The Application Of Bayesian Method In Motor Insurance

Motor vehicle insurance occupies an important position in China's property insurance.With the popularization and development of motor vehicles in China,vehicle insurance has gradually become the largest part of property insurance.For quite a long time,the insurance company has made a simple division of various risks in the determination of vehicle insurance rates.When specifically determining the pricing,it is roughly based on experience.Only the risks of the motor vehicle itself are taken into account,but the risk of the driver is not considered.On March 24,2015,China Insurance Regulatory Commission issued the“Working Plan for Deepening the Reform of the Commercial Vehicle Insurance Tariff Management System Pilot Program”,which proposed the schedule and blueprint for the reform of commercial vehicle insurance.Therefore,how to scientifically use risk pricing and rationally determine the insurance premium rate not only conforms to the trend of the times,but also plays a decisive role in the development of insurance companies.In non-life insurance actuarial,Bayesian method is applied in many aspects.Its basic principle is to use Bayesian theorem to make statistical inference on posterior information by combining sample information with prior information.In recent years,the rapid development of computer software technology has provided a simple and quick solution for solving complex Bayesian parameter estimation.The MCMC method is a dynamic simulation process,which is based on Bayesian theory.The Markov process is added to the stochastic simulation and then iteratively calculates to simulate the actual distribution of the variables.This paper mainly applies the Bayesian method to the analysis of the vehicle insurance premium determination case.We provide the claims data of a car rental company's ten-year car insurance.Then we use the premium insurance pricing method and Bayesian method,combined with the MCMC method based on Bayesian theory,to simulate the posterior estimate.After comparison and analysis,the model is proved to be feasible.By using the credibility theory,the problem of insufficient prior information such as missing partial claims and vehicle insurance claims data is studied.By using the MCMC stochastic simulation method based on Bayesian method,the number of claims is simulated using a negative binomial distribution function.A posteriori estimate of the number of claims and missing data was obtained,and the feasibility of the method was demonstrated through empirical analysis.