Search

In the first chapter of this dissertation we explore the information content of insurance company ratings by examining the effects of rating changes on stock prices of publicly traded insurance companies. Using event study methodology, we test the hypothesis that changes in A. M. Best's ratings of insurance companies contain information relevant to the valuation of these firms. We document an asymmetric reaction of stock prices to rating changes subject to a rating benchmark: rating downgrades cut share prices by approximately 3.25% but upgrades have no statistically significant effect. Furthermore, the market appears to anticipate changes in A. M. Best's ratings. Stable ratings, on the other hand, yield a slightly positive response below 1%. In the absence of this benchmark rating, our results show that any marginal change in a rating does not significantly influence share prices. In the second chapter we approximate the average ratio of prudence to risk aversion by applying cross-sectional household survey data to an expected utility model of life insurance demand. This procedure allows us to test the Pratt-Arrow hypothesis of decreasing absolute risk aversion (DARA). Additionally, we estimate the relative magnitude of prudence, the propensity to take precautions when faced with risk. We find significant evidence of prudence, but the magnitude is smaller than that predicted by the DARA hypothesis. On the basis of this evidence, we reject the DARA hypothesis. The final chapter extends the work of the second by focusing on risk aversion measures for individuals. We derive a reduced form equation for the Pratt-Arrow coefficient of relative risk aversion and employ life insurance, assets, pension wealth, earnings, mortality, and Social Security data to estimate this measure for individuals. The risk aversion measures are used to examine differences across demographic groups based on age, sex, education, nationality, race, marital and parental status, religion, health