The Market for Catastrophe Risk

Contrary to theoretical predictions, USAA's reinsurance program protects it from smaller losses and leaves it relatively uncovered in the event of the largest ones. Also flying in the face of theory is the fact that USAA paid premiums that were, according to the author's calculations, 9.1 times the actuarially fair premium.

During the era of frequent citywide fires, many fire insurance companies failed when all of their insured houses went up in smoke at the same time. Although today's catastrophic losses are more likely to result from a category 5 hurricane along the Florida coast, the risk of failure from the simultaneous loss of an entire class of insured objects continues to bedevil modern insurance companies.

Insurance companies buy reinsurance to protect themselves from catastrophic losses. In exchange for a set premium, a company offering reinsurance might promise to pay for 90 percent of any losses within the next year that exceed $450 million and are less than $600 million.

In The Market for Catastrophe Risk: A Clinical Examination (NBER Working Paper No.7286), author Kenneth Froot examines the theory that risk adverse firms would be more likely to insure against large losses and that the premiums for such insurance would be close to the value of the expected loss from catastrophes. He considers USAA, one of the 10 largest automobile and home insurers in the United States. It has large numbers of policyholders in California and Florida. After hurricane Andrew created the largest catastrophic loss in the United States in 30 years, USAA began developing a bond based alternative to traditional reinsurance programs. After examining the structure of USAA's reinsurance program, Froot concludes that, contrary to theoretical predictions, USAA's reinsurance program protects it from smaller losses and leaves it relatively uncovered in the event of the largest ones. Also flying in the face of theory is the fact that USAA paid premiums that were, according to the author's calculations, "9.1 times the actuarially fair premium."

This raises two questions. The first is whether the USAA reinsurance profile matches that of the market as a whole. Froot's analysis of reinsurance transaction data from the largest U.S. catastrophic risk reinsurer for 1970 to 1998 suggests both that other companies also self-insure against the largest catastrophic losses and that they pay higher prices than one would expect in order to insure themselves against the smaller ones.

The second question is whether reinsurance market imperfections might explain the differences between theory and observed insurer behavior. Froot considers three broad possibilities. The first is the possibility that reinsurance markets suffer from a shortage of capital, particularly after a catastrophic event occurs. He finds support for this in the higher premiums charged insurance companies with greater exposure to hurricane losses after hurricane Andrew. Scarce capital would also give reinsurance firms able to supply it greater market power, perhaps enabling them to command higher than expected premiums.

Government intervention in insurance markets is another potential source of market imperfection. Insurance commissioners are elected officials in 12 states, and to the extent that states use regulatory barriers to keep insurance costs down, "insurers must underwrite the catastrophic component of risk at prices that are well below" profitability. Insurance companies make ends meet by not insuring against catastrophic risk, with the result that policyholders and taxpayers (through state guarantee funds) self-insure whether they know it or not. Government also distorts the market with post-disaster aid that eliminates the incentive to buy insurance in the first place.

Transactions costs, moral hazard, and adverse selection also inhibit the reinsurance market's ability to spread risk. In fact, rather than act solely as a mechanism for shifting risk, it may function as a form of prepaid financing. "Often an explicit reinsurance contract contains an implicit agreement that reinsurers will charge more in the aftermath of a claim and that the cedent will continue to buy reinsurance from the same underwriter."

In an era that securitizes everything from home mortgages to high risk credit card payments, Froot concludes, evidence from the reinsurance market shows that "securitization is not automatically the lowest-cost way to transfer risk" although using bonds to underwrite catastrophic reinsurance may "lower, but not eliminate" the costs imposed by market imperfections and the barriers that keep capital out of the reinsurance market. Still, the fact that "managers of insurance companies purchase reinsurance at far above the fair price" shows that they must believe that "risk management adds value."

-- Linda Gorman