摘要
DCC(dynamic conditional correlation)模型能够简洁地刻画投资组合中资产相关性的动态结构.本文通过引入正态Copula函数,使用灵活的边缘分布来取代DCC模型中的正态假设,将DCC模型推广到边缘分布可为任意的一般形式.最后我们利用推广后的DCC模型对我国的股市进行了实证研究.
The DCC(dynamic conditional correlation) model can describe the dynamic dependency structure for a large class of assets.In this paper,by introducing the Normal-Copula function,we relax the normal assumption on marginal distributions and generalize the model to the case of arbitrary marginal distributions.By applying this generalized model to Chinese stock market as the empirical study,the effectiveness of the propose model is verified.
出处
《工程数学学报》
CSCD
北大核心
2011年第1期21-27,共7页
Chinese Journal of Engineering Mathematics
基金
中国科学院知识创新工程重要方向项目(KJCX3-SYW-S02)~~