Could Emotional Markers in Twitter Posts Add Information to the Stock Market ARMAX-GARCH Model

In our paper, we analyze the possibility of improving the prediction of stock market indicators by adding information about public mood derived from Twitter posts. To estimate public mood, we analyzed the frequencies of 175 emotional markers, words, emoticons, acronyms and abbreviations in more than two billion tweets collected via Twitter API over the period from 13.02.2013 to 22.04.2015. We found that, from 17 emotional markers frequencies with established Granger causality, six provide additional information for the baseline ARMAX-GARCH model according to Bayesian information criteria for the in-sample period of 421 days, and two emotional markers improve directional accuracy and a decrease in the mean-squared error of the model.Our analysis reveals several groups of emotional markers, such as general and specific, direct and indirect, which relate differently to the dynamics of returns