This is the accessible text file for GAO report number GAO-04-720T entitled 'Terrorism Insurance: Effects of the Terrorism Risk Insurance Act of 2002' which was released on April 28, 2004. This text file was formatted by the U.S. General Accounting Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. Testimony: Before the Subcommittee on Capital Markets, Insurance and Government Sponsored Enterprises and the Subcommittee on Oversight and Investigations, Committee on Financial Services, House of Representatives: United States General Accounting Office: GAO: For Release on Delivery Expected at 10:00 a.m. EDT: Wednesday, April 28, 2004: Terrorism Iinsurance: Effects of the Terrorism Risk Insurance Act of 2002: Statement of Richard J. Hillman, Director, Financial Markets and Community Investment: GAO-04-720T: GAO Highlights: Highlights of GAO-04-720T, a testimony before the House Financial Services Committee, Subcommittee on Capital Markets, Insurance, and Government-Sponsored Enterprises and the Subcommittee on Oversight and Investigations. Why GAO Did This Study: After the terrorist attacks of September 11, 2001, insurance coverage for terrorism largely disappeared. Congress passed the Terrorism Risk Insurance Act (TRIA) in 2002 to help commercial property-casualty policyholders obtain terrorism insurance and give the insurance industry time to develop mechanisms to provide such insurance after the act expires on December 31, 2005. Under TRIA, the Department of Treasury (Treasury) caps insurer liability and would process claims and reimburse insurers for a large share of losses from terrorist acts that Treasury certified as meeting certain criteria. As Treasury and industry participants have operated under TRIA for more than a year, GAO was asked to describe how TRIA affected the terrorism insurance market. What GAO Found: TRIA has enhanced the availability of terrorism insurance for commercial policyholders, largely fulfilling a principal objective of the program. In particular, TRIA has benefited commercial policyholders in major metropolitan areas perceived to be at greater risk for a terrorist attack. Prior to TRIA, we reported concern that some development projects had already been delayed or cancelled because of the unavailability of insurance and continued fears that other projects also would be adversely impacted. We also conveyed the widespread concern that general economic growth and development could be slowed by a lack of available terrorism insurance. Since TRIA’s enactment, terrorism insurance generally has been widely available, even for development projects in perceived high-risk areas, largely because of the requirement in TRIA that insurers “make available” coverage for terrorism on terms not differing materially from other coverage. Although the purpose of TRIA is to make terrorism insurance available, it does not directly address prices. As part of its assessment of TRIA’s effectiveness, Treasury is engaged in gathering data through surveys that should provide useful information about terrorism insurance prices in the marketplace. Despite increased availability of coverage, limited industry data suggest that most commercial policyholders are not buying terrorism insurance, perhaps because they perceive their risk of losses from a terrorist act as being relatively low. The potential negative effects of low purchase rates, in combination with the probability that those most likely to be the targets of terrorist attacks may also be the ones most likely to have purchased coverage, would become evident only in the aftermath of a terrorist attack and could include more difficult economic recovery for businesses without terrorism coverage or potentially significant financial problems for insurers. Moreover, those that have purchased terrorism insurance may still be exposed to significant risks that have been excluded by insurance companies, such as nuclear, biological, or chemical contamination. Meanwhile, although insurers and some reinsurers have cautiously reentered the terrorism risk market to cover insurers’ remaining exposures, little progress has been observed within the private sector toward either finding a reliable method for pricing terrorism insurance or developing any viable reinsurance alternatives to TRIA once it expires. What GAO Recommends: GAO recommends that the Secretary of the Treasury, as part of Treasury’s study of the effectiveness of TRIA and after consultation with insurance industry participants, identify for Congress alternatives that may exist for expanding the availability and affordability of terrorism insurance after TRIA expires. These alternatives could assist Congress during deliberations on the insurance industry’s capacity to provide terrorism insurance. [End of section] Mr. Chairman, Madam Chairwoman, and Members of the Subcommittees: I am pleased to be here today to discuss our report on the implementation of the Terrorism Risk Insurance Act of 2002 (TRIA) and the act's impact on the economy.[Footnote 1] The terrorist attacks of September 11, 2001, drastically changed the way insurers viewed the risk of terrorism. An industry that had considered the risk of terrorism so low that it did not identify or price terrorism risk separate from property and casualty coverage will ultimately pay approximately $40 billion for losses arising from September 11, according to industry experts. Responding to terrorism risk after September 11, reinsurers began excluding terrorism from coverage as contracts between reinsurers and insurers came up for renewal.[Footnote 2] Without reinsurance, insurers retained greater levels of risks than they could responsibly carry, and their reaction was to exclude these risks from commercial policies as they were renewed. In short, believing that neither the frequency nor magnitude of terrorism losses could be estimated, insurance companies withdrew from the market. In the aftermath of September 11, we reported that terrorism insurance was disappearing in the marketplace, particularly for large businesses and those perceived to be at some risk.