Academic journal article Journal of Risk and Insurance

Moral Hazard in Reinsurance Markets

Academic journal article Journal of Risk and Insurance

Moral Hazard in Reinsurance Markets

Article excerpt


This article attempts to identify moral hazard in the traditional reinsurance market. We build a multiperiod principal-agent model of the reinsurance transaction from which we derive predictions on premium design, monitoring, loss control, and insurer risk retention. We then use panel data on U.S. property liability reinsurance to test the model. The empirical results are consistent with the model's predictions. In particular, we find evidence for the use of loss-sensitive premiums when the insurer and reinsurer are not affiliates (i.e., not part of the same financial group), but little or no use of monitoring. In contrast, we find evidence for the extensive use of monitoring when the insurer and reinsurer are affiliates, where monitoring costs are lower.


Insurance companies whose book of business is exposed to high risk, such as hurricane or earthquake losses or class action product liability lawsuits, have traditionally hedged the right tail of this exposure through reinsurance. Like primary insurance, reinsurance contracts encounter moral hazard. It is costly for the reinsurer to monitor the underwriting activities of the primary insurer and how the latter settles claims with its own policyholders. Consequently, reinsurance relaxes the incentive for the primary insurer to engage in careful underwriting and loss mitigation. This problem can be especially severe after a natural catastrophe where the primary insurer is overwhelmed with flood or earthquake claims and so is able to pass on the cost of settlements to the reinsurer.

Traditional reinsurance includes price controls against moral hazard, including deductibles, co-payments, and "ex post settling up," which is a retrospective adjustment of the premium based on losses incurred during the policy period that is also known as "retrospective rating." Less formal and longer-term controls are also used. Reinsurance is usually conducted as a long-term relationship. Experience bonds parties together and increases the cost of opportunistic behavior. The primary insurer gets continuity of access to reinsurance, whereas the reinsurer can use the relationship's duration to increase the effectiveness of its monitoring, and can use experience to set future prices and terms. (1)

Controlling moral hazard via long-term relationships can be costly. Froot and O'Connell (1997) have documented the costs of catastrophe reinsurance and show that the ratio of premium to expected loss increases dramatically at higher layers of coverage (i.e., for reinsurance in the right-hand tail of the loss distribution). Since moral hazard will increase in intensity the greater the level of reinsurance, this pricing pattern is quite consistent with unanticipated moral hazard. (2) Moreover, the shear size of these premium loading suggests that addressing moral hazard in this way is expensive. (3) These large premium loads are relevant today as both insured property and insured claims have increased significantly in the past few decades. (4)

Monitoring can also redress moral hazard. (5) In his transaction-cost-based model of firms, Williamson (1985) argued that, whereas markets use price incentives to resolve agency conflicts between separate organizations, monitoring can be a more efficient way to resolve conflicts within organizations where there is greater access to information. Conflicts within organizations are not fully internalized without monitoring since it is difficult to observe the contribution that each worker makes across all affiliated groups; hence, each agent is typically compensated according to his or her readily observed output. The models of vertical integration by Riordan (1990a,b) and Cremer (1993, 1995) show that for transactions within firms, where monitoring is relatively cheap, more emphasis should be placed on monitoring and less on contractual incentives. The opposite is true for transactions between firms where monitoring costs are higher. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.