Key Coverage Issues in the TRIA Debate

Table 1: TRIA Approaches

Program Duration: How Long Is Long Enough?

The new House bill, passed by the Financial Services Committee in early August, stunned a lot of observers by proposing a 15-year extension. The Bush Administration immediately objected, pointing out that this is in direct conflict with their policy goal of eventually eliminating the federal program altogether. The committee's proposed extension is 5 years longer than the 10-year extension in the previous draft approved by the subcommittee only a week earlier. It is also 5 times the 3-year extension sought in HR 4314.

The need for a long-term extension reflects the broadly held belief that the terrorist threat is here to stay for the foreseeable future. Based on my informal survey, there is widespread support for an extension of at least 5 years and preferably 10 years or more. This will likely be the one of the most hotly contested issues as the current law approaches its natural expiration. Most stakeholders will be surprised if a reauthorization of less than 5 years is enacted.

Should Group Life Be Included?

At numerous Senate and House hearings, life insurers have argued cogently that they should be included in the TRIA program. Their basic reasoning is similar to the argument for including workers compensation in the program: Accumulations of employees at any one location present a possible large loss exposure that, as a practical matter, would be impossible to effectively monitor or control and could produce a huge loss. For example, imagine the consequences of a bomb exploding in the middle of a stadium containing 20,000 employees at a national sales conference. This kind of attack scenario could produce what many insurance company CEOs fear the most, even more than a weapons of mass destruction (WMD) attack. That would be a large, financially destabilizing loss that is not shared broadly across the market but uniquely exposes their company to the risk of ruin. In this type of scenario, the likelihood of a post-loss, market-based inflow of capital or a Congressional bailout is slim, and the protection afforded by TRIA is especially critical.

The argument against including group life is based on reactive assumptions in traditional economic thinking. In brief, since no "market failure" occurred in group life after the attacks on September 11, 2001, and since the market for group life appears robust and competitive today, despite what many agree is a continuing threat of a major attack, there is no reason to include group life in the program. This is the argument we find in numerous government documents such as the Treasury Department's 2005 Assessment report to Congress.

The rebuttal to this position, simply stated, is that 9/11 changed all the rules of extreme event risk management. Broadly applied, the classical "market failure" prerequisite for government involvement in insurance could be used to negate the entire federal program (given that only 2 foreign reinsurers failed because of the 9/11 attacks of approximately 150 insurers or reinsurers sustaining some loss). Properly understood, the TRIA program implicitly reflects the preemptive imperative of the "Bush Doctrine" itself. Applied to insurance, this doctrine would mandate the inclusion of Group Life because the stakes are too high, i.e., we cannot afford to wait to see if a "market failure" occurs if the next attack inside the United States involves a "mushroom cloud."

Including group life, however, would broaden the program and increase taxpayer risk, exactly the opposite of the Executive branch's policy goals. Few will be surprised if this change is again deleted from the final bill.

Should Domestic Acts of Terrorism Be Included?

A major conceptual shortcoming of the current program, in the opinion of many observers, is that it attempts to "fight yesterday's war" by assuming that the next major attack in the United States will be similar to the attacks on 9/11. One of the main ways it does this is by requiring that an act of terrorism subject to the law must be conducted "on behalf of a foreign person or foreign interest." Opponents of the current approach argue that domestic acts of terrorism need to be included in the next bill, pointing to two threats that are not now covered:

• The widely reported growth of "homegrown," Internet-trained sympathizers with Al Qaeda and other foreign terrorist organizations. Until fairly recently, for example, the London suicide bombers of July 7, 2005, were believed to have acted entirely independently of any direction from outside the country.
• Purely domestic large scale terrorist attacks such as the devastating 1995 bombing by Tim McVeigh of the Alfred P. Murrah Building in Oklahoma City.

To many observers, the argument for a broadened definition including domestic acts seems undeniable. The 2005 House bill, drafted within months of the London attacks, attempted to make this change. However, for reasons that I have never seen articulated, the final Extension Act continued the limitation on foreign-sponsored acts.

One can only conclude that American policymakers are averse to formally distinguishing between "criminal" acts or "vandalism" and political or religiously motivated deeds. Insurers have already implicitly conceded that this distinction does not need to be made for smaller losses. For example, the property insurance terrorism exclusion wording promulgated by Insurance Services Office, Inc. (ISO), only applies to industry insured losses of at least $25 million. Given the underwriting communities' singular focus on major attacks, including domestic terrorism seems to be much needed. But, much like group life, this important change may again be eliminated as the clock expires on the federal program.

Should Insurers Be Required to Underwrite the "Sum of All Fears"?

For most insurers, the most unacceptable proposal in the new House bill is the requirement that they separately make available insurance for attacks involving a nuclear, biological, chemical, or radiological agents or devices (NBCR, or what some would call "the sum of all fears"). The current approach requires that TRIA-related insurance be "made available" but this insurance is subordinate to all of the other terms of a given policy. The new House bill would require a preemptive coverage grant specifically for these acts, so there would be no confusion over whether the nuclear, pollution, or some other limitation applies. This change is broadly supported by policyholders.