[Footnote 3] We also reported significant concern that some development projects had already been delayed or cancelled because of the unavailability of insurance and fears that others would follow. Furthermore, there was widespread concern that general economic growth and development would be slowed by a lack of insurance availability and uncertainty in the marketplace. Because of concerns about the lack of available and affordable terrorism insurance, Congress passed TRIA, which took effect on November 26, 2002. TRIA is currently scheduled to expire at the end of 2005. Our report on the implementation of TRIA has two objectives. First, we describe the progress made by Treasury and insurance industry participants in implementing TRIA. We found that Treasury has made significant progress in implementing the provisions of TRIA, but has important work to complete in order to comply with all its responsibilities under the act. Second, we discuss the changes in the market for terrorism insurance coverage under TRIA. As requested, my testimony today focuses on the second of these two objectives. That is, how TRIA has affected the market for terrorism insurance and, more generally, the economy. Additionally, I have included appendixes to this statement that provide background information on TRIA and describe completed and ongoing engagements that GAO has undertaken for this Committee that relate to increasing the insurance industry's capacity to provide insurance for terrorism and natural catastrophe risks. In summary, it appears that Congress's first objective in creating TRIA--to ensure that business activity did not materially suffer from a lack of available terrorism insurance--largely has been achieved. Since TRIA was enacted in November 2002, terrorism insurance generally has been available to businesses. But most commercial policyholders are not buying the coverage. According to insurance industry experts, purchases have been higher in areas considered to be at high risk of another terrorist attack. However, many policyholders with businesses or properties not located in perceived high-risk locations are not buying coverage because they view any price for terrorism insurance as high relative to their perceived risk exposure. Further, those that have bought terrorism insurance remain exposed to significant perils because insurers have broadened longstanding policy exclusions of nuclear, biological, and chemical (NBC) events. Congress's second objective--to give private industry a transitional period during which it could begin pricing terrorism insurance and develop ways to cover losses after TRIA expires--has not been achieved yet. Industry sources indicated that under TRIA, insurance market participants have made no progress to date toward the development of reliable methods for pricing terrorism risks and little movement toward any mechanism that would enable insurers to provide terrorism insurance to businesses without government involvement. In conducting this work, we reviewed and analyzed relevant information concerning state legislation and publicly available and proprietary industry data and studies on the terrorism insurance market. We interviewed officials at Treasury, the National Association of Insurance Commissioners (NAIC), and state insurance regulators from six states with high insurance sales volumes. We also interviewed representatives of insurance companies, reinsurance companies, brokers for insurance and reinsurance companies, industry associations, property owners and developers, and insurance filing services and credit rating agencies.[Footnote 4] In our discussions with these organizations, we endeavored to gain an understanding of their experience in implementing TRIA requirements, obtain their views on the effects of TRIA on the terrorism insurance market, and identify developments within the industry to address terrorism risks after TRIA expires. We conducted this work in Chicago, New York City, and Washington, D.C., from January 2003 through April 2004 in accordance with generally accepted government auditing standards. Although Available, Few Are Buying Terrorism Insurance and the Industry Has Made Little Progress Toward Post-TRIA Coverage: While TRIA has improved the availability of terrorism insurance, particularly for high-risk properties in major metropolitan areas, most commercial policyholders are not buying the coverage. Limited industry data suggest that 10 - 30 percent of commercial policyholders are purchasing terrorism insurance, perhaps because most policyholders perceive themselves at relatively low risk for a terrorist event. Some industry experts are concerned that those most at risk from terrorism are generally the ones buying terrorism insurance. In combination with low purchase rates, these conditions could result in uninsured losses for those businesses without terrorism coverage or cause financial problems for insurers, should a terrorist event occur. Moreover, even policyholders who have purchased terrorism insurance may remain uninsured for significant risks arising from certified terrorist events--that is, those meeting statutory criteria for reimbursement under TRIA--such as those involving NBC agents or radioactive contamination. Finally, although insurers and some reinsurers have cautiously reentered the terrorism risk market, insurance industry participants have made little progress toward developing a mechanism that could permit the commercial insurance market to resume providing terrorism coverage without a government backstop. TRIA Has Improved the Availability of Terrorism Insurance, Particularly for Some High-Risk Policyholders: TRIA has improved the availability of terrorism insurance, especially for some high-risk policyholders. According to insurance and risk management experts, these were the policyholders who had difficulty finding coverage before TRIA. TRIA requires that insurers "make available" coverage for terrorism on terms not differing materially from other coverage.