Wikipedia Search Momentum and Stock Returns
42 Pages Posted: 2 Aug 2018
Date Written: August 1, 2018
We present empirical evidence that collective investor behavior can be inferred from large-scale Wikipedia search data for individual-level stocks. Using Shannon transfer entropy, a model-free measure that considers any kind of statistical dependence between time series, we investigate the statistical information transfer between daily Wikipedia searches and stock returns for a sample of 447 stocks and all trading days from 2008 to 2017. We entertain a hypothetical trading strategy based on Wikipedia search momentum that maps our findings to average portfolio returns. Results are in line with the notion of investor sentiment and the trading patterns of retail investors.
Keywords: Wikipedia, stock market, investor behavior, transfer entropy
JEL Classification: C14, G14, G15, G40
Suggested Citation: Suggested Citation