CLAIM RESERVE ESTIMATION FOR CORRELATED PAID AND INCURRED RUNOFFS IN GENERAL INSURANCE BUSINESS USING COPULA

• Main Author: Abraham Nugraha, Natanael
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/67450
• Institution: Institut Teknologi Bandung

The main topic in this final project is a claim reserve estimation method that uses copula. Copula is a tool which is used to model the dependency of two or more random variables, in this case they are incurred claims and paid claims. This method will be used if there is a strong dependency between incurred claims and paid claims. In this final project, the explanation about Copula focuses on the Archimedean copula family, such as Gumbel, Clayton, and Frank copula. Furthermore, the modelled dependency will be used to perform a Monte Carlo simulation. In the simulation, we will generate data of incurred runoff triangle using information from paid runoff triangle. To estimate the claim reserve from the generated runoff triangle, the Basic Chain Ladder method will be used. Using Monte Carlo simulation will result in having a predictive distribution for the claim reserve estimate. This makes this method different from other reserve estimation methods, such as Basic Chain Ladder and Munich Chain Ladder. In practice, predictive distribution for the claim reserve estimate gives a flexibility in decision making.