Study: Google Searches Can Improve Stock Market Predictions
WASHINGTON (CBS DC) — Analysis of financial terms being searched on Google can be used to predict the direction of the stock market.
According to new research from the Warwick Business School in the United Kingdom, the financial world doesn’t simply respond to online searches, instead, the Internet searches can also predict what the stock market will do.
The findings appeared to show that people do more Google searches on terms such as “stocks,” “portfolio” and “economics” when they are worried about the state of the markets, Tobias Preis, associate professor of behavioral science and finance at Warwick Business School, told Live Science.
In previous research, Preis and his colleagues found a correlation between the amount of Google searches for a company’s name and the number of times that company’s stock was bought and sold.
But that method could not predict a stock’s price.
However, in this research they looked at the volume of searches for 98 terms, such as “metals”, “stock”, “finance”, “forex”, “house”, “unemployment” and “health” as well as non-specific or neutral words, such as “ring”, “train”, kitchen” and “fun”. Next, they constructed a virtual portfolio investment in the Dow Jones Industrial Average (DJIA), with a strategy based on search volumes that occurred on Sundays.
If the search volume for that day was high compared with just a week earlier, the Dow Jones investment was systematically sold at the closing price the following day, and then repurchased at the end of the first day of trading in the week after.
On the other hand, if the volume of searches on Sunday was low in comparison with the previous week, the researchers “bought” the next day.
With the power of hindsight, trading on the basis of Google search volumes would have led to significant investment gains, Mr. Preis told Live Science. A short-term trading strategy constructed around searches for “debt”, for instance, would have returned 326 per cent between 2004 and 2011.
Google releases data each week showing the volume of searches for specific keywords, which provide the raw material for the analysis. The increasing availability of large amounts of search data has given rise to a series of “big data” attempts to forecast financial markets.
According to the Financial Times, the work was funded by a U.S. government program which was set up to study the predictive power of many different types of data. Mr. Preis said his group was in talks with numerous investment groups about practical uses for their research.