[Footnote 5] Largely because of this requirement, terrorism insurance has been widely available, even for development projects in high-risk areas of the country. Although industry data on policyholder characteristics are limited and cannot be generalized to all policyholders in the United States, risk management and real estate representatives generally agree that after TRIA was passed, policyholders--including borrowers obtaining mortgages for "trophy" properties, owners and developers of high-risk properties in major city centers, and those in or near "trophy" properties--were able to purchase terrorism insurance. Additionally, TRIA contributed to better credit ratings for some commercial mortgage-backed securities. For example, prior to TRIA's passage, the credit ratings of certain mortgage-backed securities, in which the underlying collateral consisted of a single high-risk commercial property, were downgraded because the property lacked or had inadequate terrorism insurance. The credit ratings for other types of mortgage-backed securities, in which the underlying assets were pools of many types of commercial properties, were also downgraded but not to the same extent because the number and variety of properties in the pool diversified their risk of terrorism. Because TRIA made terrorism insurance available for the underlying assets, thus reducing the risk of losses from terrorist events, it improved the overall credit ratings of mortgage-backed securities, particularly single-asset mortgage- backed securities. Credit ratings affect investment decisions that revolve around factors such as interest rates because higher credit ratings result in lower costs of capital. According to an industry expert, investors use credit ratings as guidance when evaluating the risk of mortgage-backed securities for investment purposes. Higher credit ratings reflect lower credit risks. The typical investor response to lower credit risks is to accept lower returns, thereby reducing the cost of capital, which translates into lower interest rates for the borrower. To the extent that the widespread availability of terrorism insurance is a result of TRIA's "make available" requirement, Treasury's decision on whether to extend the requirement to year three of the program is vitally important. While TRIA has ensured the availability of terrorism insurance, we have little quantitative information on the prices charged for this insurance. Treasury is engaged in gathering data through surveys that should provide useful information about terrorism insurance prices. TRIA requires that they make the information available to Congress upon request. In addition, TRIA also requires Treasury to assess the effectiveness of the act and evaluate the capacity of the industry to offer terrorism insurance after its expiration. This report is to be delivered to Congress no later than June 30, 2005. Most Policyholders Have Not Bought Terrorism Insurance: Although TRIA improved the availability of terrorism insurance, relatively few policyholders have purchased terrorism coverage. We testified previously that prior to September 11, 2001, policyholders enjoyed "free" coverage for terrorism risks because insurers believed that this risk was so low that they provided the coverage without additional premiums as part of the policyholder's general property insurance policy. After September 11, prices for coverage increased rapidly and, in some cases, insurance became very difficult to find at any price. Although a purpose of TRIA is to make terrorism insurance available and affordable, the act does not specify a price structure. However, experts in the insurance industry generally agree that after the passage of TRIA, low-risk policyholders (for example, those not in major urban centers) received relatively low-priced offers for terrorism insurance compared to high-risk policyholders, and some policyholders received terrorism coverage without additional premium charges.[Footnote 6] Yet according to insurance experts, despite low premiums, many businesses (especially those not in "target" localities or industries) did not buy terrorism insurance. Some simply may not have perceived themselves at risk from terrorist events and considered terrorism insurance, even at low premiums (relative to high-risk areas), a bad investment.[Footnote 7] According to insurance sources, other policyholders may have deferred their decision to buy terrorism insurance until their policy renewal date. Some industry experts have voiced concerns that low purchase rates may indicate adverse selection--where those at the most risk from terrorism are generally the only ones buying terrorism insurance. Although industry surveys are limited in their scope and not appropriate for market-wide projections, the surveys are consistent with each other in finding low "take-up" rates, the percentage of policyholders buying terrorism insurance, ranging from 10 to 30 percent. According to one industry survey, the highest take-up rates have occurred in the Northeast, where premiums were generally higher than the rest of the country. The combination of low take-up rates and high concentration of purchases in an area thought to be most at risk raises concerns that, depending on its location, a terrorist event could have additional negative effects. * If a terrorist event took place in a location not thought to be a terrorist "target," where most businesses had chosen not to purchase terrorism insurance, then businesses would receive little funding from insurance claims for business recovery efforts, with consequent negative effects on owners, employers, suppliers, and customers. * Alternatively, if the terrorist event took place in a location deemed to be a "target," where most businesses had purchased terrorism insurance, then adverse selection could result in significant financial problems for insurers. A small customer base of geographically concentrated, high-risk policyholders could leave insurers unable to cover potential losses facing possible insolvency. If, however, a higher percentage of business owners had chosen to buy the coverage, the increased number of policyholders would have reduced the chance that losses in any one geographic location would create a significant financial problem for an insurer.[Footnote 8] Tighter Exclusions Leave Policyholders Exposed to Significant Perils: Since September 11, 2001, the insurance industry has moved to tighten long-standing exclusions from coverage for losses resulting from NBC attacks and radiation contamination. As a result of these exclusions and the actions of a growing number of state legislatures to exclude losses from fire following a terrorist attack, even those policyholders who choose to buy terrorism insurance may be exposed to potentially significant losses. Although NBC coverage was generally not available before September 11, after that event insurers and reinsurers recognized the enormity of potential losses from terrorist events and introduced new practices and tightened treaty language to further limit as much of their loss exposures as possible. (We discuss some of these practices and exclusions in more detail in the next section.) State regulators and legislatures have approved these exclusions, allowing