P/C Industry Says Treasury Study Inadequate Gauge of Terrorism Insurance Market

Two months before the study is expected to be done, the property casualty insurance industry is getting nervous about a study on terrorism insurance being conducted by the Treasury Department, a study some in Congress are waiting to read before deciding whether to renew the Terrorism Risk Insurance Act beyond its expiration this coming December.

In a letter to Treasury officials overseeing the TRIA study, a coalition of insurance company organizations has raised extensive concerns over how the survey is being handled and whether it will accurately reflect the industry’s capability to provide terrorism coverage without an extension of TRIA.

Treasury officials have said the study will be finished in June.

The complete text of the letter follows:

April 8, 2005

Mark J. Warshawsky
Assistant Secretary for Economic Policy
United States Department of the Treasury
1500 Pennsylvania Avenue, N.W.
Washington, DC 20220

John D. Worth
Director, Office of Microeconomic Analysis
United States Department of the Treasury
1500 Pennsylvania Avenue, N.W.
Washington, DC 20220

Re: Comments on TRIA-Mandated Study

Dear Assistant Secretary Warshawsky and Director Worth:

The undersigned are part of an ad hoc industry working group that includes insurance industry organizations, insurance companies, and the following property-casualty insurance industry trade associations and their member companies: American Insurance Association (“AIA”), National Association of Mutual Insurance Companies (“NAMIC”), Property Casualty Insurers Association of America (“PCI”), Reinsurance Association of America (“RAA”), and Surety Association of America (“SAA”) (collectively, “Working Group”). Many of the Working Group’s individual members are companies that are part of the survey sample for the United States Department of the Treasury’s (“Department” or “Treasury”) study and report to Congress mandated by Section 108(d) of the Terrorism Risk Insurance Act of 2002 (“TRIA”).

That subsection of TRIA requires:

“The Secretary, in consultation with the NAIC, representatives of the insurance industry and of policy holders, other experts in the insurance field, and other experts as needed, … [to] assess the effectiveness of the Program and the likely capacity of the property and casualty insurance industry to offer insurance for terrorism risk after termination of the Program, and the availability and affordability of such insurance for various policyholders….”

Several of the undersigned organizations have been working with Treasury pursuant to this provision of TRIA as industry representatives or experts, both prior to the development of the survey instrument and during the survey process when concerns have arisen. As the final wave of survey responses have been received and Treasury develops its report to Congress, it is in this spirit that we offer the following comments for Treasury’s consideration.

Our comments focus on two aspects of the TRIA study: (a) areas of continuing concern with the 3 survey instruments, and (b) the inability of the survey to reflect the capability of the property-casualty insurance industry to provide terrorism risk insurance absent a federal backstop.

Areas of Concern With The Survey Instrument

Company respondents to the insurer and reinsurer surveys have identified three areas of concern over the course of completing the three waves of the survey. The undersigned organizations have relayed these concerns to Treasury during the course of the study, but believe that reiterating the concerns here will aid Treasury as it is developing its report.

First, we have doubts about the ability of the surveys to produce consistent and reliable responses, as reflected in both the questions and the responses. Companies have approached the survey in different ways. Some have produced data estimates. Others have tried to produce “hard” data to actually quantify their responses. Still others avoided certain questions because of confusion or fear that the response would be misinterpreted by the contractor or Treasury. We saw these differing approaches after the Wave 1 TRIA Survey for Insurers was released. Many companies, confused by how to respond to many of the questions, specifically asked some of the undersigned trade organizations to encourage Treasury to accept survey responses that included an explanatory field allowing companies to elaborate on their responses.

The difference of response is likely to be compounded by the nature of the surveys themselves. For example, the Wave 2 TRIA Survey for Insurers does not give a definite period of consideration for survey responses. For many data points, the inquiry is for policy information “to date.” As a result, insurers may have interpreted “to date” to mean either the date of the letter enclosing Wave 2 of the survey or their own respective dates of completion – a 2 _ month difference of time. Whatever the choice, the end result is likely to be an inconsistent set of responses and more difficulty reconciling that inconsistent data.

The same inconsistencies promise to impact responses to the TRIA reinsurer surveys. The reinsurer surveys use terms such as “program” and “contract,” but do not account for differing interpretations of those terms by reinsurers. Similarly, while the survey inquires about the existence or absence of terrorism exclusions, it does not account for the differing approaches that reinsurers may take to answering those questions. Reinsurance companies that take a conservative approach to capturing their own terrorism coverage data most likely will treat the lack of a terrorism exclusion in a reinsurance contract as providing terrorism coverage. However, some primary companies may exclude terrorism risk from their policies, when there is no terrorism exclusion in the reinsurance agreement. The reinsurer, in completing the survey, will most likely count the premium and limits of these contracts in their totals, although there is really no exposure from the underlying business. This represents an overstatement of coverage.

