An empirical evaluation of the influential nodes for stock market network: Chinese A-shares case

Abstract

This paper aims to rank the influential nodes for Chinese A-share market by employing complex network analysis approach. More than one hundred directed weighted stock market networks are constructed by the methods of Engle-Granger test, Granger Causality test and moving window among 847 stocks for the time period from January 2006 to June 2019. Then the identification of important nodes is investigated by using weighted LeaderRank algorithm. The results show that: (i) the average clustering coefficient and global efficiency increase sharply in the run-up to, and during the financial crisis, and decline rapidly afterwards. (ii) 66.98% of stock market networks have scale-free property. (iii) the influential companies are generally large-capitalization companies. In addition, an interesting finding is that, top 3 influential stocks are high price stocks which are so called “hundred shares” by Chinese investors.