广义自回归条件密度(GARCD)建模为描述金融资产收益的概率密度函数提供了一种工具,这对于全面、准确把握金融资产收益的动态行为具有重要的意义.在一元GARCD-JSU模型的基础上,提出了多元GARCD—JSU模型并给出其向量表示;利用动态条件相关设定,给出多元GARCD—JSU模型的简化表示;接着给出了模型参数的三阶段极大似然估计方法和诊断检验方法.最后,对中国股市进行了实证研究.
Generalized autoregressive conditional density model provides a useful tool for simulating the probability density function of financial asset returns. It is important to describe the dynamic character of financial asset return comprehensively. Based on univariate GARCD-JSU model, the multivariate GARCD-JSU model has been proposed in the paper. First, we give the vector expression of the new model. Second, the reducing expression is obtained by dynamic conditional correlation method. Third three-stage maximum likelihood estimation and diagnosis test methods to the new model have been proposed. Finally, Chinese stock markets are selected for empirical anaIysis.