Deposit Competition and Financial Fragility: Evidence from the US Banking Sector

PRELIMINARY AND INCOMPLETE, PLEASE DO NOT CITE WITHOUT PERMISSION. We develop and estimate an empirical model of the U.S. banking sector using a new data set covering the largest U.S. banks over the period 2002-2013. Our model incorporates insured depositors and run-prone uninsured depositors who have rich preferences over differentiated banks. Banks compete for deposits in the spirit of Matutes and Vives (1996) and can endogenously default. We estimate demand for uninsured deposits and find that it declines with banks’ financial solvency, which is not the case for insured deposits demand. We calibrate the supply side of the model and find that the deposit elasticity to bank default is large enough to introduce the possibility of multiple equilibria, suggesting that banks can be very fragile. We study how competition for deposits among banks affects the feedback between bank distress and deposits, and transmits shocks from one bank to the system without direct links between banks. Last, we use our model to analyze the proposed bank regulatory changes and find that some regulations could exacerbate the instability of the system. ∗We thank David Scharfstein, and the seminar participants at Chicago Booth Microeconomics Lunch, New York University, University of Tokyo, University of Calgary Empirical Microeconomics Workshop, Wharton Conference on Liquidity and Financial Crises, and the Econometric Society North America Meetings. The authors are from a: University of Chicago b: University of Chicago Booth School of Business, and NBER. All errors are our own.