This study aims to investigate the price impact of trading by institutions that differ in their connectedness in a network of financial institutions. Our primary aim is to investigate whether highly connected institutions get an informational advantage over their less-connected counterparts, which ultimately reflects in prices.
We classify institutions into more-connected and less-connected groups by examining the topological structure of the network of financial institutions, such as banks, insurance companies and investment companies. Two institutions are considered connected if they share common shareholdings. Then, the institutions are further classified accordingly using measures such as betweenness centrality, closeness centrality and network strength. Then, we regress stock returns on trading by institutions classified by their network connectedness to find the impact on market efficiency.
We find that trading by peripheral institutions improves stock price discovery, whereas trading by connected institutions neither harms nor improves the price discovery process. Moreover, we find that trading by peripheral institutions has predictive power for future earnings surprises and abnormal returns around earnings announcements, suggesting a high likelihood of information-based trading.
This research provides valuable insights to investors and policymakers regarding the price stability that has always remained a significant concern in the financial markets. This will protect the investors’ hard-earned money and suggest appropriate trading strategies to the investors to generate superior risk-adjusted returns.
The study uses a new approach to create a network of financial institutions and challenges the existing claims in the literature that connectedness gives an informational advantage. These results give rise to a new puzzle in the social network literature that overlooks the role of networking as an information blocker. We also uncover a new type of heterogeneity across institutions, i.e. connectedness.
