基于动态网络方法的中国行业板块联动效应分析

该文通过DCC-MVGARCH模型和网络分析方法,建立动态最小生成树网络,利用Tucker分解和K-均值聚类方法,将动态网络聚集成3个代表网络,并生成分层结构树图,对在2007—2018年包括次贷危机、“2015年股市波动”和中美贸易战等多个股市重大事件时间段内的中国行业板块联动性进行实证分析.研究结果表明:行业板块联动性长期处于波动状态,在股价快速上涨、下跌时,联动性的波动尤为剧烈,但在代表网络中具有重要影响力的节点变化不大,分层结构树图聚集状态变化较小,行业板块联动网络基本稳定; 在中国股票市场的网络结构中生产制造业处于核心地位,金融业处于边缘地位.

The Analysis on Linkage Effect of Chinese Industry Sector Based on Dynamic Network Method

The dynamic minimum spanning tree network is established through DCC-MVGARCH model and network analysis method,and the dynamic network is integrated into three representative networks by using Tucker decomposition and K-means clustering method and hierarchical tree diagram is generated.The empirical analysis of China's industry sector linkages is made during the period from 2007 to 2018,including the subprime mortgage crisis,the "2015 stock market fluctuations" and the trade war.The results show that the linkage of industry sector has been in a state of volatility for a long time,when the stock price rises and falls rapidly,the linkage fluctuation is particularly severe,but the nodes with important influence in the linkage network have not changed much,hierarchical tree graph aggregation state change less,the linkage network between the industry sectors is basically stable.In the network structure of China's stock market, manufacturing industry is at the core,and the financial industry is at the marginal position.