Publications/Forthcoming Papers:


 Journal of Finance, June 2024

Abstract:  Financial institutions received investments under the Troubled Asset Relief Program (TARP) in a bad state of the world but repaid them in a relatively good state. We show that the recipients paid considerably lower returns to the taxpayers compared to private market securities with similar risk over the same investment horizon, resulting in a subsidy of over $50 billion on the preferred equity investment by the government. Ex-post renegotiation of contract terms limited the upside gains received by the taxpayers in good times and contributed to the subsidy. These findings have important implications for the design and implementation of future bailouts. Our simple methodology for calculating the subsidy can be applied to evaluate the financial costs of other bailouts.


Journal of Finance, forthcoming

Abstract:  Using a novel dataset of realized syndicated loan cash flows and a risk-adjustment methodology adapted from the private equity literature, I provide a measure of risk-adjusted returns for bank loan cash flows. Banks, on average, generate 180 bps in gross risk-adjusted returns and add $75 million of value annually to their sample loan portfolios. Banks earn higher returns when they lend to financially constrained borrowers, and the risk-adjusted performance of bank loan portfolios exhibits persistence. However, banks require higher risk-adjusted returns when facing their own financing frictions, and shareholders earn nearly zero net risk-adjusted returns once bank staff are compensated for their effort in lending. Overall, these findings offer evidence that banks provide valuable services to mitigate borrowers’ financing frictions, and the present value of loan cash flows pays for the costs of the bank providing these services.

Working Papers:


Media Coverage: Bloomberg, Financial Times

Abstract: Private debt funds are the fastest-growing segment of the private capital market. We evaluate their risk-adjusted returns, applying cash-flow-based methods to form a replicating portfolio that mimics their risk profiles. Accounting for both equity and debt factors, a typical private debt fund produces insignificant abnormal returns to its investors. However, gross-of-fee abnormal returns are positive, and using only debt benchmarks also leads to positive abnormal returns, as funds contain equity risks. The rates at which private debt funds lend appear high enough to offset funds’ fees and risks but do not exceed both fees and investors’ required rates of return.


Abstract: A key function of government-sponsored enterprises (GSEs) is to insure mortgage default risk, yet little is known about whether these guarantees are fairly priced. Using a replicating portfolio and market prices of reinsured mortgage default risk to value the net cash flows from guaranteeing $5 trillion mortgages, I show that GSEs shifted from providing a subsidy of 20 basis points (bps) pre-crisis to generating risk-adjusted profits of 30 bps in the post-crisis period. Using a regression discontinuity design around the conforming loan limit, I document that this higher pricing is passed on to borrowers via higher conforming mortgage rates. Following this increase, banks reduce their reliance on GSE securitization but still securitize most mortgages through GSEs despite paying an above-market rate. Overall, these findings imply lenders value using GSEs for reasons other than government-subsidized funding in the post-crisis period. 


Presentations: 19th Annual FDIC/JFSR Banking Research Conference Poster Session, CEAR-CenFIS Financial System of the Future Conference 2019, SFS Cavalcade 2020

Abstract:  Despite plenty of anecdotal evidence of hidden losses in banks, there is no systematic study analyzing its economic drivers: we simply do not get to observe what banks are hiding. Using a regulatory change in India that forced banks to reveal their hidden losses, we show that banks with higher shareholding by distant and passive investors hide more. These effects are stronger for banks where CEOs get highly compensated for reported profits. Our findings caution against using high-powered compensation contracts as a substitute for diluted monitoring. Instead of solving the agency problem, it can result in perverse misreporting incentives.


Abstract:  How different is the cost of capital when banks finance a loan compared to the rate required by direct household investors? This paper quantifies this difference by estimating the prices both investors would be willing to pay for an identical set of loan cash flows when markets are segmented and households are subject to idiosyncratic consumption shocks. Using recently developed methods from the private equity literature, I find that banks are willing to pay 70% more than households for the same loans, which equates to a 2% pp lower cost of capital for banks. This difference in cost of capital arises because loans are more ‘diversifiable’ in the bank’s portfolio than they are in a household’s portfolio, which is subject to idiosyncratic consumption shocks. Additionally, this cost of capital advantage increases when banks lend to riskier firms and possess a larger deposit base. Overall, these findings corroborate classic theories of bank risk-sharing, in which banks invest on behalf of risk-averse households in an effectively more risk-tolerant fashion, providing a lower cost of finance.

Other Papers: