A new class of risk has been identified by the September 11 attacks. By Andrew Coburn. As the firemen and investigators slowly sift through the rubble of the World Trade Center, the analysts and claims managers have been picking their own way through the complexity of insurance claims that surround the events of September 11. The true extent of the insured loss from the World Trade Center attack will take a long time to be finalised. Very many different lines of business were affected and many of the claims are expected to take years to finally settle.
Development of the WTC loss
By 9 November 2001, 11,383 commercial claims and 7,476 personal claims related to the World Trade Center attack had been logged by the New York State Insurance Department, at a notified value of $8.4bn. These represented a small sample of the eventual claims total. Unlike a natural catastrophe, where the majority of claims are paid out within a year, some observers expect that payments will take longer and a small proportion of the claims from this event could run on for more than a decade.
Losses are now expected across more than 22 different categories of insurance business. These classes of insurance range from the property loss of the buildings damaged in the attack, to the compensation payments due under workers' compensation and other accidental death and dismemberment policies, through to various classes of business interruption and specialty contents lines. Each line of business is being explored and examined carefully to review contractual conditions, some already resulting in legal actions. The settlement for the loss of the two World Trade Center towers themselves will depend on the legal action currently underway between the insurers and the leaseholder. Liability issues are being examined in detail to determine the contractual obligations on such items as third party injury resulting from the building collapse. Business interruption claims are being received from locations all across the country, and terms and conditions are under careful scrutiny to decide which are eligible for settlement. Some losses, which had been reserved for, have now been recovered, such as the bullion recently retrieved from the rubble.
The variability of each of these components will impact the final industry loss. Taking optimistic assumptions about each line of business, RMS estimates that the loss could stabilise at under $35bn, but with more realistic assumptions, RMS estimates that losses will eventually reach just over $50bn. There are some less likely scenarios where class actions and other settlement decisions could result in the industry paying out tens of billions of dollars above that.
By any analysis this is the largest single loss in insurance history. It is not, however, completely outside the realm of catastrophe loss levels contemplated by the insurance industry. Many insurers and reinsurers manage their accumulations to between the 100- and 250-year loss level for natural hazard risks such as earthquakes, hurricanes and other windstorms. RMS models place such losses between $50bn and $70bn for the US P&C industry. By this measure, the World Trade Center disaster, while not a natural catastrophe, will generate a level of loss of the order of the PMLs anticipated by most insurers and reinsurers. Consistent with this analysis, rating agencies believe that while some specific firms may be challenged, the industry as a whole will be able to handle the claims from the World Trade Center event without significant disruption.
What is becoming clear is that this will be largely a reinsurer's catastrophe. The type of assets impacted and their insurance structures ensured that retentions were low. Around two-thirds of the loss will be paid by reinsurers, one-third by primary insurers – the opposite way round to typical natural catastrophes. As a result, insurers are focusing on their reinsurance recoverables and worrying about liquidity during the time until settlement. Other consequences are that reinsurance prices are rising rapidly and there could be a drying up of capacity in key accumulation zones.
A new class of risk
Many insurers are now reacting to the World Trade Center loss by instigating far-reaching reviews of their business. The overall approach has been to review urban risk in a broader context than purely terrorist threats. There is an obvious need to review preparedness in case there are further terrorist attacks, but the overall emphasis is in learning from the disaster about the nature of the exposures, and the interlinkages between insured lines of business in a modern urban context. Companies are reviewing their terms and conditions, rating and covers, reinsurance purchasing, accumulation management, and other portfolio management processes.
The World Trade Center loss signifies virtually a new class of risk – urban catastrophe. Understanding this new class of risk, and how to manage it, is challenging many senior insurance professionals. RMS has been working with its insurance clients to review scenarios and approaches to managing complex urban risk following the World Trade Center disaster. Workers' compensation has emerged as an important theme, with companies wanting to model risk from all causes, including earthquake and terrorism, as a result of the losses of September 11.
It is important for insurers and reinsurers to learn from the World Trade Center event about the correlation of different lines of business in a modern urban central business district, and to put into place measures that manage urban accumulations of risk. It is possible to ‘stress test' certain parts of the business through scenarios, and to examine how very different lines of business could suffer simultaneous and correlated losses.
The World Trade Center catastrophe raises several issues that challenge pre-existing measures of probable maximum loss:
For good reason, the insurance industry has focussed its catastrophe management activities on natural hazards. Since 1970, 38 of the top 40 most costly catastrophe losses have resulted from natural hazards. Natural catastrophe is the driver of catastrophe cost – RMS models suggest that the global P&C industry can expect $20bn in natural catastrophe losses in an average year, including the possibility for single event losses in excess of $100bn.
