Crowd-Sourcing Stock Picks

At the end of March, researchers from the University of California, Riverside released findings from a study of Twitter users. The scientists went looking to see if there was any sort of interrelation between Twitter activity and the stock market, and were surprised by the correlations they found. A high volume of Twitter discussion of a particular issue tends to mean a high volume of trading in that stock the following day.

That’s a provocative finding, although it’s of limited utility. More useful was the somewhat weaker finding — in that the correlation wasn’t nearly as strong — that higher Twitter activity resulted in a higher stock price. The more people tweeted about a stock, the better it performed. That held even though the researches didn’t control whether the Twitter conversations were positive or negative. As long as people were talking about a stock, it did well.

This wasn’t the first such study looking at the effects of social media on stock prices. Several others have found similar effects: the more optimistic the Twitterverse is, the better the Dow performs, and the more social media buzz there is about a stock, the better it does. Think of it as crowd-sourcing stock picks. It’s not so much a “wisdom of the crowds” situation as a measure of interest. Presumably Twitter and other forms of social media reflect how much interest the larger world has in a particular stock as well.  

Perhaps the first such study of Twitter’s effects on the markets was released in 2010 by two economists from the University of Indiana and one from the University of Manchester in the U.K. That one didn’t study stocks but rather the general mood of the Twitterverse, through analyzing nearly 10 million tweets issued between February and December of 2008. By searching for such expressions as “I feel” and “I am feeling” and “makes me,” the researchers were able to tell more or less how the Internet was feeling. The researchers used two Twitter “mood tracking tools”: OpinionFinder, which measures positive versus negative mood, and Google’s Profile of Mood States, which measures six dimensions of moods: alert, sure, vital, kind, happy and calm.

Keeping track of those feelings, the economists write, helped “find an accuracy of 87.6 percent in predicting the daily up and down changes in the closing value of the DJIA.” In other words, as goes the Twitterverse, so goes the Dow.

As a result of this study, a London-based firm called Derwent Capital Markets launched a hedge fund in February 2011 based on its findings. The so-called Twitter Fund makes its investments by evaluating whether people are generally happy, sad, anxious or tired, with the theory that the market will quickly follow the mood. The fund’s methodology examines a random 10 percent of all Twitter feeds, weighs the ratio of positive to negative comments and uses a Google’s six-mood matrix