Bank Stress Tests: An Active Treatment or a Placebo?
Posted: 9 Dec 2016 Last revised: 20 Mar 2018
Date Written: September 15, 2017
Abstract
We empirically examine the quality of accounting information that authorities utilise when conducting bank stress tests and provide evidence on how this information links to the outcome of these tests. Based on a unique dataset of tested and non-tested banks from 27 European countries, we document a clear link between stress tests and accounting discretion. We find that tested banks, in contrast to their non-tested counterparts, are penalised by regulators and markets through stress tests. Consequently, this incentivises bank managers to exercise accounting discretion over loan loss provisions in order to manage both capital and earnings. Our results suggest that stress tests exacerbate banks’ discretionary behaviour with the purpose to pass stress tests and convey a sound image to regulators and market participants. This phenomenon appears to be modest under stricter institutional environment. Overall, we show that, in certain occasions, stress tests constitute a ‘placebo’ and not an active treatment.
Keywords: Bank Stress Tests; Financial Statements; Accounting Information; Accounting Discretion
JEL Classification: C12; C20; E58; G21
Suggested Citation: Suggested Citation