Abstract

Using sentiment analysis, we extract activity volume, bullish sentiment and disagreement from SINA micro-blog. We investigate the dynamic relationships between social media and stock market returns and trading volumes in China. We found that, in Chinese stock market, the bullish sentiment on social media is positively related with returns and trading volumes but both activity volume and disagreement on social media are generally negative related with trading volumes. Moreover, the effects of social media features on returns are more sustainable than their effects on trading volumes. The results of Granger causality test show that both bullish sentiment and disagreement are useful for prediction stock market trading volumes, but not stock market returns.

Paper Details