer than the break-even time,
it would be preferable to accumulate the stockpile where holding costs
are lowest.
A stockpile factory may be very small. Consider a factory that produces
one-tenth of the corporation's worldwide demand for an item. Suppose
that on average the corporation keeps two months of inventory. The
factory would spend almost two years stockpiling its production before
the corporation would double its normal world inventory. Hence, like a
lowest unit cost factory, a stockpile factory should be assigned no near-
obsolete or faddish items.
Purchasing
It is unfortunate that the world view of most corporations is dominated
by the export orientation of sales managers and that most purchasing is
done domestically. A very different evaluation of national risk and oppor-
tunity for the corporation would emerge from a global purchasing per-
spective.
Great savings are possible by the systematic and skillful use of inter-
national purchasing. For example, most centrally planned economies
require an inordinate amount of red tape before goods of a capitalist
corporation can be imported. Exports are quite different. In centrally
planned economies there are sporadic national shortages of key currencies,
at which times a capitalist corporate purchasing manager could negotiate
a very attractive deal.
Cost
$Iunit
Ordinary factory
Stockpile factory
Production
cost
o Age of inventory
EXHIBIT 6.4
Production cost plus inventory holding cost.
Digitized by Google
196 Multinational Manufacturing
Digitized by Google
market, held at a medieval market town near a modern airport. Marketing
managers from around the world would fly in to deal with production
schedulers from factories around the world. They would haggle over
price and quantity. Such an open market would encourage initiative
and autonomy.
IJnfortunately, with such markets would
hi~tivel expense~ problem is the of deals that
nptimal to th(: unless a forced to
ELame price to (f.o.b. the le:(tory would tenh
tiharge higher markets distant factonh
competition by their higher logistics costs). This would upset comparative
accounting systems and could lead to a harmful split between marketing
and production managers. Furthermore, if only one factory supplied
a particular product, the pricing relationship would be similar to bilateral
monopoly. To avoid these problems, it is prudent to computerize the
2L2L:0ELket and thus to ELtandardize policies hn)h)~ction~moothhnh
hntie,am that alloceln) to plants manner (withjxl
bounds).
Model
Headquarters is interested in two cost components: the worldwide logistics
cost (developed in Chapter 5) and the production costs totaled over all
the factories. For any given production allocation, the sum of these two
costs gives the total cost of output.
Let us first look at just two factories. Given a worldwide demand in 100
each with logistics cost how should
,§Uf"e(,te productio:n plants? Exhihif how total c:)hi
he minimized hELoduces 60 B produces
f£:iI[~:0tillt of producJinn~ IDllowing paragrahhEL DDplain how th'::D(
are estimated Jhe smoothing by computer
many plants.
TC rc
Production
'-EL:- 2L2L----~
~-J optimal
of markets
E:;1J
t 4 - 0 - - - - - - - - - - 100%
100% · · - - - -....
Plant B " • Plant A
EXHIBIT 6.5
Production smoothing for two plants.
Digitized e
198 Multinational Manufacturing
Digitized by Google
Production Smoothing 199
/----- , .........
, ---- "\
,,
Total
/
~:;>/' ,
production '\
cost
Stow Sr., old
~--:~
Slow I
+ pos. - neg. shipment
slack slack 4
I /
I
L..-_ _L...-~'----I~_ q ...... ...- / '
-10% t +10%
old--- - - - -
_-------
shipment
schedule
Production smoothing Global logistics model in
model in subsidiary headquarters
EXHIBIT 6.6
Marginal costs are transmitted from each manufacturing subsidiary to
headquarters.
transportation tableaus, and for each factory
new shipment =old shipment + negative slack - positive slack
Enter the FIT new shipments into the right-hand side of the transporta-
tion tableau and begin a new iteration.
Step 2: Back to Each Factory. Telex the IT new shipments to each
factory, and ask that each run its local production-smoothing model. A
factory can use a linear program, a linear decision rule, a quadratic pro-
gram, or an intelligent cost accountant-as long as the solution to the
model yields marginal costs (the change in total cost if all other ship-
ments remained unchanged except the incremental change in the one
shipment under review).
The factory then telexes its IT marginal costs to headquarters. It supple-
ments this telex with possible warnings about limits on the extent to
which shipments could be increased, buttressing its argument with data
on what the marginal costs would become.
Step 3: Back at Headquarters. In each transportation tableau, update the
availability figure (to the tentative shipment schedule that emerged from
Step 1) and the costs on the positive and negative slacks (to the marginal
costs from the factories, add and subtract a little, as in Figure 6.6). Solve
the IT transportation problems.
After a few iterations as described here, the tatonnement process
stabilizes. Specifically, both slacks will be zero for all items, signifying
that no changes in shipments appear indicated. At this point reduce the
tolerances on the slacks (say from 10 percent, as in Figure 6.6, to 5
percent) and correspondingly reduce the estimated alteration to the
marginal cost.
Digitized by Google
200 Multinational Manufacturing
tentative shipment
Step 1
schedule
Step 2
Production
smooth
marginal cost
Step 3
Reallocate
production
over all
plants revised shipment
schedule
Production
smooth
marginal cost
Discover
no
reallocation
necessary
Reduce
tolerance
from 10%
to 5%
slightly revised
shipment schedule
EXHIBIT 6.7
Computational scheme.
Digitized by Google
----=-~-,=_c_:=_~ __ .....
_==:-:==-~---'
Human Problems
Massey-Ferguson had a problem with integrated logistics because market-
ing managers persistently inflated their forecasts of demand in the
interval from two to six months in the future. Their long-run forecasts
were unbiased. Their one-month forecasts were very accurate, But their
intermediate forecasts were persistently optimistic. At Massey-Ferguson
this became known as a "bow wave" forecast. The bureaucratic rem-
edy would be that headquarters scale down the received forecast. The
economic remedy would be that the marketing manager be charged
a penalty for changes he makes to his forecast, the charge being greatest
for the near term (the appropriate charges can be found by rerunning
the production-smoothing problem).
Some markets are appreciably less predictable than others. To accom-
modate that randomneBB the marketing manager is encouraged to keep a
local inventory (and bear its carrying costs) so that shipments are first in,
first out (FIFO) for that inventory. To compute the target size of that
inventory, an analyst needs, as input, the penalty for changes to the fore-
cast. Thus the penalty charge intended to avoid "bow wave" forecasts
also gives the value of more accurate marketing research, and of market-
ing methods to stabilize demand.
The final adjustment may be required in logistics. From each factory
to each market, sum all items of all the product divisions to be shipped
in each month. Add needed flows of subassemblies. With these total
logistics flows, estimate the logistics costs on each route. If a chartered
vessel now has an uneven carrying pattern, it may be neceBBary to revise
the logistics costs in certain transportation tableaus and reiterate the
problem.
At this time let us mention problems of administration.
First, if some subsidiaries are late in responding with their marginal costs,
the headquarters can nevertheleBB proceed to rerun logistics with what-
ever partial information is on hand.
Second, some subsidiaries, curious to learn how to game the system,
may report back false marginal costs in an effort to get the shipments they
want. To thwart this, the headquarters computer can store for each
factory its past history of IT shipments and IT marginal costs. The statis-
tical analysis of these data can give headquarters estimates of changes in
marginal cost if shipments are higher or lower, a basis for conducting
exploratory staff analyses without disturbing the factory, as well as a
means to detect possible errors in the cost feedback from the factories.
Finally, if one factory's marginal costs become very high because of
high wage settlements and currency revaluation, the factory will be asked
to make fewer shipments. UnleBB headquarters is planning to phase out the
factory, local management must improve productivity. To this end, they
Digitized by Google
202 Multinational Manufacturing
need an explicit target of unit costs toward which they must aim. This
analysis provides a basis for setting explicit targets.
Conclusions
In summary, certain markets will be flip-flopped between factories to ease
production smoothing. Rather than identify them ahead of time, we
allowed them to be identified on economic grounds. Think of the border
that divides the districts of each plant. In each period we may move the
border because it is defined as the point at which it is equally expensive
to deliver from either adjacent plant. The purpose of production smooth-
ing is to respond to fluctuating demands for items. The advantage of being
explicit about the cost of absorbing demand fluctuations is that it makes
explicit the advantage to adjust other parts of the global system.
Governments are concerned with employment stability and are aware of
all layoffs. The foreignness of a multinational means that its layoffs will
attract all the more attention. Thus any multinational must administer
layoffs with an eye to the long-run stakeholder concerns discussed in
Chapter 4. Offsetting this vulnerability is the fact that a multinational can
choose where to make the layoff. Demand fluctuations mean that
production-smoothing activities will be done somewhere. The essence of
geocentric production smoothing is to orchestrate where to do which
activities. In some nations the corporation will disquiet some stakeholders,
but in no nations should it cross their irritation threshold.
Digitized by Google
Production Smoothing 203
quarters who are assembling and computing the production plans
have a frightening responsibility. Their calculations determine who
will feel the grip of layoff and which divisional managers will have
their relationship with their community damaged by a layoff. Is it
better that the headquarters production planners be cool technocrats
or that they blister in the emotional ravages of their acts?
6. Corporate Emotional Normative Subsidiary. You are the production
manager in one of the subsidiaries where costs are such that employ-
ment levels vary widely. You have only a few products, so there is no
hope of transferring workers between products. For these products
the corporation builds a new plant somewhere in the world about
every two years, and so you have been experiencing a two-year
sawtooth cycle to your shipments in addition to the random fluctu-
ations of the world markets. Individuals who were foremen have been
demoted to workers and individuals who were workers have been
laid off. Write a one-page letter that job applicants can take home
with them, explaining these facts of life. Be aware that a strong union
could end your practice, and that although there is currently no
government legislation giving workers tenure after six months, there
is a national sentiment that workers have a loyalty to their employers
and employers should have a reciprocal loyalty to their employees.
7. Corporate Emotional Descriptive Global. Describe the kind of team
managerial policies you would expect to see if the production planning
team appears competent, engrossed in their work, and seemingly
effective.
8. Corporate Emotional Descriptive Subsidiary. You are a production
manager in a developing nation, struggling with all the problems this
entails. Someone at headquarters sends you this chapter, with the
implication that you will be included in the grand production smooth.
Identify accomplishments that may be undermined by such inclusion.
For each, how would you feel?
9. Social Rational Normative Global. As long as there is fluctuating
demand there will have to be some forms of production smoothing
(including rationing). List each of the cost coefficients for the one-
nation problem. Beside each coefficient indicate whether the social
cost coefficient is greater or lower than the corporate cost coefficient.
If there is fluctuating demand, from a social point of view which
kinds of nations would be least harmed by the fluctuations in
employment?
10. Societal Rational Normative Global. Suppose a national government
reduced the corporation's cost of carrying inventory in that nation
(for example, under the motivation of an oil stockpile, the govern-
ment loans money at low interest against warehouse receipts of all
kinds of energy-using product, broadly interpreted). Would the
indirect effect be to stabilize or destabilize employment?
Digitized by Google
204 MultiMtioMl Manufacturing
Bibliography to Chapter 6
Bergstrom, Guy L. and Bamard E. Smith, "Multi-Item Production P1anning- An
Extension of the HMMS Rules," Manogement Science, Vol. 16, No. 10 (1970),
pp. B614-629.
Comanor, William S. and Thomas A. Wiison, "The Effect of Advertising on Competi-
tion: A Survey," Journal of Economic Literature, Vol. 17 (June,1979), pp. 453-
76.
Crowston, Wallace B., Warren H. Hausman, and William R. Kampe 11, "Multistage
Production for Stochastic Seasonal Demand," Manogement Science, Vol. 19,
No. 8 (1973), pp. 924-935.
Digitized by Google
Production Smoothing 205
Holt, Charles, Franco Modigliani, James Muth, and Herbert Simon, PltJnning Production
Inventories and Work Force (Englewood Cliffs, N.J.: Prentice·HalI, 1960).
I.L.O., Yearbook of Labour Statistics (Geneva: International Labour Organization,
1976).
O.E.C.D., Reviews of Manpower and Social Policies, Manpower Policy in Japan, Vol.
11, 1973; ManpowerPolicy in IreltJnd, Vol. 15, 1974; Manpower Policy in FinltJnd,
Vol. 17,1977.
Topkis, Donald M., "Optimal Ordering and Rationing Policies in 8 Non Stationary
Dynamic Inventory Model with n Demand Classes," Manogement Science, Vot.
15, No. 3 (1968), pp. 160-176.
Welam, Utf Peter, "Multi·Item Production Smoothing Models," Management Science,
Vot. 21, No. 9 (1975), pp. 1021-1028.
Yuan, John S.C., Jeffrey H. Horen, and Harvey M. Wagner, "Optimal Multi-Product
Production Schedulingand Employment Smoothing with Deterministic Demands,"
Manogement Science, Vol. 21, No. 11 (1975), pp. 1250-1262.
Digitized by Google
CHAPTERSEVEN
Bell GmbH
A t the time of this case Europe was still divided into a Euro-
pean Economic Community (EEC) and a European Free
Trade Area (EFTA), but merger negotiations were well along,
so that the 9.6 percent tariff from the United Kingdom to
the EEC was about to be eliminated. Bell's products, plastic
boxes for the cosmetic industry, cost about one-tenth of a
deutsche mark each. Empty plastic boxes are bulky, so the
cost of transporting them to the filling plants of the cosmetic
companies is not insignificant.
The sales growth of 20 percent per year in countries such
as Greece, Portugal, and Spain was occu"ing partly because
citizens there were purchasing more cosmetics. but also
because the cosmetics companies were starting to moue their
filiing operations to these countries. Nevertheless, at that
time, Greece, Portugal, and Spain were not members of the
EEC and so faced the EEC tariff. Currently, Bell is not
dependent on anyone cosmetic company, because even the
largest takes only 10 percent of its output.
Plastic boxes are made of petrochemicals. As the cost of
their raw material increases all manufacturers, including Bell,
will charge higher prices; the boxes are only a small element
in the cost of a finished cosmetic item, there is no conven-
ient substitute, so industry demand will hardly be affected.
Different students emphasize different aspects of this case.
Some focus on the engineering costs; others focus on indus-
try rivalry. Yet others study the cosmetics industry so as
206
Digitized by Google
Multinational Plant Location 207
Digitized by Google
208 Multi1UJtional Manufacturing
Bell's American name derived from the fact that its prod-
uct line was produced under an exclusive license of an
American company, Bell Industries, Inc. The American
licensor, however, had no ownership position in Bell GmbH.
Although Bell GmbH was a relatively small firm, it had
already opened a plant in Great Britain in the mid-1960s.
The United Kingdom plant had a capacity of 25 million
units as compared to the 200 million unit capacity of
the main German plant, located in Hanover. Both plants
ordinarily produced at 90 to 95 percent of capacity. About
DM18.8 million of sales came from the German plant's pro-
duction; the remaining DM2.3 million was accounted for by
U.K. production. Because of the 9.6 percent EEC common
tariff, U.K. production went only to the United Kingdom.
All other European markets (including other EFTA markets)
were supplied by Germany. Indeed, the U.K. plant had been
set up only because of the 9.6 percent ad valorem EFTA
tariff facing German output. The U.K. plant employed 30
people, compared to 120 in Germany. Of the German person-
nel, 90 were women workers who assembled and decorated
packages. The remainder consisted of administrative person-
nel plus technicians who designed and built much of Bell
GmbH's machinery and who handled the very important
(indeed critical) quality-control aspects of the business. None
of Bell's employees were union members. According to
management, "We are too small to have attracted the atten-
tion of the unions."
Sales were made in all countries of western Europe and in
Yugoslavia. Under the terms of its American license, the com-
pany was restricted to selling in Europe and the socialist
countries. Exhibit 7.1 shows the total 1971 European and
U.S. markets for the kind of special cosmetics-packaging
products produced by Bell. Given normal economic condi-
tions, that is 4 to 5 percent yearly growth in GNP, manage-
ment expected 20 percent per year sales growth in countries
such as Greece, Portugal, and Spain.
Digitized by Google
Multinational PlIlnt Location 209
EXHIBIT 7.1
Digitized by Google
210 MultinatiollGl Manufacturing
EXHIBIT 7.2
Digitized by Google
Multinational Plant Location 211
EXHIBIT 7.3
Digitized by Google
212 MultirllJtional Manufacturing
Digitized by Google
EXHIBIT 7.6
Exchange Rates, Money Supply, and Prices, Selected European Countries 1965-1971 0
1966 1968 70 1971
Italy
Rate of IIH1r dollar) 624.45 1::123.50 62i'i :;!: ::::.00 581.5
Money supply (1963 = 100) 125 142 164 184 213 273
Cost-of-living index (1963 = 100) 109 112 115 117 121 128
France
Rate of exchange (francs per dollar) 4.902 4.952 4.908 4.948 5.558 5.520 5.116 b
Money supply (1963 = 100) 118 127 133 145 146 157
Cost-of-livJ:i:lmi:l:dl!:K 1.1963 = 100) 105 108 112 118 124 131
United
Rate of e:,::ch,IUl:ii)I" f,:H:HUld per dollar) 0.358 .419 0.4 0.383
!'oO) 118 140 144
!:11I113 = 100) 114 123 12'1:1'
Gennany
Rate of exchange (marks per dollar) 4.006 3.977 3.999 4.000 3.690 3.648 3.223
Money supply (1963 = 100) 117 119 131 142 150 165
Cost-of-living index (1963 = 100) 107 109 110 113 116 120
Spain
Rate of exchange (peseta per dollar) 59.99 60.00 69.70 69.82 70.06 69.72 64.47
Money SUPIP']:II!,OO) 155 198 22'1:1'
!:::1113 = 100) 130 142 14';'
°Year end.
bCommerclal
Source: IMF, l'll':lonciol Stoti.tic.,
IT ~
'< ~
c; CO
o
a
~
......:I EXHIBIT 7.6
• International Financial Data, Selected European Countries (in $ millionsr
1971
1965 1966 1967 1968 1969 1970 (Nov.)
Italy
Official reserves 4,800 4,911 5,463 5,341 5,045 5,352 6,431
Balance on goods and services b 1.883 1,779 1,273 2,336 2,013 679
Trade (goods) balance only 646 331 -21 1,048 542 -340
France
Official reserves 6,343 6,733 6,994 4,201 3,833 4,960 7,494
Balance on goods and services b 732 -238 -971 1,148
Trade (goods) balance only 356 -158 -1,223 726
United Kingdom
Official reserves 3,004 3,099 2,695 2,422 2,527 2,827 5,572
Balance on goods and services b 378 801 -115 -118 1,613 1,911
Trade (goods) balance only -664 -204 -1,446 -1,543 -338 7
Germany
Official reserves 7,430 8,029 8,153 9,948 7,129 13,610 17,371
0
cO" Balance on goods and services b -86 1,593 3,970 4,554 3,780 3,225
""
N"
(J)
Trade (goods) balance only 248 1,878 4,116 4,485 3,902 4,024
Q.
IT Spain
'<
c; Official reserves
Balance on goods and services b
1,422
-846
1,253
-983
1,100
-907
1,150
-709
1,282
-959
1,817 3,104
0
a-
( i)
Trade (goods) balance only
aEnd of year.
bNot includinl uanafer payment••
-1,759
(fore·
cast)
1967 1 ::169 1970 rO";fl
"lJ 1), 1972 b
Italy
Percent incl~'II'lm,:11111 i,1':I IrUiihtlili,l:ial output 8.5 :::!.9 4.0 «m,,:;!: ,!l:ll 8.0
Percent increase in wagesd 5.2 3.6 7.5 21.4 14.5 15.0
Ratio of output to wage increases (1.64) (1.75) (0.36) (0.14) (0.19) (0.53)
France
Percent increase in industrial output 4.3 2.6 4.1 12.7 5.6 2.5 5.0
Percent increase in wagesd 5.9 6.0 12.4 11.3 10.5 11.1 10.0
Ratio of OUli,!P'tiili: to '11I1I111,1i!1I i rllcreases (0.43) 1.13) (0.5::::1 ~;!: :;;!) (0.50)
United Kingdoli:n
Percent inclI'IIi'IIt:'IIII' -0.9 :1.4 1.6 11),111, 3.5
Percent incJ:'II'II'I,:!I,II!' 4.0 !:1I.2 9.6 1:1:, 1 12.0
Ratio of OUII,!P'lllli: to '1111':III!;l1II iJllcreases (-0.22) (0.37) (O.U'I ((:1,(]17) (0.29)
Gennany
Percent increase in industrial output 1.8 -1.7 12.3 12.5 6.3 3.2 0
Percent increase in wagesC 7.3 3.9 4.3 9.1 12.8 13.3 6.5
Ratio of output to wage increases (0.25) (-0.44) (2.9) (1.4) (0.49) (0.24) (0)
Spain
Percent inCII"!'IiII:!lIII' 6.2 14.5 7.9 6.6
Percent inc!~'II!IIIt:II'III' 15.0 !:I.O 17.0 12.0
Ratio of OU!;IP'llllii (0.42) (1.61) (O.4';r} (0.55)
aCalculated frollllll i"iconomic Indlcatorll',
bEurofinance·'!,i 1,11111 0 1'1, Vision. January :1,1:;1'11
IT
~ CHourly earninllllll"
'< ... dHourly rates.
c; C1I
o
a
~
....lI:I
~
EXHIBIT 7.8
1965 3,141 359.0 923 3.00 4.12 386 21.57 5.20 8/9
1966 3,514 389.7 1,040 3.18 4.42 401 25.13 5.58 9/3
1967 3,781 414.2 1,128 3.37 4.60 426 28.81 5.94 9/8
1968 4,018 438.7 1,283 3.79 4.79 445 31.16 6.24 10/4
1969 4,263 474.1 1,407 4.21 5.28 489 34.69 6.64 11/2
1970 5,074 4.56 5.77
~Male and female.
