A Markov chain based No-claim discount system

Abstract

This paper attempts to inspect the desirability of multilayer premium system, popularly known as No-Claims-Discount (NCD) system, particularly in the area of Automobile Insurance. The discrete time parameter Markov chainsare used as the basic tool of simulation in the analysis as the basic framework of the study. The movement by the drivers between the various levels of discounts in the NCD system is discussed under various scenarios. There is also an attempt to evaluate the reflection of the distribution of a particular group over the various levels of premium in the long run. In turn, a model is developed to evolve a NCD system as a fair system so as to cover the expected claim amount from a group by the premium collected from that particular group only.