Analysis of Important Nodes in the Complex Network of Shanghai Stock Exchange 50 Constituents based on PageRank and Louvain Algorithm

作     者：Wenhui Zhu 

出 版 物：《Frontiers in Economics and Management》 

年 卷 期：2021年第2卷第6期

页      面：132-139页

主　　题：Complex Network Pagerank Algorithm Louvain Algorithm Pearson Correlation Coefficient 

摘      要：The stock market has typical complex network characteristics. By building a network share, dividing the community and analysising major nodes, conducive to in-depth understanding of the stock market as a whole and local properties and excavate important industry and stock information. The logarithmic return data of the 50 constituent stocks of the SSE are the research samples, and the stable, medium-strong correlation and significant stock sequences are screened through the ADF test and the Pearson correlation coefficient. Use the PageRank algorithm for reference to rank the importance of stock nodes, and use the Louvain algorithm to construct a stock network and divide the community network. Then study the local important nodes and stock network structure of the divided sub-community network. The research screened out 23 stocks and divided them into 3 communities in the community network. The main industries are the financial industry represented by securities, the financial industry represented by banks, and the mining industry. On the whole, the financial industry, mining industry and construction industry are key industries, accounting for 61%, 13% and 9% of the total industry distribution respectively. Among them, the financial industry is the most important industry, and China Construction (601668) s PR value ranks first in both the overall network and the local network. In view of this, investors should put forward a rational investment, do a good job of evaluating important stock nodes when choosing investment objects; regulatory authorities should pay attention to the overall and partial market changes in important industries and stocks, make reasonable adjustments, etc.