Nearly two years after the terrorist attacks of September 11, 2001, businesses continue to evaluate the threat of another catastrophe and try to gauge their own exposure to it.


Their perception of the risk is conditioned by periodic terrorist attacks abroad, fears of reprisals for the U.S. invasion of Iraq, the federal government’s intensive work on terrorism and an underlying insecurity that may now be a permanent feature of the national psyche.


But none of that has translated into a rush for coverage, according to participants at “Assessing and Managing Extreme Events,” a roundtable for insurers, reinsurers, academics and modelers held by Wharton’s Risk Management and Decision Processes Center on April 28. The event focused on the challenges facing the insurance industry, government and modelers in seeking to manage the risk of terrorism and other extreme events such as natural disasters.


A majority of insurance brokers says that fewer than 10% of their small commercial property clients and under 20% of medium-sized accounts have purchased terrorism coverage since federal legislation required insurers to start offering it last November, according to a survey by the Council of Insurance Agents & Brokers (CIAB), a trade group, published March 24.


Even in New York, one of four cities given a “Tier 1” rating by the insurance industry for the highest risk of another terrorist attack, only about one in eight commercial clients are adding terrorism coverage to their policies, according to John DeMartini, senior vice president at Towers Perrin Reinsurance, a broker in Stamford, CT. About one in three or four New York City businesses are placing initial orders for terrorism coverage, but about half of those decide to cancel “when they see the bill,” he added.


Because it is difficult to quantify terrorism risk, some managers are not paying enough attention to the issue, noted conference chairman Howard Kunreuther, co-director of the Wharton Risk Management Center. They need to abandon the view that “I’m not thinking about this and I’m not putting my money into something I don’t think is going to happen,” he said.


Terrorism premiums nationwide add about 10% to the cost of existing insurance policies for small and medium-sized companies, and 20% or less for large accounts, according to the CIAB. But the costs are predictably higher in those cities – New York, San Francisco, Washington D.C. and Chicago – that have been classified as a “Tier 1” risk.


One broker quoted by another trade group, the Independent Insurance Agents & Brokers of America, reported a 17% loading for terrorism coverage on apartment buildings in The Bronx, N.Y. Significantly, the client declined the coverage, which would have cost $18,000 a year.


Another broker seeking terrorism coverage for a building in San Francisco received a quote that would have doubled the client’s existing $600,000 premium for earthquake insurance, according to the February survey. Again, the coverage was declined pending the receipt of further quotes. An insurance wholesaler in Wisconsin, by contrast, reported terrorism loadings of between 4% and 10% on existing policies.


“Small, relatively low-profile accounts seem to be able to find terrorism coverage at a reasonable cost but many are opting not to buy it because they don’t think they are at risk,” said CIAB President Ken Crerar in a statement accompanying the survey. “On the other hand, some of the riskier operations, with real exposures, choose to do without coverage because of the cost.”


Insurers are now required by the Terrorism Risk Insurance Act to offer coverage against foreign, but not domestic, terrorism. The federal government has agreed to underwrite most of the risk for the three-year life of the Act as a substitute for the reinsurance industry which largely withdrew from coverage after suffering about three-quarters of the $40 billion losses stemming from the September 11 attacks.


Overall, insurers seem to be following guidelines for terrorism pricing published by the New Jersey-based Insurance Services Office, an advisory company to the industry, said spokesman Dave Dasgupta. For Tier 1 cities, the guidelines recommend a premium of 3 cents per $100 of coverage. For the Tier 2 cities of Philadelphia, Seattle, Los Angeles, Houston and Boston, the premium is 1.8 cents, and for Tier 3, representing all other areas, the premium is 0.1 cent.


