Non-life Reserve Progress Method With Randomly Selection Of Factor

The liability reserve evaluation has been an important issue in non-life insurance company. reserve progress method is an important method. But future reserve progress rate and payment rate of reserves progress method are determined values, which makes the selection of these two factors is subjective. Ultimately the outstanding claims reserve is only a point estimate value, which can’t provide the information of reserve fluctuation. In order to measure reserve uncertainty, this paper presents the progress rate and payment rate randomly selected progress method.Future reserve progress rate and payment rate are as random variables,they were randomly selected from the distribution of the progress rate and the payment rate in corresponding progress year, respectively as future payment rate and payment rate, using the Monte Carlo simulation to obtain the progress rate and payment rate numerous traffic triangle, then get numerous outstanding claims reserves by progress method, and ultimately get the reserve standard prediction error and forecasting distribution. Making progress method has the characteristics of randomness, thereby improving the functionality and usability of the progress method.In the random selection method based on factors, if historical progression rate and payment rate are discrete distribution, called the random selection method based on discrete factor; if historical progression rate and payment rate are continuous distribution, called the random selection method based on continuous factor, this method uses a kernel density estimation, kernel density estimation is used to estimate the density function of a given sample points, the method does not use a priori information and the distribution of the data does not make any assumptions, it studies date distribution starting from the data itself, which is a typical non-parametric estimation methods.