PURE PREMIUM MODELING OF VEHICLE INSURANCE BASED ON NUMBER OF CLAIMS HAVING POISSON DISTRIBUTION

• Main Author: APRINALDY (20815006), DEDI
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/21622
• Institution: Institut Teknologi Bandung

At vehicle insurance companies, the determination of an appropriate pure premium will makes The business runs well. in this research, modeling process of claim frequency data is carried out by considering Characteristics of policyholders such as sex, marital status, car engine capacity, age of policy holder, and age of car. Data that used in this research is non bus and non truck vehicle insurance data that filed claims during 2013 in a general insurance company. <br /> <br /> <br /> <br /> In this research, researcher uses poisson distribution that its parameter influenced by characteristics factor and claim frequency in the previous period that have character markovian, number of claims which happen in the future depend on number of claims on the present but doesn’t depend on number of claims on the past, to calculate claim frequence in the next period. <br /> <br /> <br /> <br /> Joint distribution for this model is known by Multivariate Poisson with Artificial Marginal (MPAM) that is used to estimate model parameters. The parameter values of this model are estimated use Maximum Likelihood Estimation Method. Furthermore, The estimation result of those parameters is used to calculate pure premium in the next period.