Job Market Paper

Do Delays in Banks’ Loan Loss Provisioning Affect Economic Downturns? Evidence from the U.S. Housing Market

I study whether banks’ loan loss provisioning contributed to economic downturns by examining the U.S. housing market. Specifically, I examine the influence of delayed loan loss recognition (DLR) on bank lending and risk-taking in the U.S. mortgage market and the aggregate effects of DLR on house prices and household consumption during the Great Recession. I first examine the effects of DLR on individual banks’ behavior. Then I construct ZIP code-level exposure to banks’ DLR to examine the aggregate effects of banks’ DLR on the housing market. I find high DLR banks reduced mortgage supply, leading high exposure ZIP codes to experience larger decreases in mortgage supply during the crisis. Mortgages from high DLR banks were also more likely to become distressed, leading to more foreclosures and short sales in high exposure ZIP codes during the crisis. Consequently, banks’ DLR negatively affected house prices during the crisis, implying a significant decrease in household consumption. These findings suggest banks’ loan loss provisioning affected loan supply and risk-taking, exacerbating the economic downturn via the household channel.
The paper was invited and presented for Emerging Scholar Poster Session at the 2018 CMU Accounting Mini Conference.
Online Appendix

Publication

Economic Consequences of the AOCI Filter Removal for Advanced Approaches Banks
with Seil Kim and Stephen Ryan, The Accounting Review, Forthcoming

We examine economic consequences of US bank regulators’ phased removal of the prudential filter for accumulated other comprehensive income for advanced approaches banks beginning on January 1, 2014. The primary effect of the AOCI filter is to exclude unrealized gains and losses on available-for-sale securities from banks’ regulatory capital. We predict and find that, to mitigate the regulatory capital volatility resulting from the AOCI filter removal, advanced approaches banks increased the proportion of investment securities they classified as held to maturity, thereby limiting their financing/liquidity and interest-rate-risk management options, and they shifted their investment security holdings into safer types of securities, thereby reducing their interest rate spread. We further predict and find that these banks borrow more under securities repurchase agreements potentially collateralized by held-to-maturity securities and reduce loan supply owing to their reduced financing options, and that they invest in riskier loans to increase their interest rate spread.
Read More

Working Papers

Seeking Accounting Arbitrage: Evidence from the US Life Insurance Industry

This study examines how accounting rules induce the US life insurers to use “shadow insurance” to manage their regulatory capital, and whether they exploit shadow insurance to hold more risky assets when a new accounting rule creates more capital pressure. Shadow insurance refers to transactions ceding liabilities to captive reinsurers under less strict regulation to circumvent the regulatory capital requirement. I employ a recent adoption of Actuarial Guideline 43 (AG 43) as a shock to the strictness of reserve requirement under Statutory Accounting Principles, and conduct a difference-in-differences analysis. I find insurers subject to AG 43 engage more in shadow insurance transactions after the adoption of AG 43, and the impact of AG 43 is significantly stronger for more capital-constrained insurers and weaker for insurers domiciled in states with strong regulatory enforcement. In addition, I find insurers subject to AG 43 invest in less risky bonds on average, and only well-capitalized insurers exploit shadow insurance to hold more risky bonds. Overall, my results suggest insurers exploit an accounting arbitrage opportunity despite the fact that these transactions are subject to the regulatory monitoring. This evidence implies that when two different accounting regimes exist, changing one accounting regime may not always bring an intended regulatory result. Additionally, my evidence adds new insights to the literature on the real effects of accounting by showing the change in accounting measurement can affect firms’ capital management behavior and investment decisions.
Read More

Works-in-Progress

Tax Planning by Financial Institutions
with John Gallemore and Brandon Gipper

This study examines tax planning by financial institutions. Like non-bank corporations, the corporate tax system provides incentives and opportunities for banks to avoid taxes. However, the tax planning of financial institutions (including commercial banks, investment banks, and insurance companies) has largely been ignored. We plan to fill this gap in the literature by providing the first evidence on the tax avoidance activities of financial institutions. First, we plan to provide descriptive evidence on the tax avoidance of financial institutions, how it varies across different types of financial institutions, how it has evolved over time, and how it compares to non-financial institutions. Second, we plan to explore how previously researched determinants of tax planning differentially affect tax avoidance at financial institutions versus non-financial institutions. Finally, we plan to explore determinants and consequences of tax planning by commercial banks, especially, focusing on the role of bank regulators and liquidity/capital shocks. This study will contribute to the literature on corporate taxation by being the first to provide a comprehensive investigation into tax planning activities of financial institutions. In doing so, we will provide an empirical framework that future research can use to explore the determinants and consequences of tax avoidance in financial institutions.

Neglected Insider-Trading in Stand-alone Public Banks
with Seil Kim

This paper examines whether stand-alone banks’ insider-trading filings with bank regulators are neglected by market participants, and if so, the channels through which this happens. We find that insider-trading filings on FDICconnect by these banks have smaller market reactions in terms of returns and trading volume compared to insider-trading filings on SEC EDGAR by bank-holding companies in the short run. However, these differences largely disappear in the long run, suggesting the different market reactions in the short run is unlikely to be driven by differences in the information content of the filings. We find that filings on FDICconnect are not covered by the media or real-time data vendors. Our findings suggest that market participants are unfamiliar with disclosure by stand-alone banks and do not react to insider-trading filings on FDICconnect as they do for the same filings on SEC EDGAR.