利用高频数据和copula度量资产组合市场风险:建模与实证  

Measuring Portfolio Market Risk Using High-frequency Data and Copula:Modelling and Empirical Research

作　　者：唐振鹏[1] 黄友珀[1] 许雅妮[1] TANG Zhen-peng HUANG You-po XU Ya-ni(School of Economics and Management, Fuzhou University, Fuzhou 350116, China)

机构地区：[1]福州大学经济与管理学院,福建福州350116

出　　处：《运筹与管理》2016年第5期174-179,共6页Operations Research and Management Science

基　　金：国家自然科学基金资助项目(71171056);福建省社科基金重点资助项目(2013A017);福建省高校新世纪优秀人才支持计划项目(JA11025S)

摘　　要：利用realized GARCH模型纳入高频信息,通过copula函数描述资产收益之间的复杂相依关系,构建资产组合市场风险度量的copula-realized GARCH模型,并以上证综指和深证成指的等权重组合进行实证分析。实证结果表明,用学生t-copula函数描述上证综指和深证成指收益序列之间的相依关系比较恰当,与常用的GARCH模型和GJR-GARCH模型相比,学生t-copula与realized GARCH相结合的模型提供相对可靠、更具效率的VaR估计以及更准确的ES估计。因此,资产组合市场风险度量有必要同时考虑高频信息和复杂相依关系。Using copula to capture complex dependency structure between asset returns and incorporating high-frequency information by realized GARCH model, we develop a new portfolio market risk measuring model whichcombines copula function and realized GARCH model. An empirical analysis is carried out using a constructedportfolio with equal-weighted Shanghai composite index and Shenzhen component index. The results show thatstudent t-copula fits better for dependency relations between return series of Shanghai composite index and Shenz-hen component index. Compared with models linking commonly used copula and GARCH or GJR-GARCH, theproposed student t copula-realized GARCH model provides more reliable and efficient value at risk and expectedshortfall estimation. Therefore, we should consider both high-frequency information and complex dependency inmeasurement of portfolio market risk.

关 键 词：金融工程 市场风险度量 高频信息 复杂相依关系 realizedGARCH模型 

分 类 号：F832.5[经济管理—金融学]