A REGRESSION APPROACH FOR GENERAL INSURANCE DEDUCTIBLE RATEMAKING

• Main Author: Gunawan, Marcelino
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49720
• Institution: Institut Teknologi Bandung

In general insurance terminology, the deductible can be considered as the amount which has to be paid by the policyholder before the insurance company starts to pay any expenses. The amount of the deductible will directly affect the distribution of both claim frequency and severity, which are the two important factors that take part in the calculation of general insurance premium. This final project will apply the regression approach in the ratemaking process. The most suitable probability models for estimating claim frequency and severity for each values of deductible are to be chosen first by utilizing the natural logarithm link function based on the generated data with lognormal values of deductibles as the predictor. Then, the relativity function given a specific base deductible can be calculated and it can be used to describe the effect of deductibles on the expected value of aggregate claims. The expected value of aggregate claims obtained is also known as pure premium. The risk premium is then calculated by using the expected value principle with two types of approximations: normal and lognormal approximations.