摘要
针对中国黄金市场收益率的尖峰厚尾和波动聚集以及长记忆性等特征,利用FIGARCH模型能准确处理波动异方差性和长记忆性以及EVT方法能准确估计尾部风险的优势,提出了基于FIGARCH-EVT的动态VaR和CVaR风险度量模型。采用上海黄金期货Au1107合约进行了实证分析,其结果表明,FIGARCH-EVT模型能够较好地处理黄金市场收益率的三大特征,而且比GARCH-EVT模型风险度量的VaR和CVaR更精确。
In view of the characteristics of sharp peak and heavy tail, volatility clustering and long memory in China gold market returns. Since FIGARCH model can accurately handle volatility clustering and long memory and EVT can accurately estimate the tail risk, the paper proposed a dynamic VaR and CVaR risk measurement model based on FIGARCH -EVT. An empirical analysis was made by the Shanghai gold futures contract Au1107, the results show that FIGARCH - EVT model can handle the three characteristics of the gold market return and the VaR and CVaR are more accurate than measured by the GARCH -EVT model.
出处
《黄金》
CAS
2016年第6期8-12,共5页
Gold
基金
国家统计局科研项目(2014LZ25)