The Impact of Supervisory Stress Tests on Bank Ex-Ante Risk-Taking Behaviour: Empirical Evidence from a Quasi-Natural Experiment

Abstract

We use the Federal Reserve's stress-testing regime as a quasi-natural experiment to examine the impact of supervisory stress tests on bank ex-ante risk taking behaviour. Using a sample comprising large U.S. bank holding companies over the period from 2003Q1 to 2016Q4, we find that banks which are subjected to annual supervisory stress tests tend to reduce their overall risk by choosing asset portfolios of lower risk exposures. Nevertheless, this risk reduction happens mainly because stress-tested banks reduce the holding of low-risk assets rather than risky assets. We also find that stress-tested banks tend to reduce their on-balance sheet exposures rather than off-balance sheet exposures. Overall, our finding implies that, while supervisory stress tests can help to reduce the banks' overall risks, policy makers should also have a closer look at the mechanisms in which banks allocate risk to mitigate moral hazard and regulatory arbitrage behaviour.