This week, Derwent Capital Markets, a London investment firm, launched a $40 million hedge fund that will use Twitter to guide its investments. The world's first social media-based hedge fund will monitor a selection of tweets in real-time to feel out market sentiment before placing its bets.

"For years, investors have widely accepted that financial markets are driven by fear and greed but we've never before had the technology or data to be able to quantify human emotion," Paul Hawtin, the founder of Derwent Capital Markets, wrote The Atlantic in an email. With Twitter, he added, investors finally had a window to the world's fear.

Why would experienced investors cede their expertise to something as crude as Twitter for guiding tens of millions of dollars in bets? Hawtin won't share how his algorithm works. But a team of professors from both sides of the pond have already answered that question for us.

HOW TWITTER PREDICTS THE STOCK MARKET

If Twitter can predict the public's mood, and the public's mood can predict the stock market, can Twitter predict the stock market?

That's the question researchers at Indiana University and the University of Manchester asked when they set out to measure Twitter's forecast of stock fluctuations. The professors harvested tweets for key words and plugged them into an algorithm to determine the mood of the broader market. Using this mood index, the professors predicted the Dow's daily fluctuations in 2008 with an astounding 87 percent accuracy.

In particular, the professors found that changes in values of "calm" can predict stock market fluctuations, as you'll see in the graph below. In red, the authors produce the "score" of calm moods found in Tweets. In blue, they track Dow changes delayed by three days. Where the two graphs overlap, the calm index predicts the DJIA's closing value three days later.

calmdow.png

The red and blue lines often trace each other, suggesting that Tweets can predict the stock market. But there are some alarming divergences. Around the "bank bailout" of October 13, 2008, Twitter calmness was at a year low, but the stock market soared. The authors explain that "the deviation between Calm values and the DJIA on that day illustrates that unexpected news [of a Federal Reserve bank bailout plan] not anticipated by the public mood."

In other words, Twitter can't predict future events. But it just might be able to predict how today's moods impact tomorrow's prices in the absence of shocking, disruptive events.

THE BANK OF TWEETS

Using Twitter to follow the stock is, self-evidently, more than an academic observation. Like following the needle of a barometer to predict rain and sun, investors could theoretically follow moods of the masses to take their portfolios long or short depending on the course they expect to see the markets take, Hawtin said.

"We made the decision to set-up a quantitative Hedge Fund before we came across the academic paper," Hawtin explained when I ask if his investment strategy was inspired by the paper excerpted above. "As soon as I read the academic paper it confirmed my thoughts so I arranged to meet up with Professor Johan Bollen, one of the co-authors."

The $40 million hedge fund's doors have been open for only a few days, but the firm is already targeting an annual return of 15% to 20% for its investors. "We have been testing the system for some time and the results have been outstanding," Hawtin said. If Twitter proves a trustworthy market guide, expect the firm to rake in more social media data from Facebook, Google trends and beyond.

"The greater the range and depth of social media data we can access the better," his email concluded.


We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.