Can Twitter Proxy the Investors' Sentiment? The Case for the Technology Sector

The stock market is influenced by several factors, such as macroeconomics, regulatory, purely speculative ones, and many others. However, one of the most relevant and meaningful is the general opinion and the overall investors' sentiment, i.e., what investors think about a certain firm and, as a consequence, about the relative stock. This investors' sentiment is here proxied by the Twitter content, and the study sums up to the recent outbreak of works that exploit sentiment analysis and Twitter data for stock market predictions. The sample analyzed concerns three major technology companies over a two-months period, on a minute basis. Using microblogging activities and a scoring algorithm for each tweet, it was possible to formulate interesting forecasting models identifying a new set of variables and indicators of the stock market future movements. A selection model has been used to implement the study, and the evidences found were encouraging, since it has been possible to draw the conclusion that this new source of data may increase the explanatory power of financial forecasting models. More in detail, it looks like that the average sentiment associated to any tweet is not so relevant as expected in prediction terms, while the posting volume has a greater forecasting power and it could be used to augment the models. Although this kind of analysis are becoming mainstream and quite common, this work represents an interesting case study for the technological sector rather than advancing fundamental new techniques in the field.