Who needs expert networks when you’ve got Twitter?
London-based Derwent Capital Markets says the social-networking system is better at gauging market sentiment than just about anything else. In particular, use of “calm” motional words on Twitter, properly analyzed, lets the firm predict where the Dow Jones Industrial Average is going in the next two to six days with a remarkable 87.6% accuracy.
What to do with this nearly perfect knowledge—without a hint of insider-trading possible, given the public nature of Twitter? Why, launch a hedge fund, of course.
Derwent’s Absolute Return Fund will debut in February with £25 million. The firm has engaged a University of Manchester computer science professor to build the trading models. That professor, Xioa-Jun Zeng, worked on a paper published in October about using Twitter to predict movements in the Dow.
The hedge fund is in talks with the other two authors, Indiana University’s Johan Bollen and Huina Mao, about coming on board.
“Sentiment and mood dramatically change the impact of positive and negative new stories,” Derwent’s Paul Hawtin told Bloomberg News. “If the market’s in a very positive and bullish mood, it can shrug off bad news—bad news comes out and you expect the Dow to fall, and it doesn’t.”
Derwint doesn’t plan to limit itself to the 30 stocks in the Dow or the index as a whole. It will also trade the FTSE 100 and FTSE 250 indices, oil, gold and other precious metals and currencies.
Despite the purported 87.6% accuracy, Hawtin said the hedge fund will target relatively modest returns of between 15% and 20% annually.
“The only risk for us is if Twitter falls away and people just don’t use it anymore,” he told Bloomberg. But with the social network’s daily post rate nearly doubling since February, Hawtin said he is confident “that it can only get bigger and better, and that more and more people will be using it to express their feelings.”