EaminIS.
CRate..
Source: lLO Yearboolr o( Lobor Stat/dtc_. 1970.
o
cO"
""
N"
(J)
Q.
cr
'<
c;
o
a-
( i)
~-....,................,-------------'--'
EXHIBIT 7.9
Digitized by Google
218 Multinational Manufacturing
EXHIBIT 7.10
BeUGmbH
Sample Cash-Flow Projection for an Italian Investment O
Projected Cash Flow.: Italian InlJutment in 50 Million-unit Plant
INVESTMENT
Machinery 500,000 OM
Working capital 220,000
Total 720,000
Less: 70% debt (504,000)
Net investment 216,000 OM
Sale. reuenue
(95 X 50,000 OM) 4,750,000 OM 4,750,000 OM
E%peme. b
Building rental 48,000 43,000
Components 2,300,000 1,970,000
Direct labor 170,000 153,000
General overhead 150,000 135,000
Transport costs 75,000 68,000
Interest (5%) 25,000 23,000
Net cash flow 1,982,000 OM 2,358,000 OM
OShowln1 results _mini: (1) no devaluation and (2) a 10'.11. devaluation.
bMajor UBIlmptions:
1. All components SIlppUed locally.
2. Income tax hoUday provided by ItaUan lovemment
3. All production Is for export.
4. Depreciation not Included In overhead charles.
Digitized by Google
Multinational Plant Location 219
be obtained at this rate in the future. Would some existing
German plant capacity eventually have to be transferred
elsewhere?
Questions
la. "The product line consisted entirely of special opening,
closing, sliding, and spring packagings for the cosmetic
industry" sold to over 200 customers of which the
largest purchased 10 percent of Bell's sales. Based on
your understanding of the manufacturing technology
and the cost of transportation approximately how many
plants would it be rational for Bell to have?
lb. Given the total market size for the cosmetic packaging
industry as a whole in Europe, estimate approximately
how many plants would minimize the sum of logistics
and operating costs.
2. From the viewpoint of Firms A and B, in which nations
would Bell's next plant least threaten them, and how
might they react to the Bell move that you are recom-
mending?
3. Layout the skeleton of the cost/benefit analysis that
would be done by the Spanish or Italian government.
4. It is mentioned that "Kahler had received a phone call
from Company A suggesting that they try to counter
B's price cutting by an arrangement." What ulterior
motives might A have had and from what actions would
it want to restrain Bell?
5. Compared to the German norm, how many additional
months of inventory would it be prudent to carry in
customer nations to compensate for the probabilities
of supply interruptions if the new plant were to be
built in (a) Yugoslavia, (b) Italy, and (c) Spain?
6. Analyze Exhibit 7.4 to calculate the economies of scale
in the cost of (a) constructing and (b) operating a pack-
aging plant.
7. Extrapolate the data in Exhibits 7.5-7.10 to predict the
equivalent of Exhibit 7.4 in 1980 in each of the nations.
8. Based on your analysis of the case in the preceding
questions which models of the chapter would you rec-
ommend to Bell's management? What budget should they
authorize for an analysis and what deadline should be
established before committing the next plant?
Digitized by Google
Introduction
Until it builds a factory in a nation, a corporation has the flexibility to
consider supplying the market from anyone of a dozen different nations,
playing governments off against one another for incentives. Once built,
however, the factory becomes a hostage by which the national govern-
ment can and does influence the corporation. Anticipating this reversal of
the relationship means that the political forecasts needed by a multina-
tional corporation should be more detailed than those needed by a domes-
tic corporation in the same nation. After all, the domestic corporation
must build a factory in the nation if it is to supply its growing market-
there may be questions of capacity, timing, layout, and unionization, but
the location decision is a foregone conclusion. Because a multinational
corporation has more options, its executives need to be able to choose
among them; political information is a vital input.
220
Digitized by Google
- -- -~--~------------
Digitized by Google
222 Multinational Manufacturing
Digitized by Google
Multinational Plant Location 228
Digitized by Google
224 Multinational Manufacturing
being foreclosed from a market, and thus build a small factory. A small
nation may become the site of too many plants, none of world scale.
In Canada this phenomenon is called the "miniature replica effect."
Let us work through the situation of two rival corporations exporting to
a market. With increased market demand and/or a rise in unit cost of im-
porting, a corporation could justify the fixed cost of a local plant. How-
ever, suppose the market is still too small for either rival to warrant incurring
the cost of a plant. Nevertheless, each firm knows that its market share
will suffer as soon as its rival builds. Many executives, experienced in such
situations, know that a firm with a local plant has a competitive edge over
one that does not. By building first, a firm can hope to increase its market
share, especially its share of new customers. Only by building a catch-up
plant can the preempted rival stop the erosion of its market share. The
longer it waits, the greater the permanent erosion of market share. Of
course, if both firms build too early, the resulting capacity glut will reduce
industry profits. Scherer et al. (1975) provide an excellent discussion of
the complex issues of competitive capacity expansion. From the date of
construction of the first plant until the catch-up plant comes on stream,
the firm that builds second loses market share to the leader, as shown in
Exhibit 7.11.
Some corporations keep thorough and systematic track of each of their
rivals. Most find it more prudent to conduct ad hoc analyses. In the United
States, the first legal means of competitor analysis is the reports a cor-
poration must file with various government agencies. Pursuant to the
Freedom of Information Act, several Washington firms specialize in ob-
taining these reports for clients. Similar intelligence is available in other
nations.
The second legal means is to interview the rival's managers and engineers.
The usual practice is to contract with a medium level executive search
Market
share
Market share
---,.....-- of second firm
__--+---Market share
II of first firm
I
I I
I I
l-.---.1.t-1
I I
EXHIBIT 7.11
Market share and sequence of construction.
Digitized by Google
Multinational Plant Location 225
EXHIBIT 7.12
Best cash flow to A for given date of B's building.
Digitized by Google
226 Multinatio1llJl Manufacturing
each date of our building the first plant, we can calculate the present
value of our future cash flows and those of our rival.
Half the problem' has been solved. Now, standing in the shoes of the
rival, and for each date at which the rival builds the first plant in the
nation, calculate when we would respond. For each date of the rival's
building the first plant, calculate the present value of cash flows to the
rival and the present value of our cash flow.
The important date is when the first plant is built. Call the two corpora-
tions A and B. In Exhibit 7.13 the first graph shows the present value of
cash flow to corporation A, one line if it builds first, the other line if it
builds second. Note that after a certain date the market has grown so large
and market share counts so much that the corporation will build a plant
as soon as its rival does; after this date it shows the same cash flow
whether it builds first or second.
The posture of each firm toward the other varies through time (Rao and
Rutenberg, 1979). We can delineate five zones or intervals that characterize
the growth of demand in the new market (Exhibit 7.13). The first is a
premature zone when both firms find it profitable to import rather than
Sum of A's
profits, $ A preempts or
responds optimally
A responds optimally
EXHIBIT 7,13
The dynamics of rivalry.
Digitized by Google
Multinational Plant Location 227
build a new plant. In the next interval, one firm has an opportunity to
preempt its rival and thereby gain a stronger market position in terms of
higher market share. This is the opportunity zone. Subsequent to this is
the anxiety zone, an interval when each firm can preempt the other
"advantageously," but only if the other does not build simultaneously.
This is the prisoner's dilemma. It is followed by an interval of mutual for-
bearance, where neither firm finds advantage in preempting; each is de-
terred by knowing that the other will respond by building simultaneously
because the market has grown large enough and market shares have be-
come more valuable. This we call the zone of mutual deterrence. Finally,
at some time one firm will find it so expensive to continue to import that
it will build a local plant. The other firm will respond by building simul-
taneously to protect market share. This is the safe zone. As can be seen, in
each zone the value for competitive information is quite different.
Everything else being equal, we would expect a large firm to build a
plant first. However, the firm that has the higher import cost may be pre-
sented with an opportunity. In a competitively adverse situation it pays
not to catch up but to initiate; intelligent leapfrogging can actually result
in gains. But such opportunities occur only occasionally for a small firm.
In practice, the delineation between the opportunity zone and the
anxiety zone is crucial, whether it is delineated by argument or calcula-
tions. Most executives have felt extremely anxious about the real possibil-
ity of excess industry capacity. "Some damn fool is going to trigger a
round of capacity expansions" is frequently repeated. The first reason for
clear delineation is that, in the anxiety zone, rivals can cope by choosing
two nations, both ripe for a plant, and each rival can build a plant in one
of the nations, sized adequately to supply the marketing needs of both
corporations. If the product is homogeneous (gasoline or liquid oxygen,
for example) this twinning of transactions is called a swap. However, there
is no reason that the products must be identical (factories commonly
produce private label brands). The twinning is essential so that each com-
pany holds the other hostage. Thus each of the two firms may build in a
different market first and then exchange their outputs locally rather than
incur shipping costs. In this way, not only is excess capacity avoided but
efficient production networks are achieved. Scherer et al. (1975) discuss
the prevalence of such practices. In the anxiety zone, both parties will
welcome government intervention to prevent both from building a plant.
The second reason for clear delineation is that corporate planners must
make an argument for new plants early enough to arrange capital appro-
priations. It is vital to forecast the opportunity zone correctly, and to
communicate the opportunities to superiors and subordinates. An inability
to seize opportunity might lead managers to swapping arrangements, or
worse, capacity expansions that result in excess industry capacity.
In summary, governments pressure corporations to build factories to
replace imports. In each nation competitive dynamics also lead to early
plants. In a polycentric corporation there will be many small plants.
Digitized by Google
Ethnocentric Plant Location to Minimize Cost
Plant
construction
plus
tieeounted
p:esent value
future
eeerating
lfrglfrts
Capacity
EXHIBIT 7.14
Economies of scale.
Multinational Plant Location 229
Improving Timing
Solve the logistics problem at time zero (using the market demands,
logistics costs, and the plant's capacities available at time zero). The solu-
tion gives one point in Exhibit 7.15. In the transportation tableau allow
the demands to grow with time (use Srinivasan and Thompson [1972]
operators to keep the transportation tableau optimized). The total of 10-
Logistics
costs
L
o Time
EXHIBIT 7.15
Logistics cost on all routes at each point through time.
Digitized by Google
230 Multinational Manufacturing
gistics plus plant operating costs will rise as demands increase as shown in
curve A of Exhibit 7.15. The upper end of line A occurs at t} when de-
mand has risen to fill all available plants.
As soon as a new plant is built, it will supply local markets, so the sum
of the logistics costs drops from Lt to L~ dollars per month in Exhibit
7.15. This desirable drop is offset by the interest on the plant investment.
If the interest charge is less than the drop in logistics and operating costs,
then it will be profitable to build the new plant earlier. How much earlier
to build depends on the logistics cost, so with the new plant available, run
the transportation tableau backward in time to calculate the lower logis-
tics curve B. Stop working backward in time as soon as the difference
between the logistics curves equals the interest cost on the plant. In Ex-
hibit 7.15 this time occurs at t2 , when the logistics savings is L~ - L~ .
Timing reoptimizations can be done one at a time, starting with the
first new plant. The calculations of Rao and Rutenberg (1977) show that
this simple retiming can be very profitable. Constructing a plant earlier
than needed has two side benefits. The excess capacity is available in case
demand picks up or some other plant fails. Furthermore, as analyzed in
the polycentric section, having the plant means that capacity overhangs
the market, thereby preempting rivals.
Value
of
Dual
Variable
New
plant
EXHIBIT 7.16
Time trace of the benefit of expanding the plant by one unit of capacity.
Digitized by Google
Multinational Plant Location 231
be reallocated, and some established plants will have excess capacity. The
exhibit is plotted in terms of dollars per month. Using the discount rate
for this nation, compute the present value of this stream of benefits.
Compare the cost of an increment of capacity with the present value of
the benefit, and make a small adjustment to the capacity to bring the
values toward equality.
Recall that the polycentric corporation was pressured to build a factory
in each strong national market. Healthy economies generally have natural
resources and/or high labor productivity. Their exports usually exceed im-
ports such that their exchange rate need not fall as fast as domestic in-
flation rises. It is difficult to export from such nations, because plant
productivity improvements tend to be offset by the rising exchange rate.
In contrast, the hardheaded production approach of an ethnocentric
corporation is to convert national balance of payments analyses into U.S.
dollars. In nations whose currencies are forecast to devalue faster than
domestic inflation will rise, a multinational will face a falling real cost of
wages and local supplies. To this end the corporation should select nations
that lack natural resources and have an adequate but fragmented
manufacturing base. Often these are the nations who give the largest in-
centives to get world-scale manufacturing plants.
Digitized by Google
232 Multinational Manufacturing
Refer back to the table of nations and product divisions. For each pro-
duct division determine the best location for the new plant. Then evaluate
locating the new plant in each of the other nations. Optimize logistics,
timing, and plant size so as to determine the additional cost that would be
incurred if political pressures necessitated this location. The table of
nations and product divisions, with each cell completed with the "addi-
tional cost," becomes the "payoff matrix" to the corporation. If each
nation insists on a plant (without specifying which product), then the
corporate staff can compute the least cost assignment of plants to nations
as an orthodox assignment problem from operations research.
If governments prefer certain divisions to others (as they will if they
perform cost/benefit analyses), then the government's payoff matrix
will be different from the corporation's. It turns out that this problem
can be formulated as a weighted distribution (machine loading) problem
where each government has a goal of benefits that must be satisfied. With
computers doing the drudgery of calculations, the analysts can work to
refine what these solutions really are in the eyes of the corporation and
in the eyes of each government.
Negotiating
The first and most important step is to stand in the shoes of the national
planning agency and, as recommended earlier in this chapter, perform a
cost/benefit analysis of the plant. Despite the aura of precision implied
by a cost/benefit analysis, much of the input data are necessarily subjec-
tive. Part of the company's negotiating strategy can be to discuss with this
government reasonable ranges for the input data that are particularly
subjective.
The second step is for the marketing manager to work through the exer-
cise of withdrawing from the nation and estimate the recoverable value.
Even though the probability of disinvestment may be low, the exercise
sets a bound beyond which corporate negotiators will not be pushed.
Furthermore, the executives gain an awareness of clauses in supply con-
tracts (both local and within the corporation) that may need to be
broken.
The third step of preparation is to rehearse. Rehearsal requires that two
people spar with one another, then reverse roles and spar again to test
their negotiating arguments. The vital thing to realize is that the cor-
poration has the upper hand only until it builds a factory, and thereafter
the factory is a hostage. Hence most of the rehearsal should consider in-
cidents that might arise after the factory begins operation. Qualitative
aspects of emotional differences and bad strategies will be exposed.
Digitized by Google
Multinational Plant Location 233
sions. But such nations usually present risks, and corporate headquarters fre-
quently eliminates risky nations without serious evaluation. Thus it is
common for an ethnocentric headquarters to create for its product divi-
sion managers a list of approved nations. Product division managers are
encouraged to build in the approved nations; by strong implication,
factories proposed in other nations would be rejected and hence are
rarely proposed.
One approach to evaluating national risk is to try to express probabilities
of disaster. The other approach is to calculate the threshold of risk within
which a factory could be justified in this nation. Those who fathom politi-
cal risk are then forced to gauge whether forecast risk is within this
threshold.
Before proceeding further, it will be helpful to differentiate between
balance sheet risk and supply risk. Many balance sheet items can be
insured. Governments of exporting nations compete in granting long-
term low-interest loans for factory equipment to be exported to under-
developed nations. With the loan the corporation can usually obtain
insurance against both expropriation and damage caused by civil distur-
bance. The corporation's entire balance sheet risk could be protected by
insurance. In that case, from the viewpoint of risk analysis (though not
equity appreciation possibilities) the factory might as well be totally
owned by the local government but run by the corporation under a man-
agement service contract. But even if balance sheet risk is zero, supply
risk remains.
The risk of interrupting the supply of goods to customers is most
worrisome. Many multinational corporations have built their marketing
strategy around a consistent supply of consistent products. Their brand
carries the implication of an assured supply. The potential penalty in the
face of the risk of supply interruptions must be analyzed within the cor-
poration, because this penalty cannot be transferred to a government
insurance agency. The corporation needs contingency plans. If a plant is
lost, or if a transportation route is closed (the Suez Canal, for example),
how can the corporation adapt? There may be several combinations of
disasters and opportunities to consider. Each combination is usually called
a scenario. For the purposes of developing a plan it is sensible to select
just a few scenarios. Once the plan has been developed, it can be validated
(checked for robustness) against many less likely scenarios.
The analysis has three stages. The first stage occurs as soon as supply is
interrupted and production stops at one plant; all other plants producing
these items must step up production (by increasing shifts, delaying main-
tenance, and so on). The second stage may involve starting to produce the
items in one of the large, flexible factories. The third stage is to expedite
the next plant opening, perhaps moving its site to a different nation to
rebalance the production network.
Digitized by Google
234 Multinational Manufacturing
Digitized by Google
Multinational Plant Location 235
Digitized by Google
Questions from Other Viewpoints
1. Corporate Rational Normative Global. You just covered this view-
point in the chapter.
2. Corporate Rational Normative Subsidiary. From an analytical
model, a headquarters planner could compute the threshold cost
curve of a plant in your particular nation. If a plant could be built
below that cost, it would pass the headquarters ROI threshold. Ex-
plicitly how would your knowing the headquarters' guideline help
you search for a plant site in your nation?
3. Corporate Rational Descriptive Global. Find checklists used by cor-
porations to evaluate plant sites. Try Business International, Carl
Heye!'s The Encyclopedia of Management, or any other source.
What emphasis do you see in these checklists? What factors do you
see as ignored?
4. Corporate Rational Descriptive Subsidiary. Some national govern-
ments offer tax incentives to entice foreign-owned factories. Other
national governments threaten to withhold import permits unless
the corporation builds a factory or expands its factory to meet local
content laws. Please generalize by describing the characteristics of
each of these two groups of nations.
5. Corporate Emotional Normative Global. Whenever a task force gets
involved in investigating a factory site, they have a tendency to get
so involved in their work that they hate to say no. How would you
manage such a group so that their inevitable identification with the
site is used to the corporate advantage?
6. Corporate Emotional Normative Subsidiary. Think about the town
or city that you are now in, paying attention to the style and climate
of work. Suppose a foreign multinational corporation has selected
your town or city as a possible plant site, because the economics look
right. As an employee of your national subsidiary, you have been
sent to your town to do an on-the-spot assessment to complement
all the data that headquarters has on the town. Select an industry
that is economically viable for your town, but whose style would not
suit it at all. Write a tightly argued two-page memo to headquarters
explaining the lack of fit.
7. Corporate Emotional Descriptive Global. Describe the emotional
problems of a team of factory location analysts who make endless
studies, none of which are implemented because the corporation
successfully resists pressures that it build more plants.
8. Corporate Emotional Descriptive Subsidiary. Describe some of the
nonquantitative, rather emotional factors, that actually affect factory
location.
9. Societal Rational Normative Global. If an industry were ruled world-
wide by one monopolist (or one planning commission), would there
236
Digitized by Google
Multinational Plant Location 237
Bibliography to Chapter 7
Bradley, David G., "Managing Against Expropriation," Harvard Business Review, Vol.
55, No. 4 (1977). pp. 75-83.
Digitized by Google
238 Multinational Manufacturing
Doz, Yves L. and C.K. Prahalad, "How MNCs Cope with Host Government Inter-
vention," Harvard Business Review, Vol. 58, No. 2 (1980), pp. 149-157.
ErIenkotter, Donald, "Plant Location: Where Demand Is Sensitive to Delivered Cost,"
Ma1ll16ement Science, Vol. 24, No. 4 (1978), pp. 378-386.
