摘要
文章提出了Copula相依序列的概念,讨论了Copula相依序列的性质及与严平稳序列之间的关系。建立了一个时间序列的非线性模型——Copula自回归模型,给出了相应的参数估计方法、相依阶数的确定方法。并基于Copula自回归模型给出了一种点预测方法、一种区间预测方法以及一个时变Va R的估计方法。运用算例说明了Copula自回归模型及相关方法的有效性。
This paper presents the concept of copula dependence series and discusses the properties of copula dependence series and the relation between copula dependence series and strict stationary series. A new non-linear model about time series named copula auto-regression model(CAR) is established, and the method of parameter estimate and the determination method of the dependence order of copula auto-regression model is given. And based on copula auto-regression model, the paper also puts forward the point prediction method and interval prediction method as well as estimate method of time-varying Va R. Finally a practical example is employed to illustrate the validity of the model and proposed method.
引文
[1]Tsay R S.Analysis of Financial Time Series[M].New York:John Wiley&Sons,Inc.2002
[2]Nelsen R B.An Introduction to Copulas[M].New York:Springer,1999.
[3]Joe H.Multivariate Models and Dependence Concepts[M].London:Chapman and Hall,1997.
[4]卢颖,杜子平.基于pair-copula方法的高维相关结构构建[J].工业技术经济,2008,(15).
[5]黄恩喜,程希骏.基于pair copula-GARCH模型的多资产组合Va R分析[J].中国科学院研究生院学报,2010,(4).
[6]张高勋,田益祥,李秋敏.基于Pair-Copula模型的资产组合Va R比较研究[J].系统管理学报,2013,(2).
[7]李宁,程希骏.一种新的期货组合动态保证金设定模型与实证研究[J].中国科学技术大学学报,2012,(3).
[8]史道济等.基于Copula的股票市场Va R和最优投资组合分析[J].天津理工大学学报,2007,(3).
[9]刘志东.基于Copula-GARCH-EVT的资产组合选择模型及其混合遗传算法[J].系统工程理论方法应用,2006,(2).
[10]李述山.Copula回归及预测[J].统计与决策,2015,(18).