基于Copula-GARCH类模型的证券分类方法

发布时间：2018-04-24 17:08

本文选题：证券分类 + 模糊聚类分析　；参考：《系统工程》2017年04期

【摘要】：本文尝试提出一种全新的证券分类方法:即运用Copula-GARCH类模型来计算两两证券价格序列间的静态Copula系数、以测算其同期相依性,并据此构建各样本间的相似矩阵,再根据相似矩阵对相关证券进行模糊聚类分析,从而对相关证券进行分类。本文所提出的证券分类方法不以有效市场假说为基础,比默认有效市场假说的证券分类方法更具普适性,更贴近于证券市场的实际情况、能更真实地反映证券市场的内在机理和波动特征。本文的实证研究,量化测定了中证100指数成份股的同期相依性,并据此对中证100指数的成份股进行分类,印证了本文所提出的基于Copula-GARCH类模型的证券分类方法的有效性和实际价值。
[Abstract]:In this paper, a new security classification method is proposed: the Copula-GARCH model is used to calculate the static Copula coefficient between pairwise securities price series, to calculate its dependence in the same period, and to construct the similarity matrix among the samples. Then the related securities are classified by fuzzy clustering analysis according to the similarity matrix. The securities classification method proposed in this paper is not based on the efficient market hypothesis, which is more general than the default efficient market hypothesis, and closer to the actual situation of the securities market. It can more truly reflect the internal mechanism and volatility characteristics of the securities market. The empirical research in this paper quantifies the dependence of China Securities 100 Index component stock, and classifies the component stock of China Securities 100 Index. The validity and practical value of the proposed securities classification method based on Copula-GARCH model are verified.
【作者单位】：复旦大学经济学院;
【分类号】：F832.51

【相似文献】