Gaeteno, Lombardo and Richard C. Norris, "Facility Location - Some Practical
Applications," paper presented at the 45th Joint National Meeting of ORSA/TIMS,
Arthur D. Little, Inc., Cambridge, MA., 1974.
Knickerbocker, Frederick T., Oligopol;.tic Reaction and MultinlJtional Enterprise
(Boston: Division of Research, Harvard Business School, 1973).
Lent, George E., "Tax Incentives for Investment in Developing Countries," IMF Staff
Papers, Vol. 14, No. 2 (1967), pp. 249-321.
Rao, Ram and David P. Rutenberg, "Multilocation Plant Sizing and Timing," Ma1ll16e-
ment Science, Vol. 23, No. 11 (1977), pp. 1187-1198.
Rao, Ram and David P. Rutenberg, "Pre-empting an Alert Rival," Bell Journal of
Economics, Vol. 10, No. 2 (1979), pp. 412-428.
Rummel, R.J. and David A. Heenan, "How Multinationals Analyze Political Risk,"
Harvard Business Review, Vol. 56, No. 1 (1978), pp. 67-76.
Scherer, Frederick M. et al., The Economics of Multi-Plant Operations: An Interna-
tional Compar;'on Study (Cambridge: Harvard University Press, 1975).
Srinivasan, Venkataraman and Gerald Thompson, "An Operator Theory of Parametric
Programming for the Transportation Problem I and 11," Naval Research Logistics
Quarterly, Vol. 19, June 1972, pp. 227-252.
Wells, Louis T., Jr., "Don't Overautomate Your Foreign Plant," Harvard Business
Review, Vol. 52, No. 1 (1974), pp. 111-118.
Wells, Louis T., Jr., "Social Cost/Benefit Analysis for MNC's," Harvard Business
Review, Vol. 53, No. 2 (1975), pp. 40-50.
Digitized by Google
Summary of Multinational Manufacturing
International trade theory builds on the concept of gains from trade. For
a variety of reasons such as raw material availability, climate, and attitude
z}f "ZZwrkers, the rel}tiv"ZZ of producing products dift"ZZ}
dttferent countri"ZZb~ the case, it for each countbtt
which it has zJzdvantage, tradi"ZZg
in which g"ZZve a comparati"ZZb
Digitized e
PART THREE
Multblation8l _.".-...,~
Digitized by Google
CHAPTER EIGHT
243
Digitized by Google
244 Multinational Marketing
Digitized by Google
New Product Launch 245
.§ 36
~ 32
-;;; 28
~ 24
!!! 20
"C
-= 16
·i 12
"C
.iij 8
Q.
X 4
t\I
I-
1920 1930 1940 1950 1960
Year
EXHIBIT 8.1
Tax-paid withdrawals of cigarettes for consumption in
Canada, 1920-1960 (billions of cigarettes)
Digitized by Google
246 Multinational Marketing
c:
.2
D-
E
~ _24
III III
5~22
U
111 ..
a; 20
:!: :118
0.-
~ ~ 16
.. 0
G) III 14
o.u
iii ! 12
~ -g 10
c: ~
111 :5. 8
G)
II 6
i
c(
4
2~~~~__~__~
Digitized by Google
8.3
1944
1945
1946
1947
1948
1929 507 1949 1252
1930 493 1950 1252
1931 437 1951 1118
1932 353 1952 1234
1933 404 1953 1415
1954
1955
1956
1957
1958
1939 630 1959 1939
1940 663 1960 1925
1941 746 1961 2012
1942 879 1962 2083
1943 953 1963 2110
Year
EXHIBIT 8.4
Filter cigarettes as a percentage of the total Canadian cigarette
market, 1957-1960
248 Multinational Marketing
~... 25
1\1
E
Cij20
'0
:: 15
o
~ 10
~
~ 5
The Competition
The cigarette industry in Canada was a classic case of an
oligopoly market. As of 1960, there were only five major
firms in the marketplace, with two of them holding a com-
bined 80 percent share. These two firms, Imperial Tobacco
Company of Canada, Limited, and MacDonald Tobacco
Company, never held less than 70 percent of the market
since Imperial was incorporated in 1912. In 1960, Imperial
held a 54 percent share and was the acknowledged market
leader. MacDonald, with a 27 percent share, seemed to be
content with a defeniive strategy and spent relatively little
on advertising and promotion. (MacDonald's was established
in 1857.) The Philip Morris affiliated company, Benson
and Hedges (Canada) Ltd. had a negligible market share.
Rothmans of Canada Ltd. was considered an up-and-comer.
Since entering the market in 1957, the firm quickly got the
reputation as a company in a hurry. Rothman engaged in
marketing practices, according to some observers of the
Canadian industry, that the old-time companies considered
quite rough, tough, and even nasty. They were spending
heavily on advertising and promotion and had introduced
the first king-size cigarette, Rothman's King Size, to the
Canadian market-it was doing extremely well.
Imperial Tobacco was recognized as being the price leader
much as R.J. Reynolds was in the United States. Competi-
tion, as in the U.S. market, was not waged on price grounds,
but was instead concentrated on advertising, promotion, and
Digitized by Google
New Product Launch 249
EXHIBIT 8.6
Confectionary stores na na
Fruit & vegetable stores na na
Grocery stores (no fresh meat) 21,683 $ 25.4 3.8%
Grocery stores (fresh meat), beer na na
Variety stores 1,081 18.6 2.8%
Combination stores
(groceries & meat) 9,340 109.1 16.8%
Combination stores (with beer) na na
Eating places na na
Eating places (with alcoholic
beverages) na na
Eating places (other merchandise) na na
Department stores 74 21.7 3.4%
General stores 7,739 23.0 3.5%
General merchandise stores 702 25.6 3.9%
Drug stores (no soda fountain) 4,630 44.6 6.8%
Drug stores (soda fountain) 154 2.5 0.4%
Tobacco stores & stands 2,702 92.0 14.2%
News dealers na na
---
44,175 $654.71 100%
Digitized by Google
250 Multinational Marketing
Digitized by Google
New Product Launch 251
EXHIBIT 8.7
EXHIBIT 8.8
Philip Morris International
Final Report - Canadian Copy Claim Report
A test of two copy claims was devised for Philip Morris International.
The procedure was to have each smoker "test" a cigarette from a plain
white package, and to react to the cigarette with various attitudinal
words or phrases. Three groups of 800 people each were used in these
tests. The first group, which we call the control or base group, smoked
unmarked Parliament cigarettes from completely blank white packages
and expressed their attitudes toward the cigarette. The second group
smoked identical unmarked cigarettes from packages which had the
phrase "American Type Cigarettes" printed on the package, while the
third group smoked cigarettes from packages labeled "Cigarettes
Blended with the Darker Colored Burley Tobacco for Better Filter
Smoking and Lower Tar Delivery." The phrases were written in English
on one side of the package and in Canadian French on the other side.
Results
1. The claim "Cigarettes blended with the darker colored burley tobacco
for better filter smoking and lower tar delivery" was an ineffective
claim.
2. The cultural difference between French· and English-speaking Cana·
dians seems more significant than difference in sex, age, income, or
type of cigarette usually smoked.
Discussion
The purpose of these tests was to determine what change would occur
in smokers' attitudes toward cigarettes which were designated as
"American Type Cigarettes" as opposed to "Cigarettes blended with
the darker colored burley for better filter smoking and lower tar
delivery." After smoking one cigarette the people were asked to select
one word or another from each pair of words to describe their attitude
toward the cigarette. For example, "mild" or "strong," "cool" or "hot."
Since one group of people smoked unmarked cigarettes in a blank
package, we have a base point against which we can measure any incre-
ments or decrements of the number of people associating a particular
attitude with cigarettes smoked from the "test" packages which bore
Digitized by Google
252 Multinational Marketing
the claims. We then concern ourselves with responses to the test cigarettes
which received either a statistically significant higher or lower percentage
of responses than were given to the blank package or control cigarettes.
In other words, when an association is classified as significant, it means
that in 95 out of 100 repetitions of the study the results would not be
reversed.
While there were 14 such associations made toward each cigarette, the
strongest pair of attitudes on the Ilst is for a person to say "For me" as
opposed to "Not for me." The following table shows two things. First,
No other breakdown of the sample, such as men vs. women, young vs.
old, showed significant differences on the "For me" item.
A second method of looking at the association data is to analyze the
total number of favorable associations as opposed to the total number of
unfavorable associations expressed as a percentage. In other words, if
every smoker selected each favorable association from each of the 14
pairs, we would have 100 percent favorable associations. In a sense, the
total favorable percentage is a rough average of the percentage of people
associating anyone favorable word with the cigarette.
The following table shows that the overall reaction to the American-
type cigarette claim was more favorable among the French-Canadians
than among the English-speaking people and that the claim was also
more effective than the "blended ... burley ... low tar" claim.
Digitized by Google
New Product Launch 253
We could caution the reader that the individual items taken singularly
can often be misleading since there is a tendency for consumers to
generalize their liking for a product over a series of items. As we have
already stated, we feel that the critical item is whether people say "For
me" or "Not for me" about the cigarette. Thus the French-Canadians
liked the "American-type cigarettes" and attributed significantly higher
associations on rich tobacco navor, good, slow burning, effective filtra-
tion and nonirritation, high~uality tobacco, well made, and desirable.
The "blended ... burley ... low tar" claim was ineffective among the
French-Canadians. There were increases in associations with high-
quality tobacco, effective filtration, and slow burning, but the taste and
flavor connotations were not increased. One other item which did
show a significant increase over the blank package cigarettes was costly
as opposed to cheap.
In summary, then, to the French4peaking people the "American-type
cigarettes" had navor, fdtration, and quality without being costly.
The "blended ... burley ... low tar" cigarettes had filtration, quality,
and expense but lacked an increase in navor associations.
To the English4peaking smokers, the "American-type cigarettes"
were associated with significant decreases in association with clean,
desirable, and for me. It appears that the English4peaking Canadians
rejected the "American-type cigarettes" purely on an emotional basis.
While the English4peaking smokers rejected an "American-type
cigarette" even though it was given most of the same qualities as was
the blank package, the "blended ... burley ... low tar" cigarettes
were given some increases in qualities but no increase in preference
over the blank pack cigarettes. Associations with good navor, high-
quality tobacco, slow burning, desirable, and well made were all in-
creased for the "blended ... burley ... low tar" cigarettes; but even
with these quality increases, the copy claim failed to draw an increase
in the critical "For me" association.
To summarize the results of the English4peaking Canadian sample,
they appear to have simply rejected the idea of an "American-type
cigarette," and although the "blended ... burley ... low tar" cigarettes
showed increased associations in taste, quality, and desirability, the
concept was not sufficiently enticing to have an increase in acceptance
of the cigarette.
Digitized by Google
254 Multinational Marketing
Questions
1. Estimate the fixed cost of launching Parliament in
(a) Quebec, (b) Ontario, and (c) all of Canada.
2. How should Philip Morris evaluate the spillover in Canada
of Parliament advertising from border television and radio
stations?
3. What competitive reactions should Philip Morris anticipate
from Imperial, MacDonald, and Rothmans? Is there any
sense that the competitors could preclude Philip Morris if
it does not act precipitously?
4. What changes would you recommend to the product and
packaging in response to the "American-type cigarette"
report?
5. "I can' believe that Parliament will sell right up to the
border and not within Canada. I've travelled in Canada
since I was a boy, and those people are just like us."
Express this belief as a covariance between 0.0 and 1.0 of
American and Canadian sales.
6. In view of Mr. Weissman's position what risks to his career
is he exposing himself to by launching Parliament in
Canada? Once Parliament is launched how low would its
market share have to be for Philip Morris to order it
withdrawn? What organizational moves would you expect
Mr. Weissman to make to shift the blame?
7. In dollars, how valuable is the Claim Report at this time?
Digitized by Google
Introduction
become globel
one advertisinb
ibe world. eountries are
ferent, and the attempts of ethnocentric corporations to impose a strait-
jacket on their operations caused them to offend national stakeholders
and to fail to adapt to many opportunities.
The following three chapters deal with central coordination of market-
ing to increase profits. Chapter 8 focuses on the decision to sink cash into
lemlehing a new markets of th,:;
It,~,grettably, mo;:;1 leunched producte the launch inve:;t~
is squandernst" other hand, it succeeds, one
n~'~""~'LL not having ,;:;arlier, therebh deofits sooner
d;:;uempting competi:;m:;" anguish is :;(£;:;ult of an unsh;:;~
tematic decision-making process that precedes the launch decision. Execu-
tives rely on will-o'-the-wisp rumors, a few data, hope, and the belief that
this product will succeed if only everyone will work. Emotions run high.
It is not the purpose of this chapter to squelch the emotions that attend a
product launch but rather to provide analytical boundaries within which
intuitive power ekT,(?tions can be
Philip Morris mualyzed repre;:;-ente product with
support datn" the produd e;:;tablished in
States, frorki ebvertising spill;:; Canada. Finallh
Torporation WLL that it
stHbies in Canada, the ethnocentrism ot the executives caused
them to misinterpret the results they obtained. The launch of Parliament
cigarettes failed in Canada.
Digitized e
M ultinationa£
foreign plants, the familiar international life cycle of (a) export from the
home nation, (b) foreign production starts, (c) foreign production be-
comes competitive in export markets, and (d) import competition begins.
To introduce a new product abroad, whether it be aimed at consumers
or industrial customers, requires a major expenditure. The promotional
theme and advertieiey must be The dealers sh,)uld
Overseeing
eetivities require) eostly wages,
secretarial which we
k, is the fix,)d
It is this entire k (usually se)(:)gg} thousand
per nation) that is at stake when the product is launched. The variable
cost of the product may be low enough to allow a comfortable profit
margin, but a break-even analysis shows the unit sales necessary to recoup
the prelaunch expenditures.
This familiar break-even analysis of Exhibit 8.9 gives the unfortunate
illusion of certainty eould be furthee truth in the '''£liAA~H
new products; ~uLd this will be
the United percent of
dmwn from £lCaeny of the rem,)ynYee
ueriable cost of but never reeneennLY fixed cost of
launch. Sales cannot be forecast with certainty, but only to a probability
distribution. The distribution is skewed, with a slight probability of excep-
tionally high sales. As in Exhibit 8.10, we will assume that the skew dis-
tribution is log normal; that is, if Exhibit 8.10 plotted the logarithm of
sales it would be shaped as a normal distribution.
Given the task sales of a neAr, an alien nation,
ethnocentric 'lTAenager feels l1'Cl'ps:*"", master his feelinh)
mey turn to me,UAures such as "~~"'~UAAV"'AA per 1,000
$
Revenue inflow
Cost outflow
---T
Break
poiut
EXHIBIT 8.9
Break-even analysis may appear certain, but forecast unit sales are stochastic.
Prnduct Launch
Probability
Sales
,,",EEiFIT 8.10
distribETtiP?E is log normal
Consumer goods can also be analyzed in the same way. The 1979
Annual Report of The Seagram Co. Ltd. discusses predictable similarities
Digitized e
258 Multinational Marketing
With increased earnings comes more disposable income, and the more frequent
enjoyment of spirits and wines is traditionally an integral part of a higher living
standard. In some countries - Brazil is a prime example - fostering a growing middle
class is an avowed government policy. In other areas - Africa, for example-
Seagram stin is the only major company in the industry to have established a
modem marketing organization. In both areas sales have multiplied over the de·
cade past and are virtually certain to do so again in the 1980&.
But capitalizing on such opportunities is easier said than done. Perhaps the pri·
mary prerequisite of major success is true appreciation of the very real differences
in any given area.
"Marketing programs in Africa have to start with the realization that each country
here has different characteristics and therefore presents a separate challenge," says
Jacob R. Scott, Seagram's top marketing man in Africa since 1971.
The same point holds for vast Latin America, where there are at least as many
differences as similarities. Brazilians react quickly to trends. F1avourful vodka·
based drinks are currently a stylish destination for consumers who have left the
local spirit, cachaca, and moved on from rum. Brandy has never been much of a
factor in Brazil but is a mainstay in Mexico. Locally made champagne is a familiar
aperitif in Argentina, virtually alone among the Latin American countries. And
Argentinians, who prize old ties with Europe, have long been fond of Scotch, as
have consumers in Venezuela.
Digitized by Google
New Product Launch 259
sales data and present an adequate case, the corporation could consider
limited remodeling of the product.
Stage 4 - Alien Nations. Perhaps the easiest way to deal with alien nations
is to export the product, allowing indigenous channels of distribution
to assume the risks. No corporate decisions should be altered by these
Stage 4 sales; they are a free bonus - not cash to be depended on.
Once a corporation has tentatively staged the nations, analytical execu-
tives may wish to be more explicit in their analysis of the risks of each
national market, thus broadening their chances for success.
Four elements are required for each nation:
1. The estimation of the launch cost k.
2. The profit contribution per unit 7r.
3. The probability of breaking even -that sufficient sales will occur
by a set date T, such that the total profit contribution will cover the
launch cost.
4. The expected sales E and standard deviation 0 of the sales rate d units
per time period T. Because only a few products usually succeed and
the majority limp along, when running regression analyses of sales data
it is customary to use the logarithm of the sales rate d, and to fit a log
normal distribution to it.
Launching the product in this market is a business decision so the E
and 0 of sales must be translated into a probability distribution of net
cash position at the end of T. For the product to break even on evaluation
day, the total profit contribution Td7r must exceed k. The price of the
product is fixed, so cash inflow varies directly with total sales. Although
expected sales may be way beyond break-even sales, there is concern for
the probability of .failing to break even. From the probability density
function of total sales one can read the area of the probability of breaking
even. The ethnocentric appendix at the end of the chapter continues the
analysis in more mathematical detail.
Digitized by Google
260 Multinational Marketing
(they may well have been the driving force behind its development).
Others decide to wait and see. And other national managers view the prod-
uct as unsuitable in its present form. These laggards save the corporation
from the cash drain that would arise if all subsidiaries launched the
product simultaneously. The task of production smoothing (Chapter 6) is
also simplified. Unfortunately some new products look so good while
still under development that, long before the public announcement date,
most subsidiaries have committed themselves to a launch.
An Example
In 1963 the Eastman Kodak Co. launched its Instamatic camera simulta-
neously in 28 nations. Business Abroad (1967) commented:
Digitized by Google
New Product Launch 261
Some Mathematics
Sales are forecast before a product is launched in a market. After the
launch the actual sales can be compared to the forecast, and the dif-
ference plotted. As in Exhibit 8.11 the majority (the mode) of products
are slightly disappointing. On the other hand, offsetting these are those
Actual
- - - - - - ' - - - - - - - - - - - - - - forecasted
(-) o (+) sales
Disappointing Surprisingly
results good results
EXHIBIT 8.11
Distribution of forecast error.
Digitized by Google
262 Multinational Marketing
few products whose sales greatly exceed expectations. The distribution is
skewed.
In this chapter we will assume that the data fit a log normal form. Re-
plotted on log paper, this distribution would appear to be a normal
distribution. Hence, later in the chapter, the mean and standard devia-
tions of the sales forecast will refer to the logarithms of sales.
The difficulty faced by Kodak was that sales exceeded expectations in
almost all markets simultaneously. If one nation yields surprising results,
what is the probability that each other nation will? For example, if
actual sales are 20 percent higher than forecast in one nation, by what
percentage do you raise forecasts in other nations?
An important distinction must be highlighted. In "Ethnocentric Simil-
itude" the similarity between markets focused on similarity of per capita
sales. The focus was on the mean, not the standard deviation, and certainly
not the covariance.
Here we are assuming that the sales managers in each nation have made
reasonable sales forecasts, which the headquarters analysts have improved
upon by their regression analyses, similar to the ethnocentric executives.
But, in addition to estimating the means, the polycentric analysts are
vitally concerned with covariances, consistent patterns in the forecast
error. If a product takes off in one country and sales exceed expectations,
then the forecast error will be positive. If two countries are alike, the
forecast error in the other country is likely to be positive also. (In this
context, "likely" implies a positive covariance that is usually less than 1.)
An established multinational corporation has launched many items
worldwide. On a worldwide basis some exceeded expectations; others fell
short. These data are valuable when transformed into covariances (Staelin
and Turner, 1973). Consider nations B and C. In nation B item i was
forecast to a mean whose logarithm was E~ , whereas the actual sales were
d~. In nation C the same item i was forecast to a mean whose logarithm
was E~, whereas actual sales were d~ .
The new item seems more similar to some past items than it does to
others. Comparing the new item with each previo.us item i, express the
similarity by subjective weightings W; such that 1: W' = 1.
