Research

Research Interests: AI and Big Data, Corporate Finance, Corporate Innovation, Investments, Textual Analysis

Job Market Paper

Scientific Talent and Firm Growth: Evidence from Scientific Breakthroughs (Solo-Authored)

This paper investigates the impact of corporate scientists on firm growth following scientific breakthroughs. Utilizing a bibliographic database of 258 million papers and a text-embedding tool, I develop a measure of corporate scientific human capital. By analyzing three major university-driven scientific breakthroughs, I find that firms with core technologies related to these breakthroughs perform better afterward. The impact is more pronounced for firms with substantial pre-existing scientific human capital. Corporate scientists add value through knowledge transfer, leading to more patents, higher-impact patents, and earlier adoption of related science post-breakthrough.  This study highlights the crucial role of corporate scientists in bridging basic science with industrial innovation in an economy increasingly relying on intangible assets and human capital.

Presentations:  AFA Poster Session 2025 (Scheduled), WEFI Student-led Workshop (Scheduled), Georgia State University

Working papers

 with Manish Jha, Michael Weber, and Baozhong Yang

This paper uses ChatGPT, a large language model, to extract managerial expectations of corporate policies from disclosures. We create a firm-level ChatGPT investment score, based on conference call texts, that measures managers’ anticipated changes in capital expenditures. We validate the ChatGPT investment score with interpretable textual content and its strong correlation with CFO survey responses. The investment score predicts future capital expenditure for up to nine quarters, controlling for Tobin’s q, other predictors, and fixed effects, implying the investment score provides incremental information about firms’ future investment opportunities. The investment score also separately forecasts future total, intangible, and R&D investments. High-investment-score firms experience significant negative future abnormal returns adjusted for factors, including the investment factor. We demonstrate ChatGPT’s applicability to measure other policies, such as dividends and employment. ChatGPT revolutionizes our comprehension of corporate policies, enabling the construction of managerial expectations cost-effectively for a large sample of firms over an extended period.

Presentations: NBER-RFS Big Data Conference 2023, NY Fed FinTech Conference, UT Dallas Finance Conference, UC Irvine, George Mason University, 4th Annual Boca-ECGI Corporate Finance and Governance Conference, 2024 Weinberg/ECGI Corporate Governance Symposium, PanAgora Asset Management, University of Lausanne, Georgia State University, Financial Intermediation Research Society (FIRS), Peking University, JBF Special Issue on Generative AI in Finance, Luohan Academy Webinar, Q-Group Fall Seminars 2024

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



with Zhen Shi and Baozhong Yang

We adopt a unique approach to identify short-selling hedge funds and study their performance and trading behavior. Short-selling hedge funds outperform other funds by an average annual abnormal return of 4.9%, with this superior performance largely orthogonal to known fund skill measures. Short hedge funds are larger and have higher turnover. Short-interest, long equity trades, and put positions all reveal that these hedge funds typically trade in opposition to retail transactions, providing a source of their abnormal performance. After the 2021 Meme Stock phenomenon, short-selling hedge funds markedly shifted their strategies by scaling back their positions against retail trades.

Presentations:  Georgia State University, Renmin University, Central University of Finance and Economics



 with Manish Jha, Michael Weber, and Baozhong Yang

We use generative AI to analyze over 120,000 corporate conference call transcripts, extracting managerial expectations about economic factors, which can then be aggregated at different levels. The resulting national measure, AI Economy Score, robustly predicts future economic indicators such as GDP growth, production, and employment, both in the short term and the long run, up to 4 to 10 quarters. This predictive power is incremental to that of existing measures, including survey forecasts. Moreover, industry and firm-level measures provide valuable information about sector-specific and individual firm activities, with a forecasting power lasting up to 4 years. Our findings suggest that managerial expectations carry unique insights about economic activities, with implications for both macroeconomic and microeconomic decision-making.

Presentations:  Georgia State University, University of Oxford (Scheduled)