Study Of CSI 300 Index From The Perspective Of Complex Network

With the deepening of financial reform in China, the continuous improvement of the financial regulation and the enrichment of national financial knowledge, securities investment is getting closer attention of investors, especially the stock market. China’s stock market provides not only reliable financing channels for enterprises, but also a broad investment channels for investors. As a result, China’s stock market is closely linked with real economy and national life. In this context, to study the characteristics of the stock market is particularly important, among them, the structure characteristics of the stock market is one of the research focuses.At present, to study the structure characteristics of the stock market, many scholars have established the complex networks of the stock market from the price of the stock, yields, main controlling shareholders and other aspects by the theory of complex network. However, few scholars establish the complex networks by the stock amplitude. The stock amplitude reflects the trading activity of stocks. It can not only reflect the industry development of stock market, but also the investment direction and the mood of investors.This article selects 161 stock shares, which meet the requirements of continuous trading from June 1,2014 to June 30,2015 in CSI 300 Index, and extracts the diurnal amplitude datum as time series datum After the unit root test, the time series datum meet integrated of order 1. Then the article deals with the datum by cointegration test to get the cointegration coefficient between any two stocks. This paper bases on the 161 stocks as the nodes of complex network, and bases the cointegrattion relationship between the two stocks as the boundaries. At last, the complex network of CSI 300 Index is established. Besides degrees, the average path length and clustering coefficient, this paper further analyses the betweenness, k-core of the complex network, and comprehensively selects some stocks that own important status and influence. The research finds that the stock network is a small-world network that owns scale-free characteristic. Finally, the paper studies the robustness of CSI 300 Index complex network. The robustness of complex network refers to its ability to maintain its basic functions under attack. This article discusses the random short pattern and deliberate short pattern. And the deliberate short pattern can be divided into the initial degree short pattern, maximum degree short pattern and betweenness short pattern. And the research simulate the changes of the average path length and clustering coefficient of CSI 300 Index complex network under these four short patterns. The research results show that the network is strong to random short but failed to deliberate short.