Analyses Of Influential Nodes For The Chinese Stocks Based On Jump Volatility Spillover Networks

The stock market is closely related to a country’s politics and economy.The stable development of the stock market is of great significance to the long-term stability of the economy.However,the market situation is often changing rapidly,and it is necessary for investors to use real-time information on stock fluctuations to deal with the complex and volatile market environment,especially when there are abnormal fluctuations in asset prices in the market.With the rapid development of computer science and technology,it is easier to obtain higher-frequency data than daily data,and the cost is gradually reduced.Now,the use of high-frequency data to conduct empirical analysis on the stock market is the focus of research in the financial field an important topic.Due to the influence of many factors such as national macro policies,international geopolitics and various extreme events,China’s stock market usually exhibits high volatility with frequent sharp rises and falls,making the stock market a typical complex system.Therefore,we apply the complex network method to study the stock market.This paper uses the complex network method to construct the stock jump volatility overflow network based on the high frequency data extraction jump volatility.Existing literature points out that stock networks are generally heterogeneous,and different nodes play different roles in structure and function.Identifying influential stocks is helpful for investors to manage risk,and can provide decision makers with certain insights.The theoretical basis allows it to better regulate and supervise the stock market.Therefore,this paper selects the 5 minute high-frequency trading data of Chinese stocks from 2006 to 2018 in the Wind database to extract the jumping volatility and constructs the annual jumping volatility overflow network.important stocks to identify and analyze the factors that affect important nodes.This article expands from four aspects.The first chapter is the introduction,which first discusses the research background and research significance,then introduces the research status of jumping fluctuations,complex networks and the identification of important nodes based on complex networks at home and abroad,and finally explains the research content and research framework of this article.The second chapter is theoretical knowledge.First,it introduces the development and research work of complex networks.Secondly,it introduces the knowledge of graph theory used to represent networks,including the basic concepts of graphs and the matrix representation of graphs.And network centrality indicators are described,namely network density and network efficiency,as well as degree centrality,proximity centrality,betweenness centrality,PageRank value and eigenvector centrality.The third chapter takes the 5-minute high-frequency data of China’s stock market from January 2006 to December 2018 as a sample,extracts jump fluctuations from it,and constructs 13 jump fluctuation overflow networks by year division,from network topology changes and important nodes.Analyze the topological properties of China’s stock market from two aspects.On the one hand,using network density and global efficiency to describe the annual changes in the network structure of stock jump volatility spillovers,it is found that these two indicators have a certain role in identifying crises.On the other hand,based on five common centrality indicators(degree centrality,proximity centrality,betweenness centrality,PageRank value and eigenvector centrality),the importance of network nodes is measured,and five network centralities are found.The Pearson correlation coefficient between the measurement values is relatively high,so we use the principal component analysis method to construct the important nodes in the composite index measurement network,and identify the important nodes in the network.In addition,stocks with larger market capitalization often play an important role in the network.The fourth chapter is to analyze the influencing factors of jump volatility spillovers on the network.On the one hand,the jump volatility spillover relationship between stock pairs is used as the dependent variable to establish a Logistic regression model to study the fundamental factors of the company’s financial and market factors on the jump volatility spillover between stocks.The empirical results show that the stock price jumps and fluctuations in the Chinese stock market tend to be from higher market value,stronger profitability,higher price-earnings ratio,higher asset growth rate,higher book value,but lower leverage and turnover.of stocks spill over to stocks with lower market capitalization,weaker profitability,lower price-earnings ratio,lower asset growth rate,lower book-to-market value,but higher leverage and turnover;the impact of importance.The empirical results show that stocks with larger market value,larger price-earnings ratio,and larger book value tend to be more important.Further,the sample period is divided into crisis period and normal period to analyze the driving factors of node importance respectively,and it is found that the driving factors of stock importance are basically similar.However,in normal times,only market factors have a significant impact on the importance of network nodes.But in times of crisis,both market factors and accounting factors can affect the importance of nodes.Finally,the stock network constructed based on jump volatility can reflect the risk situation of the stock market,as well as the research on the jump risk spillover relationship between stocks driven by corporate financial and market factors,we have drawn some conclusions and discussed future research direction.