Information Transmission,Market Risk And Motif Evolution Based On Shareholders' Co-holding Networks

The stock market is a market of information.The social attributes of investors encourage investors to be state-dependent when making decisions.They observe each other to transmit information,which in turn affects investment decisions and individual choices,and ultimately leads to market volatility.After the Equity Split Reform,the major shareholder obtained the right of circulation of their shares.Due to the information superiority and special status of major shareholders,major shareholder transactions have received much attention from beginning to end.The focus of this thesis is on the impact of information interaction between shareholders with informational advantages on market risk.In recent years,the research based on complex networks has become more and more mature,which provides a novel research idea for this proposition.This thesis uses the method of complex network to build a co-holding network among shareholders,mines the economic meaning of network topology indicators,establishes the relationship between the network and the stock market,and explores the mechanism of information transmission between shareholders on market risk.Firstly,this thesis uses two different weighted network clustering coefficient calculation methods to measure the information transmission speed among shareholders.Three risk measurement methods include stock risk,company trait risk,and stock price crash to measure market risk.Three risk measures to describe market risk,including overall market risk,corporate idiosyncratic volatility and stock price crash risk.The results show that there is a nonlinear U-shaped relationship between information transmission speed and market risk.When the information transmission speed is slow,the market risk decreases as the information transmission speed increases.When the information transmission speed is fast,the market risk increases as the information transmission speed increases.Secondly,as a robustness test,this thesis verifies that the network formed by the shareholder relationship is a private information network between shareholders.Research shows that the trading behavior of shareholders is affected by the trading behavior of other shareholders in the joint network,controlling for the other types of indicators such as market public information and stock trait information.The results show that style investment is not the main factor leading to the co-holding relationship between shareholders.Finally,this thesis explores the local information of the shareholder co-holding network from the microscopic perspective.More specifically,from two aspects: motif recognition and motif evolution,this thesis reveals the basic structure of the co-holding network and its dynamic evolution.Research based on motif recognition shows that although the closed motif has a low proportion in the co-holding network,it is an important motif in the network.Strongly connected closed motifs have important influence on network structure.Research based on the motif evolution shows that all structures have a high tendency to disappear.In particular,the motifs containing financial investment companies tend to disappear,while the motifs containing general companies tend to remain unchanged during the evolution process.