缺乏先验信息条件下的对数正态分布车险索赔额定价模型

An Automobile Insurance Pricing Model for Claims with Lognormal Distribution and without a Priori Risk Information

假定车险索赔额服从对数正态损失分布,并且其结构方差和过程方差存在显著差异,通过分析全体车险保单组合的历史索赔损失数据,估计出结构方差和过程方差的先验参数,从而得到个体保单未来索赔额的最优估计模型;在此基础上,给出了一个既考虑索赔频率又考虑索赔严重性的车险经验费率定价模型。

Suppose that the size of claims of automobile insurance follows a lognormal distribution,and there are significant differences between parameters of process variance and structure variance,by analyzing loss data of full automobile insurance policy,the process variance and structure variance parameters are estimated. Then,the optimal estimate model of future claim size of individual policy is obtained.Finally,an automobile insurance pricing model is established taking into consideration of both claim frequency and claim severity.