In addition to the links in the earlier post Unintended Consequences: TRIA Renewal Said Vital to Workers Comp Market, several useful articles recently released address some of the hard questions about the Terrorism Risk Insurance Act, (TRIA), its societal value and the coverage it provides. (read more)
Prof. Kent Smetters of the Wharton School wrote "Insuring Against Terrorism: The Policy Challenge" (NBER 2005) which argues that "mostly unfettered insurance and capital markets are capable of insuring large terrorism losses, even losses 10 times larger than the $40 billion loss that occurred on September 11, 2001. A $400 billion loss in capital markets is common." He argues "if there is any 'failure,' it rests with government policies. Government tax, accounting, and regulatory policies have made it costly for insurers to hold surplus capital. They have also hindered the implementation of instruments that could securitize the underlying risks."
His paper also points out that TRIA makes no concurrent premium charge for its billions in “supercat” reinsurance, making it impracticable for private competitors to offer reinsurance. He suggests that elimination of the government subsidy would encourage the development of private alternatives.
A free copy of the 2004 draft is online at: http://irm.wharton.upenn.edu/WP-Insuring-Smetters.pdf
How does the $100 billion “ceiling” on government coverage compare to the probable maximum loss in an historic hurricane or earthquake? Some estimates are that a 100-year hurricane hitting Miami head-on, or a 100-year earthquake hitting San Francisco (again), could cost close to $100 billion in property damage, business interruption and loss of life. Why is the federal government not passing special legislation about that? Because it’s a regional hazard? Florida and California have addressed their local catastrophes with state legislation and catastrophe pools.
In February 2005, the General Accounting Office released a study "Catastrophe Risk : U.S. and European Approaches to Insure Natural Catastrophe and Terrorism Risks," GAO 05-199 (2005). It looks at terror risks and major natural disaster risks and suggests:
* that the industry may not be able to address a major natural catastrophe,
* that CAT bonds and tax-deductible reserves may have the potential to enhance industry capacity for CATs
* that European countries mix approaches to insure natural catastrophes and have national terror insurance programs
It includes statistics on the 2004 hurricane season losses ($20 billion losses in Florida) and the impact of that quadruple-cat on government windstorm “backstop” programs, plus statistics on CAT bonds outstanding 1997-2004. It compares policies embedded in French and German terror coverage schemes and the reserve policies in European countries. It looks at consumer motivations in not buying available earthquake insurance (which implies similar “won’t happen to me” forces at work re terror insurance). It also examines the structure of "Special Purpose Reinsurance Vehicle" (SPRV) structure and payment flows.
It addresses proposals to change U.S. tax law so that insurers could set aside funds on a tax-deductible basis to establish reserves for potential future natural catastrophes or terrorist attacks. Presently, insurers may not deduct money set aside as reserves for events that have not occurred and the losses from which are not probable and reasonably estimable.
The 2005 GAO study is online at: http://www.gao.gov/new.items/d05199.pdf.
Recent state insurance department and Attorney General investigations are probing the use of high-layer catastrophe reinsurance, such as that sold by Berkshire Hathaway's National Indemnity and other reinsurance subsidiaries. A recent article at MarketWatch.com reported that in 1999 the New York Insurance Department criticized "supercat" finite reinsurance agreements between Berkshire subsidiaries and Gerling Global.
The agreements' reported triggers included one if a California earthquake between 4/1/94 and the end of 1995 exceeded $10 billion, another if windstorms in Florida, Georgia and the Carolinas during the same period exceeded $7 billion. A third would be triggered by Texas windstorm losses exceeding $5 billion in the same period. The article quoted one consultant as saying that the regulatory criticism was that "the trigger mechanisms for these policies were so high that no real risk was transferred." As a result, the regulator ruled that they should have been accounted for as loans rather than as reinsurance.
Do such regulatory positions have the effect of discouraging the sort of private reinsurance that might otherwise provide a private alternative to government programs for “supercats,” as Smetters suggests.
That article (which includes a link to the insurance department's report) is Barr, "Berkshire Reinsurance deals queried : Agreements helped Gerling Global manipulate results" (MarketWatch.com April 26, 2005).
Posted by dougsimpson at May 6, 2005 07:22 PM