Tesco
News versus Sentiment
Report

News versus Sentiment

Predicting Stock Returns from News Stories

Steven L. Heston and Nitish R. Sinha
Federal Reserve Board, 2016
Get the Report

Read or listen offline

automatisch generiertes Audio
  • 0.5×
  • 0.75×
  • 1×
  • 1.25×
  • 1.5×
  • 1.75×
1×
automatisch generiertes Audio

Editorial Rating

8

Qualities

  • Analytical
  • Scientific
  • Eye Opening

Recommendation

Algorithms have been buying and selling stocks for some time now, and in the past few years, as many media outlets look for cost savings, algorithms have also begun writing news stories. So the time may be right for artificial intelligence to take over the management of investments, sending the multitrillion-dollar investment industry into a paradigm shift. An intriguing new paper from economists Steven L. Heston and Nitish R. Sinha demonstrates that artificial intelligence and algorithms can successfully make share price predictions using news stories. getAbstract recommends this thought-provoking but highly technical description of novel research on news sentiment analysis and stock prices.

Take-Aways

  • Research shows that “neural networks” – systems that can assess the interactions of various functions, much as a biological central nervous system does – and content analyses of news stories can predict stock returns, corporate distress and bankruptcy.
  • Articles containing information on companies should have a lasting impact on the value of that firm, whereas news stories including “sentiment” should have fleeting effects.
  • Positive media can forecast optimistic returns for about seven days, while negative articles foretell adverse returns for up to three months.

About the Authors

Steven L. Heston is a professor of finance at the University of Maryland. Nitish R. Sinha is an economist at the Board of Governors of the Federal Reserve System.

Comment on this summary or Diskussion beginnen