Center for AI in Business Analytics & FinTech

For our first MSFE Practitioner Seminar of Fall 2024, we were pleased to welcome Jaehyuk Choi from Peking University HSBC Business School. His talk focused on "Recent Advances in the SABR Model." Jaehyuk began with an introduction to the Stochastic-alpha-beta-rho (SABR) model. Since the 2002 paper "Managing Smile Risk" by Hagan et al., the SABR model has been one of the most popular stochastic volatility (SV) models used in financial engineering, largely due to its parsimonious and intuitive parameters that fit the smile and its analytical tractability, including an asymptotic approximation for the equivalent Black-Merton-Scholes volatility. He explored the SABR literature with the aim of obtaining equivalent volatility. He reviewed how the SABR model converges to Black-Merton-Scholes when beta = 1, to the normal Bachelier model when beta = 0, and to the Constant Elasticity of Variance (CEV) model when beta is between 0 and 1, all under the case of low volatility of volatility. He also discussed the equivalent CEV volatility of the SABR model and presented numerical tests to illustrate his points. Jaehyuk also discussed option pricing under the normal SABR model (i.e., beta = 0) and introduced the Hyperbolic Normal Stochastic Volatility (NSVh) model, which bridges the normal SABR model (lambda = 0) and Johnson’s distribution (lambda = 1). He concluded by covering efficient simulation techniques for the SABR model. Many thanks to Jaehyuk on behalf of the Industrial Engineering and Operations Research Department at Columbia Engineering and the Columbia Community. Special thanks to Cindy Borgen for coordinating and assisting with the seminar. Slides are available at this link.

On behalf of the Columbia Engineering (SEAS), the Center for AI in Business Analytics & FinTech, The Data Science Institute at Columbia University (DSI), and Bloomberg, Gary Kazantsev and I are grateful to the speakers: Martin Haugh, Giulio Renzi Ricci, Fengpei Li, Zacharia Issa, Branka Hadji Misheva, Mutisya N., Giles Shaw, David Byrd, and Camilo Ortiz Diaz, as well as the poster presenters and participants for making the 10th Annual Bloomberg-Columbia Machine Learning in Finance Conference on Thursday, September 19th, 2024, a great and insightful event. Topics covered included: (a) A Machine Learning Framework for Stress Testing Financial Portfolios with Coverage Guarantees (b) Optimizing Portfolio Construction with JAX: Leveraging Evolutionary Algorithms for Enhanced Investment Strategies (c) Early Price Trajectory Data and Efficiency of Market Impact Estimation (d) Non-Adversarial Training of Neural SDEs with Signature Kernel Scores (e) eXplainable AI for Finance (f) Leveraging Deep Graph Learning for Enhanced Decision-Making in Financial Trading Applications (g) Learning Not to Spoof: Normative Guidance for Autonomous Trading Agents (h) Building an AI/ML Pipeline to Nowcast Intraday Bond Prices Slides and posters will be posted on the event’s website later this month.

On behalf of the School of Engineering and Applied Science (SEAS), the Center for AI in Business Analytics and Financial Technology, the Data Science Institute (DSI) at Columbia University, and Bloomberg, Gary Kazantsev and I are thrilled to announce that the 10th Annual Bloomberg-Columbia Machine Learning in Finance Workshop is scheduled to take place on Thursday, September 19th, 2024. It will be held in Lerner Hall at Columbia University in NYC. Here's the link to the current agenda. https://lnkd.in/emHUR5re The list of speakers will be finalized by the first week of August, and registration will open shortly thereafter. Stay tuned -- and please spread the word! #AI #ArtificialIntelligence #QuantFinance #QuantitativeFinance #quantitativeresearch #MLinFinance #ML

On behalf of the School of Engineering and Applied Science (SEAS), the Center for AI in Business Analytics and Financial Technology, the Data Science Institute (DSI) at Columbia University, and Bloomberg, Gary Kazantsev and I are thrilled to announce that the 10th Annual Bloomberg-Columbia Machine Learning in Finance Workshop is scheduled to take place on Thursday, September 19th, 2024. It will be held in Lerner Hall at Columbia University in NYC. Here's the link to the current agenda. https://lnkd.in/emHUR5re The list of speakers will be finalized by the first week of August, and registration will open shortly thereafter. Stay tuned -- and please spread the word! #AI #ArtificialIntelligence #QuantFinance #QuantitativeFinance #quantitativeresearch #MLinFinance #ML

We have completed the Hi-Touch AI Lab for the 2023-2024 academic year and are already looking  forward to the next academic year. I would like to take this opportunity to thank Carmel Macklin and Linwood Anderson of Eagle Academy for Young Men of Harlem, Roxana Bosch and Erin Flaherty of Columbia Secondary School, and Raven James and Cassandra Armstrong of Columbia University for their assistance, help, and support throughout the academic year. Many thanks to Columbia students Anna Weinacht, John Shaw Moazami, and Riju Dey for helping me with student groups. We are very excited to add Mott Hall High School to our group of schools for the 2024-2025 academic year. The experience and learning opportunities for students, along with their engagement, have been priceless.

