Publications

Abstract: Floods are among the costliest natural disasters, yet we know little about what drives business recovery. This gap matters as climate risk rises and government bailouts may become more limited under tightening fiscal budgets. Using establishment-level data that link remote sensing inundation to FEMA flood maps, I provide novel causal evidence that flood insurance is a key driver of business recovery. I combine a triple difference design around Hurricane Sandy with a spatial regression discontinuity at floodplain borders. Flooded establishments just inside floodplains, where properties with federally backed or regulated mortgages must carry flood insurance, recover more in employment and sales than otherwise similar sites just outside. Effects are larger where firms are more likely to be insured and where policy limits can cover more losses. Establishments of firms that disclosed insurance pre-flood also recover more. These patterns reflect an insurance-liquidity channel that supports rebuilding and reallocation, allowing firms to return stronger rather than merely to baseline. In equity markets, price drops around flood news are smaller for firms with prior exposure or disclosure. Overall, the evidence indicates insurance coverage materially shapes business recovery.

Presentations: EFA 2026 (scheduled); AFA Poster 2026 (scheduled); NBER Climate Finance PhD Workshop 2025; Cornell Finance Brown Bag

Abstract: Biodiversity risk is an emerging challenge for firms and a growing concern for investors. We evaluate how companies disclose biodiversity risk exposure in their 10-K filings and how these disclosures shape investor perceptions. Using a two-step approach that combines natural language processing and large language models, we identify and classify voluntary disclosure of exposure to biodiversity risk as either direct (explicit acknowledgments of exposure) or indirect (implied exposure embedded in business discussions). We find that firms are more likely to disclose biodiversity risk exposure, particularly through direct disclosure, when institutional ownership is higher and when local stakeholder pressure intensifies. While managers tend to issue direct disclosures in response to information demand, investors react more strongly to indirect disclosures, especially when these disclosures appear for the first time. This divergence underscores a tension in disclosure preferences for exposure to emerging and rapidly evolving risks such as biodiversity: managers prioritize “reliability” and disclose only when confident, whereas investors value “relevance” and respond more strongly to timely, even if less definitive, signals.

Presentations: CICF (China International Conference in Finance) 2025; SMU SOAR Accounting Symposium 2024*; 2024 HKUST Conference Accounting Research Symposium*; Cornell Accounting Brown Bag*

Donor Advised Funds (DAFs) have grown rapidly as a charitable giving vehicle, yet concerns remain about their timely responsiveness to urgent social needs. Using Form 990 e-filings, we analyze DAF behavior during natural disasters, which typically drive donations. Contributions to DAFs rise significantly in disaster years compared to public charities, but grantmaking does not increase proportionally. This limited response holds for both pure-play and community DAFs, even in locally affected areas. The pattern is not explained by liquidity constraints or investment motives. However, DAFs do exhibit selective grantmaking, favoring media-visible disaster-relief organizations over others.

Abstract: I study how sustainable investing impacts cross-sectional equity prices and valuation with institutional investors’ heterogeneous demand and tastes internationally. To obtain a sustainability measure for companies around the world and to capture the ESG tilt in portfolios of institutional investors, I construct a reveal-preference sustainability measure for each firm instead of using a third-party ESG score. With Factset international institutional holding data from 2010 to 2021, I apply an equilibrium asset pricing framework to empirically estimate heterogeneous preference, allowing for investment portfolio choices within and across countries. I find that separately estimated investor demands are sensitive to the sustainability of firms. The demand of investors on average increases by 26% following a one standard deviation increase in the perceived greenness, but there exists huge investor heterogeneity across countries; for example, investors from mainland China would decrease their demand by 21%. With the estimated coefficients, I conduct counterfactual analyses that consider the implications when the ESG coefficient increases following realized climate risk and when a subset of ESG investors switch to holding a market-weighted portfolio to understand the significance of different groups of institutional investors.

Presentations: NFA 2024; Cornell Sustainable Environment, Energy, and Resource Economics Seminar.

This paper examines the stock market reaction to corporate apologies. We construct a comprehensive database of corporate apologies issued for major chemical disasters between 1985 and 2017. Results from event studies and cross-sectional regressions suggest that 1) on average there is no significant stock market response after a firm apologizes; but when we allow for heterogeneity in the type of apologies, we find that 2) some apologies are associated with significant abnormal returns. In particular, firms that shift the blame to others consistently experience positive abnormal returns, while firms that admit to making mistakes tend to face negative stock market reactions in the longer-term. Greater media coverage for the chemical spill and the apology is associated with a negative and positive stock market response, respectively. Our findings are consistent with a signaling theory of apologies and help shed light on corporate crisis management and understanding of what makes an effective apology.

My paper analyzes the impacts of chemical disasters on both incident firms and non-incident firms. I consider a sample of 58 explosions in chemical plants in the U.S. over the period 1990-2017. Using event study, I find that incident firms on average experience a statistically and economically significant drop in market value following chemical disasters, and the negative impacts are permanent. In addition, my paper finds that on average the equity value of competitor firms suffers significant losses stemming from contagion effects; however, competitive effects dominate when accidents are minor. Using cross-sectional analysis, I show that the loss is significantly related to the severity of the incident as measured by causalities and by chemical pollution. Therefore, my research can be used to promote and justify the cost of safety mechanisms in the petrochemical industry.