COMPARISON OF GENERAL INSURANCE CLAIMS RESERVING METHODS IN MALAYSIA: BASIC CHAIN LADDER METHOD VS. OVER-DISPERSED POISSON MODEL

Abstract

The adoption of stochastic claims reserving methods has become increasingly prevalent in the General Insurance (“GI”) business to quantify uncertainty. The research study will compare two methodologies: the Basic Chain Ladder (“BCL”) method, which is a deterministic approach, and the Over-Dispersed Poisson (“ODP”) model, which is a stochastic approach. These methodologies provide varying levels of accuracy in estimating claims reserves. Nonetheless, the absence of a thorough and comparative examination regarding implementing these methods within Malaysia’s GI sector, encompassing both conventional and Takaful insurance, provide a clear understanding of which model is suitable for distinct operational contexts and objectives within the compliance framework of GI companies. With a data analysis between a general conventional insurer and a general Takaful operator in Malaysia, the study has found that the hypotheses tested in this study have significant differences. Both methods can easily approach the prediction reserves. However, the ODP model provides statistical information, for example, prediction error (also known as root mean square error, which is also discussed in this research), an advantage the BCL does not have. This research will give decision-makers the information to make educated choices in selecting claim reserves prediction methods.