Catastrophic Events, Parameter Uncertainty and the Breakdown of Implicit Long-term Contracting in the Insurance Market: The Case of Terrorism Insurance

Following Hurricane Andrew (1992) and the Northridge earthquake (1994), insurance companies expended considerable resources on the measurement and management of the risk of natural catastrophes. Unfortunately, the next major “catastrophic loss” that the industry would face would be a man made event. After all losses are accounted for, the terrorist attacks on the World Trade Center (WTC) on September 11th, 2001 will be the costliest insured property loss in history, with current estimates of insured losses ranging from $40-$70 billion. Although the insurance industry appears to have the financial resources to absorb the WTC losses, this event has placed enormous stress on the insurance industry, created structural changes in how the market evaluates risk and return, and exposed fissures within the industry structure. Faced with a significant increase in uncertainty about the frequency and severity of future terrorist events, international reinsurers responded to the event by excluding or significantly restricting terrorism coverage from most reinsurance policies. This in turn motivated primary insurers to exclude terrorism coverage from most commercial lines insurance policies or forced primary insurers to cover terrorism risk without the benefit of reinsurance for some regulated insurance lines. At first glance, the response of insurance and reinsurance markets to the World Trade Center attacks seems to parallel the industry’s response to earlier unanticipated loss shocks, including natural disasters such as Andrew and Northridge as well as the 1980s crisis in commercial liability insurance (see, for example, Berger, Cummins, and Tennyson 1992, Cummins and Danzon 1997, Froot and O’Connell 1999, and Cummins and Weiss 2000). Following those unexpected loss events, insurance prices rose sharply and supply was restricted, an outcome that is usually explained in terms of probability updating and capital market imperfections (Gron 1994, Winter 1994, Cummins and Danzon 1997, and Froot and O’Connell 1997). Although the insurance market response to the WTC attacks exhibits similarities with the market disruptions caused by earlier large loss events, there are also indications that the WTC response may not represent merely a temporary market disequilibrium. The tendency of insurers and reinsurers to exclude terrorism coverage altogether rather than offering coverage at higher prices hints at least at a quantitative difference between this and previous catastrophic events. The nature of the event, a deliberate attack by terrorists, suggests that it will be more difficult to reduce parameter uncertainty through scientific and statistical modeling than in the case of natural catastrophes. Moreover, the mitigation of terrorism does not lend itself to domestic legal and contractual reform as in the case of the U.S. tort system. The objective of this paper is to provide evidence regarding the similarities between terrorist attacks and other types of catastrophic events by comparing the response of the U.S. equity markets to the WTC attack and earlier large loss shocks. Specifically, we conduct an eventstudy analysis of the response of equity markets to three large loss events – the WTC attack, Hurricane Andrew, and the Northridge earthquake. We differentiate the impact of eventinduced uncertainty (e.g., parameter uncertainty) and flight to quality in determining the market’s valuation of different insurance companies in an attempt to better understand the process by which the industry moves towards a new market equilibrium following a crisis. The remainder of the paper is organized as follows: Section 2 develops hypotheses based on a theoretical explanation of the relationship between loss shocks and insurance market disequilibria, drawing upon the prior literature on insurance market disequilibria. Section 3 presents an historical review of prior terrorism losses to shed light on the nature of the information communicated to the market from the WTC attacks. Section 4 discusses the sample of insurers used in the analysis and our event-study methodology. The empirical results are presented in section 5, and section 6 concludes with a discussion of implications of our findings for proposed Federal intervention in the market for terrorism reinsurance.