Googling Politics Predicts Stock Market<img class="wideThumb" src="http://www2.warwick.ac.uk/fac/soc/wbs/subjects/bsci-copy/news/stockswide.jpg" style="display: none" />
Chester Curme, Tobias Preis and Suzy Moat of WBS (along with H. Eugene Stanley of Boston University) have demonstrated that an upturn in Google searches for political terms tends to precede a fall in stock prices.
Using information from Google Trends, the researchers simulated trading over an 8 year period. Historical data from S&P 500 index was used to facilitate the experiment in which the researchers bought or sold the index based on whether the number of google searches for thematically linked groups of key words were rising or falling each week.
A control strategy of buying or selling randomly each week was used as a baseline.
The politics themed group showed the biggest net gain - a median return on investment of 38% above the control. Business related terms showed a 28% return above the control. This contrasts other thematic groups which showed no significant deviation from the control.
Subscribers to the Wall Street Journal can read that publications thoughts on the result here