TOTAL LOSS ESTIMATION USING MIXED GAUSSIAN COPULA MODEL OF VEHICLE INSURANCE
The total loss estimation in the insurance portfolio can be determined based on the average claim size and the frequency of claim. Average claim size and frequency of claim are two random variables that have a relationship. So, these two random variables will be passed to get a jo...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/29501 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | The total loss estimation in the insurance portfolio can be determined based on the average claim size and the frequency of claim. Average claim size and frequency of claim are two random variables that have a relationship. So, these two random variables will be passed to get a joint distribution model using the mixed copula model. There are various covariates that affect the average claim size and frequency of claim. The covariates used consist of: sex of policyholders, marital status of policyholders, employment status, type of vehicle, size of vehicle, state code, and location code. Based on the covariates, policyholders data can be grouped into policy groups classified on the basis of homogeneity of their covariates. The marginal model of each variable is modeled by the Generalized Linear Model (GLM) Gamma for average claim size and Zero-Truncated Poisson GLM for frequency of claim. The mixed Gaussian copula is used to get the joint distribution of average claim size and frequency of claim. Using the mixed Gaussian copula, we can get the expected total loss. In this study, we apply the model to the data of policyholders in one of the vehicle insurance company in United States for the period of 2011. |
|---|