In my opinion, including this requirement in a late December reauthorization would be hugely counterproductive and possibly quite disruptive to the market. In brief:

• Many insurers have concluded that such an attack would be the equivalent of an act of war and that an attack involving NBCR is inherently uninsurable. The new bill attempts to address this concern by requiring only a 7.5 percent deductible for NBCR acts instead of the 20 percent required for conventional weapon attacks. But even 7.5 percent may be too much for insurers who have committed to their Boards to expose none of their assets to this kind of loss. Although it is anecdotal, I am aware of several insurers who have made this commitment. Requiring these underwriters to offer this coverage would likely force them to withdraw or severely ration their capacity in the major urban centers that are generally believed to be the most likely targets.
• According to the Reinsurance Association of America (RAA), there is at most about $8 billion in global reinsurance capacity for conventional terrorist attacks and not more than an additional $1.5 billion in capacity for an NBCR attack. It is important to note that all of this capacity has already been committed and deployed. Therefore, if a new TRIA extension is passed in late 2007 requiring this constructive grant effective January 1, 2008, the huge increase in demand for new capacity will be impossible for the private sector to supply.

That said, this issue is critically important to resolve. It would be a major error to allow this change to once more die on the "cutting room floor." Possible coverage litigation in the aftermath of an NBCR attack, with insurers almost certainly attempting to invoke nuclear or pollution exclusions, would be exponentially worse than the "wind versus flood" litigation we have witnessed in the aftermath of Hurricane Katrina. It is in everyone's interest to resolve this issue.

Instead of deleting this reform or adding it without the careful consideration that is needed, policymakers should make the implementation of the NBCR coverage grant requirement subject to further review during 2008 with an effective date not sooner than 2009. As we saw in the creation of the American Nuclear Insurance (ANI) pool in 1957 (in response to the Price Anderson Act), creative risk takers can develop innovative approaches to insure even the "unthinkable" loss scenarios, but it takes time to develop these new approaches. This issue would appear to be a logical top priority for the House bill's proposed "Commission on Terrorism Risk Insurance."

Conclusion: Is Other Liability the Sole Remaining "Elephant in the Room"?

Although these proposals will again be at the center of the upcoming debate, opposition to the House bill will likely focus as much on what it does not propose. Unlike the Extension Act of 2005 or the earlier House bill, the new effort does not include any more cutbacks to the lines of insurance included in the program. One of the important policy objectives of the Bush Administration has been progressive reductions each year in federal taxpayer risk. Although much of this is achieved through changes in the program's financial parameters, the Extension Act surprised many underwriters by significantly cutting back the lines of insurance in the program.

As we see in Table 2, using direct written premium in 2006 as our base year, the initial TRIA legislation eliminated almost half of the nation's P&C premium by eliminating personal lines (totaling $223 billion in 2006 of a total $488 billion). The 2002 program also eliminated almost $30 billion (or more than 10 percent) of the commercial lines premium through the elimination of selected lines (see Table 1) clearly not related to terrorism exposures.

The 2005 Extension went even further, increasing the eliminated premium to almost $56 billion, mainly through the exclusion of commercial auto and all professional liability (other than directors and officers Liability). My best estimate is that the elimination of professional liability excluded about 25 percent of the premium booked to an eclectic line called "other liability." In 2006, this mixture of general liability, excess and umbrella liability, professional liability, D&O, and miscellaneous lines such as warranty insurance totaled $53.3 billion, with about $40 billion still included in the TRIA program. Liability premium from commercial multi-peril (CMP) added another $14.4 billion in 2006.

If yet another legislative surprise is pending this December, the sole remaining "elephant in the room" may very well be the approximate $54 billion in combined other liability and CMP liability premium. I have heard that further reductions in these lines are being considered. Needless to say, most underwriters and risk managers will argue passionately that liability for terrorism acts must be included in the program particularly in light of the completely unknown new standards for what will be considered "prudent" in our new world of risk post-9/11.

What appears to be an insurmountable problem, however, could be a unique opportunity. Simply stated, the single most glaring deficiency in the TRIA program from an underwriting perspective is the complete lack of any explicit incorporation of loss mitigation or preparedness requirements. As the first 9/11 lawsuits come to trial in September 2007 in Judge Hellerstein's New York City courtroom, federal policymakers should consider the many benefits of embracing preparedness in the new legislation.

One way to achieve this goal, already in progress on a voluntary basis, would be to require the Department of Homeland Security to develop minimum, economically feasible benchmarks for preparedness and loss mitigation in collaboration with the private sector. In addition to limiting the human casualties and the economic impact of the next attack, minimum national or regional standards could enable a more efficient liability insurance market to develop. Minimum required mitigation standards would also reduce taxpayer risk, particularly if compliance with these standards became a precondition for inclusion in the federal program. Much like the concept of requiring an affirmative grant of NBCR coverage, developing a new approach toward defining liability for terrorism attacks requires thoughtful consideration and collaboration prior to implementation. This is another logical issue for the proposed Commission on Terrorism Insurance to prioritize.