Working Paper

Abstract: I develop a quantitative dynamic general equilibrium model to explore how climate regulatory risk is reflected in cross-sectional asset pricing. The representative household has preferences for low carbon. Greener environmental preference makes the market price of climate policy risk more positive conditional on a low elasticity of substitution between green and brown capital. The quantitative implications of the model can rationalize the recent empirical evidence on the positive price of carbon transition risk and carbon intensity risk. It also highlights the heterogeneity of carbon premium over energy structure dimension. Finally, using textual analysis to measure transition risk from 10K filings, the paper shows that lower transition risk-exposed firms carry a 8.4% risk premium. Consistent with the model, the price of transition risk tends to be negative.

Presented at SoFiE 2023 Climate Finance (NYU Shanghai), Imperial College Business School

Presented at  Baruch-JFQA Climate Finance and Sustainability Conference 2026*,  FMA(2025, Semi-Finalist FMA Conference Best Paper Award 2025), Shanghai Jiaotong University (ACEM), 5th Annual Boca-ECGI* Conference 2024, University of Liverpool Finance Seminar, University of Macau, Economic Seminar, HKUST(GZ) Seminar (2024.9) 

Abstract: This paper demonstrates that municipal borrowing costs increase with biodiversity exposure, with affected counties experiencing 42-63 basis point higher yields. Using Endangered Species Act (ESA) critical habitat designations as exogenous shocks, we identify two transmission channels: employment reductions in regulated industries and declining property tax revenues. These effects are moderated by local governance quality and debt structure. Our results establish biodiversity regulation as a distinct priced risk factor in municipal finance, with important implications for public borrowing costs.

(Presented at AAEA(2025), SWUFE ESG Conference, HKUST(GZ) FinTech Workshop)

Abstract: We develop a unified machine learning framework to extract the specific topic risk from the SEC 10K filings and finish its inference with three-stage regressions. We implement both supervised and unsupervised learning algorithms in Natural Language Processing to accomplish this task.  With application into climate topic, the proposed risk exposure estimator is not only testable within Fama-MacBeth framework by compressing high-dimensional text information, but also has statistical sufficiency property. Empirically, this technique could generate a monthly abnormal return on climate-topic risk from 1.2% to 2.1%, depending on the specification of the benchmark model. The pricing finding is robust at the firm-level for Fama-MacBeth regression. Additionally, we find that the sufficient text-implied risk exposure is consistent with recent empirical literature in Engle et al.(2020), implying the rationality of our method. Moreover, a higher causal stock market reaction over Paris Agreement indicates a consistent finding as Bolton & Kacperczyk(2021).