Abstract
This thesis examines whether Elon Musk's Twitter use is associated with Tesla's stock price volatility by examining the number of tweets, retweets, likes, and replies his activity generates.
These aggregated numbers are then tested against how Google search volumes are impacted by his Twitter activity and his engagements with the public via the platform. Both of these factors are then tested against different range-based volatility estimators.
We used the range-based estimators from Rogers and Satchell (RS), Garman and Klass (GK), and Parkinson (PK). These estimators calculate the volatility regarding day-to-day changes. During the data exploration portion of the research, significant autocorrelation lags were discovered in the Google search volume for "Tesla."
The key findings were that Elon Musk's use of Twitter led the Google search volumes of both keywords "Elon Musk" and "Tesla" and that his use of Twitter was statistically significant towards the volatility estimators. However, the observed changes were smaller than first anticipated but they confirmed that there is an effect. In some specifications, we identify that there is a significant effect of social media interest, but this result is not consistent across all specifications; this might be because of how the HAR-RB captures the realized volatility.
