Demand for a new hedge fund that will use sentiment on Twitter to predict the moves of the market was so great, the a financial fail whale had to be rolled out, as the fund expanded its operations. Instead of the $40 million planned in the initial raise, the Derwent Absolute Return Fund got $100 million in commitments from investors.
The hype around this kind of trading is so far based largely on academic analysis. A study out of Indiana University from last year, which claimed to predict market moves days in advance, provided the underpinning for Derwent fund. That work was put together by an associate professor. A new study, highlighted yesterday by Matthew Ingram on GigaOm, comes from a PhD student, who claims investors that followed his model would have seen a 15 percent return in the first half of 2010, compared to a drop of 8.5 percent for the S&P over the same period.
The fact that a single academic paper tested over just six months could justify a $100 million investment shows the level of froth surrounding social media right now. Stock Twits, the mashup of Twitter’s real time follow model with financial advice, has doubled it’s monthly visitors since December. The service recently rolled out a new discovery tool, so that users don’t have to be familiar with a trader to follow their advice, but can simply pick from a curated list that seems to match their risk profile.
Ingram points out that it won’t be long till someone starts trying to game this system, much as Google’s search algorithms led to an explosion in black hat SEO. But it wouldn’t even require an automated spam bot to juice sentiment in the wrong direction. As Betabeat has written before, Twitter is a powerful platform for snowballing sentiment, given the outsized influence of celebrities. Take for example, 50 Cent pumping a penny stock up 1000% with just a few tweets. Or that mentions of Anne Hathaway may drive up shares of Berkshire Hathaway.