Insurers, government struggle to manage risk of terrorist acts

In the aftermath of the terrorist attacks of Sept. 11, 2001, insurance coverage for commercial property and casualty loss in the event of terrorism became hard to find and prohibitive to purchase. Congress responded in November 2002, enacting the Terrorism Risk Insurance Act. TRIA temporarily put the federal government in the property and casualty risk reinsurance business, agreeing to share in the compensation costs resulting from acts of terrorism. Meantime, insurance companies were to have assessed their exposure and learned how to price and underwrite terrorism policies. In December, unless Congress takes action, the TRIA will expire, but insurance companies have yet to come up with a way to provide meaningful and manageable terrorism coverage.

In the aftermath of the terrorist attacks of Sept. 11, 2001, insurance coverage for commercial property and casualty loss in the event of terrorism became hard to find and prohibitive to purchase. The worst impact, according to a U.S. Treasury Department official, was "the drag it created on businesses' ability to finance new job-creating economic activity."

Congress responded in November 2002, enacting the Terrorism Risk Insurance Act. TRIA temporarily put the federal government in the property and casualty risk reinsurance business, agreeing to share in the compensation costs resulting from acts of terrorism. Meantime, insurance companies were to have assessed their exposure and learned how to price and underwrite terrorism policies. In December, unless Congress takes action, the TRIA will expire, but insurance companies have yet to come up with a way to provide meaningful and manageable terrorism coverage.

A major stumbling block, according to Daniel Brooks, professor of supply chain management at the W. P. Carey School of Business and director of the W. P. Carey MBA - Executive Program, has been the industry's inability to create a risk assessment model for a low-probability, high-impact event like a major terrorist attack. Help may be available in the research community, however, as scholars who do risk modeling develop quantitative models that are useful in forecasting the economic ramifications of a terrorism attack.

"The issue is how to estimate the frequency of something that happens very, very rarely and for which there is very little historical data," Brooks said. Finding the answer will require insurance companies to approach the question from a new direction.

"Insurance companies understand that the future may not look like the past and as a result they have to rely on modeling rather than anchoring on a big events like 9/11," Brooks said. "That's what makes their role as the providers of contingent capital for catastrophic events really challenging."

Challenging indeed — and in response, many insurance providers have excluded catastrophic events like terrorist attacks from their coverage offerings. Those who do charge significant premiums — sometimes as much as half of the estimated cost of an attack. Recent studies by the RAND Corporation and the Organisation for Economic Co-operation and Development point out that insurance policies do not offer coverage for chemical, biological, radiological and nuclear attacks.

The RAND study concluded that the nation's terrorism insurance system is failing to offer the financial protection businesses would need to prevent economic instability in case of a major terrorist attack. Exacerbating the problem, businesses are not buying the terrorism insurance policies that are available. The OECD reports that at the end of 2004, only about half the companies in the U.S. were covered.

It's a hazardous gap, Brooks said.

"Some of the large high rises may continue to operate without that kind of coverage, but the airlines need to have contingent capital available should an event like 9/11 occur again," he said. "The government plays a role in this as a resource of last resort."The temporary protection government has offered with TRIA may in fact have encouraged the insurance industry to delay collecting the data it says it needs. However, the industry's slow speed may also be explained by a reluctance to move ahead using tools that are still under development.

According to the OECD study, improvements have been made in terrorism risk modeling, but the models are still nowhere near the point of making attacks more predictable.However, "the insurance companies aren't really sure how to deal with potential," Brooks said. "The way they've always dealt with risk is to look at long-term trends to calculate their expected cost. From there they determine a premium that would allow them to have enough contingent capital if something occurs requiring them to provide insurance."

Unlike weather patterns or seismic information, however, when it comes to terrorism, insurance companies aren't dealing with traditional actuarial-style data. They're dealing with the machinations of shadowy organizations. As Brooks pointed out, there are agendas and alliances at play on a global scale that have finally come to U.S. shores and are the calling cards of highly networked, highly patient groups such as al-Qaeda.

These groups are strategic in their focus on influencing U.S. foreign policy through the use of primarily terrorist tactics. While the U.S. may be new to such global terrorism, other countries such as Great Britain, Germany, France and Italy are not, and their experiences could pave the way for reliable risk assessment models for this country.

