When Does Digital Infrastructure Improve ESG Rating Accuracy? Evidence from State Blockchain Laws
Job market paper
Abstract: Does digital technology substitute for geographic proximity, or can proximity make technology more valuable in information production? I study this question in ESG ratings by combining staggered state blockchain-law enactments with time-varying rating-agency office locations in a matched stacked difference-in-differences design. Following enactment, ratings predict future ESG incidents more accurately only when the provider maintains an office in the rated firm's state; I detect no improvement when such local presence is absent. Cross-sectional tests further show that the effect is stronger for firms facing higher information-processing costs. These findings support a complementarity between digital infrastructure and geographic proximity: proximity can make technology more valuable in information-intensive assessment.
with Jie Cao, Tao Shu, Xuan Tian, and Xintong Zhan
Abstract: Patent grants are economically consequential certification events that shape the allocation of resources toward innovative technologies. We study how climate shocks affect the certification of green innovation by U.S. patent examiners. Exploiting geographic variation in examiners’ exposure to local climatic disasters, we find that exposed examiners grant significantly more green patents in the following year. We document that green patents granted by exposed examiners are of higher quality, and that exposed examiners devote more effort to reviewing green applications, thereby improving screening accuracy. Overall, our results identify a novel channel through which climate salience influences the certification of green innovation.
Best Paper in ESG and Sustainability, The 37th Australian Finance & Banking Conference, Sydney, 2024
CICF (2026), AEA Poster (2025), SFS Cavalcade Asia Pacific (2025)
with Jie Cao, Amit Goyal, Xintong Zhan, and Weiming Zhang
Abstract: We investigate whether firms’ workforce risk is priced in the options market. Using the U.S. opioid crisis as an exogenous shock to firms’ workforce conditions, we show that firms exposed to greater workforce uncertainty have higher option-implied downside risk. We establish causal effects using various difference-in-differences methods. The effects are stronger for firms that are labor-intensive, in low-labor-supply regions, with higher investor attention to labor issues, and without hedging activities. Overall, our findings identify workforce uncertainty as a novel determinant of downside risk, highlighting the role of human capital in firm risk management and option pricing.
WFA (2026), CICF (2026), AFA Poster (2025), Finance Down Under (2025), FMA Conference on Derivatives & Volatility (2025)
Bitcoin Option and Stock Return Predictability
with Jie Cao, Gang Li, and Xintong Zhan
Abstract: We show that the Bitcoin options market contains information about future stock returns. The implied volatility smirk of Bitcoin options, the difference between the implied volatilities of out-of-the-money and at-the-money puts, positively predicts U.S. stock market returns from 2017 to 2024 at horizons from one day to two months, both in and out of sample. Consistent with a premium for bearing downside risk, the predictability is strongest when downside risk is elevated and extends to the returns from selling crash insurance, and the predicted returns increase with downside beta across equities, bonds, commodities, and currencies.
Best Ph.D. Paper in the 9th China Derivatives Youth Forum, Hangzhou, 2025
CICF (2026)
