风险价值(VaR)计算方法的实证分析  

An empirical analysis of value at risk(VaR)

作　　者：杨杨[1] 徐文彬[1] 

机构地区：[1]北京信息科技大学经济管理学院,北京100192

出　　处：《北京信息科技大学学报（自然科学版）》2013年第4期83-87,共5页Journal of Beijing Information Science and Technology University

基　　金：北京市哲学社会科学规划基金项目(11JGB057)

摘　　要：选取101组上证综合指数等数据,用历史模拟法、方差—协方差法和蒙特卡罗模拟法,对风险价值(VaR,value at risk)进行估算。用3种方法估算出的VaR值表明:在样本数据服从正态分布的条件下,3种计算方法均能得出可靠的风险值,且运用风险值所做的预测可以较好地拟合实际值,3种计算方法得出的VaR值差距可以忽略不计。尽管不同方法各有特点,但无论运用哪种风险值计算方法,对于我国金融机构和投资者以及金融监管部门来说,都有利于我国金融市场风险的管理和控制。Based on the three basic methods including Historical Simulation Method, Variance- covariance Method and Monte Carlo Simulation Method, Shanghai composite index is selected to calculate the ＂value at risk （VaR）＂. The comparison of the results under the three methods shows that under the condition of the sample data obey normal distribution, reliable results can be achieved and the forecast matching degree between the calculated VaR and actual values is high. In addition, the difference of the results can be neglected. Although different methods have their characteristics, suitable VaR calculation method should be used by risks management ,financial institutions ,investors and financial regulators.

关 键 词：历史模拟法 方差-协方差法 蒙特卡罗法 VAR值 上证综合指数 

分 类 号：F830[经济管理—金融学]