Faculty of Economics and Business Administration Publications Database

Are Crowds on the Internet Wiser than Experts? - The Case of a Stock Prediction Community

Nofer, Michael
Volume: 84
Number: 3
Pages: 303 - 338
ISSN-Print: 0044-2372
ISSN-Online: 1861-8928
Link External Source: Online Version
Year: 2014
Keywords: Wisdom of crowds Stock prediction communities Social media Forecasting

According to the “Wisdom of Crowds” phenomenon, a large crowd can perform better than smaller groups or few individuals. This article investigates the performance of share recommendations, which have been published by members of a stock prediction community on the Internet. Participants of these online communities publish buy and sell recommendations for shares and try to predict the stock market development. We collected unique field data on 10,146 recommendations that were made between May 2007 and August 2011 on one of the largest European stock prediction communities. Our results reveal that on an annual basis investments based on the recommendations of Internet, users achieve a return that is on average 0.59 % points higher than investments of professional analysts from banks, brokers and research companies. This means, that on average, investors are better off by trusting the crowd rather than analysts. We furthermore investigate how the postulated theoretical conditions of diversity and independence influence the performance of a large crowd on the Internet. While independent decisions can substantially improve the performance of the crowd, there is no evidence for the power of diversity in our data.