基于跳跃波动溢出网络的中国股票市场重要节点分析

Analysis of Important Nodes in China’s Stock Market Based on Jump Volatility Spillover Networks

摘要: 对股票市场中重要节点识别及其影响因素分析是金融风险管理的热门话题之一。针对此，本文首先根据中国A股市场5分钟高频数据提取跳跃波动，运用格兰杰因果检验方法构建跳跃波动溢出网络；然后，针对网络中的5种中心性指标，通过主成分分析法构建复合指标测度网络中重要节点；最后，运用面板数据回归模型考察影响重要节点的因素。研究发现：1) 福星股份、上海机电、长春高新等是股票网络中的重要节点。2) 70%的重要节点具有较大市值，与“大而不能倒”的理念相吻合。另一方面，有30%市值较小的节点在跳跃波动网络中也扮演着重要角色，这也意味着不能忽视“太关联而不能倒”。3) 市值、市盈率、账面市值比越大，换手率越小的股票，在网络中往往越重要。

Abstract: The identification of important nodes in the stock market and the analysis of their influencing factors are one of hot topics in financial risk management. To achieve this purpose, this paper first extracts jump volatility in the Chinese A-share market by using intraday 5-minute high-frequency data and uses the Granger causality test to construct a jump volatility spillover network. Then, five network centrality metrics are used to construct a comprehensive index to measure the important node. Finally, we identify the major factors, which affect the important nodes with panel data regression analysis. The research found that: 1) Hubei Fuxing Science and Technology Co., Ltd., Shanghai Mechanical & Electrical Industry Co., Ltd., and Changchun High and New Technology Industries (Group) Inc. are important nodes in the stock network. 2) 70% of important nodes have a large market value, reflecting the risk of “too big to fail”. On the other hand, 30% of nodes with a small market capitalization also play an important role in the jump and volatility network, which also means that the risk of “too relevant to fail” cannot be ignored. 3) Network nodes with larger market value, higher price-earnings ratio, higher book-to-market value ratio, and lower turnover rate tend to have stronger importance.