Social media echo chambers and stock price volatility: Evidence from Chinese investor forums

This paper investigates whether social media echo chambers amplify stock price volatility. Using 42 million posts from Eastmoney Guba and 186,542 firm-month observations from the Chinese A-share market over 2014 to 2023, we construct a Sentiment Homogeneity Index (SHI) based on semantic similarity to measure opinion convergence. We find that higher sentiment homogenization significantly increases subsequent idiosyncratic volatility. The effect is asymmetric: bearish echo chambers exert a stronger impact than bullish ones, consistent with loss aversion and short-sale constraints. The effect intensifies for stocks with higher retail ownership, lower analyst coverage, and smaller capitalization, supporting the joint hypothesis of sentiment and limits to arbitrage. Mechanism analysis confirms that echo chambers amplify volatility through order flow imbalance. This study contributes to the social media and asset pricing literature by showing that opinion homogeneity, rather than sentiment level alone, provides incremental information for understanding volatility dynamics in retail-dominated markets.