Expected utility strategic decision models for general insurers

Abstract

It has been argued in previous studies that the expected utility decision criterion provides useful insights for certain insurance problems, such as underwriting, reinsurance and portfolio optimization problems. In this study three new models are developed which extend and generalize previous results. The first model analyses modified stop-loss reinsurance. The second model analyses risk pooling where both inward and outward reinsurance occur. Expected utility calculations can be used to provide insight on the attractiveness of competing reinsurance and risk pooling options. The third model is for strategic planning, where risk/reward tradeoffs for all the insurer's business activities (underwriting, investment, reinsurance) can be considered in aggregate. The simpler models can often be solved analytically however the strategic planning model is relatively complex and uses Monte Carlo techniques to determine retained earnings distributions. The expected utility approach has been found to be powerful, flexible and comprehensive as a decision aiding mechanism. From a normative viewpoint, this approach accounts very well for all the important decision elements. Recent developments in decision support systems will allow these models to be made available to practitioners in user friendly forms. © 1986, International Actuarial Association. All rights reserved.