In other words, each of these similarity weightings will be used in place
of a probability measure, in the formulas for variance and covariance as
explained in any statistics text.
and
Cov (log dB) (log de) = };; Wi (log d~ - E~) (log d~ - E~)
This example is of two nations, but to prepare for the problems of real
Digitized by Google
New Product Launch 268
Geocentric Launch
Geocentric corporations launch new products in an adaptive sequence.
For years, Unilever has used Belgium as a detergent test market for the
Digitized by Google
264 Multinational Marketing
Australia has indeed been a good target for Japanese products. The country, with a
sparse population of 14 million, has accepted technology more readily than some
larger and more advanced nations. The Japanese, too, have learned, through keep-
ing in close touch with this market,' that the Australian arena is a comprehensive
microcosm of outside world requirements.
The full story on how and why the Australian axis is enabling Japanese makers
to pry their electronic consumer goods out of their own cloistered region to com-
pete in foreign-dominated territories may never be told. The Japanese themselves
tend to be inscrutable and coy on the subject, never revealing why Australia has
consistently been one of the first countries to receive their new products.
One Australian computer distributor, Gary Blom, is more forthcoming. Blom
says this about the Japanese junction: "In effect what the Japanese do is develop
a machine for their own markets, and then we modify it for an individual or
possibly an international reqUirement. The Japanese respond quickly in giving us
anything we want, whether it is in operating systems or in some structural detail,
such as larger keyboards."
"There are many more [types of computers] than the Japanese can see at home,"
Matts [another AUstralian dealer] says. "They may send technicians out here to
study them, or we can take machines like Digital Equipment or Data General to
pieces and ask them to build revisions into their operating systems." They can
then, he points out, "come back very quickly with these revisions to make their
products more powerful and competitive."
The brash and entrepreneurial approach of medium-sized companies such as Tee
has benefited Japanese suppliers of small- to medium-6ize business equipment. But
in the past, there were too many uncomfortable failures of individual operators
and small systems houses which exploited the market and could not support the
hardware or software they sold.
Today the ground rules are tougher. The Japanese appreciate and work well with
Australian inventiveness, particularly in software.
The Australian supplier, however, has to play with a full hand, which may in-
clude such things as paying in advance for goods or showing proof of a substantial
customer support service.
Fujitsu's shrewd assessments and predictions of the Australian market have made
Digitized by Google
New Product Launch 265
it the bane of IBM Australia. The clever company has filched about a dozen instal-
lation sites away from the mighty mainframer, and it continues to present a
strong challenge.
Facom [another Australian distributer] sends about 20 engineers each year to
Fujitsu for training in advanced technology, so that the local company retains its
high degree of autonomy and independence toward users. Support services are
backed up by MART (Maintenance Assistance by Remote Telecommunications).
Throughout the industry there is little doubt that, especially in the formative
years, the close working partnership between Australian entrepreneurial marketers
and the patient and skillful Japanese adaptation of technology contributed signif-
icantly to the incursion of Japan's electronics goods into expanding world markets.
While today Japan clearly does not depend on the support of any particular coun-
try, its computer makers may still pay homage to the adage that it pays to have a
little bit of help from friends.
The important idea of the ethnocentric section was that the fixed cost
of a launch should not be undertaken unless the risk is low enough. From
the polycentric section the important idea was that of a covariance matrix.
Before launching the product in a market, a marketing manager has
estimates of expected sales and of the variance about that mean. As soon
as the product has been launched in a first market, and actual sales data
are gathered, the manager can update the sales estimate of each of the
other markets (only if the covariance between markets was zero would
there be no value in this information). Nations in which the product is
launched serve as test markets for the remaining nations.
It is desirable to preempt competitors by being first to launch a prod-
uct in a nation. Thus, the decision rule of when to launch in a market
should incorporate p:n, the expected profitability of market m if it is
launched at time t. To calculate p:n the reaction of each significant com-
petitor has to be anticipated. First, some competitors monitor globally
while others have surprisingly parochial headquarters. Represent this
assessment as a matrix of markets and competitors, in which each cell
entry is the subjective probability that launching in that market will alert
a competitor to prepare a rival product. Second, use this relationship to
estimate market shares expected after the introduction of rival products.
The market shares usually decrease with launch sequence, one reason·
able assumption being a linear relationship between log market share and
log sequence; the slope of the linear relationship varies between markets.
Third, from competitive intelligence assess each corporation's current
state of product development and the resources it is likely to devote
to competition on this product. From this, gauge how quickly the com·
petitor will respond once he learns of a launch, and how vigorously he
will buy market share despite being eclipsed by the initial launch. All of
this is required to estimate p:n .
Once the product has been launched in one market, the longer the
corporation waits to launch it elsewhere, the more likely a competitor
Digitized by Google
266 Multinational Marketin,
will be to preempt the next market. Therefore, the longer the innovator
waits, the lower are its expected profits. The rate of drop-off in expected
profits depends on the nation since nations differ in the speed at which
the structure of a new industry is established.
Advertising spills over national borders. Direct spillover is exemplified
by U .S. television commercials seen in Canada, Radio Luxembourg heard
in Holland, and British magazines read in Bermuda. Indirect spillover
occurs when tourists abroad request familiar products or return home
wanting products they sampled abroad. This spillover is apparent when an
advertising agency quotes its costs for two nations at less than the price of
two; whichever is launched second should be charged only the incremental
cost. Advertising spillover from earlier launches reduces the fixed cost of
later launches.
Some characteristics of those nations suitable for the first stage of the
launch are:
1. The international product managers stationed in the ''first-stage''
nations should be familiar with these countries so that they might
better interpret and use the emerging sales figures (or market researchers
should be accessible).
2. The nations should have well-established channels of distribution.
3. The nations should be small enough that the sum of their launch costs
k remains within the budget.
4. Sales within these countries should have high correlations with other
markets so that a given plant capacity yields the most information.
5. For first introductions, the launching corporation should look for
markets to which competitors are blind so that few will get excited.
6. Customers vary by nation in their loyalty to "first launched" brands.
It would be an advantage to enter those nations that form quick brand
loyalties; however, these are frequently those customers who are slow
to try a new product. Hence there is a tradeoff between competitive
and informational objectives.
7. It is desirable to avoid those nations with a high advertising spillover.
Customers affected by the spillover may be annoyed at their disadvan-
tage of not being able to purchase the product. The advertising theme
would no longer be new or as effective once these customers were ap-
proached and the spillover could give competition a head start.
A "real-sized" problem concerns perhaps 25 nations, of which the larger
ones, such as the United States, might be subdivided into regions, to a
total of say 30 markets, with several million different combinations of
launches possible. To find a good solution, we will use a linear program.
(Details are in the Geocentric Appendix.) The advantage of a linear pro-
gram is that most of the pertinent considerations can be considered
simultaneously. Furthermore, the opportunity cost of adopting a politically
more acceptable solution can be calculated explicitly.
Digitized by Google
New Product Launch 267
Ethnocentric Appendix
Instead of fixating on breaking even, let G be the net cash position (at the
end of T). If G is negative, think of a line G units below "cost outflow"
in Exhibit 8.9. We can calculate the probability of achieving at least a net
cash position of G.
A product manager has a certain amount of discretion. Specifically, he
can lose a certain amount of money on a product failure without serious
reprimand. Rarely is it an explicit statement, but a person aware enough
to be a product manager may sense that he can lose, say, a million dollars
(G* = -$1 million) but a greater loss will seriously endanger his job. This
threshold implies that he will launch the market only if:
Digitized by Google
268 Multinational Marketing
log nonnal distribution. Hence there is a different net cash line for each
(E, a). In Exhibit 8.12 the solid net cash line is for the (E, a) of this
market. This market fails to satisfy the chance constraint set by G* and
Q *. There are two ways to express the fact that this market fails to satisfy
the chance constraint:
1. At the threshold probability Q* the net cash flow is G* - G~ap.
2. Sales d are forecast. For the same expected sales E. there IS a standard
deviation agoal that would just satisfy the chance constraint. The net
cash line (E, agoal) is shown as a dashed line.
Both of these expressions are depicted in Exhibit 8.12. Let us develop
the mathematics for Exhibit 8.12 by following part of Charnes et al.
(1968) DEMON, but with a final net cash position of G instead of O.
The constraint from Exhibit 8.12
~~~----+- __ a*
100
-k G* o Net cash G
minimal
goal
EXHIBIT 8.12
Market forecast (E. 0) transformed into probability (l * of achieving
net cash position G.
Digitized by Google
New Product Launch 269
In (k + G/7rT) - E .,;;:: P
-------...... I - ac
a
E + PI _ ac a ~ In (k + G) -In 7r -In T
All the current information about the market is contained in (E, a).
Market research or launches in similar nations could add information that
will revise both E and a. But now, before the information comes, our
preposterior estimates of the revisions owing to market research (or
launches in similar nations) will be a reduction in standard deviation to
a'. No change in the mean E can be forecast (we guess it will rise or fall
but do not know which). The difference between the G contour through
(E, a) and the G contour through (E, a') is the value of this market
research information. For example, in Exhibit 8.12 the value of market
E
G=2 Launch market if
G = 1 G* its (E, u) is above
frontier.
G=O
Expectation G = -1 = G*
of In sales G =-2
~-------------------------------------u
Standard deviation of In sales
EXHIBIT 8.13
Parameterizing on net cash position G with threshold probability ac*
held constant.
Digitized by Google
270 Multinational Marketing
Geocentric Appendix
Expected
Sales E
Target for period 2
Target for period 1
- - - - - .. - - - - - - _ u2 initial
u 2 after
period 1
1----I-g2:--_
~ _ _ _ _ g1 _______
Digitized by Google
New Product Launch 271
Actually, this presupposes that all the information for a market becomes
instantly available. Realistically only the fraction fJn of the information
could be collected in market m and processed by market research within
one period of the launch, a further (f" within the second period from the
launch, ... , subject to 1: t fI = 1. Thus
Our concern is with the variance "gap." Thus the first period's equations
are simply
gl = 0fnitial - target
The second period equations include both the insight from fIrst period
launches and the movement of the target:
In other words,
xl I f2~ + x2 fl~ _ g2 + g3
I =target 2 - target 3
In Exhibit 8.15 the group of equations whose right-hand side is 1, con-
strains to 1 the probability that a market gets launched. The group of
equations whose right-hand side is 0 assures that a market pays a G cost
if it launches in that period, but not otherwise.
The inequalities should not be necessary but are merely a prudent cau-
tion that would flag attention and simultaneously avoid preventing earlier
launches in case some error in ~ should accumulate to a negative variance.
Production limitations are frequently important, as they were in the
Instamatic example mentioned previously. The capacity of each plant
Digitized by Google
272 Multi1UJtional Marketing
Maximize
Subject to
Xl X2 X3 X4 gl g2 g3 g4 r.h.s.
Initial Target
I
variance
- 1
-I I Target Target
fIe 1 - 2
Target Target
f2 e fIe -I I 2 - 3
Target Target
f3 e f2 e fIe -I I
3 - 4
I I I I 1
I -I 0
I -I 0
I -I 0
I -I 0
EXHIBIT 8.15
Sequencing as a mathematical program.
Digitized by Google
Questions from Other Viewpoints
1. Corporate Rational Normative Global. You just covered this view-
point in the chapter.
2. Corporate Rational Normative Subsidiary. Evaluate the suggestion
that a corporation should test market in a few cities of a nation as
part of staging nations for a new product.
3. Corporate Rational Descriptive Global. Read and summarize the
analyses of a product life cycle contained in Louis T. Wells, ed.,
The Product Life Cycle and International Trade (Boston: Division
of Research, Harvard Business School, 1972) or Louis T. Wells, "Test
of a Product Cycle Model of International Trade: U.S. Exports of
Consumer Durables" (Quarterly Journal of Economics, February
1969) or William H. Gruber, Dileep Mehta, and Raymond Vernon,
"The R&D Factor in International Trade and International Investment
of United States Industries" (Journal of Political Economy, February
1967).
4. Corporate Rational Descriptive Subsidiary. What methods can be
changed to speed the flow of accurate reports on actual customer
sales which are needed for market staging?
5. Corporate Emotional Normative Global. Most of the "data" for this
problem are subjective impressions. What mathematical and organiza-
tional weight would you give to these subjective impressions, in that
each "expert" knows only some nations and some products, and that
experts differ in their expertise?
6. Corporate Emotional Normative Subsidiary. All the discussion of
adaptive sequences, or keeping the sequence the same but advancing
or delaying launch dates, has implications for a subsidiary. Specifically,
the subsidiary sales managers do not know when or whether they will
have to launch the product - predictions are constantly being redated.
Hence when called on to act they will have to do so with little notice.
Suppose you are the subsidiary marketing executive. Layout a specific
program of preparing your sales managers. Be specific about the state
of psychological readiness in which they should hold themselves.
7. Corporate Emotional Descriptive Global. Many historians of World
War I have commented that declarations of war became irrever-
sible once mobilization had been ordered. Railroads had been
scheduled to converge troop trains, weapons, ammunition, food,
horses, and feed, all from different origins, and these schedules were
too intricately interconnected to change. Suppose someone "solved"
the launch problem for you. Describe the rules you would enforce
that would modify the schedule to make it more ready for the
unexpected and simpler to manage.
8. Corporate Emotional Descriptive Subsidiary. Describe the emotional
278
Digitized by Google
274 Multinational Marketing
Digitized by Google
New Product Launch 275
spare parts are frequently unavailable unless the design sells well.
Draft legislation to compensate your citizens.
15. Societal Emotional Descriptive Global. You are a theoretician for a
particular kind of "ism" (for example, communism or capitalism)
with global aspirations. How does your problem differ - or does it -
from the one described in this chapter?
16. Societal Emotional Descriptive Subsidiary. List the qualities needed
for a nation to be first to receive a new product. Launching new prod-
ucts brings to the nation the products themselves (with their con-
sequences) and the scouting and pressure of determined executives.
Will the long-run effect of these be a slight enhancement or a slight
diminution of the qualities that made that nation first?
Bibliography to Chapter 8
Ayal, Igal, "Simple Models for Monitoring New Product Performance," Decision
Sciences, Vol. 6, No. 2 (1975), pp. 221-236.
Boston Consulting Group, Perspective on Experience (Boston, Boston Consulting
Group, 1970).
Charnes, Abraham, William W. Cooper, James K. DeVoe, and David B. Leamer,"DE-
MON Mark 11: An Extremal Equation Approach to New Product Marketing,"
Ma1Ul(lement Science, Vol. 14, No. 9 (1968), pp. 513-524.
Cyert, Richard M., Morris H. DeGroot and Charles A. Holt, "Sequential Investment
Decisions With Bayesian Learning," Management Science, Vol. 24, No. 7 (1978)
pp. 712-718.
De Groot, Morris H., Optimal Statistical Decisions (New York: McGraw-Hill, 1970).
Dillon, William R., RogerCalantone, and Worthing Parker, "The New Product Problem:
An Approach for Investigating Product Failures," Management Science, Vol. 25,
No. 12 (1979), pp. 1184-1196.
Ehrenberg, A.S.C. and G.J. Goodhardt, "A Comparison of American and British
Repeat Buying Habits," Journal of Marketing Research, Vol. 5 (February 1968),
pp. 29-33.
Green, Paul E. and Donald S. Tull, Research for Marketing Decisions (Englewood
Cliffs, NJ: Prentice-Hall, 1978).
Heeler, Roger M. and Thomas P. Hustad, "Problems in Predicting New Product Growth
For Consumer Durables," Management Science, Vol. 26, No. 10 (1980), pp. 1007-
1020.
"How Kodak Clicked Worldwide on the Marketing Plan of its Famous Instamatic,"
Business Abroad, Vol. 31, No. 11, June 12, 1967, pp. 31-34.
Keegan, Warren, "Multinational Marketing: The Headquarters Role," Columbi4
Journal of World Business, Vol. 6, No. 1 (1971), pp. 1-6.
Kemp, Norman, "Japan's Amicable Alliance," Datamation, Vol. 26, No. 2, September
1980, pp. 74-82.
Digitized by Google
276 Multinational Marketin,
Martin, Warren S. and AI Barcus, "A Multiattribute Model for Evaluating Industrial
Customer's Potential," Interfaces, Vol. 10, No. 3 (1980), pp. 40-44.
McConnell, J. Douglas, Behaviorol Factors in World Consumer Markets (Palo Alto,
CA: Long Range Planning Service, Stanford Research Institute, 1970).
Moyer, Reed, "International Market Analysis," Journal of Marketing Research, Vol. 5
(November 1968), pp. 353-360.
Raiffa, Howard and Robert Schlaifer, Applied Statistical Decision Theory (Cambridge:
Harvard University Press, 1961).
Robertson, Thomas S., D.J. Dalrymple, and M.Y. Yoshino, "Cultural Compatibility
in New Product Adaptation," Journal of Marketing, Vol. 34, No. 1, January 1970,
pp. 70-75.
Sethi, S. Prakash, "Comparative Cluster Analysis for World Markets," Journal of
Marketing Research, Vol. 8 (August 1971), pp. 348-354.
Sommers, Montrose and Jerome Kernan, "Why Products Flourish Here, Fizzle There,"
ColumbillJournal of World Business, Vol. 2, No. 2, March-April 1967, pp. 89-97.
Staelin, Richard and Ronald E. Turner, "Errors in Judgemental Sales Forecasts: Theory
and Results," Journal of Marketing Research, Vol. 10 (February 1973), pp. 10-16.
Terpstra, Vemon, International Marketing (New York: Holt, Rinehart and Winston,
1972).
Urban, Glen L. and John R. Hauser, Design and Marketing of New Products, (Engie-
wood Cliffs, NJ, 1980).
Wells, Louis T.,Jr., "A Product Life Cycle for International Trade," Journal of Market-
ing, Vol. 32 (July 1968), pp. 1-6.
Digitized by Google
CHAPTER NINE
Pricing
277
Digitized by Google
278 Multinational Marketing
Digitized by Google
Pricing 279
Background of Genag
Although slightly larger financially than Tyler, the sales of
Genag were more concentrated both geographically and by
product.
Genag was one of a number of full-line farm equipment
manufacturers in the world. Although the company mar-
keted and/or produced farm equipment throughout the
world, most of its production, sales, and profits were derived
from the Vnited States and Canada, although the contribu-
tion from foreign sales had been growing recently, from 8
percent in 1976 to 35 percent in 1980.
In addition to its V.S. operations, Genag owned and operated
plants in Australia, Argentina, Canada, France, Germany,
Mexico, and South Africa.
A highly centralized organization itself, Genag believed
there were many benefits to this form of organization on a
worldwide basis and often extolled the merits of this approach.
In the late 1960s and early 1970s, Tyler enjoyed the major
share of Genag's grinding wheel business, having an estimated
40 to 50 percent share. By 1980, Tyler's share was down to
about 20 percent. The factors that affected this decline in-
cluded a realization by Genag that it was too heavily com-
mitted to one supplier. Also during this time, Genag, while
Digitized by Google
280 Multinational Marketing
Digitized by Google
Pricing 281
EXHIBIT 9.1
Tyler Abrasives Inc.
Actual and Potential Sales to Genag ($000)
UNITED STATES
Actual sales
to Genag 1977 1978 1979 1980 1981
Vitrified $228 $200 $258 $212 $190
Organic 154 144 180 115 102
Diamond 103 89 171 171 128
Total $485 $433 $609 $499 $420
Total potential
sales to Genag $1,680 $1,800 $2,150 $2,500 $2,550
Actual as per-
centage of
potential 29% 24% 28% 20% 17%
FOREIGN
Actual sales
to Genag 1977 1978 1979 1980 1981
Canada $39 $ 67 $ 73 $ 35 $ 4
Germany n.a. 67 74 66 61
France 22 35 47 51 57
South Africa 1 1 2
Argentina 12
Australia 19 29 23 29 25
Total $80 $198 $218 $182 $161
Potential sales
to Genag 1977 1978 1979 1980 1981
Canada $84 $112 $154 $162 $105
Germany n.a. n.a. 305 325 350
France 48 77 69 105 119
South Africa n.a. n.a. n.a. n.a. 20
Argentina n.a. n.a. n.a. n.a. 20
Australia 28 37 32 39 35
Total n.a. n.a. $562 $633 $673
Digitized by Google
282 Multinational Marketing
Digitized by Google
Pricing 283
August 1981 the pattern for the position outside the United
States could be estimated as follows:
EXHIBIT 9.2
Tyler Abrasives Inc.
Relative Cost/Price Position
Index (per unit price sold United
direct to Genag) States France Germany Australia
Price indices
Vitrified bore wheels
t
It X X t Qty. 25 100 79 90 n.a.
19A70-N6VBE Qty.150 100 76 105 104
2X 1 X t Qty. 25 100 76 101 n.a.
