The Structure of Twitter $Cashtag Networks and Market Reaction to Earnings News

Despite rising interest in Twitter-based cashtag activity, accounting research has yet to examine how the structure of these networks influences the capital markets. Drawing on investor attention and information processing theories, we examine how four structural features of cashtag ego networks condition the market’s reaction to quarterly earnings announcements. Using panel regressions on S&P 1,500 data, we test whether centralization, density, clustering, and isolates influence both the magnitude and direction of the market’s response to earnings surprises. We find that reaction magnitude is lower in networks with higher centralization, density, and isolates and higher in networks with greater clustering; all four features also strengthen the market’s directional responsiveness to unexpected earnings news. These results support our argument that network structure functions as a cognitive filter, guiding investor attention and shaping information processing. Cross-sectional tests by follower count and verified user status reveal that user prominence moderates these relationships. Additional analyses using directed networks, network typologies, and alternative dependent variables (abnormal volatility and spread) provide further evidence that network structure affects the salience and market consequences of financial information. Our findings extend social network theory into the investor attention literature, offering a novel perspective on how social media platforms influence market behavior.