This project examines the relationship between sentiments and activity on social media and the performance of companies in the video game industry. We examine whether there’s a relationship between the sentiments and activity on Facebook and a range of companies’ stocks as well as their financial reports. This project uses scripts in R to retrieve Facebook data from Facebook’s API, including Facebook posts related to the specific companies and any metrics associated with them (number of comments, likes and shares) as well as the contents of the comments associated with the posts. Using a bag-of-words sentiment analysis approach to deduce contents of the comments and simple linear regression, distributed lag models and vector autoregression models, we examine the relationship between social media user sentiment and company performance. We conclude that despite a relationship between stock prices and social media activity, there is no similar connection between the companies’ performance and the activity on Facebook. There were no similar relationships between the sentiments and stock prices, nor sentiment and company performance. We suggest this relationship (and lack thereof) could be explained by the relationship between Facebook activity (or sentiments) and company performance not being reflected in the financial reports of the companies, but instead in the owner’s required rate of return when doing valuation of the companies. Another possible explanation for the discrepancy is the prevalence of herding in the financial markets. The significance of the conclusions in the project is reduced by the fact that no control variables have been included in the regression models and not all assumptions behind the regression models are kept.
|Educations||MSc in Finance and Accounting, (Graduate Programme) Final Thesis|
|Number of pages||113|