19A70·N6VBD Qty.150 100 65 115 113
Vitrified centerless wheels
20 X 6 X 12 Qty. 1 100 43 61 49
A80-06V Qty. 10 100 48 61 57
Cost indices
Alundum grain,landed cost/wheel 100 60 77 80
Direct labor cost per wheel Spec lA 100 54 258 269
1B 100 106 198 190
2 100 47 294 192
Digitized by Google
284 MultiMtional Marketin,
We should not have to tell you what your supply problems are
or what kind of proposals you should make. This is your job as
a supplier. You must do a better and more consistent job of pricing
your products and you must also improve your standing in tenns
of service and interest before we can discuss any possibility of giv-
ing you an increased share of our business.
Digitized by Google
Pricing 285
Genag not only seemed to want improved prices, but closer multi-
national coordination on R&D and technical problems as well. Mr.
Home kept stressing that some of our multinational competitors
are already beginning to offer such coordination. He suggested that
a company such as ours can't really be multinational and still afford
to have decentralized decision making on the major points of pro-
duct and marketing policy. In my opinion, however, the critical
issue is how far to centralize and what parts of the operation to
centralize for maximum flexibility.
You see there are still a lot of customers where the buying choice
is technologically oriented. Over the years, our salesmen have built
a strong bond with the people in manufacturing in their companies.
It's more than good customer relations and it's stronger than is
needed for technical liaison, it's friendship - these people really
are good friends.
Digitized by Google
286 Multinational Marketintl
Sales Overseas
John Stinson, group vice president of the International
Division, felt that it was necessary for a multinational com-
pany to recognize the sensitivity of operating in an inter-
national climate. Too many multinationals, he said, knew too
little about their subsidiaries or the markets in which they
operated; Tyler intended to make sure they did not fall into
that category.
I've been abroad recently and begun to realize the depth of emo-
tions in this area. Originally the subsidiaries were completely auto-
nomous. Consequently, the more we try to achieve international
coordination the more they begin to kick about what they con-
sider "crazy American ideas." When a buyer in a foreign country
talks to one of our salesmen they both talk about "U.S. Imper-
ialism." We have to realize that the French and Germans still look
on themselves as French and German companies. However, we
can get them to see that it's the customer who calls the shots. It's
all a question of good communications.
Digitized by Google
Pricing 287
Digitized by Google
to:I
00
00 EXHIBIT 9.8
Tyler Abrasives Inc.
Financial Effect of Revised Farm Equipment Industry Prices
on Sales to Genag
Sale. Income (FiacaI1981; $000)
Before RelJiaion After RelJiaiona
Up to OlJer Up to OlJer
2 in. 2-14 in. 14 in. 2 in. 2-14 in. 14 in. Gain/
Diameter Diameter Diameter Total Diameter Diameter Diameter Total (Lo,,)
Germany 12 28 21 61 6 28 21 55 (6)
Canada 0 0
France 20 15 22 57 20 15 22 57
Australia 17 3 5 25 9 3 6 17 (8)
Argentina 10 2 12 10 2 12
Total 59 46 60 155 45 46 60 141 (14)
o
cO"
""
N"
(J)
Q.
cr
'<
c;
o
a-
( i)
Comparative Price Indices (United States" 100)
Before Revision After Revision
Up to Over Up to Over
2 in. 2-14 in. 14 in. 2 in. 2-14 in. 14 in.
Diameter Diameter Diameter Diameter Diameter Diameter
Germany 190 55 55 100 55 55
Canada 100 90 90 100 90 90
France 60 60 60 60 60 60
Australia 190 70 60 100 70 60
Argentina 60 75 75 60 75 75
Ratio of
Highest/
Lowestb 3.2/1 1.8/1 1.8/1 1.7/1 1.8/1 1.8/1
°Tbese estimates assume no cbange in tbe unit volume of businea with Genag.
bTbis ratio is tbe world price spread from bighest Tyler price to lowest Tyler price for any given size and specification of grinding wbeel.
0
cO"
""
N"
(J)
Q.
cr ~
'<
c; 00
<D
0
a-
( i)
290 Multinational Marketing
EXHIBIT 9.4
Tyler Abrasives Inc.
Indices of Popular Genag Prices
United
Product States Canada Argentina Germany France Australia
No. 1 100 118 197 74 109 117
No. 2 100 108 197 140 107 92
No. 3 100 123 n.a. 27 115 95
No. 4 100 156 n.•. 26 95 156
No. 5 100 120 n.a. 66 102 44
No. 6 100 119 n.•. 70 95 111
No. 7 100 120 n.•. 110 108 113
No. 8 100 120 n.•. 66 108 91
No. 9 100 111 n.•. 86 107 110
No. 10 100 109 n.•. 72 111 111
Summary
With the foregoing information in hand, Spencer and Stinson
were trying to decide whether to approve the plan proposed
by their staff or, if not, what alternative policy to adopt.
In reaching a decision, Spencer and Stinson felt they ought
to consider the potential impact on Tyler's relationships with
other multinational customers. If a worldwide price were
offered to Genag, they believed that once the word got
around Tyler would have little choice but to offer similar
terms to other customers that bought from Tyler in several
different countries and that individually accounted for at
least $400,000 in annual sales. There were approximately
50 customers, collectively accounting for about $25 million
in sales that fell in this category. Although Spencer and
Stinson had no way of knowing the exact sales breakdown
of these companies by country and product category, they
believed that the overall sales configuration might be fairly
similar to the pattern found at Genag.
One factor that complicated the information-gathering
and decision-making process in this situation was Tyler's
Digitized by Google
Pricing 291
current organizational structure. Becuase Tyler's interna-
tional operations were contained in a separate division and
because foreign subsidiaries had traditionally been run on a
relatively decentralized basis, it was difficult to coordinate
internally any analysis on a worldwide basis.
Moreover, Spencer was aware that not all subsidiary man-
agers would be enthusiastic about the idea of quoting world-
wide prices to Genag. In fact several managers had already
expressed concern. For example, the manager of Genag's
Australian distributor had indicated that he was against
the proposal inasmuch as Australia "runs a strong operation
and already has 70 percent of Genag's local business despite
relatively high prices. Consequently, any move toward world-
wide prices would probably only serve to reduce the profit-
ability of the Australian distributor operation."
The manager of the French subsidiary on the other hand
was against any worldwide move that would tend to increase
French prices since he felt Tyler's present low level of prices
in France was absolutely necessary to maintain market share
in a highly competitive market. Finally, the manager of the
German subsidiary said he could understand the reason for
a worldwide pricing agreement insofar as it applied to Genag,
but he was very concerned about the possible impact should
the agreement become public knowledge within the industry.
Consequently, he had written to headquarters as follows:
"We feel that it is of the utmost importance that knowledge
of any agreement be regarded as a highly confidential matter
and restricted to the smallest number of people within Tyler
and Genag. I would personally appreciate it if you would is-
sue very strong instructions to this effect."
Questions
1. Would you describe the Tyler organization as polycentric,
ethnocentric, or geocentric? Explain.
2. Would you describe the Genag organization as polycentric,
ethnocentric, or geocentric? Explain.
3. Do you think that Tyler is currently setting prices and
sales efforts optimally?
4. For small changes in the quantity to be manufactured,
would you expect the marginal cost to be altered substan-
tially?
5. Elasticity of demand is defined as the percentage change
Digitized by Google
292 Multinational Marketing
of quantity demanded per percentage change of price.
What is your estimate of demand elasticities in Germany,
France, and Australia?
6. Genag buys only a few of the many products produced
by Tyler. If the price of the items purchased by Genag is
changed, sales of neigh boring products would be affected
as customers switched to or from the changed prices. How
could Tyler predict the switch quantity?
7. How much money should the Tyler management be willing
to invest to study the rationalized pricing scheme so as to
intelligently respond to Genag's initiative?
Digitized by Google
Introduction
This chapter is addressed directly to corporations dominant enough to
be the price setters for certain products. The price takers who exist under
the price umbrella of the dominant competitor should clearly understand
those forces that hold up the umbrella so they can realistically assess the
risks of their environment.
In this chapter, we investigate pricing considerations from the view-
points of oligopolistic competition, legal constraints in various nations
that affect sales price, and the relationship between administrative co-
herence and the taking advantage of market opportunities. We are not
concerned with transfer prices, for they may be adjusted to maneuver
liquid assets as described in Chapter 3. Nor are we concerned with raw
material commodity prices, though some of the concepts from Chapter
2 may be useful guides.
Most of today's worldwide manufacturers were once domestic companies,
vying with their home competitors for a share of a market that seemed
too small. When one of the competitors expanded abroad, it changed the
basis for competition. As was described in Chapter 1, the remaining com-
petitors had to choose between specializing in narrow market niches or
also expanding abroad. Knickerbocker (1973) has shown that a powerful
motivation to become multinational is competition - that is, the com-
petitors have become multinational.
Oligopoly theorists would say that market share is valuable. But the
share of which market? Mason (1969) presents a case study of an abrasive
manufacturer trying to decide whether abrasives were becoming a com-
modity (in which case its strategy should be to go multinational so as to
increase its share of the worldwide market) or whether abrasives were be-
coming individually tailored for particular applications (in which case its
strategy should be to stress R&D and upgrade its sales force by hiring
technical specialists who would be able to translate peculiar cutting prob-
lems into particular chemical compositions). Either strategy can be viable;
both can coexist. Multinational corporations are those that have adopted
the former: there are benefits to standardizing their product around the
world. In the case you have just read Genag certainly adopted that strategy
and is attempting to force it onto its suppliers such as Tyler.
In most nations one sees multinational corporations in competition with
local businesses. The locals do not wither away (unless they foolhardily
pit themselves in head-on competition with the multinationals). Rather,
the two groups compete on different bases, with the local firms special-
izing in niches that require individually tailored responses to the market
so that the multinationals cannot compete. Expressed differently, each
may strive to maximize market share, but they are really working within
different markets.
293
Digitized by Google
294 Multinational Marketing
Umbrella Pricing
The price leader sets a price umbrella for the industry. Should it hold
this umbrella price high enough to let the marginal competitor survive? In
the U.S. automobile industry, Chrysler is the marginal competitor. In ad-
dition to obvious antitrust reasons, General Motors may perceive an eco-
nomic rationale for not pricing Chrysler out of existence. In contrast,
Texas Instruments chose to reduce prices of digital watches, steadily
squeezing out marginal competitors until there were only three effective
companies left in the business. Texas Instruments may also perceive an
economic rationale. How many competitors does a market leader prefer?
The price leader, Firm 1, knows that its market share will decrease with an
increasing number of competitors in the industry (Ijiri and Simon, 1977).
By umbrella pricing, the leader influences this number.
To develop an economic (not antitrust) rationale for umbrella pricing
we will make two assumptions. First, if the price umbrella is lowered be-
low the unit cost of the marginal producer, he will quit and the remaining
rivals will grab his customers in proportion to their preexisting market
shares. Second, because of the experience curve effect, the inflation-
adjusted unit cost experienced by a firm decreases over time (and decreases
faster with a larger market share). Learning by doing and economies of
Digitized by Google
Pricing 295
100
Market
share of
firm 1
" ...............
----
n The number of
2 3 4 5 6 firms in the
industry
EXHIBIT 9.5
Market share of Firm 1 decreases with number of firms in the industry.
P
Price
Unit cost
of firm 3
Unit cost
of firm 2 Industry demand
curve
~---------------------------------------d
Market rate of
Demand
EXHIBIT 9.6
Demand curve seen by Firm 1 depends on its market share of the
industry demand.
scale mean that tightly managed larger firms can usually produce at a
lower unit cost than their smaller rivals.
Industry demand increases when the umbrella price is lowered. Because
the marginal competitors will drop out if the price is set below their unit
costs, an insight into these unit costs allows the lead firm to calculate how
much its market share would increase with a given decrease in price. Thus
it may calculate its demand curve. The managers of Firm 1 can analyze
each segment of this demand curve and, knowing their own unit cost
cl' they can calculate a price that maximizes Firm 1 's profit. (Each
sloped segment of the Exhibit 9.6 demand curve of Firm 1 should be
analyzed separately.) To draw Exhibit 9.6, Firm 1 requires excellent
knowledge of the cost structures of its rivals. In corporate practice, the
leader gathers extensive intelligence about the financial health of marginal
Digitized by Google
296 Multinational Marketing
Well-managed costs not only yield better current profits but allow the
price leader flexibility in coping with rivals, both potential and existing.
The way to deal with rivals differs among national markets. It depends on
a corporation's rank in this industry, the number of competitors, and the
legal, commercial, and social means in which existing competitors vie with
one another. The government and other stakeholders in each nation set
many constraints on pricing policy.
In the United States, the Justice Department is especially watchful for
monopoly power. The Aluminum Company of America was cited in a
1945 Supreme Court decision that is one of the landmarks in U .S. anti-
trust law. Alcoa sold aluminum ingots as well as sheet aluminum, but it
set its prices for ingots - of which it was by far the lugest supplier - high
enough so that other firms could barely afford to buy the ingots, convert
them into sheet, and compete with Alcoa in the sheet metal market. This
"Price Squeeze" was held to be an "unlawful exercise of Alcoa's power."
In some other nations it is common practice.
Although U.S. law is by far the most restrictive in the world, other
nations do have similar provisions. In West Germany, retail price main-
tenance was forbidden as of January 1, 1974, and a ban on concerted
practices is left open to wide interpretation by the courts, including re-
strictions on parallel pricing. Previously, in 1972, Germany's Federal
Cartel Office had fined nine synthetic fiber companies more than $15
million for domestic and export price fixing, among other charges.
Digitized by Google
Pricing 297
Any price changes incur a fixed cost. All salesmen and other persons
who deal with sales figures must be infonned of any price change as rapid-
ly as possible. The cost of this worldwide education skyrockets with fre-
quent changes, not only because they have to keep memorizing new figures
but also because the assurance of a salesman is undermined if he cannot
remember the latest price.
Digitized by Google
Data Required for Multinational Pricing
Marginal Cost
Let us develop pricing theory for one nation. For each item in the line,
the most likely price and marketing budget is chosen and with these the
life cycle curve is estimated for each of the products. The areas under
these curves represent the expected total sales. After modeling the produc-
tion process, we can approximate the marginal cost of the product.
In this chapter, the first step is to calculate c, the landed direct cost of
each item in each nation (recall Chapters 6 and 7). This vector of costs is
the only source of interconnectedness in the polycentric pricing scheme.
Then for each nation separately, calculate optimal price adjustments for
items in the line following the steps of Appendix 1. The base case profit-
ability equals the revenue minus production and marketing costs. The
intent is to adjust the prices and marketing budgets to increase profit-
ability. To do this, we must consider price elasticities of demand among
the products as well as their responses to the marketing expenses.
Digitized by Google
Pricing 299
Sales
rates
/ I
Marketing
expenses
EXHIBIT 9.7
Demand responses to marketing budget.
been launched and has a budget for sustained advertising. Usually such
budgets are sufficiently large to be in the region of decreasing marginal
effectiveness seen in Exhibit 9.7. An estimate of the shape is required
for Appendix 1.
Digitized by Google
300 Multinational Marketing
Digitized by Google
Pricing 301
r-----------------------------,$~
Canada
Germany
Italy
~--------~~----~~~~~~--~$~
Germany Britain
Canada
France
Italy ~:...-~,,---
Sweden
Britain E~===~~~~~3..l...._ ________
.....L..__________..I$2ooo
~~ ~~ ~~ ~~
2000 3000 4000 5000
EXHIBIT 9.8
Wholesale prices for Ford tractors with eight-speed transmissions,
1966-1967.
to be only $110 for each Ford 2000 or 3000, and $144 for each Ford
4000 or 5000 (because Britain at that time was not a member of the EEC,
the cost of shipping a Ford 2000 across the channel to France would have
been $421).
Arbitrage opportunities like this sour dealer relations and invite govern-
ment attention. An analyst should be able to calculate the reduction in
model profits that would result from raising British and lowering Canadian
prices to eliminate the arbitrage "opportunity."
By raising the British price, Ford would forgo the profits shown by the
lower left to upper right curve in Exhibit 9.9. By lowering its Canadian
price, Ford would forgo profits shown by the lower right to upper left
curve. The temptation to arbitrage could be eliminated by raising the
British price and lowering the Canadian price so that the adjustments
eliminate the price gap. The combination giving the lowest total forgone
profits is at the minimum of the line called "sum of costs. "
Digitized by Google
302 Multinational Marketing
Profit foregone
by lowering
Canadian prices
Profit foregone by
raising British
prices
I
I
f - - - - - . British price rise I
I
Canadian price cut ...- - - - - - - l l
~I·----Arbltrage----.....·~I
opportunity
EXHIBIT 9.9
Model of profits forgone if arbitrage is eliminated.
Digitized by Google
Pricing 303
The standard worldwide base price is most likely to be looked upon by manage·
ment as full·cost pricing, including an allowance for manufacturing overhead, gen-
eral overhead, and selling expenses. Often ignored are (1) the necessarily arbitrary
nature of these cost allocations, (2) differences in costs from market to market
(in labor, capital, materials, and management in the case of overseas manufacture;
in shipping, crating, insurance, tariffs, taxes, internal transport, distribution, and
promotion in the case of exporting), (3) possibly lower marginal cost of goods
moving into foreign markets (particularly in reference to domestically oriented
research and development), (4) differences in competitive position within the
foreign markets, (5) differing degrees of optimum penetration for different foreign
markets, (6) price controls enforced by a government or by a dominant supplier.
Digitized by Google
304 Multinational Marketing
exports remains small. This pricing scheme may also be necessary for ex-
pensive, compact products that have unequivocal specifications and are
sold to multinational customers, whether corporations, trading companies,
or individuals. It may be a necessary pricing scheme if the product be-
comes embroiled in issues of equity.
Ethnocentric pricing helps avoid some of the problems associated with
differential pricing. For example, an intemational computer company
might sell to an intemational oil company. The latter may accept higher
prices in Bahrein, reflecting the higher cost of doing business there, but
will object if prices are disproportionately higher on some items than on
others. (The company will hardly be mollified by a lecture on the econo-
mics of a discriminating monopoly.)
For another example, a large V.S. steel corporation sustained its entry in
the European market by absorbing freight and duty on its export ship-
ments. One customer was the German subsidiary of a V.S. auto manu-
facturer. The Detroit plant of the automaker demanded the same treatment
from the steel company. The firm abandoned the European market.
A further problem with polycentric pricing is the cost of oversight. An
electronics firm, for example, might sell 10,000 products in 75 nations
(the model analysis would not be done on so large a problem). If head-
quarters is to maintain any degree of control, keeping price books be-
comes a huge data processing job as national subsidiaries experience varying
degrees of inflation. Some system is needed to preserve control and to
give guidelines for evaluating subsidiary performance.
In nation n, the initial vector of prices is P~. In nation m, the initial
prices may be different, say P9,. The difference in the prices (Pn - Pm) is,
however, fixed and predetermined (such as freight plus import duty). This
difference in prices could be set equal to zero. When the line of prices is
changed by Il.p in one nation, the Il.p change in prices has to be made in
all other markets. Appendix 3 shows how to calculate the Il.p optimal for
global profitability. With an ethnocentric corporation, the natural tempta-
tion is to optimize the home prices using home data; these home prices
then become the export prices. Sometimes a higher or lower export price
would be preferable. Appendix 3 raises or lowers the home price because
of price considerations in all the markets. It is therefore susceptible to
the accusation that "the tail is wagging the dog." This will hardly occur
if the foreign markets are tiny. If it does occur, and is significant, it may
be better to convert to the geocentric pricing scheme. Nevertheless, as
Kolde (1968, p. 403) wrote:
The conversion from rigid fUll-cost export pricing [ethnocentric] to demand based
pricing [geocentric] has been generally much slower than can be justified in the
light of prevailing market conditions. Too many executives are (self) righteously
resisting any changes in their unilateral "right" to prescribe the price for all their
markets .... They have, literally, priced themselves out of the market, not be-
cause the market price was too low, but because their pricing policies and strat-
egies were based on the wrong premise.
Digitized by Google
The Three Pricing Schemes Discussed
All three pricing schemes require identical input information - derived
from judgment, competitive insight, test marketing, and regression analysis
on similar products. Regardless of the input data precision, the three
schemes are ranked in order of their model profits, 'Ir > 'Irg > 'Ire. The
polycentric scheme has the fewest constraints. It may be that its optimal
prices will fall into line so that there is a constant percentage markup for
all goods in each nation, but this is unlikely. Imposing the geocentric
constraint of a constant percentage markup for goods sold in each nation
will reduce model profits, hence 'lrp > 'lrg • Similarly, with geocentric pric-
ing all nations might have identical percentage markups over cost, but this
too is unlikely. Imposing this ethnocentric constraint will reduce model
profits. Hence 'lrg > 'Ire. Reflecting on the differences between the profita-
bilities ('IrP - 'Irg) and ('Irg - 'Ire) may help corporate executives select a
better pricing scheme.
