Bank Stress Tests and Accounting Discretion

Abstract

We examine stress tests through the lenses of an implicit, underlying incentives mechanism for bank managers to exercise accounting discretion. Our research aim and the relevant components are examined in defined steps by reference to Propensity Score Matching analysis, which is paired with a Difference-in-Difference approach, and a fixed effects regression model. Our findings show that larger banks with an asset portfolio of relatively low quality, reduced profitability, and higher exposure to market-based activities are more likely to be stress tested. Stress tests exacerbate the effort of tested banks to recapitalize their balance sheet and improve their risk profiles. However, they do not significantly affect the key tools that bank managers use to apply accounting discretion. In this vein, tested banks are highly engaged in accounting discretion over loan loss provisions to manage both income and capital figures. Banks with low capital adequacy are found to apply discretionary practices to a greater extent. Moreover, the banks that participated in early stress tests appear to engage in accounting discretion to a higher degree compared to those participated in some later exercise, which reveals an upward movement on the regulatory learning curve. Our results also show that stricter regulatory rules, more robust supervisory regimes, and more transparent economies mitigate the impact that stress tests have on the incentives of the managers of tested banks to exercise accounting discretion even though the relevant incentives cannot be eliminated.

Keywords: OR in Banking; Bank Stress Tests; Managerial Incentives; Financial Reporting; Data Reliability

JEL Classification: C44; G01; G21; M41