While in flight to Boston on the morning of Sept. 11, 2001 and hearing news of the horrific attacks in progress, Hemant Shah's first concern was the fate of the business associates he met with the previous day at the World Trade Center.But as the founder of Newark, Calif.-based Risk Management Solutions (RMS), he must have realized the seismic change the field he helped pioneer-catastrophe modeling-would soon experience.
Last month, the introduction of two competing models that, for the first time, attempt to quantify the risk insurers face from the hands of terrorists offered a slight glimpse into a world where geopolitical issues factor into insurance calculations as never before. "For all of our catastrophe models, and particularly for terrorism, we like to approach the output with a significant degree of humility because there is so much uncertainty," Shah says.
But insurers are willing to spend hundreds of thousands of dollars for software developed by RMS and its competitor, Boston-based AIR Worldwide Corp., with the hope of gaining some insights into their new liabilities.
Both companies employed a wide range of political and military experts to calculate the damage that could be wreaked from weapons ranging from conventional explosives to a plausible range of chemical, biological and nuclear scenarios. The models are coming into the market at a time when the insurance industry is in a state of upheaval comparable to the aftermath of Hurricane Andrew 10 years ago.
While the likely passage of some sort of so-called terror risk liability "backstop" by Congress will ease some of the pressures, it could also add new ones as regulators may no longer permit the terror risk exclusions to primary writers that they were granted earlier this year.
For the past year, both firms have been working feverishly to come up with a model and process, which will be continually revised over the years, taking into accounting the kind of factors that previously have received scant attention from actuaries.
While analogies to the field of natural catastrophe modeling have been made, they fall short in one crucial regard: Tornadoes and hurricanes do not spring from the hearts and minds of human beings.
As product manager for AIR Worldwide, Jack Seaquist can look back on a year of countless late night sessions fueled by takeout fare as the pressures from their insurance clients grew for answers to seemingly unanswerable questions.
"The event generation part can be seen as the most challenging," Seaquist says. "But we have developed a process using expert opinion to scope the range of possible attacks that might occur from the different terrorist groups. Experts were asked to allocate frequencies and the relative likelihood for different kinds of targets, in what regions they might happen, and with what weapons."
The two companies took divergent approaches in scrutinizing the risks.
"In the beginning of natural catastrophe modeling, everyone used expert opinion for damage ratios," Seaquist explains. "There was no collected information about the kind of damages caused to different structures from different earthquakes."
AIR uses the "Delphi" method, developed by the RAND Corp. at the outset of the Cold War, to generate forecasts by ranking any number of educated guesses into a coherent hierarchy of probabilities.
"We are using that same approach for assessment as to what terrorists might do in the future," Seaquist says.
RMS planners, on the other hand, rely on a system known as "Game Theory," an approach popularized by the movie "A Beautiful Mind," the story of its chief proponent, John Nash.
Dr. Gordon Woo, the architect of the RMS model, says that in keeping with principles of Game Theory, the desires of al-Qaeda to maximize its attacks must be balanced with its rational response to stepped-up security and counter-intelligence efforts. "A traditional probabilistic approach, such as that used for modeling natural catastrophes, is simply not up to the challenge," he adds.
What Shah calls "swarm" attacks would be a strong likelihood in al-Qaeda planning. "One of the implications of all this is the probability for future swarm attacks because security tends to harden in the aftermath of an attack, and particularly security that is relevant to the attack that just happened," Shah says.
"The insight is that groups launching such attacks will tend to use simultaneous attacks against a similar class of targets using similar kinds of logistical arrangements, because if they do it once, the security will harden and it will be very difficult to do that again. You see a pattern of that in al-Qaeda with the simultaneous bombings in East Africa in 1998 and the hijacking of several airplanes on Sept. 11."
So far, AIR has signed two clients for its program, while Shah said several clients had placed orders for programs that can cost up to $300,000 a year.
So what are companies getting for their money? "We hope to narrow the range of uncertainty we have had in the past year, and in doing so, give clients a bit more insight into the implications of various decisions," Shah says. "Whether or not this leads to more coverage-it is too early to tell."
Seaquist says that with the development of new quantification methods, insurers may no longer have to price "for all perceived possible outcomes."
Don Griffin, director of business and personal lines for the Des Plaines, Ill.-based National Association of Independent Insurers, says whatever form the federal backstop finally takes, the law will have a much greater bearing on the pricing and availability of terrorism risk insurance. But the models will play a role.
"As for pricing, it could also lead companies to charge more once they better understand their risks," Griffin says. And with the disappearance of some exclusions forcing companies to provide such coverage, "the models will become even more essential," he notes.
J. Mark Puccia, managing director for Standard & Poor's, says the primary value for the models will lie in the process companies undertake to use them. "I'm not sure there's a way to validly model terrorism risk," he says. "The value is in collecting the necessary data to input to the models."
As for pricing and availability impact, Puccia says that once the more thorough quantification of risk takes place using the models, "there probably will be pricing differential based on the location of the property-with trophy property getting higher rates. But companies will get comfortable with the spread of risk concept and gradually increase their exposure."
Stephen Tuckey is a New York-based editor for Insurance Chronicle, a Thomson Financial Insurance Solutions publication.
Register or login for access to this item and much more
All Digital Insurance content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access