股市行业板块联动性分析——基于复杂网络和DCC-MIDAS模型

Co-movement Analysis of Stock Market Industries Sectors based on Complex Network and DCC-MIDAS Model

本文结合收益率联动网络以及最小生成树的行业定性分析以及网络中心性的行业定量分析,选取以化工、轻工制造和食品饮料为代表的第二产业以及以银行、房地产和非银金融为代表的第三产业两大行业组,运用复杂网络模型与DCC-MIDAS模型进行联动性实证分析。研究发现:(1)我国股市行业之间普遍存在联动效应,不同年份的不同行业网络联动效应不同;(2)中心性近似的行业其各自之间的长短期时变相关性均维持在较高水平,并且分别对组内另一行业具有较为一致的时变趋势;(3)时变联动性分析说明金融服务业间较之工业制造业间的联动关系将进一步增强。本文对复杂网络筛选得到的行业进行长短期联动性分析,有助于分析行业之间的长短期时变信息,进一步发掘新联动性行业与行业组,以及从不同效用期限出发制定相应的行业政策。

Combined with the industry qualitative analysis of yield co-movement network,the Minimum Spanning Tree,and the industry quantitative analysis of network centrality and centralization,the secondary industry represented by chemical industry,light industry manufacturing and food and beverage,and the representative industry represented by banking,real estate and non-banking finance were selected.The two major industry groups of the tertiary industry use the complex network model and the DCC MIDAS model to conduct an empirical analysis of co-movement.The research findings are as follows.1)There is a generally co-movement effect between the country's stock market industries,and the network co-movement effect of different industries in different years is different;2)Industry with the similar centrality maintained at a high level of long-term and short-term time-varying correlations,and they have a relatively consistent time-varying trend for another industry respectively;3)Time varying correlation analysis indicates that the co-movement between the financial services industry will be further streng thened com pared with the industrial manufacturing industry.In this paper,the long-term and short-term linkage analysis of the industries screened by the complex network is helpful to analyze the long-term and short-term time-varying information between industries,to further explore new linkage industries and industry groups,and to formulate corresponding industry policies based on different utility periods.