Some statistical investigations in general insurance
This research is concerned with the application of statistical methodology to general insurance, in particular to the motor insurance branch. The research consists of two independent investigations in which two important aspects of motor insurance claims were studied. The aim of the first phase was to measure the influence of a given set of rating factors on both average claim size and claim frequency, and the second phase aimed at estimating an average pattern of settlement delays for motor insurance claims. In the first investigation, five rating factors were considered: mileage, zone of garage, no claim bonus, vehicle make and vehicle age. No strong evidence was found that these five rating factors had a significant influence on the average claim, but the same was not true for the claim frequency, in which such an influence was not only detected, but also measured for a limited set of the observations. An average pattern of settlement delays for motor insurance was estimated in the second investigation, based on two years of claims experience. It was shown that such a delay could give rise to a marginal profit for an insurance company, provided that the corresponding reserves for outstanding claims were appropriately invested. Due to the difficulty in getting data from the insurance industry, the two phases of the research had to be carried out using data from different sources. Thus, the first investigation was based on data from the Swedish claims experience during the year of 1977 and the second was based on data from a medium sized British insurance company relating to claims payments associated with accidents which took place in the course of the years 1972 and 1973.