Emergent Sentiment in Financial Markets : Sentiment Flow from Social Media to Mass Media

Over the past two decades, substantial research has established the importance of sentiment in financial markets. We know that sentiment impacts investor decisions; because of advanced text mining and sentiment analysis techniques, automated trading strategies based on sentiment are prevalent. However, we know little about how sentiment emerges. The news media, the markets, and social media are all understood to contribute to overall sentiment, but the exact mechanisms at play remain largely unknown. Understanding the mechanism of sentiment emergence is critical to build systems to support investors' decision making and to ensure that markets function well and are resistant to attempts to manipulate sentiment. We studied one piece of the puzzle and examined sentiment flow from social media to mass media. We studied two years (2013 to 2014) of data from Sina Weibo and Sina Finance. After controlling for the direction of influence from mass media to social media and also for stock performance, we found that in the stock market, social media sentiment affects mass media sentiment the next day. Moreover, this impact is stronger for attention-grabbing stocks. The implications of these findings are discussed in relation to sentiment emergence and a broader demand-driven media bias perspective