The Analysis Of Social Network Structure And Complex Network Characteristic Of Hushen 300 Stocks

The stock market has become an important part of our country's economy.Investment activities of stock will have a significant impact on countries,companies and individuals.So the research on the stock market get more and more attention of the scholars.There are many factors that have effects on stock price,such as politics,economy,market,technology and investors behavior.So the formation mechanism of stock price is complex.Topological properties of complex network can give good explanation to the various phenomenon and behavior of the stock market.It makes the complex network theory become an important tool to study the internal structure of the stock market.Social network analysis method is an important tool for researching the nodes in the network and the relationship between them.Treat the stock network as a social network and explore the interaction and mutual influence between the stocks,so that we can get the quantitative analysis and rational knowledge of stock market.That will be the focus of this study.First of all,we treat the Hushen 300 index component stocks as the research object,and get the closing price of them from January 2014 to December 2015.Then we calculate the correlation coefficient between stocks and build the stock network according to this.Threshold value method is used to get the stock network simplified,that is,the edge whose correlation coefficient is greater than the threshold value can stay in the network edge.Second,we carry on the complex network research to the stock network,including small world,integration and scale-free feature.The study found that with the increase of the threshold,the scale of the stock network became smaller and the degree distribution of the network is similar to a straight line.This shows that the stock market network obeys the power-law distribution,So we can through the observation of the price trend of a few important stock to measure the overall level of the stock market.Third,social network analysis have been carried out on the stock network,including centrality,condensing subgroups,components,cliques,and so on.The study of centrality occupies larger space.We introduce Freeman's three common measures of node centrality:degree,closeness,and betweenness.Furthermore,a measure that combine both the number of ties and tie weights will be considered.Through the evaluation of the four algorithms' performance,the one of the best algorithm is obtained.Finally,this algorithm was applied to explore the outstanding stock.Stocks with the large centrality are the important stocks in the network.The movements of the outstanding stocks and the movements of the whole stock market were compared.We found that they are very close to a path,which is verified the rationality of the adopted method in this paper.