Financial Analysis Using News Data

This paper surveys the field of financial analysis using news data, and also our own considerable work regarding movie gross analysis to show the predictive power of news data. Since the 1990s, linguistic sources such as news have been continuously proved to carry extra and meaningful information beyond traditional quantitative finance data, and thus they can be used as predictive indicators in finance. In the first half of this paper, we analyze and compare important previous research in terms of targets, methodologies, and results. In addition, we discuss financial news analysis with Lydia, a system for large-scale news analysis. In the second half of this paper, based on the news data generated by Lydia, we build three different models (regression, piecewise, and k-nearest neighbor models) and achieve good prediction results for movie grosses. Therefore, we believe that Lydia data could be used in forecasting other financial variables as well, for example, stock returns, volatilities, trading volumes, or firm earnings.