The impact of news media and affect in financial markets

Abstract

Literature on financial analysis is increasingly focused on what market participants do with information on price changes and whether all participants get the same price influencing information. The efficient market hypothesis first posed by Eugene Fama says all influential information is incorporated into price. Doubts and limitations of this theory have been expressed, leading Fama to accept a weak form of efficiency in financial markets. A variety of data such as key announcements, trading volume, consumer surveys and qualitative information like sentiment and affect may contain information not accounted for in price. For such qualitative information and news media, the challenge is in collating, processing, and aggregating this information with traditional financial time series. The influence of news media, quantified by computing methods, is not definitively described by economic theory. As such, we rely on inferences made from the modelling of the data to evaluate any potential explanatory power. To assess whether the information from news media is fully incorporated into price, we use six different statistical models and evaluate a proxy for news information against that of returns for the Dow Jones Industrial Average and New York Stock Exchange trading volume.