Home >opinion >online views >Google and the stock market

Can Google predict the stock market?

Apparently, it can.

It was only a matter of time before the enormous data thrown up by human interaction with the Internet would be used to predict stock market trends. The theory itself isn’t new. As early as 80 years ago, economists such as John Maynard Keynes were studying the effect of investor emotions on stock prices.

Now, new means to quantify this has emerged with the rising use of the Internet, mobile phones and social networks. As The Economist pointed out here, scientists are using this data from social networks to come up with political insights or even predict epidemics using data on international flights and school holidays.

In a paper released on Thursday, three scientists have claimed that using the search volume for different terms on Google, they are able to understand large-scale collective decision making in the financial markets.

Written by Tobias Preis of the Warwick Business School, Helen Susannah Moat, a physicist from Boston University, and H. Eugene Stanley from the department of engineering at the University College of London, it is titled “Quantifying Trading Behavior in Financial Markets Using Google Trends and can be found here.

Preis and Co write in the paper that, “Google Trends data did not only reflect aspects of the current state of the economy, but may have also provided some insight into future trends in the behaviour of economic actors."

They used the search volume data for a set of 98 terms such as debt, colour, stocks and housing to predict the changes in stock prices. The conclusion: drops in financial markets may be preceded by an increase in Google search volumes for market-related keywords as investors become more concerned and search for information.

Thus, a trading strategy based on volume of the search term debt, the best performing keyword in their analysis, would have yielded a return of 326% between 2004 and 2011, the scientists show. In comparison, a buy and hold strategy would have yielded 16%.

One caveat: These trading strategies based on search volume data are more successful for US markets. That is in line with the assumption that the population of Internet users there contains a higher proportion of traders on the country’s markets, the paper says. The scientists showed that the results hold for both long-only and short-only strategies.

Words such as debt, colour, stocks, restaurant, economics, portfolio, and housing were the ones which yielded the highest relative returns. On the other hand, trading strategies based on words such as home and garden yielded negative returns compared with a random trading strategy.

A second caveat: Don’t use the search volume data for “fun" to formulate a trading strategy. After all, stock trading is serious business.

Subscribe to newsletters
* Enter a valid email
* Thank you for subscribing to our newsletter.

Click here to read the Mint ePaperLivemint.com is now on Telegram. Join Livemint channel in your Telegram and stay updated

My Reads Logout