Finding correlation of metaverse stocks via text mining

Abstract

Since the introduction of Facebook in 2004, social media has become an essential part of our life and many people use social media to share their ideas. In 2021, the influence of social media expanded into the investment sector because of the short squeeze event of GameStop. In this thesis, analysis on the correlation between stocks and social media is implemented. In particular, the correlation between metaverse stocks and social media are analyzed. Frequency of keywords appearing on Reddit, a social media platform, will be used to compare with the daily percentage change of share prices to calculate the Pearson correlation coefficient between them. Then, the correlation coefficient will be used in a linear regression model to find the mathematical relationship between correlation coefficient and fundamentals of companies like market capitalization, regions, sectors, and industries to see which factors affects the correlation the most. The result is categorized into three parts, namely regions, sectors, and industries. For regions, the result shows that the only regions with positive correlation that is statistically significant are Malta and Russia. For sectors, surprisingly, the technology sector is found to have a negative correlation with metaverse instead of a positive one. For industries, the airline industry is the one which has the highest correlation coefficient with metaverse followed by REIT – healthcare facilities industry.