中文摘要：

传统的copula模型在对二维以上相依结构建模时存在参数过少的缺陷，vine copula理论基本弥补了这一缺陷。介绍了vine copula理论以及其相对于传统多元模型的优势，尤其提出了vine copula对于时长不一致的数据进行建模具有数据利用率较高的特性，给出了这类数据vine copula的建模步骤以及基于极大似然估计的统计推断。最后对国内A股市场的五种金融股票的联合分布进行建模，并利用蒙特卡罗方法对资产组合的VaR进行了模拟。

英文摘要：

Vine copula precedes traditional copula in that it will contain more parameters. We introduces vine copula theory, especially for two particular regular vine： C-vine and D-vine. Applying vine copula to model financial data with different size will raise data utilization than traditional models. We expressed the procedure of vine copula modelling and the statistics inference based on Maximum Likelyhood. In the last two section, we show results of empirical study for five finance related stocks and simulate the VaR using MCMC.