APPLICATION OF COPULA IN CLAIM RESERVE ESTIMATION FOR CORRELATED PAID AND INCURRED RUNOFF WITHIN GENERAL INSURANCE BUSINESS

General insurance companies need to set claim reserve in order to pay IBNR (Incurred But Not Reported) claims. Various methods have been developed to estimate claim reserve. The most commonly used method within long-tail insurance business is Basic Chain Ladder method which analyzed in Mack (1993...

Full description

Saved in:
Bibliographic Details
Main Author: Sugeng, Andreas
Format: Final Project
Language:Indonesia
Online Access:https://digilib.itb.ac.id/gdl/view/67454
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Institut Teknologi Bandung
Language: Indonesia
Description
Summary:General insurance companies need to set claim reserve in order to pay IBNR (Incurred But Not Reported) claims. Various methods have been developed to estimate claim reserve. The most commonly used method within long-tail insurance business is Basic Chain Ladder method which analyzed in Mack (1993). This method is popular as it is easily applied and require no assumption for distribution. The method can be applied to incurred claim’s runoff as well as paid claim’s runoff to obtain reserve estimate from each runoff. However, the method would often yields substantial gap between the produced reserve estimates when it applied independently for each runoff. One of the reason is that the method neglects correlation between the two runoffs. In this paper, a reserving method that considered correlation between incurred runoff and paid runoff will be developed. Copula will be applied to model the dependency of the two runoffs. Based on the dependency model, predictive distribution for claim reserve estimate could be determined with the help of Monte Carlo simulation. The predictive distribution then would be used to set the amount of reserve. Results from this method will be analyzed and compared to that of Basic Chain Ladder method. We found out that this method is able to reduce the gap between the reserve estimates if the Kendall’s tau correlation measure between the two runoffs is large.