[Job Market Paper] "Separating E From ESG: How Mandatory Climate Disclosure Affects E/SG Reporting Incentives Differently?"
Abstract: Investors often evaluate firms' sustainability performance by collectively assessing the three pillars of ESG: environment (E), social (S), and governance (G). While ESG is commonly lumped together, this paper proposes, and provides evidence on, the difference among the pillars in terms of firms' reporting incentives and information quality. Specifically, exploiting variations in the adoption of a climate disclosure mandate, I find that affected firms are less likely to voluntarily disclose environmental information (substituting effect on ``E") but more likely to disclose general sustainability information (complementary effect on ``ESG"). At the same time, the difference-in-difference design shows that affected firms' information quality on environmental information significantly improves following the mandate, whereas no significant improvements are found for the overall ESG information quality. Collectively, these results suggest that comprehensive sustainability disclosure regulation plays a crucial role in firms' disclosure incentives.
Working Papers
[1] "Intended and Unintended Consequences of Facility-Level Mandatory CSR Disclosure: Evidence from the Greenhouse Gas Reporting Program" with Nicholas Z. Muller and Pierre Jinghong Liang
NBER Working Paper w28984 2021
Featured in Houston Chronicle, Tepper School of Business, The NBER Digest, Harvard Law School Forum
Presented at ARCS 2023, AAA 2022, Brookings Institute Workshop on Economic Research, EPA's NCEE seminar series, ZEW Mannheim
Abstract: We examine the real effects of the Greenhouse Gas Reporting Program (GHGRP) on electric power plants in the United States. Starting in 2010, the GHGRP requires both the reporting of greenhouse gas emissions by facilities emitting more than 25,000 metric tons of carbon dioxide per year to the Environmental Protection Agency and the public dissemination of the reported data in a comprehensive and accessible manner. Using a difference-in-difference research design, we find that power plants that are subject to the GHGRP reduced carbon dioxide emission rates by 7%. The effect is stronger for plants owned by publicly traded firms. We detect evidence of strategic behavior by firms that own both GHGRP plants and non-GHGRP plants. Such firms strategically reallocate emissions between plants to reduce GHGRP-disclosed emissions. We interpret this as evidence that the program is costly to the affected firms. Our results offer new evidence that public or shareholder pressure is a primary channel through which mandatory Corporate Social Responsibility (CSR) reporting programs affect firm behavior.
*Previously titled The Real Effects of Mandatory CSR Disclosure on Emissions: Evidence from the Greenhouse Gas Reporting Program
[2] "Did CECL Improve Banks' Loan Loss Provisions and Earnings Quality During the COVID-19 Pandemic? " with Pietro Bonaldi and Pierre Jinghong Liang
Draft available at SSRN (#4360356)
Presented at AAA 2023
Abstract: The Current Expected Credit Loss (CECL) standard took effect in 2020 during the onset of the unprecedented global pandemic. Proponents of CECL argue that the regulation can provide timelier provisions, while others are concerned about the potential for heightened reported earnings volatility. In this paper, we investigate the impact of CECL on the accuracy of banks' loan loss provisions and on earnings quality. We empirically document that starting in the second half of 2020, banks adopting CECL report larger reserve releases and are more likely to report negative loan loss provisions than non-adopters. Using a difference-in-difference design we find that during the first two quarters of 2020, the dispersion of analysts earnings forecasts and the level of discretionary earnings are both larger for adopting banks than for non-adopters. We interpret these as evidence of less accurate provisions and lower earnings quality which is consistent with greater accounting noise and reporting bias caused by the adoption of CECL during high economic uncertainty
[3] "Effects of Capital Requirements on Banks’ Balance Sheets: Causal Evidence from Restrictions on Trust-Preferred Securities" with Pietro Bonaldi and Pierre Jinghong Liang
Presented at WFA 2023, World Finance & Banking Symposium 2021, University of Chicago
Abstract: We provide causal evidence on how banks respond to more stringent capital requirements. We study the reaction to a regulatory change restricting US banks with more than $15 billion in total assets from including Trust-Preferred Securities (TruPS) in Tier 1 capital. Using a regression discontinuity design, we show that banks just above the $15 billion threshold lowered their reported Tier 1 capital and total consolidated assets for every year from 2014 to 2018. Collectively, our findings show that tighter capital requirements caused affected banks to shrink in size.