多元波动率模型的一些新进展  

Recent Developments in Multivariate Volatility Modeling

作　　者：王明进[1] 

机构地区：[1]北京大学光华管理学院,北京100871

出　　处：《数理统计与管理》2010年第2期232-247,共16页Journal of Applied Statistics and Management

基　　金：国家自然科学基金项目"基于数据降维技术的多元波动率模型(70671002)"的资助

摘　　要：对多个资产收益率的协方差矩阵建立动态模型是一个非常重要的问题。本文就近些年来该方面研究的一些主要进展进行了综述,特别地介绍了几种基于数据降维技术发展起来的能够适用于高维情形的多元GARCH模型,另外,对于多元波动率的模型诊断与比较方法以及条件协方差矩阵的预测等方面的研究成果也作了分析。Dynamic modeling for the conditional covariance matrix of a group of assets returns is crucially important for different financial decisions, such as optimal asset allocation, risk management and asset pricing etc. This paper makes an updated review on some recent developments in this field over the last years. In particular, some new multivariate GARCH models based on different dimensional reduction techniques, which are expected to be useful to high dimensional cases, are illustrated. Moreover, some newly developed model diagnostic methods and recent results on forecasting of covariance matrix are summarized as well.

关 键 词：波动率 多元GARCH 模型诊断 预测 

分 类 号：F830[经济管理—金融学] O212[理学—概率论与数理统计]