Finance
Artificial intelligence and asset pricing: The power of transformers
The rapid advancement of artificial intelligence (AI) has reshaped numerous fields, and finance is no exception. Asset pricing, a domain traditionally dominated by linear models and factor-based approaches, is now experiencing a transformative shift due to AI’s capacity to uncover persistent predictive patterns in financial data. The introduction of large-scale AI models – particularly transformer-based architectures – has significantly enhanced our capability to model complex relationships among assets and firms (Eisfeldt et al. 2023), leading to improved forecasting and risk assessment.
In this column, we explore how AI-driven asset pricing models, particularly those incorporating transformers, leverage cross-asset information sharing to reduce pricing errors and enhance predictive accuracy. These innovations offer a novel perspective on how financial markets process information and determine asset prices.
Traditional versus AI-driven asset pricing
Most traditional asset pricing models, such as the Fama and French (1993) framework, rely on predefined factors that explain asset returns. While effective, these models assume a fixed and linear relationship between asset characteristics and expected returns. AI, on the other hand, introduces a non-linear, data-driven approach, identifying patterns that are often invisible to traditional methods.
Machine learning models, including tree-based methods and neural networks (Gu et al. 2020), have improved asset return prediction by capturing complex relationships between firm characteristics and returns. However, these models typically focus on ‘own-asset prediction’, meaning that they use only an asset’s individual characteristics to estimate future returns. This approach ignores the broader context in which assets interact.
The role of transformers in asset pricing
In a recent paper (Kelly et al. 2024), we introduce the Artificial Intelligence Pricing Model (AIPM), which embeds transformer networks into the stochastic discount factor (SDF) framework. While initially developed for natural language processing, we show that transformers are remarkably effective in financial applications due to their ability to capture cross-sectional dependencies across assets.
Unlike traditional machine learning models, transformers incorporate the ‘attention mechanism’, allowing them to dynamically adjust the weight placed on different inputs based on their relevance. In the context of asset pricing, this means that the model not only considers an asset’s own characteristics but also how these characteristics interact with those of other assets. This approach significantly enhances predictive power by leveraging market-wide information.
Empirical findings: Performance of AI-based models
We evaluated the performance of the transformer-based AIPM using a dataset of US stock returns and conditioning variables. Compared to traditional asset pricing models and other machine learning approaches, the AIPM demonstrated:
- Lower pricing errors: The model achieved significantly smaller out-of-sample pricing errors compared to traditional factor-based models and neural networks without attention mechanisms.
- Higher Sharpe ratios: By integrating cross-asset dependencies, the transformer-based model outperformed existing approaches in terms of risk-adjusted returns out-of-sample.
- Scalability and complexity gains: We found that increasing model complexity – by incorporating deeper transformer layers – consistently improved predictive performance. This supports the notion that AI models benefit from higher parameterisation when applied to asset pricing.
Empirical evidence from the literature on large language models (e.g. Kaplan et al. 2020) indicates that adding more transformer blocks enhances the capacity of models to effectively represent language. Models with deeper architectures are able to capture more abstract features and longer-range dependencies than shallower models. Each additional layer refines the attention distributions, allowing the model to consider both short-term and long-term relationships. Interestingly, the same benefits of transformer complexity emerge in the context of the asset pricing model.
These findings suggest that AI-driven asset pricing models are more efficient in processing vast amounts of financial data, leading to more accurate and robust predictions.
Implications for investors and policymakers
The integration of AI into asset pricing has profound implications for market participants and policymakers. Investors can benefit from improved portfolio allocation strategies driven by AI’s capability to identify subtle pricing inefficiencies, a concept aligned with research on AI and personality traits shaping economic returns (Makridis 2025). Meanwhile, regulators and policymakers must consider the impact of AI-driven trading on market stability and efficiency.
Moreover, the adoption of AI in asset pricing challenges traditional views on market efficiency, echoing broader concerns about AI’s macroeconomic impact and productivity gains (Filippucci et al. 2024). If AI models consistently outperform classical frameworks, it may suggest that markets are less efficient than previously thought, opening the door for further research into the nature of pricing anomalies.
Conclusion
The fusion of artificial intelligence and finance is revolutionising asset pricing. Our research demonstrates that transformer-based models significantly enhance return predictions by leveraging cross-asset information sharing. As AI continues to evolve, its role in financial decision-making will only grow, offering new opportunities for investors and reshaping our understanding of market dynamics.
