Investigating football club stock fluctuations through Twitter sentiment analysis and sporting results

• Main Author: Ooi, Wei Chern
• Other Authors: Erik Cambria
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence; Engineering::Computer science and engineering::Mathematics of computing::Numerical analysis
• Online Access: https://hdl.handle.net/10356/166154
• Institution: Nanyang Technological University

This project aims to investigate the relationship between Twitter activity and sporting results on the stock market performance of publicly traded football clubs. Three clubs are used as case studies, where fan opinions on Twitter are analysed and compared to the fluctuations of the club's stock prices. While previous studies have investigated the direct impact of tweets and football game results on the stock market, this project is unique in that it compares the two factors side-by-side to determine which is a better indicator. The goal of this project is to determine whether Twitter activity or match results have a greater impact on the performance of football club shares on the stock market, and to understand the extent of this influence. Results from Granger-causality analysis shows that football match results is the overall better predictor of all three publicly traded football club’s stock performance. Unexpected match results demonstrated greater impact in influencing all three football club’s stock returns and price while sentiment features from football-related tweets are good predictors for Manchester United F.C. and Borussia Dortmund.