用矩匹配方法计算非线性风险度  

The Moment-matching Method to Calculate Nonlinear VaR

作　　者：刘汉中[1] 田新时[2] 

机构地区：[1]湖南商学院经济学系,湖南长沙410205 [2]华中科技大学经济学院金融系,湖北武汉430074

出　　处：《运筹与管理》2005年第1期90-94,共5页Operations Research and Management Science

摘　　要：矩匹配方法是用来求解非线性风险度(ValueatRisk,简称:VaR)的一种普遍性方法,它是先假定样本经验分布服从已知分布族,然后运用矩匹配估计方法估计相应的参数,得到资产回报样本的密度函数,再计算风险度VaR;本文采用的Johnson分布族是矩匹配方法的直接应用,并且计算出来的结果与局部Monte Carlo结果进行了比较,并通过实证分析认为这种方法是一种良好的计算非线性VaR方法。The moment-matching method is common to calculating the VaR(Value at Risk) of the portfolio including nonlinear positions.It is based on approximating the distribution of return by finding a distribution whose quantiles can be calculated. In this paper, the Johnson distributions family is applied to the method and fits the return data. The result is compared with the partial Monte-Carlo simulation method. Our simulation result and empirical studies confirm the newly proposed method.

关 键 词：Johnson分布族 总体矩 矩匹配方法 风险度 

分 类 号：O212[理学—概率论与数理统计] F830.9[理学—数学]