References
Gu, S, B Kelly and D Xiu (2020), “Empirical Asset Pricing via Machine Learning”, Review of Financial Studies 33 (5): 2223-2273.
Eisfeldt, A, G Schubert and M B Zhang (2023), “Generative AI and Firm Valuation”, VoxEU.org, 4 June.
Fama, E and K French (1993), “Common risk factors in the returns on stocks and bonds”, Journal of Financial Economics 33(1): 3-56.
Filippucci, F, P Gal, and M Schief (2024), “Miracle or Myth? Assessing the Macroeconomic Productivity Gains from Artificial Intelligence”, VoxEU.org, 8 December.
Kaplan, J, S McCandlish, T Henighan et al. (2020), “Scaling laws for neural language models”, arXiv preprint arXiv:2001.08361.
Kelly, B Kuznetsov, S Malamud and T Xu (2024), “Artificial Intelligence Asset Pricing Models”, NBER Working Paper No. w33351
Makridis, C (2025), “The Role of Personality Traits in Shaping Economic Returns Amid Technological Change”, VoxEU.org, 31 January.
Vaswani, A, N Shazeer, N Parmar, J Uszkoreit, L Jones, A N Gomez, L Kaiser and I Polosukhin (2017), “Attention is all you need”, Advances in Neural Information Processing Systems 30.
Submission
We invite researchers from both academic and policy institutions to submit a paper (preferable) or an extended abstract of two or more pages by May 15, 2025, by uploading here. We highly encourage submissions that showcase collaborations between researchers and policymakers.
The conference will feature a session in which policy makers present their research.
We aim to notify the authors of selected papers by June.
Academic Committee
Laura Abramovsky, Pierre Bachas, Anne Brockmeyer, Rishabh Choudhary, Lucie Gadenne, Pablo Garriga, François Gerard, Hazel Granger, Jonas Hjort, Christopher Hoy, Justine Knebelmann, Kyle McNabb, Joana Naritomi, Marina Ngoma, Oyebola Okunogbe, David Phillips, Thiago Scot, Mahvish Shaukat, Dario Tortarolo, Yani Tyskerud, Ben Waltmann, Mazhar Waseem.
MCLEAN, Va, & RIVERWOODS, Ill., April 18, 2025–(BUSINESS WIRE)–Capital One Financial Corporation (NYSE: COF) and Discover Financial Services (NYSE: DFS) today announced that the Board of Governors of the Federal Reserve System and the Office of the Comptroller of the Currency have approved Capital One’s proposed acquisition of Discover.
This approval follows approval of the transaction by the Delaware State Bank Commissioner in December 2024, and by shareholders of more than 99 percent of each company’s shares voting in February of this year.
“This is an exciting moment for Capital One and Discover. We understand the critical importance of a strong and competitive banking system to our customers and our economy, and we appreciate the thoughtful and diligent engagement of our regulators as they thoroughly reviewed this deal over the past 14 months,” said Richard Fairbank, Founder, Chairman, and CEO of Capital One. “I am grateful to the thousands of associates across Capital One and Discover who have worked tirelessly to help us achieve this significant milestone. We look forward to bringing these two great companies together with a profound sense of possibility and responsibility to deliver for our customers, associates, shareholders, and communities.”
All required regulatory approvals to complete the transaction have now been received, and the transaction is expected to close on May 18, 2025, subject to the satisfaction of customary closing conditions.
“The combination of our two great companies will increase competition in payment networks, offer a wider range of products to our customers, increase our resources devoted to innovation and security, and bring meaningful community benefits,” said Michael Shepherd, Interim CEO and President of Discover.
There will be no immediate changes to Capital One and Discover customer accounts and relationships now or in the period immediately following the closing of the transaction. Capital One will provide customers with comprehensive information regarding relevant conversion activities well in advance of any future change. Until then, customers will continue to be served through their respective Capital One and Discover customer communications channels.
Upon closing, Capital One will begin implementation of its historic, five-year Community Benefits Plan (CBP), developed in connection with the acquisition and in partnership with leading community organizations, mobilizing more than $265 billion in lending, investment, and services to advance economic opportunity and financial well-being across America.