Collecting data for a multinational marketing model is very expensive.
The first runs should incorporate data for only part of the product line,
and only some of the nations; even so, the available data generally are sub-
jective modifications of Comanor and Wilson (1979) statistical estimates.
Once the model has run, it can be used to "bootstrap" its way toward
better data by means of sensitivity analysis. One input coefficient is con-
sidered, say a price elasticity between items i and j (that is, Pjj ). Being un-
sure of the number Pij we vary it and determine the change in profit.
Ideally, while an analyst is estimating Pjj , he should also guess at the
percentage change in Pjj that might come from another, say, $250
further investigation. Then the computer can do a sensitivity analysis
over this coefficient range, so as to rank all the input coefficients as
to the cost-benefit trade-off of further investigation. This should be
repeated after each round of investigation, because the new coefficients
alter the slope and position of the sensitivity lines, and the work of inves-
tigation alters the analyst's cost estimates.
BehaIJioral Aspects
Once a pricing scheme has been chosen, each national subsidiary should
set its own marketing budget. These will probably be set subjectively, if
only because of the complexity of advertising carry over from one period
to the next. For purposes of headquarters control, marketing goodwill is
a crucial (though difficult to measure) gauge of managerial performance.
In the administration of polycentric pricing, headquarters has no direct
involvement, except for setting the landed cost c. To administer geocen-
tric pricing, headquarters must set the base prices but can leave the sub-
sidiaries to decide on their markup weights and marketing budgets. With
ethnocentric pricing headquarters issues detailed prices to each national
subsidiary .
305
Digitized by Google
306 Multinational Marketing
Digitized by Google
Pricing 307
where demand dO = t(p° ,m0), l' = [1,1,1, ... , 1), and the superscript prime means
transpose.
Incremental adjustments can now be made to the vector of prices and marketing
budgets so that these become p = pO + f¥J and m = mO + ~, respectively. The
intent is to optimize f¥J and ~m.
After adjustments f¥J and ~ the profitability becomes
fr = (po + f¥J)(~ + ~) -l'(mO + ~) - c(do + ~)
where dO + ~ = t(p° + f¥J, mO + ~m).
The inOuence of price or sales is summarized as linear demand curves and cross-
elasticity effects in the I X I matrix P in Exhibit 9.10 so that
= P
EXHIBIT 9.10
~ = M +
q
EXHIBIT 9.11
Digitized by Google
308 Multinational Marketin,
where q is an estimate of the decreasing marginal effectiveness of the marketing
expenditures portrayed in Exhibit 9.7.
Optimization
The vectors of changes /:lp and am will cause a change in daily profits tJ.:rr.
Price
Cf--------
dO + ~d dO Quantity demanded
(where ~d is a negative quantity)
EXHIBIT 9.12
Change in profits excluding marketing.
Thus
Some of these tenns are linear (such as /:lpdO), others are quadratic (such as /:lp'P/:lp)
and a few such as /:lp' am' q I::t.m are so small we discard them. Let us separate the tenns
of I::t.rr into two groups so that, Exhibit 9.13:
I
I M
P
I
---1--- +
o I (p°-c)'q
~
/:lp' I::t.m'
EXHIBIT 9.13
Digitized by Google
Pricing 309
x* = -[G + G'] -I h
On resubstituting for x, G, and h we get the computational scheme of Exhibit 9.14:
I -1
I
p+p' I M (p°-c)p
!::J.p* +do
I
~--- = 1----, ---- I
I (p°-c)'q
----
(p°-c)M
!::J.m* M' I -1'
+q'(p°-c)
I
I
EXHIBIT 9.14
The superscript -1 means calculate the inverse of the matrix, for which there are
standard computer programs. (A sophisticated analyst could reduce computation time
by finding the inverse by parts since the off-diagonal matrices [M and M] are almost
identical.)
N N N N
l; !::J.rr = l; /lpn (d: + !::J.dn) + l; (P: - c)' !::J.dn - l; l' !::J.mn
n=l n=l n=l n=l
N
= l; (w:!::J.p + !::J.wnpo + !::J.wn !::J.p)[d: + Pn(w:!::J.p + !::J.wnpo + !::J.wn !::J.p)
n=l
Digitized by Google
310 Multinational Marketing
N
+ Mnt:.m n - t:.m~qnt:.mn] + l; (w:po - cn>'[Pn(w:Ap + Ilwnpo
n=l
N
+ .£lwn.£lp) + Mnt:.m n - t:.m~qn t:.m n ] - l; 1't:.mn
n=l
As in Appendix 1 we will collect tenns to create .£l1r ="'G,, + hx, discarding the few
tenns that are higher multiples of various .£l. If there are two nations, ,,' = [Ap,Ilw,
.£lml, .£lm 2 ]' a vector of length 1+ (n -1) +I +I. The profit maximizing adjustment is
Digitized by Google
I I I
2 I I I 2
I (w!I'P. I 2 I I I
"=1
w!(w:PO-c.)p.
n=1
I I w!p·p. I WfM, I w~M,
2
where w~ = 1 I
n=1
I I +I w:d!
I I I n=1
I I I
~~-----~------~------T------- f-- - - - - - - - -
I pOw:p. I I I
.=2 1 I I I I
2
PO(w!PO-c.)p.
I 2 I I n=1
+I q! I ==lPOP.PO I POM, I POM,
n=1
2
+I POd:
2 I I I n=1
G= + I (w:PO -c.)p. I I I
~-~~---+------+------+------- h= 1-- ---
I I I
I I I
o
o II 0
II °°
-(w,p -c, Iq,
II if no advertising (wfpo-c,)M, -1
spill over
I I I
I I I
~---- --+- - -- -- +---- -- -l---- - --- 1-----------
I I I
I I I
o I I I -(w~p· -c,lq, (w~po-c, IM, - 1
0 I 0 I 0 I
cO"
""
N" I I I
I I
(J)
Q.
I
cr
'< ......
Coo
I I I
c; EXHIBIT 9.1&
0
a-
( i)
Arrangement of input data for geocentric pricing scheme (illustrated for two nations).
312 MultintJtional Marketing
N
II I
I
l: Pn I Ml I M2
n = 1 I N
+l:
: I
-- ---1---- + ----- n=1
------
I I
G= : (pr-cd •• i
I :
-----i-----,- ---,-
I I
I I (p~ -C2 )q2
I I
: I
EXHIBIT 9.16
Arrangement of input data for ethnocentric pricing scheme 3.
Digitized by Google
Pricing 313
Digitized by Google
314 Multinational Marketin,
als, they also are an affirmation of national identity that can thwart
the monolithic multinational. These three ideas can be sequenced in
several ways. For which national stakeholder (see Chapter 4) would
each sequence get the greatest response?
15. Societal Emotional Descriptive Global. Please review the tractor
pricing data. What configuration of data would excite the antitrust
interest of (a) the United States, (b) Canada, (c) the EEC Commission
in Brussels?
16. Societal Emotional Descriptive National. A particular nation has
administered price controls for five years. Now prices are to be decon-
trolled. Identify each affected party and predict their actions from
now to decontrol day and in the months immediately following
decontrol.
Bibliography to Chapter 9
BeUamy, Christopher W., Graham D. Child, and Anthony L Morris, Common Market
lAw of Competition 2 ed. (London: Sweet & MaxweU, 1978).
Boston Consulting Group, Perspectives on Experience (Boston: Boston Consulting
Group, 1970).
Cleland, David and William King, "Competitive Business Intelligence Systems," Bus-
inessHorizons, Vol. 18 (December 1975), pp. 19-28.
Daniels, John D., Emest Ogram, and Lee Radebaugh, International Business: Environ-
ments and Operations, 2 ed. (Reading, MA: Addison-Wesley, 1979), Chapter 19.
"Flexible Pricing: Industry's New Strategy to Hold Market Share Changes the Rules
for Economic Decision Making," Business Week, Vol. 25, December 12, 1977,
pp. 78-88.
Ijiri, Yuji and Herbert Simon, Skew Distributions and the Sizes of Business Firms,
(Amsterdam: North Holland Publishing, 1977).
Knickerbocker, Frederick T., Oligopolistic Reaction and Multinational Enterprise,
(Boston: Division of Research, Harvard Business School, 1973).
Kolde, Endel J., International Business Enterprise (Englewood Cliffs, NJ: Prentice-
Hall, 1968), Chapter 26.
Kotler, Philip, Marketing Decision Making- A Model Building Approach (Toronto:
Holt, Rinehart and Winston, 1971).
Mason, Richard 0., "A Dialectical Approach to Strategic Planning," Management
Science, Vol. 15, No. 8 (1969), pp. 403-414.
Mesak, Hani I. and Richard C. CleUand, "A Competitive Pricing Model," Management
Science, Vol. 25, No. 11 (1979), pp. 1057-1068.
Monroe, Kent B. and Albert J. DellaBita, "Models for Pricing Decisions," Journal of
Marketing Research, Vol. 15, No. 3 (1978), pp. 413-428.
Robinson, Richard 0., International Business Management, 2 ed. (New York: Holt,
Rinehart and Winston, 1978).
Schwartzman, Harry, Special Report on Prices of Tractors and Combines in CanadD
and Other Countries, Royal Commission on Farm Machinery (Ottawa: Queen's
Printer, 1969).
Digitized by Google
CHAPTER TEN
Product Design
315
Digitized by Google
316 Multinational Marketing
Digitized by Google
Product Design 317
This case was written by Detlev Hoch and Janet Hendry under the supervision of
Professor David Rutenbera of Queen's University, KlnlSton. and with financial
support from the National Sciences and Engineering Research Council of Canada.
"Workwheel"is a Black and Decker registered trademark.
Digitized by Google
318 Multinational Marketing
what resembling the paddlewheel of a boat where each paddle
is a flap of sandpaper supported by brushes.
As manufacturers of power tools for the consumer, B & D
Canada was aware that rising labor costs and increased leisure
time were largely responsible for growing interest in their
products. Furniture refinishing was a firmly established
hobby. Thousands of do-it-yourselfers were tediously strip-
ping and sanding antiques by hand, since no power tool
readily available to the public had yet been developed to
make the job less laborious. Overy decided that the need
would best be met by a hand tool rather than one on a fixed
base and proceeded to sketch a first draft of what would
later become the Workwheel power stripper and sander.
On that day, many questions remained to be formulated,
let alone answered. Aside from financial and marketing con-
siderations, technical problems (such as how power from a
high-speed motor would be transmitted to a low-speed
flap wheel) had to be solved. No one could foresee the
multiple applications of the Workwheel. Nevertheless, the
Workwheel had been born.
Standards Requirements
Black and Decker Canada, wholly owned subsidiary of the
Black and Decker Manufacturing Company of Towson,
Maryland, exports approximately 50 percent of its produc-
tion to B & D marketing subsidiaries in 45 countries. To do
so, it designs products to meet a great many differing national
standards which constitute effective nontariff barriers to
trade. As will be explained, not all importing nations have the
same electric current or receptacle configuration; thus a
single design is impossible from a technical as well as a legal
standpoint.
Regulations
Although many sets of standards exist, some nations have
regulations common to others' (as is frequently the case with
countries and their former colonies). Recognizing that com-
monality is mutually beneficial, international consultations
have been taking place almost from the beginning of the
century. The International Electrotechnical Commission
(IEC) was established early in the century for the purpose of
developing world standards. In the early 1930s, the CEE, a
Commission on Rules for the Approval of Electrical Equip-
Digitized by Google
Product Design 319
Digitized by Google
320 Multinational Marketing
Digitized by Google
- ............ ~ ".~.~.--
Technical Considerations
The voltage and frequency of power supply is not the same
everywhere. Hence electric motors require different winding
patterns, different commutators, and so on. The standard
in most nations is 220 volts/50 cycles; some use 220 volts/60
cycles; Canada and the United States use 115 volts/60 cycles.
Japan is the only country using 100 volts/60 cycles.
Receptacle configuration differs from country to country.
The four common configurations are shown in Exhibit 10.1.
Within some nations, different regions use different recep-
tacle configurations. In Britain it is not uncommon to find
variations within one house. Black and Decker U.K. responds
to this by selling power tools to the British retailer without a
plug. It is then the retailer's responsibility to attach the plug
of the customer's choice.
The manufacturer of power tools for export faces a chal-
lenging task: manufacturing a product which meets the
technical and legal requirements of the importing country
while satisfying consumer preferences and, at the same time,
standardizing components to realize economies of scale from
long production runs. To achieve these economies of scale
Australia, New
Zealand, Argentina, ( ; ' ..... )
Mainland China, Fiji
EXHIBIT 10.1
The shapes of plugs.
Digitized by Google
co EXHIBIT 10.1
N
N
Firet Draft of Model Numbe,., Ba.d on TechnicGl and Lqal Requirement. of Each Notional Morltet
Mod.I Teet lIoord Loc.I ea.,.: _t.rlaI, eolor,
,..1,., lA..,,;, Supprealon COlor
NUIfIHr CountrJI ....lfIlthll NqUired' 1101"... -t."'l
,....,/t (ill StroJn ",..,1,," "",. SUlI"'/t ,..quired, Pre(...wd
c; ·13
Austria, BelIiWD.
Germany
Y..
No
220
PVC. 1.0_2•
llue/Brown. 2M
SepuMewltb
eonIct.ap
CEE7XVD
Se_
tellll .....
Y. llue
a-
( i)
.o5 BrazI1 No 220
Blue/Brown. 2M COM~P Venloa 1
I'uIb-iD No 0rIII..
FFoduct Design
EXHIBIT 10.3
Black and Decker (Canada) Inc.
Model Numbers Proposed on March 23, 1979
-00 United States (120 V)
-01 ETI4JfHand (220/240
-02
-03
-04 V, bilingual nA"iliir"iliirmiliirfu
-05 France, Spain
-06 Australia
-07 United Kingdom (220/240 V suppressed)
-08 Italy
-09 Brazil (120 V)
-10 South Africa fufuhfuppressed)
-11 (220/240 V SUPI;FfufuII§fuII!PI
-12 (220/240 V ~lInThfu"~~'~:0~hfu
-13 'II!ppressed)
-14
Construction feature. have been opilmized for each country with regard to teat
board approval In order to minimize cost.
Digitized e
324 Multinational Marketing
EXHIBIT 10.4
Black and Decker (Canada) Inc.
Workwheel Model Reduction of March 29, 197~
Cost
Recomme~ed penalty of Annual
Model conce..ion ltJie.
number Country Conce..ionB (I/unit) forecast
Digitized by Google
Product Design 325
It was for this reason that Colin Overy had little difficulty
convincing his boss, Jack Beckering, to commit an initial
$1,000 to further develop his idea.
New product development within B & D Canada takes
place in an informal but not unstructured environment. The
potential for omitting a crucial step is too great for develop-
ment to proceed without a plan, and a 560-item checklist is
followed. In the case of the Workwheel, bimonthly meetings
were held of the Product Management Advisory Committee,
consisting of the president and representatives from Market-
ing, Manufacturing, Product Development, Quality Control,
and Engineering. Product costing and ROI calculations were
performed on a continual basis and weighted decisions were
made in light of these figures. Although they overlap, five
main steps in the development of the Workwheel may be
distinguished:
11977 1 1978 1 19791
1. Concept study
2. Design and building of
prototypes
3. Testing
4. Redesign for manufacturing
and engineering release
5. Production tool-up
Digitized by Google
326 Multinational Marketing
this working model was not very elegant for it was con-
structed from off-the-shelf items and a simple welded frame,
but it allowed engineers to try using the Workwheel and to
improve its operating design. They measured actual belt loads
in a working Workwheel for the third parallel study.
The purpose of the third study was to improve mechanical
design and development. Both the transmission system and
motor size had to be selected. No one had ever used a very
small belt to transmit as large a power load as was necessary,
so an extensive program of belt transmission development
and testing, lasting nine months, was carried out. When it
was finished, the appropriate belt width, pulley diameter,
and transmission speed had been determined. For example,
Colin Overy knew that twin belts would do the job but
twin belts are relatively bulky and expensive. After much
testing, both Goodyear and Uniroyal rubber companies
were able to design a single belt that worked.
Black and Decker is a multinational corporation; therefore,
from the very beginning, a new product is designed with
international markets in mind. Additional design criteria
include maintainability, reliability, safety, economy, ease of
operation, aesthetics, and modularity of subassemblies.
Testing, carried on throughout product development, has two
objectives: improvement and justification. Equipment is used
to measure electromagnetic interference, noise, vibration,
and resonance. Performance is evaluated under both normal
and extreme conditions. The results provide project engineers
with information necessary to ensure that components fit
exactly and operate within tolerance limits.
Product improvement may result from the suggestions of
the marketing, manufacturing, and quality contl"ol depart-
ments as well as from within the development group itself.
Technical knowledge enables B & D to build a better product.
Experience makes the designers aware of every feature a con-
sumer could want. But the product could easily become too
expensive to market successfully. Judgment is a critical
element in issues such as how long a product should last.
Black and Decker Canada designed the Workwheel to pass a
150-hour accelerated life test with no cooldown period. The
challenge is to achieve the right balance of features, quality,
and cost.
It was a low-risk venture for Jack Beckering to commit
funds for the construction of wooden prototypes, a working
model, and belt drive for the Workwheel, because of the size
of the budget involved. The next stage, however, was dif-
Digitized by Google
Product Design 327
Intercompany BusineSB
Black and Decker Canada starts to design "internationality"
into its products as soon as the concept proves itself. Sales in
each nation are handled by the B & D subsidiary located
there, which is under no obligation to accept the product.
Hence it was essential that B & D Canada convince sister sub-
sidiaries to buy Workwheel. Black and Decker has standard
procedures for the intercompany introduction of new prod-
ucts such as Workwheel.
The international sales pitch began with product demon-
strations held at regional meetings, using working prototypes
approximating the requirements of the importing country. A
brochure was distributed with photographs and information
regarding the date of market introduction, estimated inter-
company price, minimum volume requirements (economic
order quantities), shipping container volume, product and
accessory descriptions, as well as a request for feedback
regarding special requirements and sales forecasts. Final
production tooling and capacity had not been established
at this stage, so design adjustments could still be made. If
importing countries had underestimated forecasts, no prod-
uct would be available and they would have lost sales.
The next step was for each importing country to test a
sample Workwheel from a marketing as well as an engineer-
ing standpoint. It was for this reason that 100 working
prototypes had to be constructed. Although this appears
expensive initially, B & D believes it pays off in the long
run. The more accurate the marketing information, and the
more complete the standards requirements, the better able
B & D Canada will be to meet and take advantage of each
sales opportunity. Standards requirements were reported
Digitized by Google
328 Multi1lQtional Marketing
Digitized by Google
Product Design 329
EXHIBIT 10.6
Black and Decker (Canada) Inc.
Model Numbers Available April 3D, 1980G
-00 U.S.A.
-03 Japan
-04 Canada, Brazil (120 V, bilingual packaging)
-05 France, Spain, Brazil (240 V)
-06 Australia, New Zealand, South Africa (220/240 V suppressed)
-07 U.K. (220/240 V suppressed)
-OS Italy
·12 Switzerland (220/240 V suppressed)
·13 Austria, Gennany, Belgium, Holland, Scandinavia
(220/240 V suppressed)
GAvailable mid· to late 1981. subject to test board approval (where applicable).
Digitized by Google
330 Multinatiorud Marketing
EXHIBIT 10.8
Black and Deeker (Canada) Inc.
Actual Sales Information as of March, 1981
Canada Workwheellaunched In June 1979. Sales In 1980
were 55% above the forecast of 50,000 units/yr
(Exhibit 10.3) and still rising
United States Workwheel was launched In January 1980. Sales
in 1980 were substantially below the forecast of
65,000 units/yr.
Possible explanation of such low sales:
"Not invented here" syndrome
Workwheel was lost amonpt the numerous
other new B &: D products launched In U.S. at
that time
Many retailers (facing 20 percent prime interest)
refused to start stocking
United Kingdom, Workwheel was launched, in this sequence,during
France, Germany, the spring of 1981.
Italy
EXHIBIT 10.7
Black and Decker (Canada) Inc.
Marketing Questionnaire Results
In December 1980, 18 months after the launch of Workwheel in Canada,
users were mailed a questionnaire. Results included:
83% of those answering stated that they were either "satisfied" or
"very satisfied" with their Workwheel.
87% would recommend it to a friend.
Television appears very inftuentlal in the purchase decision, with 64%
stating that T.V. was how they learned about the Workwheel.
Workwheel owners tend to own many other tools:
44% own Radial or Table saw
62% own Workmate bench
78% own a Sander.
