基于贝叶斯方法的保费计算的稳健性质(英文)  

The Robustness for Premium Calculations Using Bayesian Approaches

作　　者：吴贤毅[1] 

机构地区：[1]华东师范大学金融与统计学院,上海200241

出　　处：《江西师范大学学报（自然科学版）》2015年第2期177-188,共12页Journal of Jiangxi Normal University(Natural Science Edition)

基　　金：国家自然科学基金(71371074);上海市哲学社会科学基金(2010BJB004)资助项目

摘　　要：探讨了3类关于保费计算原理的相关问题.首先,结合贝叶斯方法和损失原理定义了贝叶斯保费;然后,研究了2类保费计算原理的稳健性质问题:带任意污染系数的非贝叶斯保费的稳健性质和基于ε-污染方法讨论的贝叶斯保费关于先验分布的稳健性质;最后,运用Esscher保费原理分析了当污染在某个分布类变化时保费对污染的响应以及保费的值域.Three tightly related problems regarding the premium calculation principles are considered. Firstly,Bayesian premiums are defined by using of Bayesian approaches associated with loss principles. Then,two problems regarding the robustness of premium calculation principles are investigated. One is the robustness of nonBayesian premiums with respect to arbitrary contaminations. The other one is the robustness of Bayesian premiums with respect to the prior distributions by means of the ε-contamination arguments. Finally,the reaction of a premium with respect to the contaminations and the range of premium using the Esscher principle when the contamination distribution varies in a distribution class are discussed.

关 键 词：稳健性质 保费计算原理 ε-污染 贝叶斯方法 Esscher保费原理 损失原理 

分 类 号：F840[经济管理—保险] F224