LQG 12/10/21 – On-Line – AlphaPortfolio: Direct Construction Through Deep Reinforcement Learning and Interpretable AI – Prof. Will Cong

AlphaPortfolio: Direct Construction Through Deep Reinforcement Learning and Interpretable AI

Seminar by Professor Will Cong

12th October 2021 – On-Line

We innovate upon deep reinforcement learning (RL) in AI and tailor it to financial applications. Specifically, we directly optimize the objectives of portfolio management via RL—an alternative to conventional supervised-learning-based paradigms that entail first-step estimations of return distributions, pricing kernels, or risk premia. Our new multi-sequence neural network models can handle distinguishing features of economic and financial data, while allowing training without labels and potential market interactions. The resulting AlphaPortfolio yields stellar out-of-sample performances (e.g., Sharpe ratio above two and over 13% risk-adjusted alpha with monthly re-balancing) that are robust under various economic restrictions and market conditions (e.g., exclusion of small stocks and short-selling). Moreover, we project AlphaPortfolio onto simpler modeling spaces (e.g., using polynomial-feature-sensitivity) to uncover key drivers of investment performance, including their rotation and nonlinearity. More generally, we highlight the utility of deep reinforcement learning in finance and invent “economic distillation” tools for interpreting AI and big data models. Time permitting, Professor Cong will briefly discuss deep RL and inverse RL applications in corporate finance.

Links to published papers:

AlphaPortfolio: Direct Construction Through Deep Reinforcement Learning and Interpretable AI

Deep Sequence Modeling: Development and Applications in Asset Pricing


Lin William Cong

Rudd Family Professor of Management and Associate Professor of Finance at the Johnson Graduate School of Management at Cornell University

Professor Cong’s website is www.linwilliamcong.com and is good reading!

Lin William Cong is the founding faculty director for the FinTech Initiative. He is also a Kauffman Foundation Junior Faculty Fellow, Poets & Quants World Best Business School Professor, advisor to the Wall Street Blockchain Alliance (non-profit), and serves as editor or associate editor for journals such as Management Science. Prior to joining Cornell, he was an assistant professor of Finance at the University of Chicago Booth School of Business where he created courses on “Quantimental Investment,” faculty member at the Center for East Asian Studies, doctoral fellow at the Stanford Institute for Innovation in Developing Economies, and George Shultz Scholar at the Stanford Institute for Economic Policy Research. He also co-founded two global forums on Crypto and Blockchain Economics Research (CBER-Forum.org) and on AI and Big Data in Finance Research (ABFR-Forum.org).

Professor Cong’s research spans financial economics, information economics, FinTech and AI, and Entrepreneurship (theory and intersection with digitization and development). Widely recognized as a founding scholar for FinTech research, Professor Cong has received numerous accolades such as the International Centre for Pension Management Research Award, AAM-CAMRI-CFA Institute Prize in Asset Management, CME Best paper Award, Finance Theory Group Best Paper Award, Johnson Faculty Research Award, and the Shmuel Kandel Award at the Utah Winter Finance Conference. He has been invited to speak (including keynotes), teach, and advise at hundreds of world-renowned universities, venture funds, technology startups, investment and trading shops, and government agencies such as IMF, Asset Management Association of China, Ant Financial, Blackrock, SEC, and federal reserve banks. He was also a consultant for the incubation and development of multiple FinTech startups and litigation cases involving government regulators. He received his Ph.D. in Finance and MS in Statistics from Stanford University, and A.M. in Physics jointly with A.B. in Math and Physics from Harvard University.