Research

What's New:

• (new version) ChatGPT and Corporate Policies Chicago Booth Research Paper

• (new) Short-Selling Hedge Funds

• (new) Can Machines Understand Human Skills? Insights from Analyst Selection

• From Man vs. Machine to Man + Machine: The Art and AI of Stock Analyses, forthcoming, Journal of Financial Economics

Working Papers

• (new version) ChatGPT and Corporate Policies, with Manish Jha, Jialin Qian, and Michael Weber, Chicago Booth Research Paper

  • We use ChatGPT to extract expected corporate policies from managerial disclosure. A ChatGPT-based investment score forecasts corporate investment and expected returns for up to nine quarters in the future, controlling for Tobin's q and other predictors. Our approach is interpretable and applicable to a wide range of corporate policies.

  • Media Coverage: Harvard Law School Corporate Governance Forum, Chicago Booth Review, University of Chicago Becker Friedman Institute Research Brief, VoxEU, CFO

  • Presented: NBER-RFS Big Data Conference, NY Fed FinTech Conference, UT Dallas Finance Conference, Boca-ECGI Corporate Finance Conference, Delaware-ECGI Corporate Governance Symposium

  • Awards: The Panagora Asset Management Richard Crowell Prize (second place)

• (new) Short-Selling Hedge Funds, with Jialin Qian and Zhen Shi

  • We propose a unique approach to identify hedge funds engaged in short-selling and uncover that they exhibit superior abnormal performance unexplained by known skill measures. Short-selling hedge funds traded against retail investors but attenuated their behavior after the 2021 Meme stock frenzy.

• (new) Can Machines Understand Human Skills? Insights from Analyst Selection, with Sean Cao, Norman Guo, and Houping Xiao

  • This paper introduces a machine-learning approach that analyzes forecasting decisions to detect analysts' firm-specific skills. A "smart consensus" relying on machine-identified skilled analysts more accurately predicts earnings news and generates significant abnormal returns.

• Product Variety and Asset Pricing, with Florin Bidian and Ajay Subramanian

  • A general-equilibrium model of endogenous product variety explains observed equity premia and risk-free rates, and generates new predictions on product substitutability and risk premia.

  • Presented: Econometric Society North America Winter Meeting, Econometric Society China Meeting, Econometric Society Australasian Meeting, Econometric Society Asian Meeting

• Endogenous Fracturing under Partisan Voting with Stephen Yang

  • In this paper, we develop a novel model for voting and seek to explain the effects of partisan allegiances on legislative outcomes. We endogenously show a "fracturing" effect in majority parties and justify the removal of Kevin McCarthy as House Speaker by a "coalition" of Democrats and radical Republicans. 

  • Presented: Inaugural BAR Conference

• Visual Information in the Age of AI: Evidence from Corporate Executive Presentations, with Sean Cao, Yichen Cheng, Meng Wang, and Yusen Xia

  • Forward-looking visual information about corporate operations from executive presentations is associated with short-term returns and long-term operating performance. Such information is utilized by AI-equipped institutional investors but not by other investors, suggesting a potential AI divide in the era of machines.

  • Presented:  NBER SI Big Data (scheduled), Hong Kong Conference in FinTech, AI and Big Data, AAAI, CICF

  • Media Coverage: Columbia Law School Blue Sky Blog

  • Awards: The Panagora Asset Management Richard Crowell Prize (third place)

• Managerial Risk Assessment and Fund Performance: Evidence from Textual Disclosure, with Sean Cao and Alan Zhang, Invited for Resubmission, Management Science

  • We use state-of-the-art natural language processing tools to extract mutual fund managers' risk sentiment, which helps predict their future risk-taking and performance

  • Media Coverage: Columbia Law School Blue Sky Blog

• Blockchains, Smart Contracts, and Supply Chain Efficiency, with Hongguang Ma and Yusen Xia

  • Blockchains equipped with real-time production data and smart contracts can help to mitigate supply chain uncertainty and improve efficiency. The benefit of blockchain is non-monotone in product profit margin and defect rates, and is decreasing in product substitutability when there are competing suppliers.

  • Presented: AEA, 4th UWA Blockchain and Cryptocurrency Conference, INFORMS 2022, International Conference on Finance and Technology

• Do Hedge Funds Strategically Misreport Their Holdings? Evidence from 13F Restatements, with Sean Cao, Zhi Da, and Daniel Jiang

  • Hedge funds strategically use 13F restatements, which are meant to correct "honest" mistakes,  to prevent disclosure of private information.

  • Presented: SEC Regulation of Capital Markets Conference, NFA, FARS, Texas A&M Young Scholars Finance Consortium

  • Media Coverage: Columbia Law School Blog

•  Financial Reporting and Blockchains: Audit Pricing, Misstatements, and Regulation, with Sean Cao and Lin William Cong, Invited for Resubmission, Management Science

  • Collaboration among auditors in the adoption of blockchain technologies can increase auditing quality and reduce inefficiencies, but may need regulator's coordination.

  • Presented: NBER Blockchain Conference, SFS Cavalcade North America, PCAOB/JAR Conference

  • Media Coverage: Chicago Booth Review

Publications

• From Man vs. Machine to Man + Machine: The Art and AI of Stock Analyses, with Sean Cao, Wei Jiang, and Junbo L. Wang, forthcoming, Journal of Financial Economics

  • While an AI analyst that digests quantitative and qualitative information can beat the majority of stock analysts in stock forecasts, human analysts remain competitive when institutional knowledge is required. The complementarity of humans and AI suggests a future of "Man + Machine" for high-skill professions.

