Non Life Insurance Outstanding Claims Reserve Evaluation Research

Outstanding claims reserve is the most important part of the non life insurance reserve, so it is very important to evaluate the outstanding claims reserve. For non life insurance companies, the claims for the last year of an accident is usually unknown due to the long period of reserve assessment. Therefore, in the company's outstanding claims reserve assessment, we must consider various factors. Different assessment methods should be selected according to different data characteristics. In fact, in the actual insurance case, there may exist some human input errors, such as decimal point forward and backward and so on. All these will lead to great changes in the final claims. In addition, some non-human behaviors (such as natural disasters, credit fraud, economic crisis) cause the claim data to become suddenly larger or smaller. This will greatly influence the reserve evaluation results. Therefore, the commonly used deterministic models and methods for the assessment of outstanding claims reserve are firstly introduced. Starting from the model structure, assumption and data base, we comprehensively analyze the characteristics of the model. Secondly, taking case of compensation method as an example, we introduce the outliers. The residual box-plot method and two-point method are used to test the related claims data.Then a robust compensation method for the case of outliers is proposed.It modifies the choice of the progress factor and the closure rate data.The empirical proof shows that the case of robust compensation method can effectively identify and adjust the outliers and the final estimation results of claim amount is also more stable than that of traditional methods.On this basis, the stochastic method is introduced. Firstly, some stochastic models are introduced, and the advantages and disadvantages of each method are analyzed. Secondly, the outliers are introduced into the Double Generalized Linear Model. Considering the influence of outliers on the evaluation of the outstanding claims reserve, a robust estimation equation is proposed based on Quasi Likelihood Estimation Method. The Double Generalized Linear Model is established to identify and adjust the outliers. This helps the model get a more robust assessment results in dealing with outliers. Then, a deeper Double Generalized Linear Model is studied. It can adjust the development mode for the settlement of loss and improve the linear forecast of Double Generalized Linear Model. It introduces the internal factors and external factors that are not easy to be quantified into the model, so as to broaden the application scope of the model, and improve the prediction effect of the model. Finally, the R software is used to give an example analysis of the proposed model, and the related conclusions are obtained.