Tim Baker’s Post

Previously I posted on the topic of the BloombergGPT product. This article on Medium displays how BloombergGPT performed overall and on what they call "external financial evaluation" tasks against FinGPT, an OpenSource project in the financial space. The thing to note in the article is that the BloombergGPT LLM was not evaluated for how it did on internal financial tasks (probably tasks that work on internally generated and licensed data for Bloomberg terminal customers, along with the general public data sets referenced in the article). BloombergGPT did very well on external tasks relating to public data but FinGPT did better in a couple of categories. BloombergGPT did very well overall but how it works on public datasets is not that relevant. The important thing is that Bloomberg has arguably the best financial data sets outside the public domain. The importance of BloombergGPT is that for Bloomberg terminal customers it can yield two important things that an open source version of a financial LLM cannot: 1) there are vast datasets available to Bloomberg terminal customers that are extremely powerful in the hands of terminal users (these are not available to the general public); 2) the terminal customers can take advantage of allowing Bloomberg to train models for them with their own private data. As I noted in previous posts, Bloomberg is extremely good at keeping customer data secure and by allowing customers to integrate their data into learning models, they can benefit from Bloomberg's immense infrastructure and gain advantage by leveraging their own data with the LLM interface. This is an extremely important utility Bloomberg is putting in the hands of their customers. The article: https://lnkd.in/ec54zYAu

What Quant Trading Firms Need in the Race for Alpha Over the past decade, quantitative trading strategies have been subject to increasing scrutiny regarding their ability to seek, discover, and retain alpha, the measure of an investment’s performance relative to a benchmark, writes Michael Russo, Senior Director Global Alliance Partnerships, Financial Services at Pure Storage. As financial markets become increasingly complex and global, the competition to generate alpha becomes more intense. This makes access to diverse, accurate and timely data crucial, Mr. Russo explains. #quants #quantitiativetrading #alpha #marketdata #data https://lnkd.in/d4S38Fk6

What Quant Trading Firms Need in the Race for Alpha Over the past decade, quantitative trading strategies have been subject to increasing scrutiny regarding their ability to seek, discover, and retain alpha, the measure of an investment’s performance relative to a benchmark, writes Michael Russo, Senior Director Global Alliance Partnerships, Financial Services at Pure Storage. As financial markets become increasingly complex and global, the competition to generate alpha becomes more intense. This makes access to diverse, accurate and timely data crucial, Mr. Russo explains. #quants #quantitiativetrading #alpha #marketdata #data https://lnkd.in/d4S38Fk6

Bloomberg introduces IBVAL Front Office, using machine learning for real-time pricing of 30,000 U.S. corporate securities. IBVAL predicts prices every 15 seconds, offering unprecedented transparency for traders and investment managers. Global expansion planned for IBVAL, starting with European credit instruments in 2024, reflecting Bloomberg’s commitment to democratizing fixed-income market pricing. https://lnkd.in/djGjGv9S

TabbFORUM at STA: Bloomberg’s Larry Tabb Dissects Pending U.S. Regulation In this segment of TabbFORUM at STA, Larry Tabb, Head of Market Structure at Bloomberg Intelligence, discusses regulation in U.S. equities, particularly a proposal on predictive data analytics that has raised questions about what would fall under the rule. In this interview, conducted by TabbFORUM Editorial Director Michael Scotti, Mr. Tabb outlines other pending U.S. Securities and Exchange Commission regulation, while highlighting new trends in trading, including innovation among ATSs (alternative trading systems) that has raised the competitive bar in U.S. equities. Mr. Tabb spoke with TabbFORUM at the Security Traders Association's 90th Market Structure Conference held in Washington, D.C. from October 11-13. #regulation #equities #ats #darkpool #capitalmarkets #ai #equitytrading https://lnkd.in/dhVvJ5ZW

Dear friends. I have just seen this video about Bloomberg-gpt-50B. It is a LLM specialized in Finance. They have trained the model with private data, starting in 2007 until the summer of 2022. With an amount of data 20 times larger that wikipedia-en. Although it is not recent, Bloomberg presented it's model on March 30 2023. This talk is amazing. It is the first time I see, with such detail, all the steps you have to follow to build and train a LLM model. - Understanding of the problem. - Design. - Dataset selection. - Test and debugging. - Evaluation (comparison to other models). - Benchmarking. The speaker is David Rosenberg, Head of ML Strategy at Bloomberg. . https://lnkd.in/eNxAxg-D

