According to a recent study of Google search volumes published in the journal Scientific Reports, the answer is a qualified “yes.” Professors Tobias Preis of the Warwick Business School, Helen Susannah Moat of University College London and H. Eugene Stanley of Boston University analyzed correlations between Google Trends data and movements in the Dow Jones Industrial Average from January 2004 to February 2011.
As detailed in the research paper “Quantifying Trading Behavior in Financial Markets Using Google Trends,” certain financial terms functioned as leading indicators for stock prices over those seven-plus years. Specifically, the trio studied changes in the number of query searches for 98 investment-related words, such as unemployment, credit or stock market. Trading on one term in particular — debt — would have generated a return of 326% versus 16% for a buy-and-hold strategy using the DJIA.
But there also was a wide disparity in the usefulness of those 98 terms — and plenty of statistical noise as well. Trading on search volumes for “restaurant” or “color,” for example, would have generated significantly better results than trading on virtually all relevant financial terms.
As always, turning the theoretical into the practical is highly problematic. The authors note that the process requires further study and refinement. “It’s not as simple as just taking what we described and applying it in exactly the same way,” says Moat. “We would need a system that detects emerging trends (in financial searches), rather than just the terms we have in that paper. It would be necessary to work out which topics are becoming important.”
Preis concurs: “There are possibilities to apply these findings, but it requires that we come up with the right terms. My advice is not to do this.”