APLICATION OF PEAKS OVER THRESHOLD APPROACH FOR PROBABLE MAXIMUM LOSS ESTIMATION ON VEHICLE INSURANCE CASE IN AUSTRALIA
Claims data may have extreme values depending on the types of the insurance businesses. In this Final Project, claims data are analyzed using the Extreme Value Theory (EVT) with the Peaks Over Threshold (POT) approach. With the POT approach, the excess data over a threshold ???? may be modeled by a...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/42461 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Claims data may have extreme values depending on the types of the insurance businesses. In this Final Project, claims data are analyzed using the Extreme Value Theory (EVT) with the Peaks Over Threshold (POT) approach. With the POT approach, the excess data over a threshold ???? may be modeled by a Generalized Pareto Distribution (GPD). The threshold may be determined using: the mean excess function plot; the stability parameter plot; and the Getensgarbe and Warner plot. In this Final Project, the methodology is applied to analyze a vehicle insurance data in Australia. It was found that the excess of the claims data over a threshold of $ 968,1935 follows a GPD with parameters ????? = 0.55274 and ????? = 880.71618. The Probable Maximum Loss (PML) of the vehicle insurance claims, with risk error limits of ? = 1%, 5% and 10%, are ????????????0,01=$ 509.819,????????????0,05=$ 206.710,8 and ????????????0,10=$138.875,5, respectively. These estimates of PML may be used in determining the maximum risk an insurance company is willing to cover. |
|---|