The Research Of Multi-factor Pricing Model Under The Marketization Reform Of China's Commercial Auto Insurance Rates

Since the reform and opening-up,the Chinese economy has been growing rapidly over the past decades.In the positive economic environment,the number of vehicles continue to escalate,leading to the business growth of auto insurance.As a main business line of property insurance companies,auto insurance plays a big role in the annual profit of those companies.Since the new round of commercial vehicle insurance reform was launched in June 2015,by March 8,2018 three reform measures had been enacted by the insurance regulator.To carry out the reform incrementally,the marketization on commercial auto insurance rate and terms were tentatively introduced in the provinces,which gives insurance companies the autonomy to adjust their premium rates within a given range.Nonetheless,the reform further increases competition between insurance companies.A majority of companies either try to reach clients by lowering the premiums or expand channels of insurance agents by paying higher service fees,ignoring the risk that is faced by clients and enterprises themselves,which has led to the annual loss of most small enterprises.Therefore,it is imperative that insurance companies develop a more scientific and reasonable pricing model for auto insurance according to different degrees of client risks.Involving theoretical knowledge of insurance,statistics,actuarial science,and economics,the research in this paper discusses the utilization of the generalized linear model in determining commercial auto insurance premiums with methods of quantitative analysis,comparative analysis,and empirical research.Starting with an introduction to the previous reforms and current situation of commercial auto insurance premiums,it firstly analyzes the problems in China's commercial auto insurance premiums,which is followed by detailed information of the generalized linear model and minimum deviation model as well as a summary of their advantages and disadvantages.Then,it describes the process of using the generalized linear model to determine auto insurance premiums and explains how to manage each section of the process.In the most important part of this paper—the empirical analysis section,the generalized linear model is built based on the frequency and severity of claims,which is later used to do an empirical analysis on the data of vehicle damage claims of a Chinese insurance company in 2017.Finally,it summarizes the results of the empirical analysis,explores the possibility to use technological means to develop new products of vehicle insurance and improve the accuracy of premium determination,and offers policy recommendations.