Hall分布族在巨额索赔数据的极值分位数估计与风险度量中的应用

The Application of Hall Distribution Class to the Extreme Quantile Estimation and Risk Measurement of the Data of Large Claims

基于指数回归模型构造了厚尾分布的一个子族——Hall分布族的极值分位数估计,并将该方法应用于巨额索赔数据进行了实证分析。然后,对巨灾保险进行了风险度量,得到了该数据的极值分位数,并作出了合理解释。

A high quantile estimator of a heavy-tailed distribution subclass Hall distribution class is proposed based on the exponential regression model. Empirical analysis is made by applying this method to large claim data. Risk of catastrophe insurance is measured. The high quantile estimation of the catastrophe insurance data is derived. Finally ,rational illustration for the result is presented.