insurers to restrict the terms and conditions of coverage for these perils. Moreover, because TRIA's "make available" requirements state that terms for terrorism coverage be similar to those offered for other types of policies, insurers may choose to exclude the perils from terrorism coverage just as they have in other types of coverage. According to Treasury officials, TRIA does not preclude Treasury from providing reimbursement for NBC events, if insurers offered this coverage. However, policyholder losses from perils excluded from coverage, such as NBCs, would not be "insured losses" as defined by TRIA and would not be covered even in the event of a certified terrorist attack. In an increasing number of states, policyholders may not be able to recover losses from fire following a terrorist event if the coverage in those states is not purchased as part of the offered terrorism coverage. We have previously reported that approximately 30 states had laws requiring coverage for "fire-following" an event--known as the standard fire policy (SFP)--irrespective of the fire's cause. Therefore, in SFP states fire following a terrorist event is covered whether there is insurance coverage for terrorism or not. After the terrorist attacks of September 11, 2001, some legislatures in SFP states amended their laws to allow the exclusion of fire following a terrorist event from coverage. As of March 1, 2004, 7 of the 30 SFP states had amended their laws to allow for the exclusion of acts of terrorism from statutory coverage requirements.[Footnote 9] However as discussed previously, the "make available" provision requires coverage terms offered for terrorist events to be similar to coverage for other events. Treasury officials explained that in all non-SFP states, and the 7 states with modified SFPs, insurers must include in their offer of terrorism insurance coverage for fire following a certified terrorist event because coverage for fire is part of the property coverage for all other risks. Thus, policyholders who have accepted the offer would be covered for fire following a terrorist event, even though their state allows exclusion of the coverage. However, policyholders who have rejected their offer of coverage for terrorism insurance would not be covered for fire following a terrorist event. According to insurance experts, losses from fire damage can be a relatively large proportion of the total property loss. As a result, excluding terrorist events from SFP requirements could result in potentially large losses that cannot be recovered if the policyholder did not purchase terrorism coverage. For example, following the 1994 Northridge earthquake in California, total insured losses for the earthquake were $15 billion--$12.5 billion of which were for fire damage. According to an insurance expert, policyholders were able to recover losses from fire damage because California is an SFP state, even though most policies had excluded coverage for earthquakes. Reinsurers Have Cautiously Returned to the Market, but Many Insurers Have Not Bought Reinsurance: Under TRIA, reinsurers are offering a limited amount of coverage for terrorist events for insurers' remaining exposures, but insurers have not been buying much of this reinsurance. According to insurance industry sources, TRIA's ceiling on potential losses has enabled reinsurers to return cautiously to the market. That is, reinsurers generally are not offering coverage for terrorism risk beyond the limits of the insurer deductibles and the 10 percent share that insurers would pay under TRIA. In spite of reinsurers' willingness to offer this coverage, company representatives have said that many insurers have not purchased reinsurance. Insurance experts suggested that the low demand for the reinsurance might reflect, in part, commercial policyholders' generally low take-up rates for terrorism insurance. Moreover, insurance experts also have suggested that insurers may believe that the price of reinsurance is too high relative to the premiums they are earning from policyholders for terrorism insurance. The relatively high prices charged for the limited amounts of terrorism reinsurance available are probably the result of interrelated factors. First, even before September 11 both insurance and reinsurance markets were beginning to harden; that is, prices were beginning to increase after several years of lower prices. Reinsurance losses resulting from September 11 also depressed reinsurance capacity and accelerated the rise in prices.[Footnote 10] The resulting hard market for property- casualty insurance affected the price of most lines of insurance and reinsurance. A notable example has been the market for medical malpractice insurance.[Footnote 11] The hard market is only now showing signs of coming to an end, with a resulting stabilization of prices for most lines of insurance. In addition to the effects of the hard market, reinsurer awareness of the adverse selection that may be occurring in the commercial insurance market could be another factor contributing to higher reinsurance prices. Adverse selection usually represents a larger-than-expected exposure to loss. Reinsurers are likely to react by increasing prices for the terrorism coverage that they do sell. In spite of the reentry of reinsurers into the terrorism market, insurance experts said that without TRIA caps on potential losses, both insurers and reinsurers likely still would be unwilling to sell terrorism coverage because they have not found a reliable way to price their exposure to terrorist losses. According to industry representatives, neither insurers nor reinsurers can estimate potential losses from terrorism or determine prices for terrorism insurance without a pricing model that can estimate both the frequency and the severity of terrorist events. Reinsurance experts said that current models of risks for terrorist events do not have enough historical data to dependably estimate the frequency or severity of terrorist events, and therefore cannot be relied upon for pricing terrorism insurance. According to the experts, the models can predict a likely range of insured losses resulting from the damage if specific event parameters such as type and size of weapon and location are specified. However, the models are unable to predict the probability of such an attack. Even as they are charging high prices, reinsurers are covering less. In response to the losses of September 11, industry sources have said that reinsurers have changed some practices to limit their exposures to acts of terrorism. For example, reinsurers have begun monitoring their exposures by geographic area, requiring more detailed information from insurers, introducing annual aggregate and event limits, excluding large insurable values, and requiring stricter measures to safeguard assets and lives where risks are high.