Finance
IC Group Holdings Announces Promotion of Matan Gamliel to Vice President (Finance), Stock Option Grant, and Share for Debt Settlement with a Former Director
Toronto, Ontario–(Newsfile Corp. – April 17, 2025) – IC Group Holdings Inc. (TSXV: ICGH) (“IC Group” or the “Company“), a technology-enabled consumer engagement company that helps Fortune 500 brands simplify and amplify connections with consumers both nationally and internationally, is pleased to announce the promotion of Matan Gamliel, CPA, CA, to Vice President, Finance. Mr. Gamliel has been with IC Group for over six years, recently serving as Director of Finance. A Chartered Professional Accountant with extensive experience in financial accounting, managerial finance, and M&A transactions, Mr. Gamliel has held senior roles across the marketing, insurance, and construction sectors. Before joining IC Group, he worked with Deloitte in their M&A Transaction Services group and held finance roles at Insured Creativity and Rajotte Capital Group.
“Matan has played a critical role in building the financial strength and discipline of IC Group over the past several years,” said Duncan McCready, CEO of IC Group. “His promotion to Vice President, Finance reflects the leadership he brings to our team and our confidence in his continued contributions as we scale the business.”
In conjunction with his promotion, the Company has granted 75,000 stock options to Mr. Gamliel under the Company’s Stock Option Plan. The options have an exercise price of $0.65 per share, expire April 9, 2035, and vest in two equal tranches: 50% on the first anniversary and 50% on the second anniversary of their grant date.
Additionally, the Company has negotiated a debt settlement pursuant to which it has agreed, subject to acceptance by the TSX Venture Exchange (the “TSXV“), to issue 66,666 common shares at a deemed price of $0.75 per share to Mike Svetkoff to settle an aggregate of $50,000 owing to Mr. Svetkoff.
All securities issued under the debt settlement (or upon exercise of the options granted to Mr. Gamliel) are subject to a four-month hold period in accordance with applicable securities laws. The debt settlement with Mr. Svetkoff is subject to the approval of the TSXV.
About IC Group Holdings Inc.
IC Group (TSXV: ICGH) is transforming how brands engage with audiences across live events. It uses digital and social platforms to drive sales, capture valuable first-party data to fuel ongoing marketing initiatives, and build customer loyalty. The Company does this by simplifying and managing the technology, regulatory, data security, and financial risks of engaging with consumer audiences on a global basis. Its solutions span digital engagement, mobile messaging, and specialty insurance for Fortune 500 brands and their agency partners in international jurisdictions.
Detroit Pistons vs Knicks Prediction Roundup Ahead of Game 1
Southwest Airlines customers sound off over end to free checked bags for passengers
From Oakland Grit to Chinatown Glam: Bay Area Artists Shine at SF Art Fair | KQED
50501 protests planned across Dallas-Fort Worth on Saturday
Miami Takes Game One, Upsets No. 14 Georgia Tech
Why Marjorie Taylor Greene was ‘kicked out’ of the Freedom Caucus according to Rep. Buck
Colorado Rockies game no. 116 thread: Zac Gallen vs José Ureña
See it: Tesla crashes into Columbus convention center at 70 mph
Fox News Politics: Georgia the whole day through
Death of missing Oregon girl found in stream ruled homicide
Supreme Court blocks new deportations of Venezuelans in Texas under 18th century Alien Enemies Act
Iran seeks Russian support ahead of nuclear deal talks with US
Trump administration has $15M deal with El Salvador to accept deportees, MD senator says
Video: The Conservative Christian Network Inside the White House
Biden's vax-focused COVID-19 website obliterated by White House, replaced with 'true origins' virus guide
-
Sports1 week agoMen’s college basketball way-too-early Top 25: Purdue’s No. 1; Florida isn’t going anywhere
-
News1 week agoTrump Asks Supreme Court to Let Him Fire Agencies’ Leaders
-
Health1 week agoThe Three States That Are Especially Stuck if Congress Cuts Medicaid
-
Wisconsin1 week agoCoordinator Mike Tressel shares his goals for Wisconsin’s defense, including thoughts on team’s bigger defensive line
-
Movie Reviews1 week agoThe Amateur Movie Review – Rami Malek’s Gritty Spy Thriller
-
Science1 week agoScientists Map Miles of Wiring in a Speck of Mouse Brain
-
World1 week agoFor U.S. and China, a Risky Game of Chicken With No Off-Ramp in Sight
-
Dallas, TX1 week agoDallas Stars take on Winnipeg Jets in massive Western Conference clash on FOX 4