LONDON (CNNMoney) — Trying to predict when stocks will go up or down? Google may have the answer.
New research published in the journal Scientific Report shows that you can use Google Trends to track the search volume of important financial terms, which can indicate whether markets are set to rise or fall.
Researchers from the U.S. and U.K. found that, between 2004 to 2011, when search volumes rose for terms such as "debt," "money," and "unemployment," markets generally fell. When search volumes for those words declined, markets often rallied.
The report's co-author Helen Susannah Moat, a social scientist from University College London, explained that when people are Googling financial terms such as "inflation," "economics," "and "NASDAQ," it indicates that they're getting concerned about the markets and are likely to start selling.
When they're not searching those terms, they're probably feeling content and markets will push higher.
The researchers looked at nearly 100 different search terms and found that "debt" was one of the best terms for predicting market moves between 2004 to 2011, leading the researchers to a 326% profit when they tested out their hypothetical trading strategy on the Dow Jones Industrial Average.
Co-author Tobias Preis, a behavioral finance professor at Warwick Business School, said that when search volumes for the word "debt" started rising, the researchers would short the market for a week.
Then when search volumes fell, they'd go long for a week, Preis told CNNMoney. This weekly long-short trading strategy led them to the 300%-plus gain, but a long-only or short-only strategy would have also yielded some good returns, he added.
But investors shouldn't start trading with this strategy immediately, the researchers warned.
Preis said that times have changed and words like "debt" might no longer be useful at predicting market movements. The strategy needs updating as new keywords become more relevant to the markets.
Furthermore, Moat pointed out that now that their research paper has been released, other investors will try to use the strategy, diluting the impact in the long run.
This study follows research from 2010 that showed that weekly transaction volumes for S&P 500 companies were correlated with online searches related to those companies.