TabbFORUM at STA: Bloomberg’s Larry Tabb Dissects Pending U.S. Regulation In this segment of TabbFORUM at STA, Larry Tabb, Head of Market Structure at Bloomberg Intelligence, discusses regulation in U.S. equities, particularly a proposal on predictive data analytics that has raised questions about what would fall under the rule. In this interview, conducted by TabbFORUM Editorial Director Michael Scotti, Mr. Tabb outlines other pending U.S. Securities and Exchange Commission regulation, while highlighting new trends in trading, including innovation among ATSs (alternative trading systems) that has raised the competitive bar in U.S. equities. Mr. Tabb spoke with TabbFORUM at the Security Traders Association's 90th Market Structure Conference held in Washington, D.C. from October 11-13. #regulation #equities #ats #darkpool #capitalmarkets #ai #equitytrading https://lnkd.in/dhVvJ5ZW

Why #XBRL Data Matters for Investors (10 minute video). Sandy Peters, head of Financial Reporting Policy at CFA Institute (a global association of 190,000 investment professionals), gives the views of financial data users on the development of XBRL. Worth a watch as we enter into FY23 #ESEF season and thanks for sharing Corporatings.

In light of the recent horrific events in the middle-east, it's essential to explore innovative approaches to address the complex challenge of tracking the sources of #funding for various organizations, including those with controversial associations. Graph databases have proven to be a valuable tool in this context, offering an unbiased and efficient means to trace financial flows and gain valuable insights. In particular, the application of graph databases can enhance our ability to understand the #financial #networks that may be linked to organizations such as Hamas, without bias or preconceived notions. By representing financial transactions and relationships as interconnected nodes and edges, these databases enable us to create a comprehensive and visual representation of the #money trail. Some key benefits of leveraging graph databases include: Network Visualization: Graph databases allow us to create a visual map of financial connections, making it easier to identify hidden patterns and relationships that traditional databases might miss. Real-time Analysis: These databases can provide real-time insights, enabling organizations to stay ahead of evolving financial networks and adapt their strategies accordingly. Cross-referencing: By integrating various data sources and cross-referencing transactions, we can gain a more accurate understanding of the financial ecosystem surrounding any entity. Predictive Analytics: Machine learning algorithms applied to graph data can help predict future transactions and highlight potential risks. While it's important to stress that the application of graph databases should be done with transparency, privacy, and respect for due process, they represent a promising technology to bolster financial transparency efforts. As we continue to navigate the complexities of global finance, it's critical to remain open to new and unbiased tools and methodologies that can help us better understand the financial landscape. If you're interested in exploring this topic further, please feel free to reach out for a conversation. #graphdatabases #terroristfinancing #fatf #aml #bsa #kyc #cdd #edd #networkofaccounts #neo4j #ctf #ml #machinelearning #algorithms #outliers #behavioranalytics #followthemoney

📍 etcvenues 360 Madison Avenue 📅 Wednesday, October 16th | 9:00 AM – 6:00PM EDT The FIX Trading Community invites you to attend the Americas Trading Conference on Wednesday, October 16th at etc.venues 360 Madison Avenue, New York. Join us for a series of panel discussions across a one-stream agenda covering the following topics: AI Insights: Revolutionizing Financial Trading Explore how Artificial Intelligence (AI) transforms financial trading. Join industry leaders to discuss AI's impact on algorithms, risk management, and execution decisions. Streamlining Trades: Reviewing T+1 Trade Processing in Financial Markets Join industry experts and regulatory leaders for an insightful panel discussion on the evolution of trade processing in financial markets, including the transition to T+1 (or even T+0) settlement cycles. Digital Assets Unveiled: Navigating the Future of Trading Explore the dynamic world of digital assets with our expert panel as they unveil the opportunities and challenges in this rapidly evolving space. From cryptocurrencies to tokenized assets, the discussion will delve into how these digital innovations are reshaping traditional trading landscapes. Election Economics: Unpacking the Financial Markets' Response to the US Presidential Election Join us for a timely panel discussion on the intricate relationship between US presidential elections and financial markets. FIX at 30: Reflecting on Three Decades of Market Transformation Celebrate FIX Protocol's 30-year legacy of revolutionizing global markets. Join industry pioneers and technologists as they trace FIX's evolution, its impact on trading, and where we think the next 30 years will take us. Forging a Path: Women in FIX Explore the achievements and challenges of women in the FIX community. Join industry leaders as they discuss navigating careers in finance and technology, promoting diversity, and driving innovation. #FIX, #AI, #Innovation, #Technology, #Economy, #Economics Register here: https://lnkd.in/efEewMCp Lead Sponsors: AMD, ION Markets, ITRS Sponsors: A10 Networks, Inc, Broadridge, Imandra, Linedata, Rapid Addition, TradingHub Event Partner: Women in Financial Markets (WIFM) Trade Association Partner: SIFMA Official Media Partner: Global Trading