Twitter as a tool for forecasting stock market movements: A short-window event study

Abstract

In order to explore the relationship between politics-related sentiment and FTSE 100 movements, we conducted a short-window event study of a UK based political event. We collected a sample of over 60,000 tweets using 3 key hashtags during the period of 6 days including before, during and after the 2016 local elections. The study involved performing a collection of correlation and regression analyses to compare daily mood with daily changes in the price of the FTSE 100 at the market level. The findings suggest that there is evidence of correlation between the general mood of the public and investment behavior in the short term; however, the relationship is not yet determined as statistically significant. There is also evidence of causation between public sentiment and the stock market movements, in terms of the relationship between MOOD and the daily closing price, and the time lag findings of MOOD and PRICE. Overall, these results show promise for using sentiment analytics on Twitter data for forecasting market movements.