| With the rapid development of China's economy,the stock market known as the"economic barometer" is also growinging rapidly.However,China's stock market is not perfect enough.It is still affected by a large number of uncontrollable factors,and paradoxical phenomena occur frequently.The problems of excessive speculation,management irregularities and policy-driven negative effect in the stock market are most conspicous.Therefore,building up stock complex networks is of great significance to understand the correlation between stocks and the changes of stock market.Stock market is the place where stocks are issued and circulated.Listed companies issue stocks to raise funds from the public for the development of enterprises and return the operating profits to investors in the form of dividends.However,stock market returns and risks coexist at the same time,and stock prices are affected by many factors.this paper,in order to study the impact of financial indicators on stock prices,we chooses three indicators which is the profit margin of total assets,the return on net assets and the return on investment to reflecting profitability.In order to make suggestions for shareholders'investment We build stock network using the daily closing price of stocks,and study the relationship between the total asset profit margin,net asset return rate,investment return rate and the topological structure characteristics of the stock Association network.In the first chapter,we provide sufficient analysis of background of the correlation characteristics of Chinese stock market and the studies progressing.In addition,the methods that used in this paper are introduced in detail.In the second chapter,we introduce the theoretical basis of complex networks,the statistical characteristics of complex networks and how to build networks.In the third chapter,we use the threshold method and the minimum spanning tree to construct the network.Empirical research shows that:1.the 12 static stock networks from 2006 to 2017 have small world characteristics,that is,the stock networks have a high clustering coefficients,small average path lengths,and a strong correlation between stocks.2.After studying the degree distribution of the stock network,we find that the degree distribution of the stock network under different threshold levels basically does not obey the power law distribution,but in the stock network constructed by MST method,the degree distribution of the stock network in 2006,2007,2009,2013,2015 and 2016 obeys the power law distribution,that is,there exists "hub"node.3.Using the threshold method,we study the clustering coefficient,degree and feature vector centrality of stock networks at different threshold levels.After statistics,we find that in the 12 static stock networks from 2006 to 2017,the central stocks are mainly concentrated in manufacturing,financial and information industries,of which 600177 and 600074 are two stocks.Basically,votes are always in the key position of the stock market.In the Chapter 4,we makes the total assets profit margin,net assets return rate,investment return rate and the topological structure of the stock network which includes degree centrality,proximity centrality,eigenvector centrality and degree.The results show that:1.There is no unit root in the series of profit margin,degree,degree centrality,eigenvector centrality and proximity centrality of total assets,that is,the sequence is stable,and we can make a conclusion that there is no pseudo-regression.2.There is no linear correlation between return on net assets and return on investment and degree,degree centrality,proximity degree centrality and eigenvector centrality in the stock Association network.3.Total asset profit margin is proportional to degree centrality and inversely proportional to proximity centrality,eigenvector centrality and degree.When the degree centrality and proximity centrality of the network are the same,we find that the degree centrality has a stronger impact on the return on total assets. |
PDF Full Text Request