After much speculation, the US federal terrorism backstop - TRIA - was renewed until December 2007. But is that enough time to find a viable long-term alternative? asks Nick Thorpe.
"The standalone market for casualty terrorism insurance, absent TRIA, is virtually nonexistent." This was Marsh's conclusion in its recent report, "Marketwatch: Terrorism Insurance 2006", reflecting many of the concerns currently felt throughout the insurance and reinsurance industry. Despite the extension of the Terrorism Risk Insurance Act (TRIA) last year until December 2007, there is growing unease in an already besieged industry as to what happens when the sun finally sets on a piece of legislation that has, in many eyes, enabled a viable terrorism insurance market to operate for almost five years.
The extension, like the original act, was passed by Congress because the insurance industry had not amassed sufficient capital to insure catastrophic terrorism losses without some form of federal support. Like the original legislation, the extension is intended as a short-term solution. However, concerns remain that the two years will expire without a permanent resolution having been found. "A long-term solution is essential," says Aaron Davis, director of national terrorism and property resources at Aon. "The failure to implement something before the act expires in 2007 will not only impact insurance buyers by vastly decreasing risk transfer options, it will also expose the US economy to potentially devastating uninsured economic loss in the event of another catastrophic terrorism attack."
Terrorism exploded into worldwide consciousness in 2001 with the al-Qaeda attacks on the World Trade Center and since then barely a day passes without some mention in the media of new, thwarted or planned attacks somewhere in the world. Indeed, although Risk Management Solutions' (RMS) outlook in its annual US Terrorism Risk Model update maintained that 2006 would be a "quieter year" with the insurgency in Iraq "possibly peaking". It also pointed out that "despite fewer attacks being committed, more people were being killed in deadly, macro attacks" and revealed that the likelihood of a chemical, biological, radiological or nuclear (CBRN) attack over the next 12 months had increased to 5% from 3.5% a year ago. RMS also revealed it expected a decrease of 3% in the combined average annual loss for workers' compensation and property to $1.23bn.
When TRIA was signed into force on 26 November 2002, the industry breathed a collective sigh of relief that a terror cover solution had been provided. Three years later Bush was putting pen to paper once again to sign a similar document, the Terrorism Risk Insurance Extension Act of 2005, which extended the original arrangement by two years until midnight on 31 December 2007. Like a weary parent scolding a child, Congress had given the industry a further two years to adapt to a new age of terror while a private market solution was sought.
Since its inception in 2002, TRIA has helped to stabilise the terrorism insurance market, making coverage more available and affordable. It has now become part of the furniture, a backstop that is heavily relied upon by US primary insurers. "The capacity that exists in the primary terrorism insurance market today exists because of TRIA," says Property Casualty Insurers Association of America's Gregory Heidrich, senior vice president of policy development and research. "We see no evidence that reinsurers or any other capital sources would be able to fill the gap created by the expiration of the law."
The act has been very successful in what it set out to do as a public private solution to the lack of terror cover available after 11 September 2001. Take-up rates have more than doubled since 2003 to 58% on average according to Marsh, with real estate and financial institutions leading the way with a 79% increase in the amount of cover purchased.
Like the original act, the extension offers a short term solution, a federal backstop in the event of a "certified" act of terrorism, ie an event committed by or on behalf of a foreign person or interest, occurring within US borders. It is significant to note that acts of domestic terrorism are not covered. But the revised act has introduced some major modifications designed to ease the withdrawal of TRIA in 2007 (see table 1).
Effective from 31 March 2006, the US government is now restricted from outlaying federal funds until a programme "trigger" has been reached. This trigger is $50m in 2006 and rises to $100m in 2007 meaning, for example, that while an event may cause losses of $25m and be deemed certified, no federal reinsurance would be available. The Extension Act has also been modified to reduce government participation in 2007. Currently, the federal government will cover 90% of certified losses once an insurer's deductible is reached, with the other 10% remaining the insurer's liability. But this is set to decrease to 85% in 2007. In addition, TRIA caps the total liability of the programme and of the insurers at $100bn in any one year. The Extension Act also mandates that the government recoup payments if the industry's aggregate retention is less than $25bn in 2006 and $27.5bn in 2007. This is up from $15bn in 2005.
Like the hastily constructed splint strapped to an injured soldier, TRIA is not considered to be a sustainable or viable long-term option for the US insurance industry. But there are several issues preventing the industry from adopting a longer-term option. For instance, despite providing the ultimate safety net, the high cost of terror reinsurance and the retention of risks below the trigger levels means insurers are still cautious about committing to terrorism exposure. The rating agencies are also likely to revisit the capital requirements for those providing terror cover when TRIA expires.
Then there is the issue of modelling. By their very nature, acts of terrorism are random and unpredictable, making it difficult to quantify the economic and human losses from an event. Options for modellers include exposure-concentration analysis (identifying and quantifying concentrations of exposures around potential terrorist targets), deterministic modelling (establishing an event "footprint" at a specified target), and probabilistic modelling (estimating losses based on a large number of events and their associated probabilities). But despite the advances in this area, the capabilities fall short of those available to insurers in nat cat events. "Reinsurers do not have the same technical capacity to model terrorism risk as they do for natural catastrophes," agrees Chris Waterman, senior director at Fitch. "The inherent unpredictability of terrorism makes prediction and modelling very difficult."
Some within the industry are voicing concern over these hiccups on the road to a viable successor to TRIA. Joe Plumeri, Willis group chairman and CEO, points out: "There simply is not enough capacity in the world to endure a worst-case scenario. Whether we establish a national disaster insurance programme that resembles Pool Re in London, or have carriers set aside additional surplus on their balance sheets with some tax benefit or some other mechanism - we need a programme that we can all agree upon and will all live by as we enter this new era of risk. This is not an insurance issue; and it is not a political issue - it is a matter of national economic security."
If a successor to TRIA cannot be found before 2007, there is a real risk that capacity could disappear and rates could spiral out of control. Industry observers point out that the terrorism insurance market would become increasingly unsettled without TRIA or an alternative, something that would reflect negatively on the economy as a whole. "The commercial P/C insurance market has made it very clear that they will look to either exclude or severely sublimit terrorism exposures beyond 2007 if TRIA is not extended," explains Aaron Davis. "This will create an immediate imbalance in terms of available supply for terrorism capacity versus demand at the primary insurance level. It will also drive up pricing, potentially to levels that are no longer economic for many buyers."
Key leaders in Congress rejected a move towards a scheme similar to the UK's Pool Re in 2002, and there is nothing to suggest that the reception would be any different in 2007. Pool Re highlights the benefits of an initial public private partnership with a progressively greater transfer of responsibility to the private sector, but in its 12 year history it has not been tested by losses that approach those of the magnitude of 9/11. In light of this, and considering TRIA's overall success in achieving its stated aims, some believe a long term solution modelled on TRIA might be more appropriate. "The long-term solution needs to address all aspects of terrorism risk currently dealt with by TRIA, and needs to provide certainty and stability to insurers about what their risks and role will be," says Gregory Heidrich. "It also needs to be based on the coverage already provided by insurers and not create additional liabilities or exposures that they have never insured before."
"Terrorism is a global phenomenon and here to stay," says Chris Waterman. "TRIA is not an ideal solution to terrorism risk and I think the market is looking for a safe, long-term alternative." One thing everyone agrees on is that the current cycle of renewal or extension is not sustainable, nor is it providing the all-important stability the industry needs.
- Nick Thorpe is senior reporter of Global Reinsurance.