The risk management profession is in need of some fresh air. Aside from the emergence of ERM, the changes to the risk management framework have been minor over the last 10 years. The good news is that some very impressive work regarding risk has been done in other disciplines, such as psychology and decisions analysis. This work can readily be adapted for use by most every risk manager. However, these new concepts do not often find a ready audience in the risk management community outside of Wall Street.
For example, rarely will one hear mention the work of Daniel Kahneman and Amos Tversky. Kahneman and Tversky won the Nobel prize in 2002 for work on how people make decisions regarding risk and uncertainty. A brief look at it will illustrate why their work is important to the risk management profession.
Kahneman and Tversky set out to discover how people deviated from rational choices when making decisions involving risk. For example, they found that if they offered people two choices, often they would choose the option that, mathematically, would result in a lower reward. For example, in their paper on Prospect Theory, they specifically sought an answer to why people would buy insurance policies even when doing so would result in a loss. What they found challenged accepted concepts around how people decide.
To explore decision-making, they gave test subjects choices between different gambles. For example, they offered the choice of either having an 80 percent chance of winning $4,000, or a 100 percent chance of winning $3,000. In standard probability analysis, one would multiply the award by the probability and conclude that the bet on $4,000 at 80 percent is worth $3,200. This is clearly more than $3,000, so, rationally, the subject should choose to take the bet for $4,000. However, the test subjects overwhelmingly chose to take the $3,000. Likewise, if they gave subjects the choice of losing $4,000 at 80 percent or being guaranteed to lose $3,000, they overwhelmingly chose to gamble on losing $4,000 rather than surely losing $3,000.
In other words, when contemplating a risk involving an award, people will become irrationally risk averse; when the gamble involves potential loss, people become overly risk seeking.
In their work, Kahneman and Tversky also looked for an answer to the question: "Why . . . would people spend so much money to purchase insurance policies at a price that exceeds the expected actuarial cost?" A very good question indeed, and one that should have been answered long ago by the risk management profession. What they found was that the majority of test subjects valued reducing a loss much less than eliminating a loss to the point that they would pay far more than they rationally should for broader insurance coverage.
There are other numerous conclusions drawn by Kahneman and Tversky. For example, they explain how probability gets ignored in favor of severity for low-probability losses. It is all of crucial importance to all risk managers. Applying their work to the commonly used risk map, one can readily conclude that risk maps exacerbate the biases outlined in the researchers' first paper, published in 1979. Yet in the 27 years since that publication, and in the four years since Kahneman and Tversky won the Nobel Prize, there has yet to be a new risk map version widely circulated that addresses these biases.
Others besides Kahneman and Tversky have done the hard work of solving risk management problems. The work is done. We only need have the desire to learn and apply it. And we must in order to energize the field of risk management.
BEAUMONT VANCE, senior risk manager for Sun Microsystems Inc., is a columnist for Risk & Insurance®.
September 1, 2006
Copyright 2006© LRP Publications