Research On Customer Lifetime Value—Based On The Adjustment Of "Beta Factor" About Compensate Risk

Utilize Customer Lifetime Value(CLV)as an indicator to measure the value contribution of auto insurance customers,and conduct effective renewal business management,thereby transforming auto insurance short-term transactions into long-term value,which has far-reaching significance in the future development of auto insurance business.Therefore,in order to distinguish between different value customers,it is very important to establish a precise customer value evaluation model in the customer value division.First,on the basis of the traditional CLV model,this paper draws on the method of risk correction of individual assets in the financial sector,introduces the customer compensation risk factor?_c as an adjustment factor of the discount rate,and adjusts the customer discount value by modifying the risk discount rate,thereby creating RCLV(Risk-adjust Customer Lifetime Value)model.The RCLV model has greater risk tolerance and be more flexible.The purpose of creating RCLV is to solve the problem that the CLV model cannot distinguish the potential high-value customers under the same profit cash flow level created by customers with different compensation risks.This is the core idea of this paper.Second,the RCLV model focuses on how to determine the form of the coefficient?_c.This paper defines the customer compensation risk as the ratio of the single customer's loss rate to the overall customer's loss rate,and actively explores the degree of interpretation of human factors such as the NCD coefficients of Motor Vehicle Insurance,NCD Coefficients of Motor Vehicle Accident Liability Compulsory Insurance,and limitation of Third-party Liability Insurance of Motor Vehicles.Finally,based on the situation that the individual claims ratio is consistent with the cumulative loss trend,this paper refers to the method of quantifying the cumulative loss using the exponential distribution family model.Due to the great influence of the deductible and NCD system on the claim,the probability of the zero-point claim expansion was expanded.In the selection of the quantification model of?_c,a zero-adjust regression model in the mixed distribution was used to make a reasonable predic tion of the individual loss ratio.In the empirical part,this paper uses the auto insurance actuarial data of a property insurance company in Beijing in 2015 to perform the prediction and assessment of expected NCD coefficient,continuous renewal rate an d individual loss ratio(including claims probability).The empirical results show that the impact of human factors on individual claims ratio is relatively greater than the vehicle factors.At the same time,a typical sample is taken for RCLV measurement.The larger the coefficient?_c of a customer,the greater the risk discount rate and the more adjustment to CLV.Comparing this result with the original CLV model,it shows that the established RCLV model can indeed avoid the misjudgment of potential high-value customers.