基于Copula的财险公司经济资本计量

作为预防非预期损失的有效方法,经济资本管理逐渐成为资本管理的新趋势。本文以中国人民财产保险股份有限公司2005-2020年上半年五条业务线的赔付率为研究对象,首先考察了业务线赔付率呈现出的不同水平和其随时间的变化规律;然后为了计量经济资本,用核密度函数拟合各业务线的边缘分布,从单条业务线出发,分别运用VaR模型和TVaR模型计量了每条业务线所需要的经济资本;再逐渐拓展到机动车辆和意外险这两条业务线,考虑他们之间的相依性,选用Gumbel Copula模型和蒙特卡罗方法计量经济资本,并与独立情形的结果进行对比和分析。

As an effective way to prevent unexpected losses,economic capital has gradually become a new trend of capital management.Taking the loss ratio of five business lines of PICC from 2005 to the first half of 2020 as the research object,this paper first investigates the different loss ratio levels of business lines and their changing rules over time;then,in order to measure economic capital,it uses kernel density function to fit the marginal distribution of each business line,starting from a single business line,it uses VAR model and TVAR Model respectively,and then gradually expands to the two business lines of motor vehicles and accident insurance.Considering the dependence between them,Gumbel copula model and Monte Carlo method are used to measure the economic capital,and the results are compared and analyzed with the results of independent cases.