极值理论在风险度量中的应用——基于上证180指数  

The Application of Extreme Value Theory to Risk Measurement Based on SSE-180 Index

作　　者：田新时[1] 郭海燕[1] 

机构地区：[1]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》2004年第1期106-111,共6页Operations Research and Management Science

摘　　要：精确度量风险是金融风险管理的关键问题。本文引入广义帕雷托分布代替传统的正态分布等,精确描述金融收益的厚尾特征。并将基于广义帕雷托分布的VaR模型和其它模型方法,如GARCH(1,1)、GARCH(1,1)-t、历史模拟法、方差-协方差方法,进行比较分析。实证研究表明,基于广义帕雷托分布的VaR模型比传统的模型方法更适合厚尾分布高分位点的预测,并且其预测结果比较稳定。这使得基于广义帕雷托分布的VaR模型成为VaR度量方法中最稳健的方法之一。The accurate estimation of Value-at-Risk(VaR) is essential for risk management.In this paper,we use the General Pareto Distribution(GPD)instead of Normal distribution to describe the heavy-tailed characteristic of financial time series.We compare this model with other well-known model such as GARCH(1,1),GARCH(1,1)-t,variance-covariance method and historical simulation.Our studies indicate that the model based on GPD fit better than traditional methods for heavy-tailed distribution on forecasting high quantiles, furthermore, the forecasting results of this model is more stable. This shows that this model is a robust tool to forecast quantiles,which is practical to implement and regulate for VaR measurements.

关 键 词：极值理论 风险度量 金融风险管理 VALUE-AT-RISK GARCH模型 

分 类 号：F830[经济管理—金融学] F224.7