Social media engagement and cryptocurrency performance.

Area of Science:

• Financial markets
• Computational social science
• Cryptocurrency analysis

Background:

• Cryptocurrencies are volatile assets, posing investment risks.
• Predicting cryptocurrency performance is crucial for investors.
• Existing methods using volume and sentiment have limitations.

Purpose of the Study:

• To predict cryptocurrency future performance using social media data.
• To introduce a novel model for measuring user engagement with social media topics.
• To analyze the impact of engagement and bot activity on cryptocurrency returns.

Main Methods:

• Developed a new model to quantify user engagement based on social media interactions.
• Estimated engagement coefficients for 48 cryptocurrencies using Twitter data.
• Measured bot post activity associated with cryptocurrencies.
• Correlated engagement coefficients and bot activity with future cryptocurrency returns.

Main Results:

• Cryptocurrency future returns are dependent on engagement coefficients.
• Extreme engagement coefficients (very low or very high) correlate with lower returns.
• High engagement may indicate artificial activity from bots.
• Increased bot posts generally lead to lower future returns.
• Engagement coefficient is the strongest predictor of short-term returns.

Conclusions:

• Social media engagement, particularly when extreme or bot-driven, can predict cryptocurrency performance.
• Investment strategies focusing on cryptocurrencies with moderate engagement coefficients show promise.
• The developed model offers an alternative to traditional volume and sentiment analysis for cryptocurrency evaluation.