Google searches may provide warning signs of stock market moves
26 Apr 2013
Analysis of Google searches reveals patterns that according to research may be early warning signs of stock market moves.
In a joint study, researchers at University of Warwick, Boston University and University College London, analysed Google searches from 2004 to 2011 for studying the information-gathering process that preceded trading decisions. A report on the study was published by the journal Scientific Reports this week.
According to the researchers, there was a relationship between changes in search volume and changes in stock market prices. The researchers found that notable drops in financial market came before periods of investor concern, which was evident from the Google search terms.
The researchers said in such periods, investors might search for more information about the market, before eventually making a buy or sell decision.
The study found that people would search for terms like "debt" before the sold stock at a lower price.
According to researchers, Google Trends data not only reflected aspects of the current state of the economy but could also provide some insight into future trends through the search behaviour of share traders.
They added, the results further illustrated the exciting possibilities offered by new big data sets to advance understanding of complex collective behaviour in society.
They found that as the search volume on generic terms such as "debt", "portfolio" and "stocks" fell, the Dow Jones average tended to go up - and vice versa.
According to their study an investment strategy based on the search volume data between 2004 and 2011 would have made a profit of 326 per cent.
The analysis in Scientific Reports which was based on publicly available data from the Google Trends service, joins an ever-increasing array of "big data" studies in which aggregated data are beginning to provide striking insights into behaviour.
Web searches had become integral to decision-making, and thanks to its dominance among search engines, Google data had already proven their worth in big-data studies.
According to Google's own researchers, searches can track the spread of influenza and more recently showed that they "predict the present" as regards economic indicators.
The Bank of England, in 2011 determined that searches for relevant terms could even predict house prices.
According to the new report straightforward analysis of interest in general finance-related terms could be a good predictor of overall market health.
According to Suzy Moat of University College London, co-author of the paper, the researchers were intrigued by the idea that stock market data served as a really large record of the actions people took in the stock market but did not necessarily tell about how people decided to take those actions.
