The Impact of Artificial Intelligence Discourse on Short-Term Trading Volume: Evidence from S&P 500 Earnings Calls

Abstract

Artificial intelligence (AI) is heavily influencing corporate strategies and investor sentiment across financial markets around the world. Industry and corporate discussions regarding AI during quarterly earnings calls have become a metric that is closely monitored by investors looking for any signs of true technological adoption. Whether mere rhetorical emphasis on AI by a company actually does drive any anomalous trading volume immediately after its earnings disclosure remains debated. This study aims to clarify the relationship between the AI mentions and trading volume and its magnitude. We use natural language processing to capture the mentions of AI in 4,892 earnings transcripts among S&P 500 companies from 2022 to 2024. We also track cumulative abnormal volume in the four days after each earnings announcement and control for standardized unexpected earnings and firm size. Empirical analysis demonstrates that companies that landed in the top quartile for AI mentions saw a real and statistically significant jump in cumulative abnormal volume right after earnings (p < 0.05), while those in the bottom quartile showed no real movement in abnormal volume. The multivariate regression model, which yields a strongly positive and significant coefficient for AI mentions (beta = 0.052, p = 0.003) and standardized unexpected earnings (beta = 0.034, p < 0.01), alongside a statistically insignificant coefficient for firm size (beta = -0.008, p = 0.31). These findings highlight a market vulnerability where AI language used by corporates triggers short-lived liquidity events and have an immediate impact on their financial valuations.