Member-only story
To VaR is Human
Originally appeared in the June 2009 Edition of Journal of Risk (Defunct).

Risk management is relatively simple in its essence: what is the probability of an event occurring, and, if it happens how badly will it hurt? However, if we consider how we estimate probabilities and losses, the answers to these seemingly simple questions become complicated very quickly. To estimate the size of losses, risk managers have increasingly turned to valuation models, some of which are well entrenched and are taught as gospel at many MBA and CFA programs[1] despite their deficiencies. For probabilities, on other hand, we now rely heavily on statistics to make guesses at estimation. Unfortunately, though, the human brain is incapable of distinguishing probabilities of less than 1%, rendering the comprehension of rare events rather difficult.
For both probability and loss estimation, it is important to understand the limits of theory. Traditional engineers, who often use the same techniques, usually build the highways that fulfill their utility functions, with the full understanding that they may collapse during stress tests (like the ones in the Northridge earthquake in California in 1994). Interestingly, the engineers who have designed collapsed structures have not received much blame for their lack of insight, because it was understood that these structured would fail (eventually). One can’t stop thinking of Roman’s practice of standing the engineer under the bridge as their legions crossed it. In short, while engineers hold the social responsibility to design structures that are well thought out and safe, they do fail[2] from time to time.
While traditional engineering dates back almost 3,000 years, financial engineering is in its infancy (it’s just 50 years old) and has much to learn from its mistakes. In traditional engineering, most structures are overdesigned by a factor of three, a number reached through many years of experience and failure. Why is the factor of safety set to three (even though the actual overdesign number is a multiple of that)? No one knows, but it is assumed that over the years, the rule of thumb (expertise) was developed.
``We would like society to lock up quantitative risk managers before they cause more damage,’’