Type of store where Workwheel was purchased:
39% Canadian Tire (a Canada wide discount chain that began as an
automotive supply house, and has since expanded.)
26% Department Stores
15% Hardware Stores
11% BuDding Supply
9% Other
Digitized by Google
Introduction
The scope of this book is limited to tactical planning, so we will take
the corporate product strategy as given. In General Electric the strategic
business units include jet engines, nuclear reactors, light bulbs, small
appliances, and about 100 others. This chapter is limited to analyzing
the depth of anyone product line of one business unit. Kotler (1967,
p. 298) calls this the product mix problem. "The company's current
product mix is said to be optimal if no adjustment would enhance the
company's chances of achieving its objectives. If the company's objective
is primarily profit maximization, then the product mix is optimal if
profits could not be improved by deleting, modifying, or adding prod-
ucts." For example, the decision problem for this chapter is how many
different toasters to sell and how to design each.
A polycentric attitude to product design is that it is most profitable to
understand the subtle intricacies of the consumer purchase decision in
each nation and to design products to satisfy the consumers' specific
preferences. Hence no standardization should be imposed. An ethno-
centric attitude about product design follows from a belief that people of
all nations basically want a serviceable and inexpensive product; this can
best be achieved by standardizing design worldwide and optimizing
production to achieve the lowest cost worldwide manufacturing network.
This chapter proposes that it is possible to reconcile these two views. In
a geocentric corporation, the product is designed in modules, most of
which are standard worldwide in order to cut costs, but some of which are
tailored to particular desires. The optimal balance depends on both the
national markets and the particular product. Let us begin with the poly-
centric view that national differences do affect sales.
Digitized by Google
332 Multinational Marketing
tive person who has experienced a rich multinational life, his own intui-
tion will not provide sufficient input for a design strategy. He will have to
plod consciously through many factors which he would have handled
intuitively in his home nation. Isn tt this too much to ask of one central
design team? Shouldntt each nation design its own products?
Climate is a factor that should be considered when designing some prod-
ucts. For example, an automobile in Singapore or Hong Kong must have
protection from tropical humidity, whereas a car in Australia or South
Africa would require dustproofing. Both features are required in Ghana,
where the climate ranges from desert heat to rain forest humidity.
Where there are considerable differences in economic well-being and
standards of living, consumers have different needs and different means of
satisfying these needs. Motor scooters in Thailand satisfy a need for
reliable, affordable transportation; in America, they satisfy a desire for
fun and pleasure. To a polycentric corporation this implies two different
scooters. A geocentric designer would want to start out designing a basic
motor scooter body, with options to be added, depending on the target
market: for the American model, flashy chrome, white-walled tires, and
an electric start, for the Thai market, a kick start and no frills. These
tangible geographic and economic factors are easy to visualize and to
consider when designing a product.
Cultural factors are more subtle. They include the values and attitudes
of a society and thus, in part, they determine behavior. The preferred
cycles of a washing machine vary with culture. In the United States several
rinses are desired. The English require a heater capable of boiling the load
at the end of the wash cycle. In France and Italy several wash cycle
options are called for since the number of options has become a measure
of the quality of the machine.
Tradition is another cultural variable that affects a consumer's needs
and wants. Even within one nation, a more industrial segment of culture
may value novelty, while the less industrial segment may cherish con-
sistency. In Australia, the traditional sign of a good homemaker is the
ability to bake a sponge cake from scratch. No seller of sponge cake mixes
has yet succeeded in Australia. The norms of tradition prevent the prod-
uct's acceptance by the consumer.
Consumer tastes vary. The Nestle Corporation has chosen to make over
40 varieties of instant coffee to satisfy international tastes. Heinz ketchup
is manufactured to a different recipe in each nation.
Attitudes concerning nationality are important to product design. In
some market segments of some countries, the label "Made in U.S.A."
or "Made in Switzerland" symbolizes a way of life to which the consumer
aspires, and therefore its presence will aid the sales of the particular
product. Usually, however, these attitudes prevent a foreign product's
successful introduction in a country no matter how excellent the design.
Most tourists shopping for souvenirs prefer those made locally rather than
a superior product made in Japan.
Digitized by Google
Product Design 333
Digitized by Google
334 Multinational Marketing
The idea of a product may seem intuitively obvious until one analyzes the
consumer's motivation to buy. Kotler (1967, p. 288) illustrates the dif-
ference with the example of a camera.
The enormous number of Japanese Buddhist temples built between AD 700 and
1600 were built by flexible mass production methods. Each of the temples looks
Digitized by Google
l'roduct De,ign
qyite different. And yet each is put to«ether out of essentially standardized parts,
such as beams standardized as to width and length; standardized roofing and roof
tiles; standHdized intervals between the various levels of a pagoda, and so on. The
individually distinctive features such as the doors, iron grills, or the ornamentation
of the tiles on the roof's edge, were added only at the very end, thus creating
brilliant diversity based, however, on true mass production, that is, on standard·
parts assembled drearranged nattr;r;,,,,
Market Rert'ftrftE
multinational n,ay be dealing as 50 nationr,
the larger of which are subdivided, to yield a total of perhaps 100 market
aegments. Each segment should be researched separately. Care must be
taken to avoid problems of translations and to avoid erroneous conclu-
sions due to different national perceptions of the research questions being
posed. This market research is focused on designing the most suitable
mml'ules for many
1 (at D;:,cide N, the """"'"'''''' zmmber of item;:
mmEeting region Dlnwed to sell.
2 (by the mmEeting managem), the chance to
N items, deciEft would design l'hftm. IntzospectiG?"1,
unstructured group interviews, and an analysis of complaints are called
for. Step 2 identifies attributes desired in each region.
Step 3 (at headquarters). Cluster the regional designs into like groups. A
particular design may be wanted in many nations leading to a lower cost
of production. Create one design typical of each group, and estimate a
and upper bZ'znnl' unit cost.
4 (in each Dnrftmble a small l'Gtential custom,,;:;:
panel. Each the panel is u;:der, in terms
EGftdftrence, a serie;: Ezzch card shoulE product samh1ft,
indnding products and from Step
Dhese data are anz;:lyzeE in terms of rationD/physicD and irrationftl/
emotional preferences that would affect purchasing behavior. The regional
marketing manager should also interview in depth a few panelists, so that
his report to headquarters will clearly articulate the attributes perceived as
important. He must also estimate the sales forecast of each of the designs
from Step 3.
5 (at DEgregate the for each desihn
Step 4 and "'"'''''''''''' nnrts for this volm'filie Again consiEilifr
e;:ilin;:Glidating the seem similar nfnnomies of
"stimate the ne,';: for each of tEn fGnfnHHGted sales voluny,,;:
Exhibit 10.8).
Step 6 (in each region). The consolidated design incorporates design
compromises which reduce the product's utility to consumers of each
nation. Conduct market research on the attributes each region views as
compromised to determine how much the price would have to be lowered
to maintain sales volume.
Digitized e
336 lIultintJtiomJi Marketing
Regions IP=====I=t=+
1,11=t=====+=F===4=,II,1
"
,I
'~
\
..... /
/
EXHIBIT 10.8
Consolidating typical designs for modular design determination.
Digitized by Google
Product Design 337
Digitized by Google
338 Multinatio1UJI Marlteti,..
Percentage
loss of
capacity
$
Inventory
savings
Number of
alternate sources
EXHIBIT 10.10
Inventory savings.
Digitized by Google
----------
Conclusions
If products are designed for the home nation, and then adapted for export,
the adaptation itself may be expensive and customers abroad will perceive
a disutility. This disutility of a product which is far from the customers'
ideal point, presents a ready-made market for a competitor who is sensi-
tive to local desires.
The geocentric theme of this chapter is that a product can be designed
in modules, some standard to achieve economies of scale, and others
tailored to local desires to reduce customer disutility. The concept of
modular design is commonplace, but extending it to multinational market-
ing requires many years of diligent work.
The case study of Black and Decker Canada's Workwheel describes
engineers surmounting diverse technical standards and market preferences.
During the 1980s and 1990s a great deal of work will have to be done
developing market research procedures that will yield comparable infor-
mation from peoples of many nationalities, and then learning how to
present the information in a form that design engineers can use.
339
Digitized by Google
340 MultinatiollGl Marketing
Miles per hour
EXHIBIT 10.11
Input the rankings.
Cost 0
fuel economy, and speed give additional support to the Exhibit 10.12 price versus
speed trade-off.
This analysis has to be repeated for each individual respondent, from the sample of
respondents chosen to represent the society. Usually there are competing products
already on the market, each with its own characteristics. We have to assume that each
respondent will select the product that maximizes his utility. For each of the possible
designs, we can count the number of respondents who prefer that design. Our designs
differ in cost, so we should produce the one that would maximize our profits.
The second market research technique, nonmetric scaling, can measure products
with emotive attributes. The method of conjoint measurement presupposes the prod-
uct attributes were concrete enough to be described (100 miles per hour, six seats,
etc.) or that a sample could be displayed (the color of lemon chartreuse). With non-
metric scaling, the attributes of different brands of beer could be as disparate as taste
and social acceptance. (Urban and Hauser, 1980). For example, "Dow" had a strong
taste and an old fashioned image, "Labatt" had an intermediate taste and was modem
and young, while "Molson" was very strong in taste and evoked an image of average
sociability. In Chapter 4 we used nonmetric scaling without even naming dimensions.
For nonmetric scaling there must be at least nine items if the results are to have
statistical significance, but rarely should there be more than 15 items if the panelist
is to remain alert. Fifteen items means 105 pairs of items. Each pair of products
Digitized by Google
Product Design 341
is portrayed on a separate care!. The panelists sequence these pairs, ranging from the
pair most similar to the pair least similar. AB before, a separate map is made for each
respondent. The challenge then is to design a product and an advertising campaign
such that the product is perceived to have preferred attributes of a particular product
space. The several design attempts can be verified only by repeated panel testing.
Products have both physical and emotional appeals. So it seems wise to use both con·
joint measurement and nonmetric scaling, their strengths and weaknesses being com·
plementary. Conjoint measurement is easier to administer and less expensive than
nonmetric scaling. It can easily handle noncontinuous characteristics such as two,
four, or six car seats, and it can handle a large number of attributes, limited only by
the respondents' becoming tired. Nonmetric scaling begins with preferences between
product pairs, and It is the computer program that calculates the attributes (which
still remain to be named). It can depict an ambiance and guide the way to attributes
never thought of before (but only two or three). Finally, nonmetric scaling is less
dependent on language and the authenticity of translation because it could use physical
objects or pictures.
Digitized by Google
342 Multinational Marketin,
Digitized by Google
Product DeBign 343
Bibliography to Chapter 10
Bongers, C., SttmdIJrdization (Boston/The Hague: Martinus Nijhoff, 1980).
C1aycamp, Henry J. and William F. Massey, "A Theory of Market Segmentation,"
Marketing Research, Vot. 5 (1968), pp. 388-394.
Drucker, Peter, Manogement (New York: Harper and Row, 1974).
Green, Paul E. and Donald S. Tull, Research for Marketing Decision. (Englewood
Cliffs, NJ: Prentice·Hall, 1978).
Keegan, Warren J., "Multinational Product Planning: Strategic Alternatives," Joumal
of Marketing, Vol. 33 (1969), pp. 58-62.
Kienzle, Otto, Normungzahlen (Berlin: Springerverlag, 1950).
Kolde, Endel J., Intemational Business Enterprise (Englewood Cliffs, NJ: Prentice·
Hall, 1968}, pp. 317-321.
Kotier, Philip, Marketing Manogement (Englewood Cliffs, NJ: Prentice·Hall, 1967).
Rutenberg, David P. and Timothy Shaftel, "Product Design: Subassemblies for Multiple
Markets," Manogement Science, Vol. 18, No. 4 (1971), pp. 220-231.
Sadowski, Wieslaw, "A Few Remarks on the Assortment Problem," Manogement
Science, Vol. 6 (1959), pp. 13-24.
Stan, Martin K., "Modular Production - A New Concept," Harvard Business Review,
Vol. 43, No. 6, November-December 1965, pp. 131-142.
Wolfson, M.L., "Selecting the Best Lengths to Stock," Operations Research, Vol. 13,
No. 2 (1965), pp. 570-585.
Urban, Glen L. and John R. Hauser, Design tmd Marketing of New Products (Engle·
wood Cliffs, NJ: Prentice·Hall, 1980).
Digitized by Google
Summary of Multinational Marketing
344
Digitized by Google
Product Design 345
The high fixed cost of developing advertising compaigns could then be
spread over wider markets. During the 1970s the disadvantages of com-
mon promotion became apparent as the significance of nationalism
became apparent. During the 1980s common promotion is used for only
a few products.
Digitized by Google
PART FOUR
International Executive
Development
349
Digitized by Google
The Road to Hell ...
This case was prepared by Mr. Gareth Evans and Is reproduced with bls permission.
350
Digitized by Google
International ExecutilJe DelJelopment 351
Digitized by Google
352 Multinational Executive Development
Digitized by Google
International Executiue Deuelopment 353
Digitized by Google
354 Multinational Executive Development
impression to others so at variance with what one intends? I
can only assure you once again that my disputes with Jackson
- and you may remember also Godson - have had nothing at
all to do with the color of their skins. I promise you that if a
Barracanian had behaved in an equally peremptory manner
I would have reacted in precisely the same way. And again, if
I may say it within these four walls, I am sure I am not the
only one who has found Jackson and Godson difficult. I
could mention the names of several expatriates who have felt
the same. However, I am really sorry to have created this
impression of not being able to get on with Europeans - it
is an entirely false one - and I quite realize that I must do all
I can to correct it as quickly as possible. On your last point,
regarding Europeans holding senior positions in the company
for some time to come, I quite accept the situation. I know
that Caribbean Bauxite-as they have been doing for many
years now-will promote Barracanians as soon as their ex-
perience warrants it. And, finally, I would like to assure you,
John - and my father thinks the same too - that I am very
happy in my work here and hope to stay with the company
for many years to come."
Rennalls had spoken earnestly and, although not convinced
by what he had heard, Baker did not think he could pursue
the matter further except to say, "All right, Matt, my impres-
sion may be wrong, but I would like to remind you about the
truth of that old saying, 'What is important is not what is
true but what is believed.' Let it rest at that."
But suddenly Baker knew that he didn't want to "let it
rest at that." He was disappointed once again at not being
able to "break through" to Rennalls and having yet again
to listen to his bland denial that there was any racial pre-
judice in his makeup. Baker, who had intended ending the
interview at this point, decided to try another tack.
"To return for a moment to the 'plus and minus technique'
I was telling you about just now, there is another plus factor
I forgot to mention. I would like to congratulate you not
only on the calibre of your work but also on the ability you
have shown in overcoming a challenge which I, as a European,
have never had to meet.
"Continental Ore is, as you know, a typical commercial
enterprise - admittedly a big one - which is a product of the
economic and social environment of the United States and
Western Europe. My ancestors have all been brought up in
this environment for the past 200 or 300 years and I have,
therefore, been able to live in a world in which commerce
Digitized by Google
International Executive Development 355
Digitized by Google
356 Multinational Executive Development
14 August 19xx
From: Assistant Engineer
To: The Chief Engineer, Caribbean Bauxite Limited
Assessment of Interview Between Messrs. Baker and Rennalls
It has always been my practice to respect the advice given me by
seniors, so after our interview, I decided to give careful thought
once again to its main points and so make sure that I had under-
stood all that had been said. As I promised you at the time, I had
every intention of putting your advice to the best effect.
It was not, therefore, until I had sat down quietly in my home
yesterday evening to consider the interview objectively that its main
purport became clear. Only then did the full enormity of what you
said dawn on me. The more I thought about it, the more convinced
I was that I had hit upon the real truth - and the more furious I
became. With a facUity in the English language which I-a poor
Barracanian -cannot hope to match, you had the audacity to insult
me (and through me every Barracanian worth his salt) by claiming
that our knowledge of modem living is only a paltry 50 years old
whilst yours goes back 200-300 years. As if your materialistic com-
mercial environment could possibly be compared with the spiritual
values of our culture. I'll have you know that if much of what I saw
in London is representative of your most boasted culture, I hope
fervently that it will never come to Barracania. By what right do
you have the effrontery to condescend to us? At heart, all you
Europeans think us barbarians, or, as you say amongst yourselves,
we are "just down from the trees."
Far into the night I discussed this matter with my father, and he
is as disgusted as I. He agrees with me that any company whose
senior staff think as you do is no place for any Barracanian proud
of his culture and race - so much for all the company "clap-trap"
and specious propaganda about regionalization and Barracania for
the Barracanians.
I feel ashamed and betrayed. Please accept this letter as my resig-
nation which I wish to become effective immediately.
c.c. Production Manager
Managing Director
Questions
1. Evaluate John Baker with respect to each of IBM's six
psychological dimensions of a good international execu-
tive as outlined in this chapter.
2. Which of Erikson's eight stages of man's development best
describes the behavior of Matthew Rennalls? Of John
Baker?
Digitized by Google
International EzecutiIJe DeIJelopment 857
Digitized by Google
Introduction
As has been shown in Chapters 2 to 10, the functional activities offinance,
production, and marketing require agenda to mesh them smoothly into a
global network. For a matrix organization, the geocentric archetype of
this book, the agendas become particularly complex. Yet, as pointed out
in Chapter 1, the intricate coordination of a geocentric organization de-
pends on effective and capable international executives. The personal
contacts necessary to make a matrix organization work take years to nur-
ture. Therefore, the corporation must develop international executives
adequate for its overall long-term objectives.
Newly recruited employees rarely move directly into the grooming pro-
cess because the risk of their leaving the company is too great. If systematic
development begins at age 25, with the hope of producing a capable inter-
national executive at age 55, the process of development can be only 30
years in duration. If assignments average three years in duration, there is
room for only 10 major assignments in the grooming of an international
executive. In this chapter, the series of assignments an individual uses
will be likened to steppingstones across a river.
This chapter sketches the future management needs of the corporation
and describes a model for an executive development program. The chapter
begins with a consideration of how the future management needs of the
corporation are projected and continues with a discussion of program
development - the selection and assignment of chosen individuals to the
steppingstones of their careers. Compensation and assessment of per-
formance are also discussed.
Quantitative Forecast
Few corporations, particularly young multinationals, have conceptualized
their need for executives 30 years hence. Actually, most personnel man-
agers are more aware of their pension fund liabilities in 2010 than they are
of their need for executives then. To implement a program of executive
development, the initial step is to forecast the needs.
From the corporate planning group of your corporation or the corpora-
358
Digitized by Google
InterntJtional Executive Development 359
tion you are considering, get an estimate of the rate of growth of the
corporation as a constant percentage over the next 30 years. Because
productivity can be forecast to increase, let g be the growth in sales minus
productivity growth.
To calculate future management needs assume, for simplification, that
the organization will remain hierarchical in structure, with one chief exec-
utive officer and a constant span of control. His span of control is e, and
each of these people in turn controls e people in the level below, and so
on. From the top to the bottom, each level is ranked 1, 2, 3, ... , where
n + 1 is the bottom level (n + 1 levels are used for simplicity in calculations).
Since there is 1 person in level 1 and e people in level 2, at the bottom
level n + 1 there should be en workers. Call this number of workers
W = en. The growth factor g is applied to these W workers. The need for
future managers for all the levels above can be calculated.
log e = log Wo
no
Thus to find the span of control
e = antilog log Wo
no
Positions in each of the original levels of the organization will grow. Be-
cause this is compound interest, the number of years to double size, mul-
tiplied by the growth rate, will equal 70. For example, if the number of
workers is forecast to grow at 5 percent per year, the number of positions
in each existing management level will double every 14 years (70/5 = 14).
Thus in addition to the expansion of existing positions, new levels will
have to be created: As the organization grows, the number of levels will
have to increase. The manager of key personnel has the immediate task of
forecasting steppingstones over the next year as well as the long-run task
of envisioning steppingstones 10, 20, and 30 years hence. In the long run,
newly created layers of the organization will increase the rate of growth
of executive positions. The number of levels in the organization at time
t can be projected using the forecast number of workers (W t = gfW o ).
Digitized by Google
360 MultinatiolUll ExecutilJe DelJelopment
or
At the present (t = 0), there are Wo bottom-level employees [Wo = c" (0)].
The number of managers needed for the organization is the sum of a geo-
metric series approximated as follows:
Current number of managers:
The new positions evolved would be the difference of this value and the
value at time 0_
Digitized by Google
International Ezecutille Dellelopment 361
within the corporation, and a few positions will exist into which capable
outsiders can be recruited and still have time to master the corporation.
