Financial markets around the world breathed a sigh of relief last month when the vast majority of European banks (84 out of 91) passed their stress tests.
U.S. equities are up, high-risk sovereign debt is rallying, credit-default swap spreads are tightening and the euro and the pound have hit multimonth highs against the dollar.
The same phenomenon occurred after the landmark U.S. bank stress tests of May 2009, which ushered in a 12-month stock market rally. Like a middle-aged man who celebrates a successful cardiology check-up with an extra-cheese and pepperoni pizza, the markets have taken the good news as encouragement to get back to their normal ways.
But as any cardiologist will tell you, it is not enough to pass just one stress test. You must exercise every day to be healthy. Stress tests have received a lot of attention lately, but it is crucial that investors realize this risk management practice is not a one-time pass/fail procedure only to be used in an emergency. Stress testing should become a part of our everyday decision-making. It should inform not just risk managers but also asset managers and senior management. In fact, if the practice of stress testing had gotten the attention before the crisis of 2008 that it is getting today, we might never have gotten into the predicament in the first place.
Stress testing is not a crystal ball. It is not a predictive tool and was never supposed to be. It is a means to get a glimpse of risk under unknowable future situations that are only imagined today. Of course, no one knows what those situations will be, so lots of possibilities are tested with the goal of getting close to the real risks.
That means frequently testing for stresses that never occur. A good analogy is planning for a fire only to be washed away by a flood. To avoid that, a good risk manager plans for both.
The practice of stress testing requires a robust imagination that postulates a variety of situations that could conceivably happen. All these situations can then be evaluated on a scale from most probable to least probable. Once these baseline risks are plotted, they must then be tracked regularly to determine whether they are becoming more or less probable. The key is to have the ability to construct scenarios that are plausible even if they are not all that pleasant to think about.
The second part of stress-testing discipline is to map to current scenarios actual historical events and the fundamental effects these had on valuations and fundamentals when they occurred. This is much more than a one-to-one comparison; the historical analysis requires a deep understanding of the often unpredictable correlations within the universe of assets being stressed, both on a historical and present-day basis. These relationships must be measured, updated and stored every day.
Not long ago, this level of analysis would never have been possible. When stress testing was introduced in the early 1990s, it was essentially a worst-case scenario tool developed to augment J.P. Morgan's value-at-risk, or VaR, methodology. VaR was designed to predict the effects of scenarios that occur 95% of the time, and stress testing was developed to help risk managers prepare for the other 5%, those rare occurrences that can send shock waves through a portfolio. But it was a very manual process, heavy on complex mathematical calculations.
Today, theories can be applied more widely in fractions of a second thanks to the firepower built in to current technologies. It is now possible — in large part due to technological development that grew out of the crisis — for investors to track the relationship between individual balance sheet items and overall, companywide financials and then stress those relationships against a multitude of scenarios without an army of PhDs poring over spreadsheets.
For all the ease of access to analytics, though, the bigger concern is whether market participants will take the lessons of the U.S. and European bank stress tests as a call to action or simply an invitation to make the same mistakes twice. To avoid the latter scenario, it is essential that stress testing become a continuing, integral and aggressively pursued endeavor by policymakers both public and private as a way to better understand and estimate risk.
