您的位置:首页  > 论文页面

基于条件Copula-GARCH模型的VaR估计

发表时间:2010-01-15  浏览量:1870  下载量:708
全部作者: 段福林,李元
作者单位: 广州大学数学与信息科学学院
摘 要: 针对传统分析在险价值(value at risk,VaR)模型的不足,结合Copula技术和GARCH模型,提出了条件Copula-GARCH模型。该模型不仅能够得到金融市场间的非线性相关性,而且可以得到更灵活的多元分布,进而用于资产投资组合VaR分析。在详细探讨了条件Copula-GARCH模型的估计方法后,运用具有不同边缘分布的条件Copula-GARCH模型对上海股市和深圳股市进行了实证研究,研究表明:对于不同的阿基米德Copula,利用Frank Copula和Gumbel Copula可以比较准确地估计VaR.
关 键 词: 应用数学;GARCH 模型;条件Copula;在险价值;多重分布函数边际分布推导法
Title: Estimation of VaR based on conditional Copula-GARCH model
Author: DUAN Fulin, LI Yuan
Organization: School of Mathematics and Information Sciences, Guangzhou University
Abstract: By combining Copula techniques with GARCH model, conditional Copula-GARCH model is provided to avoid defects of classical risk analysis models. Not only the non-linear dependence between financial markets is able to be caught, but also the more flexible multivariate distribution which can be use to analyze portfolio VaR is able to be obtained from this model. The estimation of conditional Copula-GARCH model is fully discussed in this paper. The empirical results getting from Shanghai and Shenzhen stock markets indicate that employing Frank Copula and Gumbel Copula can estimate VaR more accurately than Clayton Copula.
Key words: applied mathematics; GARCH model; conditional Copula; value at risk; multi-distribution function; inference functions for margins
发表期数: 2010年1月第1期
引用格式: 段福林,李元. 基于条件Copula-GARCH模型的VaR估计[J]. 中国科技论文在线精品论文,2010,3(1):41-46.
 
0 评论数 0
暂无评论
友情链接