Quantifying the Unquantifiable: Some Perspective on Terrorism Risk

A brief commentary from a terrorism model expert on classifying terror risks after September 11, 2001.

In Jack Seaquist’s article Perspectives: Trying to Quantify Terror Risks at BusinessInsurance.com, the terrorism model expert, and Assistant Vice President at the catastrophe modeling firm AIR Worldwide Corporation, gives a brief overview of what he considers the most significant issues in terrorism risk assessment, as well as a look at the impact of terrorism on the insurance industry in general. Seaquist describes the terrorism threat in the United States as "highly dynamic" considering the "evolving political situations" overseas.[1]

The impact of Sept. 11, 2001, on the insurance industry was immediate. Once covered in most standard all-risk commercial policies, reinsurers either refused to renew terrorism coverage or began charging exorbitant rates. Unable to purchase reinsurance or to otherwise raise sufficient capital, insurers adopted new policy forms with terrorism exclusions. For a time, terrorism coverage was virtually nonexistent.[2]

Until the devastation from Hurricane Katrina in 2005, the 9/11 attacks in 2001 were "the largest cumulative claims payout in global insurance history." The attacks brought about insured losses of $32.5 million across multiple lines of insurance "including property, business interruption, aviation, workers compensation, life and liability."[3]

But terrorism insurance per se did not exist. Insurers paid claims on a loss for which they had collected no specific premiums. Because of its nature, terrorism was a risk considered impossible to underwrite, and insurance dried up for areas—such as Manhattan and Washington—deemed likely targets for future attacks. As a result, risk managers, insurers and business groups pushed for some sort of federal terrorism insurance response.[4]

Seaquist explains that the U.S. government responded by passing the Terrorism Risk Insurance Act ("TRIA") in November of 2002. TRIA was meant to stabilize the insurance market and it established the Terrorism Risk Insurance Program ("TRIP"), which provides "government-furnished reinsurance for direct terrorism losses" for most commercial lines above the company’s deductible. The Terrorism Risk Insurance Program Reauthorization Act, passed in 2007, extended TRIP through 2014 to give insurers "the sense of stability needed for a viable market."[5]

The future of the federal backstop for terrorism coverage is set to expire in 2014 as the administration considers limiting its exposure as part of deficit-reduction efforts. While the current appetite for terrorism coverage is healthy, many insurers have begun to make longer-term plans for terrorism risk management in the absence of the TRIP.[6]

As Seaquist points out, the insurance business model functions best when losses are relatively small, uncorrelated and random, even if those losses are relatively frequent. However, catastrophic losses are generally the opposite: "large, infrequent and highly correlated." Additionally, while catastrophe modelers have been able to use historical data and other methods to overcome many of the obstacles to estimating future losses with respect to natural disasters, estimating losses from terrorist attacks is much more challenging.

Historical data on terrorist attacks is much more limited and may not be representative of the current threat. Even more importantly, while scientists and engineers can achieve mastery over the physical science underlying natural catastrophes and their impact on the built environment, terrorist activity resists scientific quantification. In addition, while natural catastrophe risk remains relatively stationary over time, terrorist threat is highly dynamic.[7]