“Opportunistic Pricing”

As primary insurers look to spread their risk, reinsurers are seeking to offset their September 11 losses with some “opportunistic” pricing, the Wharton conference heard. A sample cover quoted by DeMartini as being representative of the market showed one insurer agreeing to pay a premium of $900,000 for $9 million worth of terrorism coverage. This 10% “rate on line” is dramatically higher than the generally agreed probability of a terrorist attack occurring, DeMartini said. That’s more likely to be about 0.5%, on which a reinsurer would normally charge a small technical margin.


“The reinsurers don’t want to look to their clients like they are running away from the problem,” he said. “Their investors are probably saying, ‘We really don’t want to take on the risk but if we do we had better charge a good price.’”


For companies seeking to buy primary insurance, the sticker shock is coupled with a reluctance to believe that the coverage is necessary despite the events of September 11, even in areas classified as high risk.


“Transfer of low-frequency, high-impact risk is a difficult sell,” according to conference presenter Jim Ament, vice president, operations, at State Farm Insurance Companies. “The fact that the risk is unlikely to occur, even if it does have the potential for devastating consequences, results in a reluctance on the part of potential risk transferors to incur the cost associated with the risk transfer.


“The challenge is to help stakeholders understand the nature of low-frequency, high-impact events and appreciate the means available to mitigate or transfer those risks,” Ament said.


For clients, insurers, reinsurers and modelers who seek to simulate the risk, the fundamental challenge of terrorism insurance is one of uncertainty. While historical data on natural disasters such as earthquakes and hurricanes provide a reasonably clear risk profile, there is no comparable record for the nature, location, frequency and impact of terrorist attacks.


A further source of uncertainty is the future of the Terrorism Risk Insurance Act, which expires on December 31, 2005, and is based on the assumption that the insurance industry will by that time have found ways to offer coverage to its clients and cover its own risks at reasonable rates.


There’s little chance the act will be extended, said DeMartini. “The word is that Congress doesn’t have the appetite. They saw it as a bailout for the insurance industry, and they’re not inclined to extend it.”


For now, the industry is reacting to the random nature of the terrorist threat and pricing its coverage accordingly. “A big part of the availability and price of terrorism coverage is a function of the inability to understand the risk,” noted DeMartini. “They don’t know when, where or how the next terrorist attack will occur and how much damage it will do.”


The Challenge for Modelers

But modelers are working hard to assess the risk. Applied Insurance Research (AIR), for example, has incorporated more than 300,000 U.S. landmarks such as corporate headquarters, industrial and energy installations, transportation facilities and government buildings into its model. The model includes a Group Threat Index for known terrorist organizations which includes assumed targets, weapons and attack locations, based on what is known of each organization’s objectives, history and capabilities, Jack Seaquist of AIR told the conference.


In building its model, AIR is challenged by a lack of detailed information on target areas, Seaquist said. Building data, for example, should include the address, construction type, age and number of floors, but often the only information initially available is the company location.


Another model produced by Eqecat incorporates a “severity assessment” that includes a database of probable target sites, simulation of the “footprint” of a chemical, biological, radiological or nuclear attack, and a vulnerability assessment in which the intensity of the hazard is converted into property damage and injury rates. The Eqecat model covers more than 250,000 sites and is based on information from government and academia.


Despite the wealth of data being incorporated into models, there is a risk that planners are not focusing on the right targets for another attack, Eqecat’s Jim Johnson told the Wharton conference. While there is an assumption that big cities such as New York and Washington are at greatest risk, planners should consider the possibility of an attack on the American heartland in which terrorists would seek to demonstrate that “nowhere is safe,” Johnson said.


Common assumptions about the means of attack may also be off the mark, Johnson added. “The surprise will be multi-mode attacks (such as chemical or biological devices), not just bomb blasts. And if there is another airliner attack it’s not likely to be a passenger jet; it’s more likely to be a plane filled with explosives.”


The magnitude of the problem requires new thinking about managing risk, said Paul Kleindorfer, co-director of Wharton’s Risk Center. “We felt that we have to expand our horizons, and to bring together representatives of academia, government and industry. It was really important to try to understand some of the tremendous challenges we face.”