Investigating the relationship of stock micro blog sentiments in analyzing stock price movements: A case of the Philippine blue chips

• Main Authors: Ellescas, Bianca Ysabel C., Santos, Johanna Polina L., Yu, Beatriz Airah D.
• Format: text
• Language: English
• Published: Animo Repository 2016
• Subjects: Stock price forecasting > Philippines; Investments > Psychological aspects; Finance and Financial Management
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/9034
• Institution: De La Salle University

Over the last few years, online-based communication platforms have been used to obtain indicators of economic and social trends. With the growing popularity of Twitter and Finance Manila to the investors, its relevance in predicting stock movements was examined. In this study, the proponents aimed to identify the role of micro blog sentiments for the randomly selected blue chip stocks of the Philippine market. In particular, sentiment analysis was performed for over 15,000 online posts using Chung & Liu's Bag of Words (BOW) lexicon model. Results revealed a weak to moderate significant relationship between sentiment features and stock parameters suggesting that sentiments do provide valuable information for most of the stocks sampled. Furthermore, Granger Causality Analysis demonstrated that majority of the causal direction points out to sentiments greatly affecting the stock parameters, which supports the behavioural finance assumptions in proving the presence of irrational noise traders in the market.