Did the Cost of Risk Change on 9/11?

by Oakley E. "Lee" Van Slyke

Twelve actuaries met at the Special Interest Seminar, "The Changing Insurance Market," in Dallas, on April 16, 2002, to discuss the question, "Did the Cost of Risk Change on 9/11?" I served as facilitator of the group.

Some participants noted that the "cost of risk" did increase on 9/11. As Barry Franklin, of Aon, put it, "Risk transfer is more expensive today." Shaun Wang, in a paper presented at AFIR, has shown that the cost of reinsurance has increased even for earthquake cover, for which the probabilities of loss had not changed.

There was a great deal of discussion about the meaning of the question. To my personal surprise, there was a consensus that the question should be worded, "Did the cost of uncertainty change on 9/11?"

The group noted that the markets changed over time following the events of 9/11. Referring to a hardening market, Franklin said, "9/11 amplified a lot of things that were going on anyway." In many cases, he noted, "Prices skyrocketed." In some markets, the cost of risk changed on 9/11, but then changed back. Specifically, the group observed that although the property catastrophe market is still tight, the broad capital markets recovered within a month or so.

Bruce Bradley, FSA, of UICI, a health insurer, reported that UICI's stock price had gone down, then recovered and risen to new highs after 9/11. Apparently the capital markets initially punished all stocks, then backed off to punishing all insurance stocks, then finally settled on those parts of the insurance industry that really were affected by the events, as hindsight informed us about those events. This process is still going on, according to the discussion that followed.

Greg Cote, of The Travelers, pointed out that changes in probabilities have arisen in two ways. Certainly, the events of 9/11 changed the explicit probability that certain events will happen in the future. In addition, the subjective probability that the underwriters and actuaries consider in their world outlook had also increased. As someone had stated at an earlier session in the day, "what before had been unthinkable was now the subject of serious thought."

After a lengthy exchange, the group reached a consensus regarding how to use the terms "risk" and "uncertainty," a conclusion that may surprise most actuaries as much as it did me. Although for more than a century actuaries have used the term "risk" to denote uncertainty, the group agreed that this made for poor communication. Underwriters, risk managers, and regulators all use the term "risk" to mean insurance obligations, not just their uncertainty. The group recommended that casualty actuaries refer to the cost of the capital that supports underwriting and investment outcomes that are uncertain as "the cost of uncertainty."

The attendees then turned their attention to the question, "Did the cost of uncertainty change on 9/11?"

It was easy to agree that premiums increased by more than the increase in the expected value of loss payments. "I can think of no large corporate risks where the per unit `cost of uncertainty' has not increased," stated Mark Ames of MMC Enterprise Risk Management.

The market's charge for any given amount of uncertainty has changed within a specific market (such as property reinsurance) as a response to events. The market's charge per unit of uncertainty changes across markets indirectly, and perhaps 9/11 was not a very large dollar event in the context of the larger capital markets.

As it was for the cost of risk transfer, timing was a key factor in the changing cost of uncertainty.

Stephen Philbrick of Conning Asset Management observed that stock price data suggested that the cost of uncertainty, as reflected in the broad capital markets, has not increased much. "Our stock markets performed well compared to what might have happened," he said.

All agreed that the capital market rewards consistent earnings growth. As a result, risk transfer can add value to both transferor and transferee. One person commented that there are risks for which the cost of transferring the uncertainty is many times the expected value of the losses.

Gary Venter of Guy Carpenter Instrat wrote in his article "Allocating Surplus_Not!" (The Actuarial Review, February 2002): "A given line of business could look extremely profitable or a waste of effort, depending on the method chosen [to evaluate the cost of uncertainty]." Ames made a similar point, stating that there is no consensus in the investment community on a theory of risk transfer that encompasses all types of risks, and there is no one financial theory to make all types of financial decisions.

The group generally felt that it is reasonable to analyze the "cost of uncertainty" into the product of "an amount of uncertainty" times "a market price per unit of uncertainty." There was a diversity of opinion, however, about the meaning of a "per unit cost of uncertainty." Also, the group was divided on whether the economic concept of equilibrium of supply and demand could be used to determine the "per unit cost of uncertainty."

Other participants in our group discussion were Walt Jedziniak of The Travelers, Bob Wolf of MMC Enterprise Risk Management, Alan Kaliski, of Royal Sun Alliance, Dave Powell of Tillinghast, Bob Conger of Tillinghast, and Al Commodore of Deep South Surplus, Inc.

***

On the plane to Dallas, I clipped the following quote of Jacques-Lucien Monod from Discover magazine: "Personal self-satisfaction is the death of the scientist. Collective self-satisfaction is the death of research. It is restlessness, anxiety, dissatisfaction, [and] agony of the mind that nourish science." If this article has surprised you as much as moderating the discussion surprised me, or if it has stimulated you to search for answers as to the question of how to price the risk of uncertainty, then it has served its purpose.