Authors: Alexandre Bénétreau
This thesis discusses the claim of 16 computational finance articles according to which it is possible to predict the stock market using sentiment analysis of social media data. The purpose of this paper is to investigate whether this is indeed true or not. In economic theory, the efficient market hypothesis states that markets are not predictable, that they follow a random walk and that irrational behaviour cancels out in the aggregate. However, behavioural economics research shows that investors are in fact subject to predictable biases which affect the markets. This study uses data from the WeFeel project that analyses tweets in English to infer social mood on a world scale. It also uses data from the Wilshire 5000 index from June 2014 to March 2015. The hypothesis is that changes in aggregate mood arousal mediate stock market fluctuations. Yet linear regression shows that there is no relation between emotional arousal and the stock market, nor between primary emotions and the stock market. Hence, the conclusion is that global social sentiment as derived from social media has no relation with stock market fluctuations. Further research may better focus on social media specialised in the stock markets, such a finance micro-blogging data. Keywords: sentiment analysis, efficient market hypothesis, social networks, computational finance, behavioural finance, stock market, emotion recognition, stock market prediction, social sentiment, behavioural economics
Comments: -The thesis is in English, apart from the title page which contains bits of French (it is a thesis defended at a French uni).-Pages: 27. -Licence: This thesis is licensed under a Creative Commons Attribution 4.0 International Public License.
Unique-IP document downloads: 39 times
Add your own feedback and questions here:
You are equally welcome to be positive or negative about any paper but please be polite. If you are being critical you must mention at least one specific error, otherwise your comment will be deleted as unhelpful.