However, as a man-made catastrophe, the World Trade Center attack serves as a reminder for companies to consider additional risk scenarios beyond the predominant natural hazards. Even prior to September 11, man-made events such as industrial accidents, aviation losses, large urban fires and explosion have contributed 20% of the total catastrophe losses over the past five years. Accidental explosions, such as the factory explosions in Toulouse in France ten days after the World Trade Center attack, have caused multi-billion dollar insured losses. Some of the worst man-made disasters in recent years such as Bhopal and Chernobyl could have cost the industry billions if they had occurred under different circumstances.
Incorporating terrorist attacks and other man-made super-catastrophes into an overall understanding of catastrophe risk is not straightforward. Some of the events cause a geographical area to be impacted, such as a triggered building collapse (like the September 11 attack) or an explosion or aerial impact. These are ‘vertical' events, where the correlation is predominantly how the spatial geography of the loss affects buildings, people and infrastructure in close proximity. Other events are more ‘horizontal' in nature, affecting more dispersed targets or economic sectors, where the correlation occurs through networks or dependencies. Horizontal scenarios encompass biological or chemical hazards (accidental or otherwise), IT network disruptions (such as virus attacks), and events that would close transport networks or cause mass evacuations of personnel.
Urban catastrophe scenarios
Despite the complexities, generalised analyses can rapidly identify the orders of magnitudes of losses generated by different scenarios and test assumptions about the correlation of losses across multiple lines of insurance business. These scenarios ‘stress test' a business, examining the robustness of the individual silos of risk and the overall balance sheet. It demonstrates where additional data is needed to inform a company's picture of its own risk. The range of outcomes is used to supplement the more rigorous and probabilistic assessments of natural hazard risks and informs the management of risk across the business.
Part of the debate also hinges on the approach to pricing coverage for this type of event. To assess risk pricing methodically, the severity of loss needs to be augmented by assessments of the frequency of occurrence. Data is scarce, thus defining frequencies of events of these classes is more judgmental. Companies need to justify their approach and to communicate it to clients, reinsurers and regulators. However, methodologies do exist for constraining the range of likely probabilistic assessments and incorporating the uncertainties into the overall assessment of risk.
One of the important lessons from the collapse of the twin towers is the concentration of insurers' property exposures in dense urban environments and the need for improved data on such exposures. RMS databases of insured value in major cities show that there are many cities in US with dense concentrations. In several major cities, there is over $30bn in insured buildings and contents property value within a one kilometre radius of a downtown location. Other major US cities have lesser but significant densities of insured value. The complexities of a modern urban ‘matrix' and the challenge of defining insured values for a portfolio of highly customised risks, brokered and underwritten through multiple channels on a global scale is daunting. Companies are seeing claims emerging from exposures that were not in their main exposure register prior to the event. It is possible the industry underestimated its exposures in and around the World Trade Center by as much as 40%.
Property, contents and equipment, and business interruption tend to drive the risk from natural perils. Losses from workers' compensation, casualty and liability comprised only 5% of the losses from 1989 Loma Prieta and 1994 Northridge earthquakes. In the WTC loss, RMS estimates that over 35% of the total claims will relate to casualty, liability and life insurance exposures. Highly destructive urban catastrophe scenarios are also likely to have significant losses in these lines of business. There is a close correlation between concentrations of workers' compensation exposure and commercial property concentrations. While modern building codes are explicitly optimised for life-safety considerations, catastrophic destruction of structures are possible in earthquakes, terrorist attacks, fire conflagrations and industrial accidents. In such circumstances, property damage is correlated with casualty risk, which is in turn correlated with liability exposures.
To manage this new class of risk effectively, the insurance industry will need to consider spatial correlation of risk across various lines of business, presenting a new challenge.
Many of these lines are managed independently, and there is often little visibility across the enterprise where high multiline accumulations may exist.
To be effective, information needs to be gathered on the locations and interrelationships between the various classes of business. Geographical information on workers' comp accounts and other exposures will need to be captured by companies' accumulation control systems. Given the scale of the loss and the potential correlation with what may already be significant property exposures, the stakes should warrant the necessary investment in business processes, systems and risk analytics.
The World Trade Center disaster has changed the insurance industry's view of its risk profile. Its view of its probable maximum loss prior to September 11 is now a lower bound estimate. In addition to natural hazards, it is now recognised that a new class of risk exists from urban catastrophes. Some of these are intentional – such as terrorist attacks – and some are accidental. The World Trade Center disaster demonstrated that these have the potential to cause losses rivalling or even exceeding those that occur in major earthquakes and hurricanes. Insurers of tomorrow are already making rapid progress in understanding and beginning to manage this broader spectrum of risks.