Research On Ratemaking Optimization Of Automobile Insurance Based On The Risk Characteristics

With the development of economy and the improvement of living standard, the numberof residents who own automobiles has been increased year by year. Therefore, the motorvehicle insurance has got a rapid development which has played an important role in thepremium income of property insurance. Although the insurance market has been developed,the situation of automobile insurance is not optimistic that is reflected in following threeaspects: the payment rate is higher, the rate making is unreasonable and the loss risk control isinadequate. So it is an important problem for the development of auto insurance to identifyrisk factors scientifically, improve risk technology methods creatively and make rate for autoinsurance reasonably, all of which has a great significance for the development of Chineseauto insurance market.On the theoretical basis of the economics, statistics and mathematics, the paperintroduces the basic principles of auto insurance pricing and risk characteristics. According tothe characteristics of risk factors faced by automobiles, the paper classifies the riskscientifically. The paper analyzes and discusses the compositions, influence factors, ratemaking methods and optimization of vehicles insurance rate systematically by usingqualitative and quantitative analysis, theoretical research as well as empirical study. The paperadds random effects to generalized linear models to form the generalized linear mixed models.In the part of empirical study, the claim intensity and claim frequency are analyzed by usinggeneralized linear mixed models. At last, the paper discusses the advantages anddisadvantages of rate making system and points out some practical problems and alsoproposes some specific suggestions. The rate making of vehicles insurance is a developingprocess. The paper pays more attention to discuss the risk factors and apply model systemswhich is an exploratory attempt to optimize the rate making methods.