Prediction of Future Insurance Premiums When the Model is Uncertain

Abstract

It is an important task for an actuary in determining an appropriate premium price for each customer with different risks and characteristics. The purpose of this study is to determine the best model for pure general insurance premiums and variables that can affect the amount of pure premiums. One of statistical analyzes that can be used to model insurance premiums is Generalized Linear Models (GLM). GLM is an extension of the classic regression model that can accommodate the flexibility of its users to use multiple data distributions, but is limited to the exponential family distribution. In the GLM model the premium is obtained by multiplying the conditional expected value from frequency of claims and cost of claims. Based on the research that has been done, it is found that frequency of claims follows the Poisson distribution. Meanwhile, cost of claim follows the Normal distribution. From the two models, it is found that the variables that affect the pure premium are the type of work, the reason for the claim, the location of residence, the marital status and the class of the customer's vehicle. It indicates that the GLM model is a representative model and useful for the insurance company business.