[Footnote 12] And as discussed previously, almost immediately after September 11 reinsurers began broadening NBC exclusions beyond scenarios involving industrial accidents, such as nuclear plant accidents and chemical spills, to encompass intentional destruction from terrorists. For example, post- September 11 exclusions for nuclear risks include losses from radioactive contamination to property and radiation sickness from dirty bombs. As of March 1, 2004, industry sources indicated that there has been little development or movement among insurers or reinsurers toward developing a private-sector mechanism that could provide capacity, without government involvement, to absorb losses from terrorist events. Industry officials have said that their level of willingness to participate more fully in the terrorism insurance market in the future will be determined, in part, by whether any more events occur. Industry sources could not predict if reinsurers would return to the terrorism insurance market after TRIA expires, even after several years and the absence of further major terrorist attacks in the United States. They explained that reinsurers are still recovering from the enormous losses of September 11 and still cannot price terrorism coverage. In the long term and without another major terrorist attack, insurance and reinsurance companies might eventually return. However, should another major terrorist attack take place, reinsurers told us that they would not return to this market--with or without TRIA. Conclusions: Congress had two major objectives in establishing TRIA. The first was to ensure that business activity did not suffer from the lack of insurance by requiring insurers to continue to provide protection from the financial consequences of another terrorist attack. Since TRIA was enacted in November 2002, terrorism insurance generally has been widely available even for development projects in high-risk areas of the country, in large part because of TRIA's "make available" requirement. Although most businesses are not buying coverage, there is little evidence that development has suffered to a great extent--even in lower-risk areas of the county, where purchases of coverage may be lowest. Further, although quantifiable evidence is lacking on whether the availability of terrorism coverage under TRIA has contributed to the economy, the current revival of economic activity suggests that the decision of most commercial policyholders to decline terrorism coverage has not resulted in widespread, negative economic effects. As a result, the first objective of TRIA appears largely to have been achieved. Congress's second objective was to give the insurance industry a transitional period during which it could begin pricing terrorism risks and developing ways to provide such insurance after TRIA expires. The insurance industry has not yet achieved this goal. We observed after September 11 the crucial importance of reinsurers for the survival of the terrorism insurance market and reported that reinsurers' inability to price terrorism risks was a major factor in their departure from the market. Additionally, most industry experts are tentative about predictions of the level of reinsurer and insurer participation in the terrorism insurance market after TRIA expires. Unfortunately, insurers and reinsurers still have not found a reliable method for pricing terrorism insurance, and although TRIA has provided reinsurers the opportunity to reenter the market to a limited extent, industry participants have not developed a mechanism to replace TRIA. As a result, reinsurer and consequently, insurer, participation in the terrorism insurance market likely will decline significantly after TRIA expires. Not only has no private-sector mechanism emerged for supplying terrorism insurance after TRIA expires, but to date there also has been little discussion of possible alternatives for ensuring the availability and affordability of terrorism coverage after TRIA expires. Congress may benefit from an informed assessment of possible alternatives--including both wholly private alternatives and alternatives that could involve some government participation or action. Such an assessment could be a part of Treasury's TRIA-mandated study to "assess…the likely capacity of the property and casualty insurance industry to offer insurance for terrorism risk after termination of the Program.": Recommendation for Executive Action: As part of the response to the TRIA-mandated study that requires Treasury to assess the effectiveness of TRIA and evaluate the capacity of the industry to offer terrorism insurance after TRIA expires, we recommend that the Secretary of the Treasury, after consulting with the insurance industry and other interested parties, identify for Congress an array of alternatives that may exist for expanding the availability and affordability of terrorism insurance after TRIA expires. These alternatives could assist Congress during its deliberations on how best to ensure the availability and affordability of terrorism insurance after December 2005. Mr. Chairman, Madam Chairwoman, this concludes my statement. I would be pleased to respond to any questions that you or other members of the Subcommittees may have. Contacts and Acknowledgements: For further information regarding this testimony please contact Richard J. Hillman or Davi M. D'Agostino, Directors, or Lawrence D. Cluff or Wesley M. Phillips, Assistant Directors, Financial Markets and Community Investment, (202) 512-8678. Individuals making key contributions to this testimony include Sonja Bensen, Rachel DeMarcus, Tom Givens III, Jill Johnson, Barry Kirby, Caitlyn Lam, Tarek Mahmassani, Angela Pun, and Barbara Roesmann. [End of section] Appendix I: TRIA Background: Under TRIA, Treasury is responsible for reimbursing insurers for a portion of terrorism losses under certain conditions. Payments are triggered when (1) the Secretary of the Treasury certifies that terrorists acting on behalf of foreign interests have carried out an act of terrorism and (2) aggregate insured losses for commercial property and casualty damages exceed $5,000,000 for a single event.