Prediction Model Of The Total Losses Of Regional Non-Operating Small Vehicles

With insurance industry in China gradually joining in the international markets, competition in it is getting tougher. Scientifically rating of the vehicule insurance is the main approach of increasing competition capability of vehicle insurance product. That is to apply actuarial technology in a more effective way and design products according to area characteristic.This article mainly focuses on problems in the loss prediction of vehicles, based on historical statistics of the loss, by the means of the combination of judgment forecasting and statistical forecasting, a model of research on the losses of non-operating vehicles within the area is suggested.While evaluating statistically the losses of the vehicle, by using the collective risk model and the Bayesian Estimation, the frequency and the severity could be estimated. Therefore, the total losses could be calculated through integrating those two variables. At the same time, a risk assessment on all the vehicles within the area should be carried out, and the weight of different loss factors should be evaluated through analytical hierarchy process. Later, according to the factors that could possibly change rapidly, a correction on the total losses evaluation could be finished.Thanks to this model, an estimation of greater precision during the next year could be completed. The estimation lays the foundation for rating of the vehicle insurance. At the end, the feasibility of this method will be discussed in details by using the example of Shenyang area.