"Transferring the experience of other countries would aid in the creation of good risk models in this country," Brooks said. "These models would help us develop practices that could protect against terrorism. At the same time, the models could aid insurance companies with decisions about how to design coverage that would provide companies with contingent capital for operations vital to the economy on the one hand, and controls their exposure on the other."

Brooks said that up to now, risk modeling in the United States has been focused on military targets and those related to national security, with relatively little attention paid to publicly accessible buildings or so-called soft targets. That would include targets that have national significance, are tied to national identity, and create an impact that is felt across the country - such as the World Trade Center. According to the RAND study, al-Qaeda is still pursuing soft targets that carry enormous economic consequences.

While the government may be out of the insurance business if TRIA is allowed to expire in December, Brooks said there are still things the government can do to make insurance companies more willing to offer meaningful terrorism insurance. One area of focus is the development of guidelines that would aid in allocating federal funds in general, and Homeland Security funds in particular, in a way that maximizes the risk-reduction benefit achieved by those funds.

This entails consideration of two complex issues: identifying where the risks are greatest and identifying the risk-mitigation activities that are most effective in addressing those risks. Currently, such monies are primarily allocated along traditional guidelines such as population. The emphasis, Brooks said, should be on risk-based prioritization. In other words, if al-Qaeda is looking to attack a target with maximum impact, those targets are found in major cities like New York, Washington, D.C., or Los Angeles. Therefore, those cities should receive more Homeland Security money than smaller communities that don't house the "attractive" targets.

"Risk based prioritization of resource allocation is a key focus in this," he said. "Is there a reasonable, auditable, rational way of allocating resources to address this kind of risk that is based on a risk-based allocation scheme? Insurance companies would obviously care about that."

How those major cities use their funds also plays a role in how insurance companies would view their risks. Recently, New York City's Metropolitan Transportation Administration awarded at $212 million contract to Lockheed-Martin to provide state-of-the art security on the city's subway system. Brooks said insurance companies would look favorably on the city investing money in an effective manner to protect a high-risk location with national and international economic importance.

Along with assessing the risks of a terrorist attack, insurance companies also must be able to quantify the impact of such an event, said Stephan Dieckmann, a finance professor at the W. P. Carey School of Business.

"If you would be asked to write a contract for terrorism insurance, what would this contract look like?" he asked. "What is the basis for judging whether the insurance contract has to payoff? Is it the amount of dollar loss in the sense of capital market loss? Is it the amount of people who died in such an event? Or can it be quantified in terms of a terrorism risk index? It is difficult to measure."

"We have to think about the impact of a catastrophic event on the aggregate economy," Dieckmann said. The price for such insurance will consist of the expected loss plus an additional insurance risk premium, linked to the aggregate impact. To illustrate, Dieckmann used an example that is not connected to terrorism, but is still pertinent. A small Florida city being battered by a hurricane would certainly affect the lives of local citizens, he said, "but the impact on the overall U.S. economy could be small."

"Sept. 11 was an event that did matter in aggregate," he said. "It's seems unclear if it mattered in aggregate in terms of loss of productivity, but it certainly mattered as a structural change or structural break in how we think about these rare events, and how aware we are of them." Participation in rare event insurance as a buyer or a seller has generated interest in catastrophe bonds — capital market securities that have an insurance contract embedded in the bond itself. If the catastrophe does not occur, it works like a regular bond.

But if the catastrophic event does occur during the lifetime of the bond, investors do not get principle and/ or interest payments in return. The principle will then be used to payout the insurance claim. Such a bond linked to terrorism risk has recently been written against the 2006 Soccer World Cup in Germany not taking place, said Dieckmann.

Finding and perfecting new ways to handle the financial exposure of terrorist attacks is becoming increasingly critical as the government shows signs it may be stepping out of the insurance business by the end of the year. If that happens, the insurance industry, indeed the entire economy, will be left still facing the daunting challenge it first encountered on Sept. 11, 2001.

"In this area of terrorism, because we are an open economy in an open culture, we can't afford to nor do we want to put a fence around everything with guards with guns in front of them," Brooks said. "So is there a way to reduce the risk of terrorism in a cost-effective way that doesn't impact the culture but does provide real risk protection? That's the area where modelers are working to try to come up with ways to address risk without changing the culture and without incurring debilitating costs."