What Matters Most in the Design of Stress Tests? Evidence from U.S. and the Europe

In the aftermath of the global financial crisis, supervisors in Europe and the U.S. have undertaken a series of bank stress tests to restore market confidence. In this paper we use event study methods to compare the market impact of all U.S. and EU-wide stress tests performed from 2009 to 2013. We find that, typically, the publication of stress test results has a positive impact on stressed banks’ returns. However, while the 2009 U.S. stress test had a large and positive impact on stressed banks, the impact of subsequent U.S. exercises decreased over time. Contrary to anecdotal evidence, we find that the 2011 EU exercise is the only EU-wide stress test that resulted in a significant negative market reaction. Comparing EU-wide stress tests among themselves and with U.S. stress tests highlights the importance of the governance of the stress tests. Governance turns out to be more important for the success of stress tests than technical elements, such as the minimum capital adequacy threshold or the level of disclosure of bank-by-bank data.