Improved Payments Per Claim Method And Its Application In Non-life Insurance

In recent years,human and natural catastrophic losses occur frequently and bring new challenges and opportunities to the development of non life insurance companies.In order to accurately estimate the benefits of insurance companies,we should estimate the amount of money that should have been claims but not claims.In order to ensure the accuracy of the non life insurance reserve estimation,this paper first uses the residual box line graph to find the abnormal value in the reserve and adjusts the abnormal value,and then uses a certain method to evaluate the non life insurance reserve.The general evaluation methods include the chain ladder method,the case average indemnity method,the reserve progress method,etc.Although these methods are all indeed of the method It is feasible,but because of the deterministic methods,the stochastic factors are not taken into account,and the accuracy of estimation is not high.With the development of China's insurance business,a number of new random methods have been introduced,such as dynamic linear method,nonlinear stratification method,robust Bayesian estimation and so on.It is worth affirming that the stochastic method improves the accuracy of the estimate of non life insurance reserve,but it is not enough because the amount and case of claim are not taken into account.The correlation between the number of years of development varies,so the accuracy is not perfect enough.The improved case claims method can solve the above problems and make the accounting method of non life insurance reserves more practical and scientific.The innovative research of this paper is mainly shown in two aspects.Firstly,this paper uses the existing fixed data to identify the annual claim amount and the abnormal value in the number of cases.If there is an abnormal value in the data of the first column of the flow triangle,then the outliers in the first column are replaced by the median of the first column data.After the adjustment of the first column of data,there is no exception in the first column of data.We calculate the new progress factor based on the first column data of the number of incremental cases,and then calculate the number of new increments in the case number of the cases which have already been reported,and the same amount of the claim amount is the same.The method takes the place of outliers,and combines them to get the outliers of traffic triangle.Secondly,The claim method uses the information in the data to take into account the number of cases and the two factors of the amount of claims.It only takes into account the advantages of the amount of claim amount than the chain ladder method and the reserve claim method.However,the case average claim method does not take into account the correlation of the year of the case and the correlation of the year of the claim.In view of this shortcoming in the case claims law,this paper proposes an improved case claim method.The improved method takes into account the correlation of the progress year through the correlation Bayesian model,which is not only considering the correlation between the amount of claims,but also considering the correlation of the number of cases to improve the prediction accuracy of the non life insurance reserve.The paper makes the corresponding numerical experiments on the abnormal value adjustment method of the non life insurance reserve and the improved case claim method respectively.The numerical results verify the feasibility and accuracy of the abnormal value adjustment method and the improved case claim method.Finally,the paper summarizes the research work and proposes further research directions.