Publications
​
Investor-paid Credit Ratings and Managerial Information Disclosure (doctoral thesis)
Management Science, accepted
​
Do Political Connections Affect Stock Price Crash Risk? Firm-level Evidence from China (master's thesis), with Lihong Wang
Review of Quantitative Finance and Accounting, April 2017, 48(3)
​
​
Working Papers
​
CEO Hometown Favoritism in Corporate Environmental Policies, with Qiping Xu and Qifei Zhu
​
Revise and Resubmit, Management Science
​
FMA (2021); ABFER (2022); MIT Asia Accounting Conference (2022)
​
Exploiting within-firm variations in plant-level toxic releases, we document the effect of managerial hometown attachment on corporate environmental policies. Pollution intensity is 20% lower for plants near CEOs' hometowns, achieved by conducting more costly waste management activities such as source reduction, recycling, and energy recovery. Analyses using CEO turnover events provide causal inference. Hometown emission reduction is stronger for poorly-governed firms, and is significantly weakened following the exogenous reduction in agency conflicts driven by the 2003 Tax Reform Act. In addition, hometown emission reduction is most salient in firms with worse CSR performance or tighten financial constraints. Our findings reveal that CEOs' personal motives affect corporate pollution abatement, which is manifested as a form of agency problem.
​
Discontinuous Distribution of Test Statistics Around Significance Thresholds in Empirical Accounting Studies, with Xin Chang (Simba) and Huasheng Gao
​
Revise and Resubmit, Journal of Accounting Research
​
​Nanyang Business School; Fudan University*; University of Adelaide*; Shanghai University of Finance and Economics*; FMA Asia/Pacific (2018)*; MIT Asia Conference in Accounting (2019)
​
We examine test statistics (e.g., p-values and t-statistics) reported in experimental and archival studies published in six leading accounting journals, finding evident discontinuities in their distributions around conventional significance thresholds (p-values of 0.10, 0.05, and 0.01). Specifically, we find an unusual abundance of test statistics that are just significant. Further analysis reveals that these discontinuities are more prominent in experimental than in archival studies. The difference in discontinuity between experimental and archival studies is related to several proxies for researcher degrees of freedom. Overall, we provide suggestive evidence that some researchers use questionable research practices to obtain and report statistically significant results and show how such behaviors differ according to research method in the accounting literature.
​
Social Connectedness and Information Acquisition: Evidence from EDGAR Searches, with Boluo Liu and Yuan Zhang
​
​
Joint Accounting Research Workshop (2023)
​
Exploiting the Social Connectedness Index constructed based on Facebook friendship links, we document that social connectedness is positively associated with investors’ acquisition of public information as captured by their downloading behaviour on EDGAR. Using M&A announcements as a shock to the extent of potential private communication through social networks, we find that our baseline result intensifies in the weeks leading to the public M&A announcements, consistent with private communication via the social network driving our results. Additional cross-sectional analyses show that our results are stronger when there is a potentially greater scope for complementarity between private communication and public information acquisition. We also find that social connectedness mitigates the local bias and affects other dimensions of investor activities. Overall, we highlight the important roles that social connectedness plays in investors’ information acquisition about firms both in local proximity and along the social network.
​
​
The Opioid Epidemic and Local Public Financing: Evidence from Municipal Bonds, with Qifei Zhu
​
Midwest Finance Association Annual Meeting (2020)
​
This paper examines the impact of the opioid epidemic on the financing costs of local governments. We find that a higher county-level drug overdose death rate raises offering yield spreads for local municipal bond issuers, while lowering their debt capacity. A causal interpretation is suggested by a 2SLS approach, using the distances between counties and their closest opioid manufacturer as instruments. Additionally, counties in states that introduced must-access prescription drug monitoring programs (PDMP) experienced a reduction in borrowing costs afterwards. The opioid crisis hurts issuer creditworthiness by reducing future governmental revenues and increasing police and criminal justice expenditures.
​
Local Gambling Preference and Corporate Cash Holding
​
CICF (2017, Hangzhou); FMA Asia/Pacific (2018, Hong Kong)
​
This paper studies the relation between local attitudes towards gambling and corporate liquidity. I find that firms exposed to stronger gambling preference hold significantly more cash, because cash enables them to take more risk while staying away from financial distress. Furthermore, I show that gambling-prone firms hold more cash only when they are financially constrained, and these firms also have higher marginal value of cash holding. Corporate headquarter relocation confirms the causal impact of gambling preference on cash holding. Taken together, these findings suggest that greater cash holding in gambling-prone firms is a value-increasing policy due to their strong risk-taking incentive.
​
(* indicates presentation by co-author)
​