“The business of insurance centers around the pooling of risks individual participants might not be able to handle on their own. Two main parameters govern the pricing and availability of insurance: the frequency of the risk event (or the probability of loss), and the size of the loss given the occurrence of the event. In statistical terms, event frequency or probability of loss is modeled as a “loss distribution”. The most common distribution is a “normal” one, which has the familiar inverted “U” shape, with a peak in the middle, and “tails” at the two extremes. For routine events, like auto accidents, and most risks to homes (fire and tornadoes), the loss distribution is well known. Much of the probability is centered on the peak in the middle, and the tails at the end - the low frequency events - are spread out rather thinly. Insurers have relatively little difficulty pricing the insurance for such events, since the average loss can be readily predicted (it is generally the “peak” of the loss distribution), and there is not a lot of uncertainty or “variance” around the average. “The events or risks that create difficulties for insurers are those where the loss distribution is spread out over a large range, reflecting large uncertainties around the likely costs in each year. A further complication is that the shape of the loss distribution itself may be uncertain. In these cases, insurance prices are more difficult to determine, and in some cases, insurers may be reluctant to provide any insurance coverage at all. “Large natural catastrophes - hurricanes and earthquakes - present these sorts of difficulties. They are difficult to predict and when they occur they can generate claims that substantially deplete, or in a worst case exceed, the accumulated capital that the insurer has built up over time. In technical terms, large mega-CATs therefore pose “timing” risks (the event happens before sufficient premiums have been collected to fund payment of claims) and “ambiguity” risks (the shape of the loss distribution itself is not well known or understood).77 Terrorism risks are especially difficult for insurers to price and to cover, since it is essentially impossible to statistically model the likelihood of terrorist incidents or the magnitude of losses, should they occur”. (Financial Services Roundtable, Nation Unprepared for Mega-Catastrophe, 2007, 45). “Higher insurance premiums understandably are unwelcome to homeowners and owners of commercial properties. But to the extent they reflect, or are permitted to reflect, the true risks of future damage from all sorts of catastrophes, then high premiums do serve a socially useful function: they send appropriate signals to individuals and businesses of the social costs of their decisions to locate in hazard-prone areas and to the consequences of investing, or not investing, in various mitigation measures that may be available. Indeed…if the marketplace, rather than regulators, set premiums, the effect will be to lower future disaster losses - suffered by individuals and absorbed by taxpayers through disaster relief - by inducing private actors in invest in mitigation measures, or in some cases to move, to reduce their exposure to catastrophic losses. “But higher insurance rates also can lead to public policy problems and challenges, and these constitute additional concerns with the current system. As homeowner insurance premiums continue to rise in areas facing above-average catastrophe risks, increasing numbers of 77 The individuals may choose either to drop their insurance or buy policies with considerably higher deductibles. Where this occurs, homeowners are exposed to higher and potentially devastating losses in the event of future catastrophic events. Because the federal government always responds out of humanitarian concern to catastrophes by providing disaster relief to help cover uninsured losses of victims, any reduction in private insurance coverage induced by higher premiums most likely will raise future federal disaster relief costs. “Any increase in disaster relief payments caused by insufficient insurance coverage leads to a second problem with the current financing system. Unlike insurance premiums, which are paid by homeowners directly exposed to catastrophe risks, disaster relief costs are borne by taxpayers (either currently or more likely in the future, because the costs of relief typically are financed by additional federal borrowing) who live outside the damaged regions. This raises an issue of fairness: why should residents of the Midwest, for example, bear disaster costs incurred by those who choose to live on the coasts There are also potential efficiency costs. To the extent that homeowners do not insure or under-insure because they expect at least some disaster relief, then they will have less incentive to invest in mitigation. “Mounting insurance premiums for catastrophe risks, even if actuarially appropriate, lead to a third policy-related concern. As insurance rates rise, so does political pressure on state insurance regulators to suppress them artificially - that is, to not permit insurers to charge premiums based on actuarial experience or the best available scientific evidence, or to not allow insurers to pass on fully the costs of reinsurance. In either of these events, more insurers will find it unprofitable not only to write new policies, but to renew existing ones. The net result will be a reduction in the availability of privately-supplied insurance, aggravating one of the main problems that now exists in coastal communities along the Gulf and East coasts. Further, if insurers cannot charge actuarially and scientifically appropriate premiums, then there is a higher risk of insurer insolvencies in the event of future Mega-CATs (thus placing greater burdens on state guaranty funds, and on the surviving insurers who finance these funds, and their policyholders who most likely will ultimately bear these costs)”. (Financial Services Roundtable, Nation Unprepared for Mega-Catastrophe, 2007, 48-49)