  • Presented: NBER Economics of AI Conference, AFA, AI and Big Data in Finance Research Forum

  • Media Coverage: Harvard Law School Forum on Corporate Governance

  • Awards: Annual Conference on Digital Economics Best Paper Award, MFA Best Paper Award in Investment, China Finance Research Conference Best Paper Award, Global AI Finance Conference Best Paper Award, SFS-Asia Cavalcade Best Paper Award

• How to Talk When a Machine is Listening: Corporate Disclosure in the Age of AI, with Sean Cao, Wei Jiang, and Alan L. Zhang, 2023, Review of Financial Studies 36(9), 3603-3642.

  • Managers reshape the style, sentiment, and emotion of their disclosure in response to the rise of machine and AI readership.

  • Presented: NBER Economics of AI Conference, NBER Big Data and Securities Conference, AFA

  • Media Coverage: NBER Digest Lead Article, Bloomberg, Financial Times, The Guardian, Quartz, CNBC, Columbia Law School Blog, Oxford Business Law Blog

  • Awards: CAPANA 2021 Conference Best Paper Award

• Copycat Skills and Disclosure Costs: Evidence from Peer Companies' Digital Footprints, with Sean Cao, Kai Du, and Alan L. Zhang, 2021,  Journal of Accounting Research 59 (4), 1261-1302.

  • Digital footprints allow the identification of copycat investment firms and their targets. Copycat investment firms are selective in their imitation of peer firms' trading strategies, generating profits and imposing costs on disclosing companies.

  • Media Coverage: Columbia Law School Blog

• Dynamic Prudential Regulation, with Ajay Subramanian (appendix), 2020, Management Science 66 (7), 3183-3210.

  • We propose a dynamic framework of capital regulation of financial institutions that admit closed-form solutions. The calibrated model suggests that optimal prudential regulation should take a multi-faceted approach, in addition to imposing a higher capital requirement.

  • NBER Law and Economics Summer Institute Program

• How Valuable is FinTech Innovation?, with Mark Chen and Qinxi Wu, 2019, Review of Financial Studies 32, 2062-2106.

  • We use supervised machine learning to classify FinTech patents of seven categories (including blockchain, data analytics, etc.) for all US Patents between 2003 and 2017 and study the value and disruptive impact of FinTech innovation.

  • Review of Financial Studies FinTech Special Issue in 2019, Highly cited paper collection published by RFS

• Blockchain Architecture for Auditing Automation and Trust Building in Public Markets, with Sean Cao, Lin William Cong, Meng Han, and Qixuan Hou, 2020, IEEE Computer 53(7), 20-28.   

  • IEEE Computer Special Blockchain Issue

• Communication and Comovement: Evidence from Online Stock Forums, with Lei Jiang and Jinyu Liu, 2019, Financial Management 48 (3), 805-847

  • Awards: Yihong Xia Best Paper Prize, 2015 CICF 

• Optimal Dynamic Risk Taking, with Ajay Subramanian, 2017, Mathematics of Operations Research 42, 599-625 

• Mandatory Portfolio Disclosure, Stock Liquidity, and Mutual Fund Performance, with Vikas Agarwal, Kevin Mullally, and  Yuehua Tang, 2015, Journal of Finance 70, 2733-2776  (supplementary appendix) 

• Uncovering Hedge Fund Skill from the Portfolio Holdings They Hide, with Vikas Agarwal, Wei Jiang, and Yuehua Tang, 2013, Journal of Finance 68, 739-783   (internet appendix)  

  • NBER Law and Economics Conference

  • Awards: Chicago Quantitative Alliance Annual Academic Competition, 3rd Prize 

• The Mystery of Zero-Leverage Firms, with Ilya A. Strebulaev, 2013, Journal of Financial Economics 109, 1-23 (Lead Article) 

  • Awards: Emerald Citations of Excellence Award, 2016 

• Dynamic Capital Structure with Heterogeneous Beliefs and Market Timing, 2013, Journal of Corporate Finance 22, 254-277  

• Communication, Excess Comovement and Factor Structures, 2013, Journal of Business Finance and Accounting 40 (9-10), 1304-1325

Book Chapter

• Green Energy, Emissions, and Blockchain Technology, with Tony Erwin, 2023, Book chapter in  FinTech and Sustainability by Palgrave MacMillan

  • Survey on the applications of blockchain technology to green energy and sustainability

• Analyzing Textual Information at Scale, with Lin William Cong, Tengyuan Liang, and Xiao Zhang,  2020,  Book Chapter in Information to Facilitate Efficient Decision Making: Big Data, Blockchain and Relevance, edited by Kashi Balachandran, World Scientific Publishers

  • Survey on new, machine learning methods in textual analysis

Publications in Mathematics Journals

• Compactification of the moduli spaces of vortices and coupled vortices, with Gang Tian, 553, 2002, Journal für die reine und angewandte Mathematik, 17-41.

• Removable singularities of Yang-Mills connections in higher dimensions, 209, 2003, Pacific Journal of Mathematics, 381-398.

• The uniqueness of tangent cones for Yang-Mills connections with isolated singularities, 180, 2003, Advances in Mathematics, 648-691.