Bank Stress Tests: An Active Treatment or a Placebo?

Abstract

The smooth operation of the banking sector can largely contribute to the stability of the financial system and to that of the entire economy. After the eruption of the global financial crisis, regulators worldwide introduced stress tests as a key supervisory tool to assess the resilience and soundness of systemically important banks on a regular basis. In this paper, we empirically examine the quality of the accounting information that bank authorities use in the conduct of stress tests and how this is related to the outcome of the tests as perceived by market participants and other economic agents. Analysing banks from 27 European countries, we document the treatment effect on the treated and the link between bank stress tests and accounting discretion. In particular, stress-tested banks apply discretionary loan loss provisioning to manage capital and to signal soundness, and thus convey to market participants a placebo than an active treatment.