[Footnote 13] TRIA specifies that an insurer is responsible (that is, will not be reimbursed) for the first dollars of its insured losses--its deductible amount. TRIA sets the deductible amount for each insurer equal to a percentage of its direct earned premiums for the previous year.[Footnote 14] Beyond the deductible, insurers also are responsible for paying a percentage of insured losses. Specifically, TRIA structures pay-out provisions so that the federal government shares the payment of insured losses with insurers at a 9:1 ratio--the federal government pays 90 percent of insured losses and insurers pay 10 percent--until aggregate insured losses from all insurers reach $100 billion in a calendar year (see fig. 1). Thus, under TRIA's formula for sharing losses, insurers are reimbursed for portions of the claims they have paid to policyholders. Furthermore, TRIA then releases insurers who have paid their deductibles from any further liability for losses that exceed aggregate insured losses of $100 billion in any one year. Congress is charged with determining how losses in excess of $100 billion will be paid. Figure 1: Prerequisites for and Limits of Coverage Under TRIA: [See PDF for image] [A] The percentage of direct earned premiums increases each year: 7 percent in 2003, 10 percent in 2004, and 15 percent in 2005. [End of figure] TRIA also contains provisions and a formula requiring Treasury to recoup part of the federal share if the aggregate sum of all insurers' deductibles and 10 percent share is less than the amount prescribed in the act--the "insurance marketplace aggregate retention amount." TRIA also gives the Secretary of the Treasury discretion to recoup more of the federal payment if deemed appropriate.[Footnote 15] Commercial property-casualty policyholders would pay for the recoupment through a surcharge on premiums for all the property-casualty policies in force after Treasury established the surcharge amount; the insurers would collect the surcharge. TRIA limits the surcharge to a maximum of 3 percent of annual premiums, to be assessed for as many years as necessary to recoup the mandatory amount. TRIA also gives the Secretary of the Treasury discretion to reduce the annual surcharge in consideration of various factors such as the economic impact on urban centers. However, if Treasury makes such adjustments, it has to extend the surcharges for additional years to collect the remainder of the recoupment. Treasury is funding the Terrorism Risk Insurance Program (TRIP) office operations--through which it administers TRIA provisions and would pay claims--with "no-year money" under a TRIA provision that gives Treasury authority to utilize funds necessary to set up and run the program.[Footnote 16] The TRIP office had a budget of $8.97 million for fiscal year 2003 (of which TRIP spent $4 million), $9 million for fiscal year 2004, and a projected budget of $10.56 million for fiscal year 2005--a total of $28.53 million over 3 years. The funding levels incorporate the estimated costs of running a claims-processing operation in the aftermath of a terrorist event: $5 million in fiscal years 2003 and 2004 and $6.5 million in fiscal year 2005, representing about 55 - 60 percent of the budget for each fiscal year. If no certified terrorist event occurred, the claims-processing function would be maintained at a standby level, reducing the projected costs to $1.2 million annually, or about 23 percent of the office's budget in each fiscal year. Any funds ultimately used to pay the federal share after a certified terrorist event would be in addition to these budgeted amounts. [End of section] Appendix II: Summary of Completed and Ongoing GAO Work Related to Insurance for Terrorism and Natural Catastrophe Risks: Terrorist attacks and natural catastrophes---such as hurricanes and earthquakes---pose unique challenges to insurers. Forecasting the timing and severity of such events is difficult and the large losses associated with catastrophes can threaten insurer safety and soundness. Insurers also frequently respond to catastrophic events by cutting back coverage significantly or substantially increasing premiums for policyholders. Over the years, several approaches have been suggested to expand the capacity of the insurance industry to cover catastrophic events. These approaches include securitization of catastrophe risk, changing tax and accounting treatment of catastrophe risk, and permitting risk-retention groups to cover property as well as liability exposures. At the request of the Chairman of the House Financial Services Committee and others, we have completed reports that address some of these issues or have work ongoing in these areas. Our work may assist the Committee in its oversight of the insurance industry and consideration of the industry's ability to both insure against and respond to catastrophic events. Securitization of Catastrophe Risk: Given the enormous financial losses associated with catastrophic events and questions about insurers' ability to cover such losses, interest has been generated in transferring some of these risks to the capital markets, which had total value of about $29 trillion as of the first quarter of 2003. Since the mid-1990s, some insurance companies, reinsurance companies, and capital market participants have developed various financial instruments, the most prevalent of which are catastrophe bonds. [Footnote 17] These bonds offer a relatively high rate of return to investors that are willing to accept some of the substantial risks associated with catastrophes. In two previous reports, we assessed the development of the catastrophe bond market.