Had a great and insightful time moderating a panel of diverse executives from Banking, FinTech, and Card Services as part of NY Tech Week, organized by New York City Economic Development Corporation (NYC EDC) and the Data Science Institute (DSI) at Columbia University. We discussed "AI & ML: Novel Use Cases in Financial Services" with a great group of professionals. Many thanks to the panelists: Claire Le Gal, Senior Vice President, Cybersecurity Products, Mastercard, Diana Meditz, Director, Product Management, BNY Mellon's AI Hub, and Anju Kambadur, Head of AI Engineering, Bloomberg. Their insights were invaluable. Special thanks also to Sharon Sputz from DSI and Brian Shoicket and Daria Siegel from NYC EDC for organizing and their support.

Have you ever wondered about the core aspects of document evaluation beyond downstream tasks like analyzing, paraphrasing, sentiment analysis, etc.? The R2-T2C framework delves into precisely these aspects, offering insights into document completeness, adherence to guidelines, peer comparison, and evolution over time at the corpus level. Key questions to consider include: (a) Is the document complete in terms of context, filler sentences/words, and embellishments? (b) How does it align with established guidelines, if any? (c) How does it compare with peer documents and the entire corpus? (d) Does it demonstrate evolution over time? The R2-T2C framework, part of the grading system, provides valuable perspectives on these aspects. As always, the focus is what is under the AI hood. Here is the link to the paper. https://lnkd.in/exvQdJFR

We were delighted to welcome back our alumni, Tuğçe KARATAŞ, Ellen Sun, and Bas Jaspers, as panelists at the Financial Engineering Practitioner Seminar on Monday, April 29th, 2024. Hosting a diverse group representing both the quantitative research and asset management sectors proved to be a rewarding experience. The discussions began with an exploration of their professional roles and how they have evolved over time in response to industry changes. They shared intriguing challenges their groups have encountered and the successful strategies they employed to overcome them, along with adaptations of skill sets to meet evolving market demands. The conversation delved into current job market trends and the specific skills now essential in their fields, which may not have been as crucial in the past. Additionally, we explored the composition of skills within their teams and how this diversity impacts their work and projects. A significant aspect of the discussion involved comparing traditional analytical approaches with modern machine learning and deep learning techniques in their respective work domains. We also examined the fungibility of skill sets in the current environment and the impact of handling unstructured big data. As always, the session concluded with candid recommendations for MSFE students, providing valuable insights to prepare them for successful careers in the industry. On behalf of the Industrial Engineering and Operations Research Department at Columbia Engineering, along with the Columbia community, we are very thankful to these alumni for sharing their knowledge and experiences at the seminar. For more insights from our alumni or to get involved in future seminars, stay connected with the Industrial Engineering and Operations Research Department. Your engagement and participation are what make these events a success.

I had an engaging and interactive presentation at Goldman Sachs as part of their AI series on April 24th, 2024. The theme of my talk was 'Under the Hood of AI'. As we explore AI and its applications, especially in asset management, it is crucial to thoroughly examine all components to ensure their temporal stability. The process cannot merely be a matter of copy-pasting or expecting data alone to work miracles. Occasionally, a complete rethinking of our approaches is necessary. Many thanks to Amal Moussa for the invitation and her insightful Q&A session, and participants for making it incredibly engaging with their excellent questions.

We were very delighted to welcome back Emanuel Derman, former colleague and dear friend, as the speaker for our Financial Engineering Practitioner Seminar on Monday, March 18th, 2024. His presentation, entitled "A Stylized History of Quantitative Finance," offered a deep dive into how the quantitative approach to finance has evolved through many small but significant steps, alongside occasional large epiphanies. Over the past seven decades, financial models have brought precision to the concepts of derivatives, diffusion, risk, volatility, the riskless rate, diversification, hedging, replication, and the principle of no riskless arbitrage. As Feynman famously summarized physics in one sentence - "Everything is made out of atoms," Emanuel encapsulated Modern Finance with, "if you can hedge away all correlated risk and you can then diversify over all uncorrelated risk, then you should expect to earn only the riskless rate." This insight leads to the foundations of CAPM (Capital Asset Pricing Model), APT (Arbitrage Pricing Theory), Black-Scholes, and beyond. He discussed the shortcomings of CAPM and explained why the Black-Scholes-Merton (BSM) model is better, exploring the expansion of quantitative finance beyond the simple world of BSM to where we are today. His talk was both insightful and engaging, sparking a lively Q&A session that highlighted the audience's keen interest in the future of quantitative finance. Many thanks to Emanuel on behalf of the Industrial Engineering and Operations Research Department at Columbia Engineering and the entire Columbia University community.