As if you needed a reason to follow your favourite celebrity on Twitter, but here’s one anyway: it might turn you into a savvy investor. Now that the random-walk theory and efficient market hypothesis have outlived their usefulness, Twitter is being called upon to work its magic in predicting stock market fluctuations. Researchers Bollen, Mao and Zeng, using behavioural economics, found that they were able to predict the daily up and down movements of the Dow Jones Industrial Average during a period in 2008 with 87% accuracy. They analysed 9.8 million tweets to develop an algorithm that categorises the mood of the twitterers. This allowed them to predict ups and downs by signalling sentiment.
But before we get swept away with excitement about finally having found a way to predict the whims and vagaries of the market, it’s worth remembering that such links have been made before—the Super Bowl, the Mets winning the play-offs and gains/losses in the month of January. Here too, there is the possibility that sentiments on Twitter are caused by the same factors that impact the stock market or it could just be a outlandish correlation. Regardless of whether Twitter will ever be used to cheat the market, the study does show that social media is emerging as an extensive yet cheap mining ground for data that may serve as a useful basis for scientific exploration.