A Textual Analysis Algorithm for the Equity Market: The European Case

Abstract

This article explores the use of the crowd to interpret text and the power of that interpretation to predict future events. This topic is addressed through an experiment in which news sentiment is evaluated by crowds and experts in different configurations. Their classifications are used as training sets for machine-learning algorithms. The testing is done based on Reuters news stories and the returns of the stocks mentioned right after the stories appear. This article explores a simple trading strategy in which the trader takes a long position when the forecast of the asset return is positive and liquidates the position after one minute. Using this trading strategy, a support vector machine trained with the sentiment of several human groups shows the highest Sharpe ratio after transaction costs and outperforms the STOXX 50 Index.