Statistical Characteristics Of Chinese Stock Market In View Of Complex Networks

The stock market can effectively collect social idle funds,promote financing and investment activities,and play a very active role in promoting the development of the capital market.As a barometer of a country's economic status,it has been widely concerned by the supervision department,financial sector and academia.At present,the total market value of China's stock market has been in the forefront of the world,however,because of the complexity and volatility of the stock market,it can also bring unexpected negative consequences,such as a huge change in stock prices can affect investor confidence and a virtuous cycle of financial markets.In 2015,Chinese investors witnessed and experienced the thousands of shares limit up,thousands of shares limit down and thousands of shares halt;early 2016,the fuse mechanism was urgently halted and the frequency and amplitude of fluctuations in the market record highs.It's show that China's stock market system is still very fragile,the effective regulation of the stock market and efficient management is still in the exploratory stage.Therefore,it is of great significance for investors and regulators to understand the changes and development of the stock market,especially to study the characteristics of China's stock market.In this paper,we synthesize the theory and method of complex network,using PageRank algorithm and Louvain Method algorithm to excavate important nodes and classify societies separately for the 2009-2015 HS300 stock market network,and analyze the influence of these nodes and associations in the market.According to the 2003-2015 SSE(Shanghai stock exchange stock market)and SZSE(Shenzhen stock exchange stock market)concentration analysis,research centralized status and development trend of Chinese stock market,to provide empirical reference for regulators and investors.This paper consists of the following four chapters:In chapter one,first of all we briefly introduce the background and significance of important node assessment and the concentration study in the stock market,then we sketch out the method and theory in this paper.In chapter two,we introduce the graph theory and the used knowledge of complex network,then give the definition of average distance,cluster coefficient and degree distribution.In chapter three,this paper uses the csi300 data for 2005-2015 to analyze the statistical features.The csi300,a set of large-scale and strong liquidity stock,can basically represent the whole A-share market and reflect the overall trend of the Shanghai and Shenzhen exchanges.We abstract stock share to node and the relationship between shares to edge to build a global coupled network,building a global coupling network,simplifying by denoising method of optimal threshold value,digging the important node by Louvain Method algorithm and PageRank algorithm.The study found that mining,manufacturing and financial industry play a important role in the csi300 market in which there is close relationship with internal stock and external.In the network there is exist clustering phenomenon between similar stock and significant mutual influence.In chapter four,we employ data of Shanghai Stock Exchange(SSE)and Shenzhen Stock exchange(SZSE)in 2003-2015 to analyze the statistical features.Using the data of shares,shareholders and market value to structure a directed cross-shareholding network and analyze the change of concentration of SSE and SZSE by the concentration metrics.Finding that,during the crisis,the SSE market value significantly concentrated in the hands of power shareholders,and the market control capacity of SZSE is superior to the SSE in the crisis.For centralized analysis,in front of the crisis to ease,the concentration of SSE is global dispersion and local focus which reverse during ease;the index of SZSE is regular monotone decreasing change,both to more dispersed.