Influencing cryptocurrency: analyzing celebrity sentiments on X (formerly Twitter) and their impact on bitcoin prices

This paper explores the dynamic relationship between Bitcoin prices and the sentiments expressed by celebrities on X (formerly Twitter), employing a Vector Autoregression (VAR) model. The aim of this study is to understand the impact of a diverse range of sentiments (positive, negative, valence, arousal, dominance) extracted from the posts of 30 influential celebrities in the Bitcoin community on future Bitcoin prices and volumes. The analysis reveals that negative sentiments have a statistically significant impact on Bitcoin price fluctuations. Furthermore, while the shocks to price caused by sentiments converge within approximately 2 weeks, it was found that in the short term, spanning just 2 days, positive sentiments and valence sentiments positively influence price fluctuations, whereas negative sentiments exert a stronger negative effect. This finding underscores the direct influence of celebrity sentiments on the short-term emotions and actions of Bitcoin market participants, supporting psychological researches that indicate a strong influence of negative information on individual cognition and behavior. The significance of this study lies in its broad analysis of Bitcoin price fluctuations using the sentiments of various influencers on social media, not limited to globally recognized figures, and shows one possible way to respond to an immature market. This study offers valuable insights for investors and market analysts in refining investment strategies and risk management by considering market sentiment fluctuations.