The Analysis Of Clustering Structure On Stock Market’s Complex Networks

In today’s economic integration, economic cooperation among various industries is more andmore closely, the influence of each other is more and more intimately. As the nationaleconomy barometer of the Stock Market, the security problem is more and more prominent,the risk management is more and more difficult. In such big background, study therelationship between each sector has a theoretical and practical significance.In1998, along with the Collective dynamics of―small-world‖networks and Emergence ofscaling in random network this two pioneering thesis published, complex networks theoryresearch has become one of the highlights in agro-scientific research in the academic circles.For more than ten years, scholars at home and abroad by using complex network theory andcombined with other disciplines cross research has achieved substantial results. Academicshave proved that the stock market is a complex system, Distinguish from traditional finance isbased on the research method of the efficient market hypothesis (EMH), using complexnetwork theory to study the stock market is from the macro-perspective to research itstopology structure and the overall features. It is a practical significance to use this new pointof view to research China stock market.After the economic crisis in2008, from2011to2013, China stock markets as a whole is stillall the way down. In order to further study the internal structure of China stock markets,discuss the operation mechanism, In this paper, the author using the theory of complexnetwork and R software and its igraph software packages to study China stock markets and dothe following works:1) Combining the theory of complex network and using the minimumspanning tree algorithm (MST) has constructed china stocks associated network model, andthen exploratory analyzed all stocks on the distribution of the yield of different time period,the overall findings presents Leptokurtic distribution.2) By using Newman fast algorithm toanalyze the stock associated network community, and comparing the average correlationcoefficient between stock within society, found in severe stock market falls the averagecorrelation coefficient significantly greater than the stock market stability, and industryaggregation degree increased significantly.3) By means of comparative analysis degree ofeach node in the network and betweenness, closeness and eigenvector, found that the financialand energy stocks are in a central location in the network. This is also in accordance with thereality that the two industries occupies an important position in the economic system. Inaddition, also found two important statistical characteristics, one is the node with highbetweenness centrality is one of the important nodes connect to multiple communities, the other one is the node which at the same time with high betweenness centrality andeigenvector centrality has strong correlation.