Multivariate Evolutionary and Tweedie GLM Methods for Estimating Motor Vehicle Insurance Claims Reserves

Abstract

An insurance policy is a contract agreement between the policyholder and the insurance company. For the contract agreement to run, policyholders need to pay premiums to insurance companies. On the other hand, the insurance company must underwrite the risk if the policyholder does submission of claims. It is necessary to estimate the reserves of claims for the company insurance accurately to prepare several funds for settlement of claim. Generalized Linear Model (GLM) can be used to estimate the claim values in a univariate form which only consists of 1 LoB (Line of Business). In practice, almost every insurance company has various types of LoB which depends on one another. Therefore, the GLM can be expanded to a multivariate GLM which can be used to estimate the claim data with more than one LoB. The researcher also wants to compare between an estimated reserve calculations of Swiss Re Group’s claims using the Multivariate Evolutionary GLM Adaptive Simple Method and GLM with the Tweedie Family Distribution Approach to find a more accurate method of finding claim reserves for each line of Swiss Re Group’s business data.