中文摘要：

主要研究了中国创业板市场与沪市日收益率间的相依关系。将时变 t-copula 模型和采用有偏误差分布的 ARMA-GARCH 模型结合起来构成了一个复合模型，并对所有的参数给出了贝叶斯估计。估计所得的两市场间 Kendall 秩相关系数的图形显示了创业板和主板市场之间确实存在着一种时变的正相依关系。进一步，对两市场在不同行业板块中日收益率的相依关系进行了类似的建模。结果显示两市场在工业板块中的相依结构与市场整体十分相近，而在信息技术板块中则存在着更强更稳定的正相依关系。

英文摘要：

The dependence of daily returns between the grow th enterprise market ChiNext and the Shanghai Stock Exchange （SSE ） in China was investigated .A composite model consisting of time-varying t-copula function and ARMA-GARCH models with skewed errors was built .Then full Bayesian estimators for the parameters were calculated .The shape of the temporal Kendall rank correlation coefficient of the two obtained markets indicates high positive and time-varying dependence between the two markets .Similar models were built for dependence structures in sectoral levels .It turned out that the dependence structure in the industrials sector is similar to the one in the entire market ,and that there was higher and more stable positive dependence in the information technology （IT ） sector .