The Topology Property Of China’s Stock Correlation Network Based On The Overall Network Analysis

In recent years, the performance of China’s stock market is not good, the experts and scholars in different academic fields have doned a lot of research on China’s stock market from different perspectives. With the deepening of the study on the stock market, people realize that the stock market is a complex system, and begin to study it using the complex theory method.This thesis uses the complex network method to establish China’s stock correlation network based on daily return correlation coefficient, which can be treated as a social network, and then uses the overall network analysis method to study the network’s small-world effect and the centrality of stocks. The study on the network’s small-world effect is conducive to understand the stock price volatility between conduction velocities and the coherence between stocks, while the study on the centrality of stocks is helpful to recognize the power structure of China’s stock market. The main content can be divided into three aspects:1, Construct the China’s stock correlation network by using an improved threshold method. More specifically, we consider the239component stocks of HS300index and analyze daily changes of their prices over a period of238consecutive trading days in2013.2, Verify the small-world character of China’s stock correlation network. We first point out the features of small-world network after the discussion of previous research, and then use the overall network analysis method to judge the small-world character of China’s stock correlation network. The result shows that China’s correlation network is a small-world network which demonstrates that the conduction velocity of stock price fluctuations is fast and the coherence between stocks is relatively high.3, Analyze the centrality of nodes in the China’s stock correlation network based on the overall network analysis method. The three main measures are the degree centrality, the betweenness centrality and the closeness centrality. The main conclusions are:(1) The top three stocks are ZGDJ (601669), JXTY (600362) and NFHK (600029) in China’s stock correlation network according to the corresponding values of the degree centrality measure, the degree of which is58,57and54respectively. This indicates that the fluctuation of their stock price will have a direct and quick impact on the price of58,57and54stocks respectively. On the other hand, they are more vulnerable to fluctuate.(2) The top three stocks are ZGJZ (601668), NFHK (600029) and DFHK (601668) in China’s stock correlation network according to the corresponding values of the betweenness centrality measure, which shows that they are in a "good" position in the network and the gap between values is rather big which indicates that only minority stocks are highly dependent on the network.(3) The top three stocks are ZGDJ (601669), NFHK (600029) and ZGJZ (601668) in China’s stock correlation network according to the corresponding values of the closeness centrality measure. This demonstrates that the fluctuation of their price is able to affect all other stocks and the price index.