Second, and more troubling, the surveys – standing alone – may not produce relevant data. This result may occur either because some questions make incorrect assumptions about the operation of TRIA or the effect of terrorism exposure, or the questions are not sufficiently detailed to provide an accurate picture of a particular company’s experience.

Wave 1 of the TRIA Survey for Insurers included a number of questions that fell into the former category. For example, several questions in that survey assumed that insurers had market flexibility under TRIA to cover terrorism risk on terms, conditions, or limitations that differ from those available for other covered types of loss. (See C11, C23, C25 & C35-C37). Other questions incorrectly assumed that differences in average rates were due solely to increased terrorism risk, rather than other circumstances. The most obvious errors occurred with some Wave 1 workers’ compensation insurance questions, which called for insurer responses in the context of evaluating “availability” of terrorism risk coverage for workers’ compensation insurance policies, (See, e.g., C3a(b)), and with Wave 1 questions calling for data on “umbrella or package” insurance policies. While these questions were brought to Treasury’s attention prior to the release of Wave 2 of the TRIA Survey for Insurers, other problematic questions remain. For instance, Questions B26 through B28 of Wave 3 of the TRIA Survey for Insurers are incorrectly premised on an insurer’s ability to refuse to offer certified terrorism coverage on the same terms, conditions, and limitations as other covered perils. As a result, we have concerns about the reliability of the data for those questions.

Of equal concern is the insufficient level of detail in the questions themselves – a concern that applies equally to all waves of the surveys. For instance, the surveys ask a number of questions designed to determine a company’s management of terrorism risk for “high risk” properties. But none of the questions are detailed enough to provide data about an insurer’s accumulation management – i.e., what any individual company does within a high-risk zone with small, mid-size, or high-risk properties or with high-risk policyholders in non-urban geographic zones.

Similarly, the reinsurer surveys ask generalized questions that may not unearth reinsurer appetite (or lack of appetite) for terrorism risk. Section D of the reinsurer survey captures general approaches to the management of terrorism risk. But, for a number of questions, there is no breakdown among lines of business or even geographic type of risk, so answering questions about “high-risk locations” and “NCBR” risk applies to all lines and all treaties. Therefore, if a company has written some small, rural treaties that include coverage for NCBR risk but have taken a stance to avoid this risk on treaties with urban exposure, a broad-brush answer would be given that the company did, in fact, write contracts with NCBR exposure.

Third, certain aspects of the surveys are burdensome. We have heard from many company respondents that the questions do not coincide with the way data is captured by company operations. As a result, company responses may be omitted, incomplete, or data reconfigured to attempt to respond to the questions posed. More importantly, perhaps, Wave 3 of the TRIA Survey for Insurers asks for “updated” data regarding policies written or renewed in January and February of 2005. To the extent that insurers were not able to provide policy data from 2004 (which is likely since full year 2004 data was not available when many companies responded to Wave 2 of the survey), they should have been encouraged to do so. In addition, Wave 3 imposes a practical impossibility on insurers, most of whom may not have been able to provide any significant 2005 data by the mid-March response deadline.

The Working Group has brought these issues to Treasury’s attention because we believe that the surveys themselves cannot answer all questions about the impact of TRIA, and that the survey data must be correlated with other information that has been collected outside the Treasury process, including publicly-available financial data, take-up and premium data collected and released periodically by the insurance brokerage community and others, rating methodologies utilized by the various insurance rating/service organizations, and other terrorism risk insurance studies. All of this data should be viewed in the context of the state insurance regulatory environment, which does not provide a “free market” for insurers and instead creates a patchwork of differing state government price and product controls that have led to artificially suppressed rates for terrorism risk insurance and limited insurance product choices for commercial customers. We believe that this correlation process will produce a more accurate picture of TRIA’s impact.

The Study’s Assessment Of Private Market Capacity After TRIA

While we have focused thus far on important gaps in the surveys’ ability to assess TRIA’s impact, the Working Group believes that the surveys will only provide highly generalized information about the private insurance market without TRIA. Our belief stems from the inability to accurately assess the future market through TRIA coverage data and the relatively small percentage of the survey questions devoted to future capacity issues. In this context, we urge Treasury to consider a number of elements outside the survey instruments.

