How Influential is Elon Musk? An Event Study Analysis of Tweets on Auto Stock Returns

Abstract

In the past few years, the social media platform Twitter has become a popular space for speculation on stock prices. Elon Musk, the CEO of Tesla, has become notable for his opinions on a variety of stocks and cryptocurrency. With his large following of 80 million users on Twitter, many of his tweets about Tesla have gone viral. Some of these viral tweets turn out to be misleading "troll" messages, but they still draw investor attention, and Tesla's stock price often moves substantially as a result. In this paper, we apply the event study methodology to analyze the effects of Elon Musk's tweets on abnormal returns of Tesla and other auto companies. We perform sentiment analysis using NLP techniques and present three models: the Abnormal Returns Model, Multivariate Regression Model, and Panel Event Study Model, which are based on OLS linear regression. The first two models reveal that Elon Musk's tweets have a statistically significant effect on Tesla and electric vehicle rival companies. There are event-induced abnormal returns on the days of events, but there is limited to no evidence that returns on the days before and after are affected. We find that some abnormal returns are confounded by other firm-specific events. Our findings have implications on the reach of Musk's influence: while his tweets affect abnormal returns of firms in the electric vehicle sector, his Twitter activity does not impact the abnormal returns for non-EV auto firms.