Do near misses with a disaster affect individual demand for credit? A large-scale study of credit profiles. (with Michal Grinstein-Weiss) [2022]
We investigate the credit demand of individuals in areas narrowly missed by severe tornadoes. Leveraging detailed credit bureau data and a difference in differences matching model, we find that, compared with similar individuals far away from tornadoes, individuals living in areas neighboring tornadoes decrease credit demand following a tornado even without suffering from tornado-induced financial distress. Within one year after a tornado, the aggregate spillover effect relative to direct impact is roughly 49% for mortgages, 115% for personal loans, and 336% for student loans. These decreases are partly driven by individuals closing down accounts on the extensive margin and more prominent among individuals with more financial flexibility pre-tornado, despite them being better guarded against disaster fallout. Taken together, such behavior suggests elevated risks perception and salience-driven decision making. Finally, we confirm the results by developing a type of interpretable semiparametric deep neural networks. The findings inform us about the aggregate impact of adverse shocks by highlighting an often neglected psychological spillover effect. They also imply that studies using neighboring areas as control groups may misestimate a shock’s impact.
Are overconfident CEOs less resilient to bad news? Evidence from negative news events and hurricanes. Under peer review at Journal of Banking & Finance. [2021]
This paper examines the implications of CEO overconfidence for firm distress risk and investment-related activities in the unique context of adverse incidents. I use an options‐based proxy for CEO overconfidence and study two types of adverse incidents: negative news reports and hurricanes. Using a difference in differences matching model, I find that firms with overconfident CEOs have higher levels of capital expenditure, bonds issued, expected default frequency, and stock volatility than other comparable firms following an adverse incident. However, in the absence of these adverse incidents, the patterns above reverse, i.e., firms with overconfident CEOs have lower expected default frequency, stock volatility, capital expenditure, and bonds issued than other firms. Furthermore, an orthogonal random forest model reveals more granular and nonlinear impacts of CEO overconfidence on firm outcomes. In contrast to previous studies, my findings suggest that in normal times CEO overconfidence results in positive outcomes for the firm, while under adverse conditions CEO overconfidence results in negative outcomes for the firm.
A real options perspective on corporate social responsibility: evidence using causal forest. (with Maggie Zhou) Under peer review at Management Science. [2020]
This paper uses real options theories to examine how two types of uncertainty influence Corporate Social Responsibility (CSR) activities based on these activities' unique features of enhancing revenue growth and insuring against negative events. In particular, we capture uncertainty over firms' negative social/environmental/political events (negative event uncertainty) using sentiment embedded in news reports and uncertainty in firms' product or financial market (market uncertainty) using crude oil price shocks. With the help of the machine learning method causal forest, we find that publicly-traded U.S. firms increase CSR activities when facing negative event uncertainty, but reduce CSR activities when facing market uncertainty. These relationships are moderated by firms' innovative productivity, which influences the opportunity cost of CSR investment, firms' brand equity, which influences the damage of negative events, and the competitiveness of firms' product markets, which influences CSR's value to the firm. The findings offer practical advice for corporate executives in CSR engagements and policymakers in encouraging CSR activities.
Why does general deterrence exist? A Study of Drinking Water Quality Regulation. (with Thomas Lyon) Under peer review at Journal of Environmental Economics and Management. [2019]
This paper studies the general deterrence effect associated with regulatory enforcement actions, using data on the quality violations and regulatory enforcement records of all 141,290 active drinking water systems in the U.S. between 1980 and 2017. We find that general deterrence among drinking water systems is much more prominent when the enforcement actions are coercive as opposed to cooperative in nature, and that it only exists among water systems that are not repeat violators. Assuming general deterrence is a response to new information revealed about the regulator, we postulate this information has to do with either (1) the regulator’s type, i.e., how pro-corporate vs. pro-environment the regulator is, or (2) the regulator’s level of resource constraints, since regulators become more strategic in enforcement when they are subject to resource constraints. Contradicting (1) but in support of (2), we find that general deterrence is unchanged or disappears during regulator turnover, when uncertainty about the regulator’s type is high, whereas general deterrence is stronger for regulators with tighter budget constraints.
Does mortgage forbearance work? Evidence from credit profiles. (with Taylor Begley, Radha Gopalan, and Michal Grinstein-Weiss) Results analysis phase. [2021]
Under a forbearance program, borrowers temporarily suspend or reduce mortgage payments while recovering from financial distress. While forbearance may help avoid foreclosures and their macroeconomic spillover effects, there are critiques that it merely kicks the can down the road. This paper explores mortgage forbearance's impact on individual credit performance. We control for omitted-variable bias in entering forbearance by exploiting the heterogeneity in agency conflicts and information asymmetries across servicers of non-agency vs. government-sponsored enterprise (GSE) mortgages. We train a deep neural network model that predicts forbearance in a sample of non-agency mortgages. We then use the model to predict forbearance among GSE mortgages and use mortgages in this sample that are predicted to be in forbearance but do not actually get it as the control group for our treatment group of non-agency mortgages with forbearance.
The impact of financial institutions on PPP loan effectiveness: evidence from Mastercard transaction data. (authorship to be finalized) Data cleaning phase. [2021]
During the COVID-19 pandemic, the U.S. government launched the Paycheck Protection Program (PPP), which offers loans that help businesses keep their workforce employed and avoid the macroeconomic spillover effects of layoffs. However, the agency problems and institutional frictions faced by the financial institutions distributing these loans mean that the same amount of loans may have varied effectiveness in boosting local business activities. Leveraging the heterogeneity in information asymmetries between community and national banks, this project studies the impact of financial institution types on the effectiveness of PPP loans in boosting local business activities, with anonymous credit and debit card transaction records provided by Mastercard.
Preventing another Flint: ownership, local community characteristics, and drinking water quality. (with Thomas Lyon and Wren Montgomery) Draft available upon request. [2020]
Drinking water quality, long taken for granted in the U.S., has emerged as a serious concern after Flint. This paper presents the first large-scale study of the relationship between ownership changes and regulatory compliance among drinking water systems in the U.S. between 1980 and 2017. We find that privatization consistently produces quality improvements, while municipalization does not. Drawing on the comparative advantages of for-profit and non-profit governance forms in providing service quality, we hypothesize and empirically prove that the impact of ownership form depends on both regulatory pressure and local community characteristics.
From Excess to Entropy: Municipal Water Services Governance in the United States. Simon Porcher and Stéphane Saussier (eds.), Facing the Challenges of Water Governance. Palgrave Studies in Water Governance: Policy and Practice, London: Springer Nature. (with Wren Montgomery) [2019]
Not a Drop to Drink? Drinking Water Quality, System Ownership, and Stakeholder Attention. Forrest Briscoe, Brayden King, and Jocelyn Leitzinger (eds.), Social Movements, Stakeholders, and Non-market Strategy. Research in the Sociology of Organizations, Volume 56, Emerald Group Publishing Limited. (with Thomas Lyon and Wren Montgomery) [2018]
A Change Would Do You Good: Privatization, Municipalization, and Drinking Water Quality. Academy of Management Best Paper Proceedings. (with Thomas Lyon and Wren Montgomery) [2017]
Use of Public Benefits Over the First Year of Pandemic. Social Policy Institute Research. (with Sam Bufe, Stephen Roll, Katie Kristensen, Mat Despard, and Michal Grinstein-Weiss)[2021]