您的位置:首页  > 论文页面

基于AR-GARCH-EVT模型的CVaR估计及应用

发表时间:2013-01-15  浏览量:2274  下载量:1620
全部作者: 谢绍魁,严定琪
作者单位: 兰州大学数学与统计学院
摘 要: 为解决金融资产收益序列数据的波动集群性和厚尾性,结合极值理论解决厚尾性的优势和广义自回归条件异方差(generalized autoregressive conditional heteroskedastiaty, GARCH)类模型解决集群波动性的优点,并且引入条件在险价值(conditional value at risk, CVaR)的概念,得到一般性的AR-GARCH-EVT-CVaR模型。通过实际计算和比较,发现提出的方法给出了比VaR更好的一天期的CVaR估计。
关 键 词: 概率论与数理统计;条件在险价值;极值理论;广义帕累托分布;风险控制
Title: CVaR estimation and application based on AR-GARCH-EVT model
Author: XIE Shaokui, YAN Dingqi
Organization: School of Mathematics and Statistics, Lanzhou University
Abstract: In this paper, in order to solve the volatility clustering and fat tails of financial asset sequence data, we combined the advantages of extreme value theory for solving fat tails and the advantages of generalized autoregressive conditional heteroskedastiaty (GARCH) class models for solving clustering volatility, and introduced the concept of conditional value at risk (CVaR). Then, we got the general model of AR-GARCH-EVT-CVaR. By actual calculation and comparison, we found that the proposed method gave a better 1-day CVaR-estimation than CVaR estimation.
Key words: probability theory and mathematical statistics; conditional value at risk; extreme value theroy; generalized Pareto distribution; risk control
发表期数: 2013年1月第1期
引用格式: 谢绍魁,严定琪. 基于AR-GARCH-EVT模型的CVaR估计及应用[J]. 中国科技论文在线精品论文,2013,6(1):36-41.
 
0 评论数 0
暂无评论
友情链接