[Footnote 18] We found that some insurance companies view catastrophe bonds as an important component of their overall strategy for managing natural catastrophe financial risks. In addition, representatives from some institutional investors we contacted expressed positive views about catastrophe bonds because they offer attractive yields compared to traditional investments and help diversify investment risks. However, other insurers and investors are not willing to issue or purchase catastrophe bonds because they are more costly than traditional reinsurance, too risky, or illiquid. We also reported that developing catastrophe bonds to cover terrorism risks in the United States is considered challenging for many reasons, including the difficulties associated with developing computer models to predict the frequency and severity of terrorist attacks. Sophisticated models have been developed to predict the frequency and severity of natural catastrophes--particularly hurricanes--that have facilitated the development of catastrophe bonds covering such risks. Tax and Accounting Treatment of Catastrophe Risk: We are currently conducting follow-up work on potential tax and accounting issues raised in our previous reports that might affect the use of catastrophe bonds. As we reported in September 2002, most catastrophe bonds are issued offshore--for example, in Bermuda--rather than in the United States due to favorable tax considerations. Some insurance industry groups have argued for changes in U.S. tax laws to encourage insurers to issue catastrophe bonds onshore to lessen transaction costs and afford regulators greater scrutiny over these activities. As part of our ongoing work, we are reviewing the tax treatment of catastrophe risk coverage in selected European countries. Furthermore, in 2003 we reported that the Financial Accounting Standards Board had issued guidance that may require insurers or investors to list catastrophe bond assets and liabilities on their balance sheets. We reported that this guidance had the potential to limit the appeal of issuing catastrophe bonds but that insurers and financial market participants were not certain of the impact of this guidance. We are continuing to investigate developments on these tax and accounting issues and will discuss them in an upcoming report. Some believe removing accounting and tax barriers that prevent U.S. insurance companies from establishing tax-deductible reserves to cover the financial risks associated with potential natural catastrophes and terrorist attacks would supplement private-sector capacity. Under current U.S. accounting standards and tax law, insurers must build any reserves for events that have not yet occurred from after-tax income (retained earnings). As a result, insurers do not usually establish reserves in anticipation of catastrophic events, such as hurricanes. Therefore, insurers attempt to limit their exposure to catastrophic risks through the underwriting process, the purchase of reinsurance, or issuance of catastrophe bonds, among other alternatives. There is considerable disagreement about the appropriateness and effectiveness of tax-deductible reserving. Advocates believe that allowing insurers to establish such reserves would provide increased capacity at lower cost. On the other hand, critics of tax-deductible reserving have argued that, in addition to lowering federal tax receipts, there is no assurance that insurers would actually increase their catastrophe insurance capacity, but rather either shield existing capital from taxes or substitute tax-deductible reserves for reinsurance. At the Chairman's request, we are currently reviewing the tax treatment of catastrophe risk reserves in selected European countries---France, Spain, Germany, Switzerland, and Italy. We continue to review these practices and will comment on them in a forthcoming report. In addition to discussing reserving practices, we are gathering information on general approaches to insuring against catastrophic risks in these countries. Implementation of the Liability Risk Retention Act: Congress enacted the Liability Risk Retention Act of 1986 (Act) to facilitate the formation of risk-retention groups (RRGs) and risk- purchasing groups (RPGs), insurance entities initially established to increase the availability and affordability of liability insurance during the 1980s.[Footnote 19] As authorized by the Act, these groups may only provide commercial liability insurance. An RRG is simply a group of businesses with similar risks that join to create an insurance company to self-insure their risks. An RPG, on the other hand, is a group formed to purchase insurance as a single entity from a traditional insurer. The majority of our ongoing work focuses on RRGs because, as insurers, they have the potential to provide new insurance capacity. A wide variety of groups, such as professional groups (doctors, attorneys), institutions (universities, hospitals), and businesses (trucking firms, homebuilders) have created RRGs. As of mid- April 2004, about 150 RRGs were operational and approximately 72 were formed in the last year and half.[Footnote 20] In contrast to most other insurers, an RRG can sell insurance in as many states as it chooses but is to be regulated by only one state--the state in which it is chartered. [Footnote 21] Our ongoing work focuses on assessing the extent to which RRGs have met the Act's intent that they increase the availability and affordability of liability insurance. We will also assess how the unique regulatory structure of RRGs--where only one state serves as regulator--has promoted the establishment of RRGs and if this structure has resulted in uneven or ineffective regulation. The recent failure of four RRGs has resulted in some regulators questioning the efficacy of having a single-state regulator and the standards used by some states to charter and regulate RRGs. If we identify any problems as part of our work, we will evaluate what legislative or regulatory changes might be needed. These changes, if needed, could be important whether or not, as some advocates have suggested, the Act is expanded to include commercial property insurance. Finally, in the event the Act were expanded to include property insurance, we also are exploring the potential of RRGs to provide additional capacity for terrorism insurance. [End of section] Appendix III: Prior GAO Reports and Testimonies Related to Insurance for Terrorism and Natural Catastrophe Risks: U.S. General Accounting Office, Terrorism Insurance: Implementation of the Terrorism Risk Insurance Act of 2002, GAO-04-307 (Washington, D.C.: Apr. 23, 2004). U.S. General Accounting Office, Catastrophe Insurance Risks: Status of Efforts to Securitize Natural Catastrophe and Terrorism Risk, GAO-03- 1033 (Washington, D.C.: Sep. 24, 2003). U.S. General Accounting Office, Catastrophe Insurance Risks: The Role of Risk-Linked Securities, GAO-03-195T (Washington, D.C.: Oct. 8, 2002). U.S. General Accounting Office, Catastrophe Insurance Risks: The Role of Risk-Linked Securities and Factors Affecting Their Use, GAO-02-941 (Washington, D.C.: Sep. 24, 2002). U.S. General Accounting Office, Terrorism Insurance: Rising Uninsured Exposure to Attacks Heightens Potential Economic Vulnerabilities, GAO- 02-472T (Washington, D.C.: Feb. 27, 2002). U. S. General Accounting Office, Terrorism Insurance: Alternative Programs for Protecting Insurance Consumers, GAO-02-199T (Washington, D.C.: Oct. 24, 2001). U.S. General Accounting Office, Terrorism Insurance: Alternative Programs for Protecting Insurance Consumers, GAO-02-175T (Washington, D.C.: Oct. 24, 2001). U.S. General Accounting Office, Insurers' Ability to Pay Catastrophe Claims, GAO/GGD-00-57R (Washington, D.C.: Feb. 8, 2000). FOOTNOTES  U.S. General Accounting Office, Terrorism Insurance: Implementation of the Terrorism Risk Insurance Act of 2002, GAO-04-307 (Washington, D.C.: Apr. 23, 2004).  