Over a million developers have joined DZone.

Can Google predict the stock market?

DZone's Guide to

Can Google predict the stock market?

· ·
Free Resource

Back in 1998, long before Google was born, the movie Pi was released.  In it a mathmatician believed he had unlocked the secrets to the stock market, and could therefore predict its fluctuations.  Suffice to say once Wall Street became aware of this they were hot on his heels.  As an indie movie it didn’t make much of a splash.  15 years later however, researchers believe they may have uncovered the secrets of the markets.

They believe that analysis of Google searches can provide accurate predictions for the movements of share prices.  For instance when searches decreased for terms such as debt, portfolio and stocks, the Dow Jones tended to go up, and vice versa.  If you had invested based upon their findings, you would have made a profit of 326%.  Nice.

The research itself was based upon data from Google Trends.  They tracked the volume of searches for 98 finance related keywords between 2004 and 2011.  They then compared this with movements in the stock market over the same period to see if they could uncover any patterns that revealed early warning signs of major market fluctuations.  They then tested these trading strategies to see how effective they proved.

“We were intrigued by the idea that stock market data serves as a really large record of all the actions people take in the stock market, but don’t necessarily tell us much about how people decided to take those actions,” said Suzy Moat of University College London, co-author of the paper.

“We wondered whether by looking at Google, we could get some insight into some early information-gathering stages of how people make decisions,” she continued.

The basic trend saw drops in search volume for financial keywords correspond with a rise in stock prices.  Testing this theory in the market saw them earn a profit of 326%.  If they had merely followed tracker style indices over the same period, the gain would have been just 16%.

Authors believe this shows very real evidence for loss aversion, a well known phenomenon whereby people are reluctant to cut their losses because it showcases failure on their part.  So they hang on in there in the hope that things will recover and they can sell at a profit, which of course it seldom does.

Just as in Pi, the financial industry are pretty keen to get their hands on this research.  The researchers have secured a grant to develop a software platform to take advantage of the findings.

Suffice to say of course, once people become aware of this trend then that in itself will change behaviour, and new models will be needed to take account of the new landscape, but for a short time at least this will provide an interesting new look at the vagaries of the markets.

Republished with permission


Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}