Discussion

Bill Murray of Chubb and Son discussed reputation as a major motivator in the decision not to invoke the war exclusion after September 11. The decision to cover those losses was voluntary, but the post-September 11 world left insurers vulnerable. While reinsurers have the option of closing shop, the primary business is bound by cancellation and nonrenewal restrictions. Debra Ballen of the American Insurance Association agreed that the Terrorism Risk Insurance Act (TRIA) should be reauthorized for three main reasons. First, a lack of precedent hinders informed projection of future risk. Second, preexisting regulatory constraints do not allow a "free-market" solution. Finally, truly turning risk evaluation over to the free market is politically unacceptable, as seen in the reception of the Pentagon's exploration of a futures market in terrorism.

Thomas Holzheu of the Swiss Economic Research Unit challenged this apparently contradictory faith in financial markets combined with distrust in the free insurance market. The government should play a role if it has better information, as a third-party player in what would otherwise be a symmetric game of noninformation. In some sense, the government may even be a risk maker. On the other hand, the federal government must be very careful about framing an incentive structure that prevents moral hazard. There should be some incentive for individuals to take precautions to reduce the probability or magnitude of damage suffered.

Neil Doherty of the Wharton School called for a corporate view to evaluate insurance. For a firm, an insurance payment provides contingent capital used in some investment; it only makes sense if the net present value is positive. In the event of another attack, the event itself will lower the internal rate of investment, perhaps to the point where there may be no demand. Even if there is demand for contingent capital, insurance may not be the best means of providing it when a credit risk is associated with the event itself.

Mike O'Malley of Chubb challenged Smetters's high-risk examples. Earthquake insurance is heavily subsidized through the California Earthquake Authority in two ways. The amount of the premium tax is allowed to go into the fund annually to accumulate surplus, and the Internal Revenue Service considers it tax exempt, as a political subdivision of the state. In addition, coverage sold by the entity has been cut by about two-thirds. Hurricane insurance in Florida also is heavily subsidized.

Warren Gorlick of the U.S. Commodity Futures and Trading Commission called attention to Basel II. Nominally a set of banking standards, Basel II will ultimately affect securities and futures firms as part of the banking structure. The rules require a reserve against operational risk, which explicitly includes terrorism. Basel rules also prohibit deducting insurance premiums from the amount by which a firm self-insures.

Robert Litan of Brookings lauded the paper for highlighting the lack of real-world evidence arguing for the Terrorism Risk Insurance Act. Agreeing that government tax policies constrain insurers from accumulating capital, he speculated on how recent administrations may have considered the issue. Under the Clinton administration, a pay-as-you-go rule prevented a change in policy to allow tax-deductible reserves for catastrophe insurance. The Bush administration did not support a tax cut for this industry either, presumably for political reasons. Given the limitations of reality—no good tax treatment of the issue exists—arguing for a Panglossian solution may be less useful than working within the regime, providing some backstop insurance where none might otherwise exist. Finally, even without regulatory impediments, there is a limited appetite for catastrophe bonds, which may be explained best by considering that institutional buyers are comprised of individual actors for whom losing on a catastrophe bond would probably end their career.

Smetters concluded by responding to several points. The ambiguity of risk, lacking historical data, should not be so great an impediment. Investors poured money into Amazon.com even though there were no time-series data on the company. While insurance has a lower risk-reward ratio than other industries, this is attributable largely to the difficulty of securitization and the difficulty of diversifying the risk. Smetters discussed the differences...