First, terrorism risk is different than other risks and is uninsurable for a number of reasons. We detailed those reasons in comments in advance of Treasury’s “make available” determination, and attach and incorporate those comments by reference here. The reasons are particularly relevant with respect to the possibility of nuclear, biological, chemical, or radiological attacks, which carry the same “lack of frequency predictability” characteristics as conventional terrorism, but whose long-term severity characteristics may be largely unknown.

Second, the risk of terrorism on U.S. soil is not likely to recede any time soon. We have been at an elevated state of alert for many months now, and are constantly reminded by Administration officials that al-Qaeda and other terrorist groups would like nothing more than to attack hard or soft targets within the United States.

Third, information about a future without federal involvement in securing the U.S. against terrorism risk can be gleaned from private market affordability and availability during the time following September 11 and preceding Congressional enactment of TRIA. As that experience showed, the market was unstable and the negative effects on the commercial real estate market and, by extension, the construction labor market, were undeniable. Other predictive studies, such as Professor Glenn Hubbard’s September 2004 “The Economic Effects of Federal Participation in Terrorism Risk,” demonstrate a significant negative economic impact from the withdrawal of public participation in the management of terrorism risk exposure.

It is important to note that, with or without TRIA, the capital markets have not shown any significant interest in providing additional terrorism risk capacity. Again, we direct Treasury to our June 2004 comments on the “make available” determination, which outline the shortcomings of reliance on other capital market instruments to fill any perceived void or to add significant capacity.

Fourth, recent “conditional” policy form filings by insurance services organizations underscore the reaction of the property-casualty industry to the impending expiration of TRIA. Those filings, which are available for use in all U.S. jurisdictions except Florida, Georgia, and New York, provide flexibility for insurers to limit or structure terrorism coverage to more effectively manage existing capacity in the absence of a federal backstop.

Fifth, we again request that Treasury consider the future ability of the private market to provide terrorism risk insurance without a federal backstop in light of the existing state regulatory system. While conditional policy forms are in place in most states, the important premium-volume states listed above have refused to approve any conditional forms and insurers have no ability to limit terrorism coverage for workers’ compensation insurance in all states or commercial property coverage in the approximately 19 states with statutory fire policies that have not allowed terrorism exclusions. These “free market” obstacles make reliance on the private insurance market to address terrorism risk unworkable.

Lastly, as Treasury finalizes its report to Congress, we recommend that the Department interview a few reinsurance brokers to get the most current assessment of the market as it begins to react to the pending expiration of TRIA on December 31, 2005, given the uncertainty about whether Congress will act on this issue before that expiration date. The reinsurance brokers’ experience with the market might also serve as a good “reality check” comparison against Treasury’s survey findings. Reinsurance brokers are in a unique position to know the present status of private market capacity for terrorism risk and have expressed, through their participation with the Reinsurance Association of America, the willingness to serve as a resource to Treasury on this important issue.

We appreciate Treasury’s consideration of the Working Group’s comments on the TRIA study process, and hope that Treasury will incorporate our suggestions into its analysis as it reviews the survey responses. We also encourage Treasury to contact any of the designated Working Group representatives if you have any questions or need additional information. We look forward to working cooperatively with Treasury to ensure that Congress fully considers the state of the terrorism insurance marketplace under TRIA and in the future.

Respectfully submitted,

J. Stephen Zielezienski
Vice President & Associate General Counsel
American Insurance Association
1130 Connecticut Avenue, N.W., Suite 1000
Washington, DC 20036
Phone: 202-828-7175
Fax: 202-293-1219
szielezienski@aiadc.org

Dated: April 8, 2005

cc: Lucy Huffman

Designated Trade Association Contacts:

David A. Winston
Senior Vice President – Federal Affairs
National Association of Mutual Insurance Companies
122 C Street, N.W., Suite 540
Washington, DC 20001
dwinston@namic.org

Donald L. Griffin
Vice President, Personal Lines
Property Casualty Insurers Association of America
2600 South River Road
Des Plaines, IL 60018-3286
Phone: 847-553-3743 Direct
Fax: 847-759-4319 Direct Fax
donald.griffin@pciaa.net

Cynthia Lamar
Vice President and Assistant General Counsel
Reinsurance Association of America
1301 Pennsylvania Avenue, NW, Suite 900
Washington, DC 20004
Phone: 202-638-3690
lamar@reinsurance.org

Edward G. Gallagher
General Counsel
Surety Association of America
1101 Connecticut Avenue, NW, Suite 800
Washington, DC 20036
Phone: 202-778-3622
Fax: 202-463-0606
egallagher@surety.org