Abstract
Social media has become a popular platform for conversations about, and with, companies and their brands. Increasingly, customers are using social media platforms to express their experiences and emotions with an organization’s products or services. Recently there has been interest in the role that social media sentiment can play in predicting stock returns.
Social media sentiment data can, for example, reveal customer preferences, customer satisfaction, and customer feedback on product ratings. In some cases, customer satisfaction can lead to increased investor returns. Added to this, strong links can be found in the literature between customer satisfaction levels and subsequent stock price returns. Therefore, our study explores the predictive power of social media sentiment on stock returns, using a logic where social media sentiment acts as an early indicator of customer satisfaction.
We draw upon publically available data to test our hypotheses. We source our customer satisfaction measure from the American Customer Satisfaction Index (ACSI) for a 2-year window. The data for social media sentiment were scraped from Twitter. Due to the qualitative nature of tweets, the data were transformed to quantitative data for analysis using sentiment dictionaries. Finally, for each of the firms in our data set, we calculated daily stock returns for the 2-year observational window.
We find that global social media sentiment is positively associated with stock returns. While financial sentiment performs even better than global sentiment in predicting returns, customer sentiment does not predict returns. Further, we find no relationship between social media sentiment and customer satisfaction. Finally, we find no relationship between customer satisfaction and stock returns.
We conclude that, while social media sentiment does indeed predict stock returns, this effect occurs wholly through financial, rather than consumer, channels.