Reinsurance is a mechanism that insurance companies routinely use to spread risk associated with insurance policies. Simply put, it is insurance for insurance companies. Reinsurance is a normal business practice that satisfies a number of needs in the insurance marketplace, including the need to expand capacity and obtain protection against potential catastrophes.  U.S. General Accounting Office, Terrorism Insurance: Alternative Programs for Protecting Insurance Consumers, GAO-02-175T (Washington, D.C.: Oct. 24, 2001), and Terrorism Insurance: Rising Uninsured Exposure to Attacks Heightens Potential Economic Vulnerabilities, GAO-02-472T (Washington, D.C.: Feb. 27, 2002).  Filing services perform many services for insurance companies, including submitting to state insurance regulators the documents required to sell a line of insurance.  TRIA defines "make available" to mean that the coverage must be offered for insured losses arising from terrorist events and that coverage not differ materially from the terms, amounts, and limitations applicable to coverage for losses arising from other types of events. However, TRIA gives Treasury the option of determining whether the "make available" requirement should be extended through 2005, the third year of the act, and gives the agency until September 1, 2004, to do so.  According to industry experts, the insurers that provided "free" terrorism insurance likely did so for policies already in place at the time TRIA was enacted and may have deferred operational changes and difficult pricing decisions because they lacked the resources to do so.  Howard Kunreuther, Erwann Michel-Kerjan, and Beverly Porter, Assessing, Managing and Financing Extreme Events: Dealing with Terrorism (National Bureau of Economic Research: December 2003), 13.  Casualty Actuarial Society, Foundations of Casualty Actuarial Science, 4th ed. (United Book Press, Inc.: 2001), 51, 86.  According to the National Association of Mutual Insurance Companies, Louisiana, Michigan, Minnesota, Nebraska, New Hampshire, Oklahoma, and Virginia have amended their standard fire policies to allow for exclusion of terrorism from statutory fire coverage. State legislators in Massachusetts have introduced a similar bill.  Capacity is the amount of reinsurance or insurance that is available for a defined risk.  U.S. General Accounting Office, Medical Malpractice Insurance: Multiple Factors Have Contributed to Increased Premium Rates, GAO-03-702 (Washington, D.C.: June 27, 2003).  Christian Brauner and Georges Galey, "Terrorism Risks in Property Insurance and Their Insurability after 11 September 2001," (Swiss Reinsurance Company: 2003), 25.  Aggregate insured losses are the sum of insured property and casualty losses from all commercial policyholders that result from a certified act of terrorism.  Section 102(4) of TRIA defines direct earned premiums as "a direct earned premium for property and casualty insurance issued by any insurer for insurance against losses …" Treasury provided further clarification that direct earned premiums are "earned as reported to the NAIC in the Annual Statement in column 2 of Exhibit of Premiums and Losses (commonly known as Statutory Page 14)" and cover all risks, not only for risks from terrorism. The percentage of the direct earned premium allowed as an insurer deductible varies over the program years: 7 percent in 2003, 10 percent in 2004, and 15 percent in 2005.  According to Treasury officials, the formula for the mandatory portion of the recoupment is intended to ensure that the insurance industry is financially responsible for a prescribed level of the first dollars of losses. The prescribed loss levels are as follows: $10 billion in 2003, $12.5 billion in 2004, and $15 billion in 2005. Therefore, if the sum of insurers' aggregate payments for deductibles and the 10 percent share--the amounts paid by industry--is less than the level prescribed for that year, then a recoupment would be required to collect the difference. On the other hand, if the amounts paid by industry exceed the prescribed level, then a recoupment would not be needed.  "No-year money" is budget authority that remains available for obligation until expended, usually until the objectives for which the authority was made available are attained.  Catastrophe bonds are an example of risk-linked securities. This statement focuses on catastrophe bonds that are privately placed securities sold to institutional investors under Securities and Exchange Commission Rule 144A. In general, a qualified institutional investor under Rule 144A owns and invests on a discretionary basis at least $100 million in securities of issuers that are not affiliated with the investor.  See U.S. General Accounting Office, Catastrophe Insurance Risks: Status of Efforts to Securitize Natural Catastrophe and Terrorism Risk, GAO-03-1033 (Washington, D.C.: Sep. 24, 2003) and Catastrophe Insurance Risks: The Role of Risk-Linked Securities and Factors Affecting Their Use, GAO-02-941 (Washington, D.C.: Sep. 24, 2002).  The Product Liability Risk Retention Act of 1981 established these groups but limited them to providing product liability insurance. In 1986, in response to another shortage of liability insurance, Congress amended the Act and allowed these groups to offer most types of commercial liability insurance.  Data estimates were provided by the Managing Editor, Risk Retention Reporter.  The McCarran-Ferguson Act of 1945 left regulation and taxation of the insurance market to the states. (15 U.S.C. 1011) As a result, states have primary responsibility for regulating the insurers operating in their states and each state has its own insurance department.