Academic journal article AEI Paper & Studies

Regulatory Stress Tests and the Blind Scales of Justice

Academic journal article AEI Paper & Studies

Regulatory Stress Tests and the Blind Scales of Justice

Article excerpt

Key Points

* One of the newest fads in banking regulation is the use of econometric models to simulate a bank's performance over a hypothetical multiyear economic stress scenario and then using the "stress test" results to calibrate the bank's minimum regulatory capital requirement.

* This regulatory approach for setting capital requirements has become popular in the US and Europe, yet there is no evidence that regulators' econometric stress test models generate accurate forecasts of bank performance under stress scenario conditions.

* Regulators keep the specific details of their stress test models confidential, inhibiting independent verification of their models' accuracy.

* These and other features of regulatory stress tests raise important legal issues and create the potential for costly misallocations of banking resources.

Around 1750 BC, Babylonian king Hammurabi codified a single set of laws to replace the patchwork of tribal laws across his kingdom. Hammurabi's laws, inscribed on stone tablets, set legal standards for tort, inheritance, contract, and family law that established principles that are still relevant today. For example, when you read one of Hammurabi's laws, you immediately understand how to comply with the law. Citizens were presumed innocent until proven guilty, and false accusations carried penalties. If Hammurabi were still alive today, he would be shocked that his basic principles of jurisprudence are ignored in the modern legislation that mandates regulatory stress tests.

The success of the 2009 Federal Reserve Supervisory Capital Assessment Program (SCAP) in the US spawned a new paradigm of bank regulation built around econometric forecasts of bank performance under stressful conditions. Since the SCAP stress tests, banking regulators in the US and Europe have adopted dynamic stress testing as an important component of their large financial institution supervision process. In the US, the 2010 Dodd-Frank Act requires the Federal Reserve Board (FRB) to perform annual stress tests. In Europe, Article 23 of European Union Regulation No. 1095/2010 requires the European Banking Authority (EBA) to develop "an adequate stress testing regime."

Hammurabi would be deeply disappointed if he read either the US or EU law mandating regulatory stress tests. Neither directive explains what is required for institutions to pass an annual stress test. Indeed, these directives never define what comprises a "stress test." Instead, legislators delegated the definition and design of the stress testing requirement to banking regulators who are free to specify the terms and conditions that banks must meet to fully comply with the law. Unlike a Hammurabi law, it is impossible for a banker to read either directive mandating regulatory stress tests and have any idea how to comply with it.

While specific details differ, bank regulators have interpreted these stress testing directives to mean the use of econometric methods to assess a bank's compliance with minimum regulatory capital standards over one or more severe multiyear economic stress scenarios. Banks are required to use their own models to project their performance under specified economic conditions, and regulators evaluate the banks' estimates by comparing bank forecasts to projections from their own supervisory stress test models.

If the stress testing exercise indicates that a bank would suffer losses that deplete the institution's capital, the supervisor is required to impose remedies. For example, in the US, if an institution fails its stress test, the FRB may require the bank to raise capital by prohibiting dividend payments and capital repurchases. The FRB may also require the bank to improve its stress test modeling practices if, in the regulator's judgment, the bank stress test model projections are inferior compared with an unspecified minimally acceptable standard. The acceptability of a bank's proprietary model is likely determined by comparison with the structure and output of the regulator's own confidential stress test model. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.