Research On The Complex Network Topology Of Shanghai And Shenzhen Stock Markets

The stock market is an ever-changing complex system with obvious characteristics of volatility,complexity and clustering coupling.It is essential to study the structural characteristics of stock market and the relationship between individual stocks in order to understanding the changing mechanism of the securities market.With the development of complex network theory,which can well abstract all the complex factors in the stock market into network topology and provide important theoretical basis for studying the development of the stock market.What is more,it is of great significance to give investors and regulators some indirect advice.Therefore,in this thesis,based on the data of Chinese stock market,we synthesize the theory and method of complex network to build the correlation between stocks and analyze the characteristic of the network topology.Further empirical analysis is made on the characteristics of network structure and the changing trend of the industry of important node,in order to understand the development characteristics and changing rules of the stock market.The details are described as follows:In chapter one,we briefly summarize the background of the application of complex network in the stock market and the research status of stock network market,then we sketch out the research content of this paper.In chapter two,we introduce the basic knowledge of graph theory,the network of several basic static characteristics and some kinds of common network mechanism models in complex network.In chapter three,in order to excavate the important node in stock network market.we make use of the CSI300'daily closing prices data from 2010 to 2017,describe the correlation between stocks with correlation coefficient.The threshold Method based on Pearson correlation coefficient was used to denoise the network,then applying the Louvain algorithm to classify each stock network and mine important nodes in the community for annual stock network.The empirical results show that the manufacturing industry play a significant role in the CSI300 market.In the network,there is a phenomenon of agglomeration in the community,and similar stocks were closely linked.In addition,China's stock market lacks stocks with relatively large influence,indicating that the current development of the stock market is still not perfect.In chapter four,the aim to understand the stability differences between SSE and SZSE in Chinese stock market.To begin with,the daily return rate data of the SSE100 index and the SZSE 100 index from 2006 to 2017 are divided into 12 time periods,and the correlation coefficient between shares to edge to build network.Secondly,applying the p-median Problem(PMP)method to construct the star cluster structure of SSE100 index and SZSE100 index,which obtained the star network graphs of 12 different time periods respectively.Finally,the similarity measurement values are introduced to discuss the change trend of the clustering structure of two stock networks respectively,and we also analyze the industry of the central stock in the clustering structure.The empirical results show that the structural stability of SSE is superior to the SZSE.As for the emphasis of industry development planning,apart from manufacturing industry,SSE tend to supply industry and SZSE tend to finance industry.In addition,the structure of the stock network will change significantly before the financial crisis,which has an early warning effect on the occurrence of the crisis.