Abstract:
The paper utilizes empirical data to investigate finfluencers – key actors in financial social media – and their impact on crowd sentiment. Despite its importance for various stakeholders such as hedge funds and regulators, the antecedents of crowd sentiment remain largely unexplored. The study analyzes 80 million posts on stocks and cryptocurrencies, grouping actors by their social networking potential (SNP) to distinguish between those with potential to influence and the broader crowd. The authors construct sentiment time series employing transformer-based models for both groups and apply panel vector error correction models (PVECMs) to examine their relationship. Drawing on herd theory, the findings confirm that finfluencer sentiment positively predicts crowd sentiment in the short term, without a reciprocal effect. This effect is stronger for cryptocurrency assets than for stocks and is positively moderated by the uncertainty of the crowd. These insights contribute to the fields of information systems (IS) and marketing research by exploring the role of social media influencers (SMIs) in contexts outside traditional consumer marketing. The study’s findings have significant implications for regulators, firms, social media platforms, and investors, supporting the development of regulatory frameworks, marketing strategies, platform policies, and investment decisions.
