| The insurance mechanism is an efficient tool for managing risks that meet the insurable risk requirements of a large number of independent exposure units for which good information exists for calculating premiums. This research examines the case in which risks do not meet these requirements and risk bearing is costly. Two main areas are addressed. First, the effect of catastrophic losses on the insurance market ex post is examined to determine how insurers respond to large, unanticipated changes in surplus. Second, the effect of catastrophic losses on the insurance market ex ante is examined in order to assess how the cost of risk determines the supply function for individual insurers. The main hypothesis tested is whether or not insurers with a lower cost of risk bearing, or lower Variance Related Costs (VRCs), will assume a higher exposure to risk. To test the hypothesis, measures of risk and Variance Related Costs are needed. Risk is measured as the proportion of an insurer's surplus at risk from a catastrophic earthquake, and since Variance Related Costs are not observable, proxies are used. The proxies used for VRCs are organizational form, marketing system, leverage, profitability, tax position, and diversification. Tests are performed to verify whether or not insurers with lower VRCs assume more risk. The results are very supportive of the VRC model: Firms with a higher cost of bearing risk assume less risk. The existence of costly risk bearing implies that not all insurers are equally equipped to insure high variance lines of insurance. Instead, an efficient market solution requires that those firms with a comparative advantage in risk bearing should assume a greater share of the risk, and the results indicate that this does occur in the market for earthquake insurance. |
