BAYESIAN APPROACH OF PREMIUM CALCULATION IN BONUS MALUS SYSTEM WITH PARETO DISTRIBUTED CLAIM SEVERITY OF VEHICLE INSURANCE

• Main Author: NANDA PERMANA (NIM: 20816021), BAYU
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/26044
• Institution: Institut Teknologi Bandung

The vehicle insurance will be fascinating if the fixed premium in every year also considers the submitted claim. The system bonus malus is one of premium establishment method which accommodate it. This system discussed about premium establishment based on each policyholder's history claims. One of the advantages of bonus malus system is a premium reduction for policyholders which not submitting a claim and is called as a bonus. Meanwhile for policyholders which submitting a claim, there is a premium increase in next period, called as a malus. <br /> <br /> <br /> The distribution of claim frequency which is used in this study is mixed Poisson distribution with prior distribution of Gamma. Meanwhile, the claim severity is using mixed Exponential distribution with prior distribution is Gamma, such that the unconditional distribution of claim severity is Pareto distributed. By using Bayesian approach, the predictive distribution of frequency and severity claim will be negative binomial and Pareto distributed. <br /> <br /> <br /> The results of study is an optimal bonus malus table which containing pure premium and weighted premium. Pure premium calculation is based on the expectation of total loss, meanwhile the weighted premium is based on the probability of aggregate loss towards premium value. <br /> <br />