Stochastic reserving using individual claims data in non-life insurance

• Main Author: Ang, Siew Ling *
• Format: Thesis
• Published: 2019
• Subjects: HG Finance; QA Mathematics
• Online Access: http://eprints.sunway.edu.my/2368/

In the thesis, numerous methods are proposed to estimate the reserve in non-life insurance. The claims data of individual customers are used. These data include the sum insured, the claim and payment records until the present time and the outstanding claims liabilities. The multivariate power-normal mixture (MPNM) distribution is used to overcome the problem posed by the large number of zero values in the individual payment amount and outstanding amount. The performance of the MPNM distribution is found to be better than that of the chain-ladder procedure for the dataset under consideration. The mixture of multivariate power-normal (MPN) distributions and a degenerate distribution is used to overcome the problem posed by the occasional extremely large claims. The extra information given by the subclasses of insurance is used to improve the estimation of the reserve. The methods proposed are found to be also applicable to the end of year individual data.