Robert F. Weber, A Theory of Stress Testing of Financial Institutions as a Deliberative Exercise, University of Tulsa Legal Studies Research Paper No. 20013-01 (June 2013), available at SSRN.
Reading about the fatal flaws and failures of financial reform day in and day out can make you forget things. Like the actual contents of financial reform…and the fact that it is happening under our noses even as we curse statutory nonsense and the glacial pace of rulemaking. Robert Weber’s article on stress tests is a healthy reminder that financial regulatory methods have changed in important ways since 2008, and that we have a lot of figuring-out to do about them.
Stress tests in finance do three things. First, they help firms identify and manage risk from adverse shocks—a spike in interest rates, a collapse in housing prices, a sharp slowdown in economic growth, or a government debt default. Second, they help regulators judge the resilience of individual firms and financial systems. Third, they help communicate information about risk and resilience to the markets and to the public at large, as well as to narrower constituencies of financial firms and their regulators.
The basic idea of stress testing is not new; I learned that depending on how you count, it goes back at least to Leonardo da Vinci or to the U.S. savings and loan crisis. Financial regulators have used stress tests here and there for over two decades to gauge the riskiness of thrifts’ derivatives plays, the resilience of banks’ trading portfolios, and the stability of entire national financial systems, among others.
In retrospect, the year 2009 may come to look like a turning point, when stress tests began to subsume and eclipse key methods of regulating financial institutions. Consider this: in the old days, national supervisors told banks to keep their capital above the more-or-less arbitrary minimum level agreed in Basel. For U.S. banks, breaching magic-number thresholds brought on “prompt corrective action” (PCA)—at least in theory, restrictions on dividends, acquisitions, asset growth—and at the extreme, resolution. But beginning in 2009, the number that really matters for biggish banks, bank holding companies, and systemically important financial institutions is the capital they have left in a “severely adverse” stress test scenario. The ad-hoc stress tests used in 2009 to decide whether the largest U.S. banks needed capital from TARP have morphed into a statutorily mandated annual cycle of supervisor- and firm-conducted exercises. A parallel process is underway in Europe. While capital adequacy still occupies the core (and the bulk) of institutional regulation, questions of who designs and administers the stress tests, how, and to what end become all-important.
In stark contrast to the old PCA regime and the scattered use of stress tests pre-crisis, today’s regime is designed in important part for public consumption. The tests’ job description has expanded from internal risk management and supervision, to serving as vehicles of market discipline for banks and political accountability for their regulators. The 2009 and 2010 U.S. tests are credited with the return of public confidence in the U.S. financial system, which benefited both banks and their regulators. On the other hand, European stress tests became the source of considerable political embarrassment, with concrete implications for the design of Europe’s banking union.
Robert Weber’s article is a wonderful resource for anyone seeking to develop a view on what stress tests are and what they should do in financial regulation. It has thorough genealogies, un-gimmicky typologies, theory, prescription, and even a bit of poetry between the lines.
I learned about stress tests and adjacent risk management methods, from their origins in engineering (how will the bridge collapse?), to their unlikely emergence as part of U.S. thrift deregulation, and their relationship to Value at Risk methodologies. I was especially glad to get frequent reminders of the transplant quality of stress tests: tools that help manage failure in physical systems may not work the same way for social systems. Throughout, Weber treats complex technical material with freshness and clarity, which makes the article enjoyable and teachable.
It would be unfair to present the project as mere exposition (though you can tell I really loved that part). The author’s overarching purpose is to recast what stress tests should do, and how. He draws the distinction between “stress tests as assurance” and “stress tests as deliberation.” The former is a static, top-down, outward-looking snapshot audit exercise. The latter is dynamic, interactive, and soul-searching—staring failure in the eye and seeing a reflection of your soul. The distinction maps onto the familiar descriptions of old (bad) and new (good) governance. The policy proposal is to get away from the old and double down on the new.
The distinction between stress-tests-as-assurance and stress-tests-as-deliberation is very nicely developed in the paper and useful in practice. It highlights the perils of stress testing, now on display in Europe, where firms that passed with flying colors folded in the ensuing months. And it puts Weber in good company. But the line is too sharply drawn for my taste. I suspect that assurance and the associated communicative functions—telling investors and voters that all is well, or not—are responsible for the political salience of stress tests, which in turn explains their rise to the top of the regulatory toolkit.
A system that runs on public confidence is hungry for assurance. In this world, stress tests will be used to tick boxes and score political or market points no matter what. In a good state, this can even help fuel demand for regulation. The trick is to figure out how to do stress tests so they provide real assurance—make them dynamic, iterative, multi-dimensional—which requires genuine humility and intense collaboration among regulators and firms around the world, all of whom must change in the process.
This is a tall order and a long path. Robert Weber shows what is at stake, and sets us on the way.