Research On An Application Of GLM Model To Automobile Usage-Based Insurance Ratemaking

China's auto industry is in a phase of rapid development and the auto sales has ranked first in the world for many years.China's car ownership has grown in recent years and become the world's second-largest car holding country after the United States.It is expected to be the world's largest car ownership by 2019.Because of the rapid development of auto industry and car ownership in C hina,the corresponding auto insurance market is developing rapidly.However,with the emergence of new technologies such as telematics and big data,new automobile insurance ratemaking models get breakthroughs in technology and innovation to achieve.In addition,with the development of car insurance rate reform,C hina's current automobile insurance ratemaking models can not meet the requirements of the new situation,accelerating the auto insurance rate differentiation and personalization has become the inevitable trend in the development of the insurance market.In recent years,domestic and foreign insurance companies are actively promoting the new motor vehicle insurance products of UBI(Usage-Based Insurance),and the risk level of driving behavior is determined based on the analysis of driving behavior and driving range.UBI can help to provide fair and reasonable car insurance rate and differentiated and individualized car insurance policy and value-added service for auto insurance policyholders.This study attempts to identify the risk level of UBI car insurance and to grade driver's d riving behavior by applying Generalized linear model.O n the basis,the rates of auto insurance are determined.The main contents of this article are as follows:(1)Firstly,the current patterns of the rate making of auto insurance are introduced and the advantages and disadvantages of the traditional methods are analyzed respectively.Then the developments and applications of the vehicular networking technology and theory,as well as the actuality and development trend of UBI are introduced.On the basis of this,the contents and ways of the effects of driver's driving behavior on driving safety or risk are discussed.(2)In the environments of the vehicular networking,this study establishes a model by using the data of driving behavior and police's information by applying Generalized linear models,based on the study of driving behavior influence on risk.The scores of each type of driving behavior are obtained.Furthermore,the models of ratemaking are established by using Generalized linear models again and adding the scores obtained above as a new factor,together with traditional pricing factor.So the discount factors and the pure premiums are determined.