Cryptocurrency ecosystems and social media environments: An empirical analysis through Hawkes’ models and natural language processing

• DOI: 10.1016/j.mlwa.2021.100229
• Publication Date: 2021-12-08T17:16:09Z
• Authors: Marco Ortu; Stefano Vacca; Giuseppe Destefanis; Claudio Conversano
• Keywords: Cryptocurrencies; 330; Fundamental analysis; cryptocurrencies; QA75.5-76.95; 02 engineering and technology; Hawkes model; 01 natural sciences; Forecasting price movements; Electronic computers. Computer science; forecasting price movements; 0202 electrical engineering, electronic engineering, information engineering; Q300-390; Social media analysis; social media analysis; 0101 mathematics; Cryptocurrencies; Social media analysis; Fundamental analysis; Forecasting price movements; Hawkes mode; Cybernetics; fundamental analysis

We analyse, using a mixture of statistical models and natural language process techniques, what happened in social media from June 2019 onwards to understand the relationships between Cryptocurrencies’ prices and social media, focusing on the rise of the Bitcoin and Ethereum prices. In particular, we identify and model the relationship between the cryptocurrencies market price changes, and sentiment and topic discussion occurrences on social media, using Hawkes’ Model. We find that some topics occurrences and rise of sentiment in social media precedes certain types of price movements. Specifically, discussions concerning governments, trading, and Ethereum cryptocurrency as an exchange currency appear to negatively affect Bitcoin and Ethereum prices. Those concerning investments, appear to explain price rises, whilst discussions related to new decentralized realities and technological applications explain price falls. Finally, we validate our model using a real case study: the already famous case of ”Wallstreetbet and GameStop”11 https://www.economist.com/finance-and-economics/2021/02/06/how-wallstreetbets-works. that took place in January 2021.