Step One: Follow lots of people.
If you want to make millions of dollars and beat the stock market using Twitter – the social media king of mini-messages – you are going to have to read a lot of tweets. About 100 million a week. That’s what a London hedge fund Derwent Capital Markets, which launched earlier this month, is doing. And it has research to prove it could work.
(LIST: The 140 Best Twitter Feeds)
The fund isn’t out trolling the twittersphere looking for stock tips or leaked information from insiders or hoping to catch some piece of corporate news before others. In fact, the fund is not even specifically following tweets that are about the economy or stock market. Instead, it is trying to gauge the mood of twitter users generally and use that information to guess whether the stock market is headed up or down. What the fund and its founder Paul Hawtin are basically saying is that all that stuff that most people look at – economic data, corporate profits, price-to-earnings ratios, doesn’t really matter. Or at least not nearly as much as most investors think. Here’s why:
I remember a few years ago when I used to cover the market and stock investing exclusively running across an odd study. It found that the stock market generally did better on sunny days than on cloudy ones. The theory was better weather put traders in a better mood. And therefore they were predisposed to react positively to market news. In effect, what the actual news is doesn’t matter. What you really need to predict is how people or the market are going to react to the news is what matters. That’s basically the same strategy that Derwent Capital has. Except instead of watching the weather, it reads tweets.
(LIST: 10 Ways Twitter will Change American Business)
Derwent Capital’s strategy is based on research that was led by Johan Bollen, who studies behavioral economics at Indiana University. Bollen found that there was a strong correlation between the mood on twitter and the stock market. When tweeters seemed generally upbeat and calm, stocks tended to go up for the next few days. When the opposite was true, the market tended to head south. Bollen determined if you were to buy stocks when the twitter mood was improving, and sell them when the twitterverse turned gloomy, you would be able to beat the market, and therefore the majority of investors out there. And that’s what Derwent is trying to do.
(LIST: Top 10 Twitter Controversies)
Step 2: Feel the emotional side of a tweet.
The hedge fund looks at 100 million tweets a week – through a specialty data provider – I think this is more than TweetDeck can handle – and feeds those tweets into its computers. It then uses a formula, created by Bollen, to score each tweet on six different measures of positive attitudes, which he has labeled calm, alert, vital, sure, kind and happy. It comes up with an overall measure of what Bollen calls subjective well being for each tweet, and then adds up all those scores to get a number that gauges the mood in general on twitter. Bollen won’t say exactly what phrases gives you a high or low score. He says his computer looks at sentence structure as well as specific words. But in a paper from earlier this year he did give some examples of tweets he has captured and analyzed in his research. Here’s a tweet that got a high SWB score:
So….nothing quite feels like a good shower, shave and haircut…love it
And a tweet that got a low SWB score
She doesn’t deserve the tears but I cry them anyway
And that might bring up what is the problem with the strategy. Selling stocks because people (teens probably) are tweeting about their break ups seems crazy. You could argue that the effect of one break up tweet, or even dozens will get lost among 100 million tweets, but imagine some kind of cultural event that causes a lot of people to express sorrow, something on along the lines of Michael Jackson dying. You would expect an event like that to produce a huge number of low SWB tweets. Yet, do we really think the death of a pop singer or celebrity would make the stock market go down?
But whether you can or can’t make money in the market following twitter, the fact that a hedge fund is trying to is reigniting the debate about a market theory called the Wisdom of Crowds. In the 1990s and mid-2000s, there was a growing belief that the market in general was smarter at determining the right value for things, like stocks or mortgage bonds, than any one individual could be. Markets sprung up that were supposed to predict elections and even terrorist attacks. The financial crisis seemed to put an end to that theory. We now know that leading up to the financial crisis the crowd seems to have acted pretty dumb. Markets in fact may just be random.
I called some of the economists and academics who had poked holes in the wisdom of crowd theory, and they seemed to be surprisingly interested in the twitter fund. They didn’t think Derwent’s strategy was as dumb at all. Bollen, himself, is a former wisdom of crowds guy. So perhaps the crowd has something to say, we just haven’t been listening right. Either way, the launch of Derwent Capital and social media in general is raising interesting questions about how markets work. And for that reason I hope Derwent sticks around for a while.
Stephen Gandel is a senior writer at TIME. Find him on Twitter at @stephengandel. You can also continue the discussion on TIME‘s Facebook page and on Twitter at @TIME.