The manager of key personnel needs to forecast flows on these five tribu-
taries to his mainstream goal of developing 50 vice presidents. A variety
of analytic techniques have been used - from guesses to Markov chains
(Vroom and MacCrimmon, 1968) which transform masses of data into
the five tributaries.
After the number of positions has been forecast and the turnover esti-
mated, the next step is to census the corporation's available cadre of po-
tential executives of each age group. Most corporations have peculiar
voids in the demographic age distributions of their managers, often
quite different in each major national subsidiary. This data analysis can
be performed quietly from existing personnel records.
The manager of key personnel can thereby assess the severity of the
corporation's problem, and can forecast years of particular strain. This
information may be valuable feedback to the corporate planning group,
which provided the original growth forecasts g.
Certain age cohorts of future executives have been identified as being
in short supply. One remedy is to hire outsiders now and acculturate them.
The second remedy is to search harder for people of the right age within
the corporation who could conceivably function as executives. The third
remedy is to work systematically to reduce the turnover of this cohort.
Once current executives agree that these steps are necessary, one can
justify thinking explicitly about investing in human capital, articulating
the attributes of executives and jobs, and ascertaining how particular
positions help people develop.
Qualitative Forecast
The -qualities that a good international executive might need in 2010 must
be forecast for this corporation. In general, he or she must possess certain
levels of interpersonal skills, technical or functional skills, and personal
skills. Acquiring these skills depends on the individual's own initiative in
becoming involved in necessary experiences and in developing appropriate
attitudes and aptitudes. The corporation can remove impediments from
the individual's growth path and provide the appropriate opportunities or
steppingstones. Even this corporate work requires incredibly hard thinking
because of the number of executive candidates. Nevertheless, the essential
work has to be done within the individual's mind.
The technical and functional skills come from experience. The precise
qualities vary by corporation, so let me create an example of an experienced
cosmopolitan executive for the year 2010. She or he must have managed
at least four different types of product (one competitively dominant and
one competitively weak each in a high-growth industry and in a stagnant
industry). She or he must have assimilated the pressures of different
Digitized by Google
Multinational Deedopment
The person who runs away from his family of origin is as emotionally dependent
as the one who never leaves home. They both need emotional closeness, but they
allergic to it. nms away geograp)iiieET) inclined to imp,,!,
behavior. He problem as ETiin"nts and runnint
as being a methn) independence fro:m The more intenshY)
cutoff, the mnl,' "'1i!iilrable to duplicatunt with the parenn,
the first availnhln )lmon. He can get marriage. Whit'"
)"yblems develop in he tends also trmm those. Exaggnz:'
ated versions ... occur in relationship nomads, vagabonds, and hermits who either
have superficial relationships or give up and live alone. (Bowen, 1978,pp. 382-383)
Program Development
Digitized by Google
Multinational
computer data input than it has to be for human use. Thus it makes sense
to begin with a pilot study of perhaps 10 high-potential executive candi-
dates who could work through perhaps 60 positions.
A position usually has a job description. Analyze the 60 positions to
develop a list of attributes (both skills and attitudes) necessary. For each
position, summarize as a set of scores how much of each attribute is neces-
sary. In other words, express each job description as a vector of require-
ments rj. When thinking in terms of a replacement table, usually some
candidates cannot yet meet all the required attributes of the position.
Associated with the vector of requirements is another vector Cj of the an-
nual cost to the corporation if the occupant falls one unit short of meeting
the requirements. It is a measure of the corporate vulnerability to incom-
petence for this position.
Individual executive candidates have to be scored on the same attributes.
For each of the 10 candidates, summarize their present state of readiness
(skills and attitudes) in the vector Si using the same list of attributes.
In the static world of a replacement table, we would match individuals
to jobs in order to minimize the sum of total incompetence, where incom-
petence is the gap between the job requirement r and the individual's state
of competence s. We should weight different forms of incompetence to
reflect the corporate vulnerability to a gap in each attribute, using the cost
vectors Cj.
A dynamic model is needed to depict the growth of an individual as he
learns from experience. At the start of period t, individual i has attributes
that are summarized by the vector s~i. Work on jobj will transform him to
s~t 1 , a different result for each job. A particular job adds to some of his
attributes and subtracts from others. A few moments of introspection will
convince any reader that he will learn less from a job if he is grossly over-
qualified. Similarly, if he is grossly underqualified, he will find the job
overwhelming and will not be able to cope, let alone master the experience
and make it part of himself. An individual who is competent in most
attributes required for the position will be able to proceed to remedy any
weaknesses and master the job. This movement toward adequacy depends
on the individual's starting state and on the job. For a dynamic model, the
simplest approximation of this process is to use a transition matrix Tj
which describes how the particular job j affects its incumbents. The indivi-
dual's interaction with the job is depicted as a multiplication of the job
transition matrix Ti and the individual's previous state:
stt
Q/
1 = Ti st.
Q/
Digitized by Google
366 Multinotional Executive Development
ample problem of only three attributes. First we show how the transition
matrix T works, and then we discuss patterns in the coefficients of T.
For some particular job suppose the transition matrix T were
Let us work through how this job would affect three different individuals
A, B, and C. Each comes to the job with different attributes, and so would
develop differently by working in the position.
The executive A has only one strong component in his starting attri-
butes, which is weakened by the attempts to remedy the other weaknesses.
Digitized by Google
International ExecutilJe DelJelopment 367
Digitized by Google
368 Multinational Executive Development
quick trajectory to reach the top, but most people have personal reasons
for preferring a slightly longer route. To aid their decision, they would
like to be able to forecast the lifetime career consequences of their per-
sonal decisions. These consequences may be insignificant or serious. In
some corporations an aspiring executive is allowed to refuse one transfer;
refusing a second transfer is deemed a statement that the individual wants
no further promotion. What unnecessary nonsensical anxiety this causes!
How much better if the aspiring executive could plot sequences of stepping-
stones and in private see their likely outcome.
For example, professionals often marry professionals, so each has to
consider a move's effect on a spouse's career. Many women executives
wonder when and for how long they can take time to have and start
raising children without imperiling their long-term career prospects.
Personal life may impose constraints on the executive aspirant's willing-
ness to transfer. There are four major periods of family life:
1. Before children - no restrictions on transfers.
2. Preadolescent children - restricted by the need of the children for
proper schooling, but family can still move frequently.
3. Adolescent to college age - it is best that the family home not be
moved, but the executive is able to change functional responsibility
or product line in the same city. During this period, he can still depu-
tize or second in other nations to stay current.
4. All children college age or older-no restrictions on transfers, though
the executive now has less desire to move.
A computer model of executive development can include these constraints
on feasible steppingstones.
Digitized by Google
International ExecutilJe DelJelopment 369
Digitized by Google
370 MultiMtional Executive Development
especially important that there be stable growth in the underlying reality
of the executive's position. Hence the importance of a steady sequence
of steppingstones.
Evaluation
An individual's performance must be evaluated in both its technical and
its psychological dimensions. Technical proficiency measured against
sales, production, and accounting data is important, is fairly well under-
stood, and can be fairly explicit. This explicitness tends to overshadow
the more subjective aspects of the executive's psychological development.
Yet the more senior the international executive, the more important is
his personality. (Expressed another way, the more destructive are his
frailties of personality.) Psychoanalyst Erik Erikson (1964) classified eight
stages in the adult development. As condensed in Exhibit 11.1, they pro-
vide a basis for fathoming the current stage and a diagram for articulating
the future development of an individual.
The attributes of an executive (and the attributes necessary for success
in a position) include personality descriptions such as these eight stages of
development. Occupying a position affects not only the executive's tech-
nical skill, but also his self-perceptions of his personality. If the distance
between steppingstones is too great, the attributes that will diminish most
will be those pertaining to his personality. Individuals find that regression
of self is both painful and disconcerting unless they have robust self-
knowledge and the counseling support of an insightful superior.
The concept of a multinational executive failing to perform a job needs
to be clearly thought through. Inevitably, the environment will change in
unforeseen ways from what had been envisioned for this position. Execu-
tive talent is insight into determining which changes are important to the
global system, how to redeploy assets under control, and how to inform
relevant parts of the global system that their interface agreements are
undergoing change. An executive whose subsidiary has severe internal
problems may foresee a need to have the subsidiary temporarily opt out
of the global networks in manufacturing, marketing, or finance. On the
other hand, such interface changes are very expensive to the corporation
as a whole, and the executive has responsibility to redeploy assets to strive
to maintain commitments. Usually this entails sacrificing the long run for
current performance. Thus an executive has quite a choice of how to fail
technically and how to manage the failure psychologically.
Digitized by Google
Depressed and morose about
life, emphasizing failures; Happy and content with life,
would change life or career if work, accomplishments; ac-
had another chance; fears get- cepts responsibility for life;
ting older and death. 8 maximizes success.
Despair vs. Integrity
Awkwardly selfconscious of
own ideas. Stays within Confident to hold own opin-
familiar ways; needs the ions and to do what he or
approval of others and avoids she feels is best. Owns his
asserting self against group. 2 own attitudes.
Shame and doubt vs. Autonomy
EXHIBIT 11.1
Erikson's eight stages of adult development.
Digitized by Google
Multinational Deuelopment
focus on the content of the issues, and to focus on the process. One of the biggest
hazards in the principle that says "be responsible for yourself and the emotional
issue will resolve itself' has to do with the inner orientation of self. It is easy for
a person in such a position to say the situation is his "fault" and to accept the
"blame" without being responsible. There is a fine line between accepting the
responsibility for the part self plays in a situation and accepting the "blame" for
(Bowen,1978,
Bidliography to
Grinold, Richard C. and Kneale T. Marshall, Manpower PlIJnning Models (New York:
North-Holland, 1977).
Gullotta, Thomas P. and Kevin C. Donahue, ''The Corporate Family: Theory and
Treatment," The Journal of Marital and Family Therapy, Vol. 7, No. 2 (1981),
pp. 151-158.
Harari, Ehud and Yoram Zeira, "Genuine Multinational Staffing Policy: Expectations
and Realities," Academy of ManogementJourool, Vol. 20, No. 2 (1977).
Harrison, Roger and Richard Hopkins, "The Design of Cross Culture Training: An
Alternative to the University Model," Journal of Applied Behavioral Science,
Vol. 3, No. 4 (1967), pp. 431-460.
Kaufman, Herbert, The Forest Ranger: A Study in Administrative Behavior (Baltimore:
Johns Hopkins Press, 1960).
Lee, Fred E., "Planning the Pecking Order for the Company Totem Pole," PlIJnning
Review, Vol. 5, No. 1 (1977), p. 14.
Levinson, Harry, The Exceptioool Executive: A Psychological Conception (Cambridge,
Harvard University Press, 1968).
Levinson, Harry, The Great Jackass F,dllJcy (Cambridge: Harvard University Press,
1973).
Niehaus, Richard J., Computer-Assisted Human Resources PlIJnning (New York:
Wiley-Interscience, 1979).
Roberts, Karlene H., "On Looking at an Elephant: An Evaluation of Cross Cultural
Research Related to Organizations," Psychological Bulletin, Vol. 74, No. 5 (1970),
pp. 327-350.
Sofer, Cyril, Men in Mid Career: A Study of British Manogers and Technical Specialists
(Cambridge: Cambridge University Press, 1970).
Vroom, Victor and Kenneth MacCrimmon, "Toward a Stochastic Model oC Managerial
Careers," Administratwe Science Quarterly, March 1968.
Wilson, A. Thomas M., "Recruitment and Selection for Work in Foreign Cultures,"
Human RellJtions, Vol. 14, No. 1 (1965), pp. 3-21.
Wilson, Godfrey and Monica Wilson, The Aoolysis of Social Change: Based on Observa-
tions in Central Africa (Cambridge: Cambridge University Press, 1968).
Zaleznik, Abraham and Manfred F. R. Kets de Vries, Power and the Corporate Mind
(Boston, Houghton Miffiin, 1975).
Digitized by Google
Summary of the Book
When you first work for a corporation you will probably work in one
function, perhaps financial accounting, manufacturing, or sales. Nobody
will encourage you to ponder how your tasks interconnect with other
functions and the pressure of work will mean that any pondering will
have to be done on your own time.
And then, like an aircraft pulling up above the cloud cover, you will
suddenly find yourself expected to maintain liaison among the functions.
Project management, however small the project, requires interconnected
functions. This summary will therefore be more useful if it connects the
functions.
We will first examine short-term decisions which pull together the first
finance chapter (exchange risk), the first production chapter (logistics),
and the first marketing chapter (product launch). After we have looked
at this tier of short-term decisions we will turn to the tier of medium-
term and finally the tier of long-term decisions. The three tiers can be
visualized as layers of an elegant wedding cake. Before proceeding to inter-
connect the nine central chapters you will find it helpful to review
Chapter 1.
Short-Term Decisions
Short-term finance, marketing, and production plans have to be made
with an eye to flexibility. For example, the corporation's ability to re-
spond to exchange rate variation and its ability to stage the launch of a
new product are critically dependent on logistics.
As exchange rates vary, so do the costs of manufacturing an identical
item at each factory in the global production network. As explained in
Chapter 2, it is unfair to penalize or reward a manager for variables over
which he has no control-such as exchange rates. Rather, an effective
evaluation should center on how adroitly the manager responds to the
new exchange rates. The manager needs an immediate understanding of
logistics costs and feasibility in order to reassign products to the new
lower cost plants, lower cost in terms of the headquarter's numeraire
currency.
Exchange rates vary for a reason and when they do the corporate mar-
keting managers must often re-evaluate the value of the national market.
In theory this is the discounted present value of the stream of net revenue.
Like changes on a barometer, unanticipated changes in an exchange rate
are a symptom of the need to assess implications.
The final short-term decision to be considered is new product launch.
Confusion abounds during the launch of a new product. Chapter 8 recom-
mended that the new product be launched in a sequence of nations. First
874
Digitized by Google
Summary of the Book 375
launches provide test market data for the rest of the world, and the sub-
sequent sequence can be adapted to whatever new information is revealed.
The logistics implications are staggering. Firstly, the early markets are in-
tended for market research so must never run dry of product even though
the demand forecast was quite hazy. If sales exceed expectations there is
a furious scramble for the product. Secondly, the sequence is adaptive.
This means that nobody will know, for sure, which markets will be launched
in the second phase until data from the first launch has been analyzed.
One key input to the sequencing in a staged launch is whether to pre-empt
competitors in each nation. The value of this depends on the value of
greater market share, which depends on what the managers expect govern-
ment policies will be. Thus the logistics problem is that a lot of inventory
has to be accumulated and prepared for shipment to not-yet-known desti-
nations by a not-yet-known deadline.
Medium-Term Decisions
Maneuvering liquid assets, production smoothing, and pricing policies are
a tier of connected medium-term decisions. A solution to each one can
sharpen the solution to the other decisions.
Maneuvering liquid assets is an attempt to get needed liquidity to each
subsidiary, mindful of taxes and interest to be paid. Because this problem
was formulated as a linear program, the solution will produce dual vari-
ables which give the value of an increment more cash in each subsidiary.
This is vital input to production smoothing.
Digitized by Google
Summary of
and Long-TGr?n
ht~hheholders of a hivision in a nation may fEttectnd either visibly or,
sometimes, invisibly by some of the liquid asset maneuvers presented in
Chapter 3. If a critical stakeholder has difficulty with one maneuver, it
would certainly be prudent to evaluate the next cheapest maneuver.
Decisions of some stakeholders are affected by financial statements, yet
subsidiary financial statements are the artificial result of the maneuvers,
t:lHnversely, if hostility meanL L:orporation i>H:JsU~s:
,"sLithdraw from HHe such method to
investment, so as to witht:LG'SG
appear ....
:SHSS
Long-Term Decisions
The analysis of stake holders in each nation as part of multidivisional ex-
pansion, the decision to close down or open plants, and the design of
one's products are strategic decisions on which all else depends.
The analysis of stakeholders articulates the risk facing the corporation's
activities, and identifies the strategic alternatives necessary to contain that
risk. One alternative might be to build a plant for one of the corporation's
product divisions in that nation, keeping in mind that each plant opening
or closing makes a political statement of long-run significance.
A plant opening can also be used as a competitive weapon in the battle
for market share (Chapter 7), it can mollify stakeholders, and intimidate
competitors. Political and competitive statements can be made in many
combinations but most such statements cause a slight increase in the cor-
poration's cost of goods. Therefore, it is vital to model the global produc-
tion network so as to calculate the cost of manufacturing and logistics.
A global production network is a meaningful concept only if the same
product (or the same subassemblies) is manufactured at several locations.
Hence, the long-term importance of designing products with an eye to
both cost and the special needs of target market segments in each nation.
An artful standardization of subassemblies makes possible both produc-
tion smoothing, and maneuvering liquid assets. Only if the product is
somewhat standardized in design is it possible for actual sales data in one
nation to be useful in revising sales forecasts in other nations. And only if
a subassembly is so standardized that the total world demand is great
enough to justify more than two plants, do corporate executives dare
build a plant in the world's impoverished nations.
This summary has emphasized interconnections between decision mod-
ules. Whereas a manager is to do something, an executive is to care for the
interfaces between managers, and adapt these interfaces as he or she fore-
sees needs. From what one can now foresee for multinational corpora-
tions, they will need to nurture many multinational executives, viewing
positions now as stepping stones to yet more demanding positions in the
future.
Digitized by Google
Index
379
880 Index
Digitized by Google
Index 381
Digitized by Google
882 Index
Digitized by Google
Index 883
Digitized by Goog le
884 Index
Product design, 315-43 ScblaiCer, Robert, 276
ethnocentric guideline, 17 Schwartzman, Hany, 294,300,314
geocentric guideline, 32 Screening, 126-27
polycentric guideline, 22 Seagrun Co. Ltd., 257-58
Production smoothing, 174-205 Seasonal factory, 193-94
ethnocentric guideline, 17 Section 482 of U. S. Internal Revenue
geocentric guideline, 32 Code, 88-89
polycentric guideline, 22 Seghers, Paul, 107
Product life cycle, 10, 298 Selznick, Philip, 138,142
Purchasing, 195-96 Senbek, Lemma, 50, 68
Senchak, Andrew, 58,68,69
Sethi, S. Prakash, 276
Quirin, David, 89, lOO, 106 ShaCtel, Timothy, 343
Quotas, 161, 224 Shapiro, AJan, 53,60,69,107
Shawky, Hany, 50, 68
Shubnan, Junes, 107
Siegel, Sidney, 20, 33
Radebaugh,Lee,314 Simon, Herbert, 138,142,205,294,314
RaiCCa, Howard, 276 Smith, Bamard, 204
Rao, Run, 226,230, 238 Snake of European currencies, 52
RCA,16 Sobey, Frank (oCIEL), 112
Reier, Sharon, 106 Sofer, Cyril, 373
Replacement tables of managers, 364 Solnik, Bruno, 69
Requirements, demanded by govern- Sommers, Montrose, 276
ments, 222-23 Span of control, 359-60
Ricks, David, 50, 68 Special drawing rights (SDR), 52
Risk spreading, 18-22 Spencer, Williun I. (President of Citi-
Robbins, Sidney, 81,106 bank),7
Roberts, Karlene, 373 Spot exchange rate, 41-44
Robertson, Thomas, 276 Spread in banking, 5
Robinson, Richard, 34, 303, 314 Srinivasan, V. 229,238
Robustness, of plant location, 232-35 Staelin, Richard, 125, 141, 262, 276
Rodriguez, Rita, 106 Stakeholder, 131-38
Roll, Richard, 69 Stansell, Stanley, 53, 68
Rosow, Jerome, 20,34,131-32,142 Starks, Laura, 69
Rothschild, Richard (of Dozier), 38 Starr, Martin, 343
Royalties and fees, 90-92 Stewart, R. J., 172
Rubber Association of Canada, 115-19 Stobaugh, Robert, 81, 106
Rummel, R. J., 142, 238 Stockpile
Rutenberg, David, 34,47,53,60,65,68, factory where stockpiling inventory is
69,106,125,141,226,230,238, cheap,194
343 purchasing, 196
Stonehill, Arthur, 106
Stopford,John, 34,141
Strategic business units, 331
Saaty, Thomas, 133-35, 142 Subcontracting for production smooth·
Sachtjen, Wilbur, 21,33 ing, 186
Sadowski, Wieslaw, 343 Subsidy, 58, 135-36, 194, 228, 368
Sapolsky, Harvey, 28,34 Suri, Rajan, 33
Sarnat, Marshall, 68 Swap, 55
Saudi Arabia, 15, 148
Sayles, Leonard, 27, 30, 34
Scanning the environment, 122-24 Tatonnement computational scheme of
Scherer, Frederick, 224, 227, 238 tentative bids and acceptances
Digitized by Google
Index 385
Digitized by Google