I provide an overview of financial big data and how machine learning and AI models can be tailored for applications in economics and finance. I start with Cong, Tang, Wang, and Zhang (2021a) to briefly discuss the development of deep sequence models, introduce their applications in asset pricing, and discuss their advantages and limitations.
About the Speaker:
Lin William Cong is the Rudd Family Professor of Management and Associate Professor of Finance at the Johnson Graduate School of Management at Cornell University, where he is the founding faculty director for the FinTech Initiative. He is also a Research Association (Asset Pricing) at the National Bureau of Economic Research (NBER), Kauffman Foundation Junior Faculty Fellow, Poets & Quants World Best Business School Professor, and editorial board member for top business and finance journals such as the 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 advised companies such as String Lab/Dfinity, DataYes, and currently holds the Senior Economist position at ChainLink, and advises Blackrock, Modular Asset Management, among other industry leaders in FinTech and asset management.
Professor Cong’s research spans financial economics, information economics, FinTech and AI, Entrepreneurship (theory and intersection with technology and development), digital economy, and China. Widely recognized as a founding scholar for FinTech research, Professor Cong has received numerous accolades such as the AAM-CAMRI-CFA Institute Prize in Asset Management, the CME Best paper Award, Finance Theory Group Best Paper Award, and has also been invited to speak or teach at hundreds of world-renowned universities, venture funds, investment and trading shops, and government agencies such as IMF, Asset Management Association of China, Alibaba, SEC, and federal reserve banks. 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.
Professor Cong’s website is www.linwilliamcong.com.
We perform a comparative analysis of these methods using data on U.S. equities. I then focus on Cong, Tang, Wang, and Zhang (2021b) and Cong, Feng, He, and He (2022) which introduce new frameworks for direct construction of portfolios using deep reinforcement learning and for asset pricing using panel trees. In the former, we develop multi-sequence, attention-based neural-network models tailored for the distinguishing features of financial big data, while allowing interactions with the market states and training without labels. Such AlphaPortfolio models yield 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 market conditions and economic restrictions (e.g., exclusion of small and illiquid stocks). We further demonstrate AlphaPortfolio's flexibility to incorporate transaction costs, state interactions, and alternative objectives, before applying polynomial-feature-sensitivity analysis to uncover key drivers of investment performance, including their rotation and nonlinearity. In the latter, we apply a new class of tree models with global split criteria (P-Tree) to split the cross section of asset returns under the no-arbitrage condition, generating a stochastic discount factor model and diversified test portfolios for asset pricing. P-Tree visualizes nonlinear feature interactions, accommodates time-series splits, and allows interactions between macroeconomic states and asset characteristics. In an empirical study of U.S. equities, data-driven P-Tree reveals that long-term reversal, volume volatility, and industry-adjusted market equity interact to drive cross-sectional return variation, and that inflation constitutes the most critical regime-switching when interacting with firm characteristics. P-Tree models consistently outperform known observable and latent factor models at pricing individual asset and portfolio returns, while delivering profitable and transparent trading strategies utilizing characteristic interactions.
The links to the three papers are as follows:
Meeting ID：812 4396 7663
Your participation is warmly welcomed!