This article was originally post on StrexGroup.com. Click here for the original article or keep reading below:

A couple of weeks ago, during the New York Fashion Week, seeing glamorous models at the Lincoln Center Plaza brought to mind the beauty of models. Unfortunately, the conversation at a meeting soon after actually highlighted the ugliness of models.

At that meeting, frustrations were evident that regulators are always complaining that, despite the beauty of their sophistication, bank models underestimate the capital regulators believe is needed to support risk at financial institutions, and thus can lead to ugliness in times of financial crises similar to the crash of 2008.

The Beauty of Models

Financial investing is based upon the theory of probability. Responsible investing requires taking into account the likelihood of favorable as well as unfavorable results. If the portfolio investing is structured in such a way that the financial return on instruments can cover unfavorable hits more times than not, then over time this will yield profitable returns. So how does one predict what the likelihood of possible outcomes is?

Previously, bankers and investors employed gut feel, developed from their historical experiences, and a few quantitative indicators to assess transactions. This had limitations as only so much information could be processed in one’s head or on paper to draw conclusions to make decisions. Quantitative models, driven by almost unlimited computing power, changed that. Now all the available data from historical experiences, even going decades back, can be turned into precise models providing key decision inputs.

This is the beauty of models. They have made possible what was unthinkable not that long ago. Today revenue engines of financial institutions are driven by the outputs of sophisticated, proprietary quantitative models.

But there is another side of models. If used improperly, they can produce numbers that convey a sense of precision, but may actually be useless to misleading in relation to how the bank may fare in financial crises, almost always understating the impact on the bank’s financials.

The Ugliness of Models

Leveraging the financial-return principle stated above, almost all risk models are designed to structure, preserve or protect revenues from risk. What they do to support this objective is pretty well understood. However, the ugly side of the models, or why people think “they don’t work,” is not always understood and appreciated.

Until 2008, most models were designed to deal with the everyday expected ups and downs of the market, or “normal risk,” and left extreme circumstances unaddressed as the models didn’t have much data to address stressful environments. In the absence of any significant problems with revenue models in the years preceding 2008, a complacent mindset prevailed that kept banks from focusing on the dangers underlying assumptions and what models don’t do. This shortcoming was highlighted by the blindside blow experienced by most financial institutions.

Since 2008, significant attention and resources have been devoted to addressing these shortcomings. However, these models still underestimate the impact of stressful environments. This is the cause of the gap between the results of these models and what regulatory stress-test models imply during severe down cycles.

Actually, the gap between what these models say and what a bank may experience in extreme crises may be worse because none of these models can address black swans, or unexpected events that can be highly consequential. All current models, including the ones used for risk management and stress testing, are “event centric” or require being able to define events. Black swan events can’t be envisioned. Hence the ugly truth of these models is that they work beautifully in normal times but leave institutions vulnerable to extreme risk. (See “What Stress Testing is Not”).

Need to Focus on the Ugly Impact of Financial Crises

Therefore, a different approach is needed to address the impact of extreme risk. This is not easy as there are not even any metrics with which to measure it. Without such metrics, banks and regulators rely on extensions of normal-risk management and remain unaware of big vulnerabilities (think London Whale).

We need an objective way to measure extreme risk, along with specific regulatory and institutional guidelines to manage and contain it effectively. Financial crises are part of economic cycles, and we will face one again. Unless banks and regulators address the root of the problem, and alter their exclusive reliance on traditional risk-management and stress-testing models, we may be in for a repeat of the ugliness of 2008.

About the Author

Karamjeet Paul, Managing Principal of Strategic Exposure Group, has developed expertise in extreme tail risk. His commentaries have been widely published. He is the author of Managing Extreme Financial Risk: Strategies and Tactics for Going available for purchase on the Elsevier Store. Use discount code “STC3014” to save 30%!

he has over 30 years of operating, finance, treasury, and exposure managment experience, giving him unique expertise in identifying and addressing critical risk-exposure-reward and other strategic issues at the highest levels in large organizations. Mr. Paul’s perspective has been gleaned from hands-on experience in large global organizations, startups in entrepreneurial settings, and consulting assignments.