"Financial Machine Learning"
July 28 – August 1, 2025
The SoFiE Financial Econometrics Summer School is an annual week-long research-based course for PhD students, new faculty, and professionals in financial econometrics. For the first two years, the Summer School was held at Oxford University’s Oxford-Man Institute and in 2014 it moved to Harvard University. Since 2017, the SoFiE Financial Econometrics Summer School has taken place in North America, Asia, and Europe. Continuing this successful tradition, the 2025 Summer School at Yale School of Management will foster a rich academic environment for participants to explore and advance the field of financial econometrics.
Course Description
This intensive program is intended for PhD students and researchers in statistics, econometrics, and finance. It covers machine learning and artificial intelligence methods and their application to asset pricing research. The course will discuss the critical role that ML/AI already plays in improving our understanding of finance and economics and discuss the various research growth areas where ML/AI will play a pivotal role in years to come. It will cover theoretical and empirical aspects of high-dimensional models, including the "virtue of complexity," "double descent," and "benign overfit." Next, we will use the problem of return prediction to introduce modeling tools ranging from penalized regression to deep neural networks, followed by a discussion on integrating ML/AI into models of the risk-return tradeoff including applications to factor pricing, stochastic discount factors, and efficient portfolios. Lastly, it will discuss NLP in financial applications using both traditional models (e.g., topic models/LDA) and state-of-the-art large language models.
Course Outline:
- Mathematical Foundations of Financial Machine Learning
- The Virtue of Complexity
- Textual Machine Learning, NLP, LLMs
- Machine Learning Factor Pricing Models
- Portfolio Tangent Kernels and Large Factor Models
- Expected Returns and Alternative Data
- Convolutional Neural Networks for Finance
- The Limitations of Financial Machine Learning
- Artificial Intelligence Pricing Theory
- AI (Transformer) Asset Pricing Models
- The final day of the program will consist of a conference featuring new research by guest speakers Jianqing Fan (Princeton), Martin Lettau (Berkeley Haas), Stefan Nagel (Chicago Booth), Andreas Neuhierl (Wash. U.) and Guofu Zhou (Wash. U.).
