Wall Street & Technology is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Infrastructure

11:40 AM
Connect Directly
Google+
Twitter
RSS
E-Mail
50%
50%

Buy, Sell or Hold With Twitter?

As a U.K. hedge fund taps the social network to predict market moves, industry participants argue that Twitter may be ineffective - or even dangerous - as a sole forecaster.

A Dangerous Game?

But even as experts concede that Twitter can be useful as a market barometer, Derwent's doubters still question whether the algorithm can be profitable. While their chief concern stems from the simple fact that random occurrences and their impact on stocks can't be predicted, they also are concerned that stock prices can be manipulated.

"Would I put my $40 million on Twitter as a sole indicator? No," says Aite Group research director Adam Honore. "There's just too much unpredictability. You just don't know who your influencer is going to be. If that's your indicator and your sole predictor, it wouldn't be rocket science to game that."

Gaming via Twitter became a concern when hip-hop artist 50 Cent's tweets urging followers to invest in a firm he co-owned netted the rapper $8.7 million in profits. In order assuage investors concerned about gaming, the researchers are developing algorithms that can weed it out, according to Bollen. In addition, the Twitter network itself has safeguards in place to eliminate spamming. But the first line of defense, Bollen says, lies in Twitter itself.

"To really effect these kinds of data feeds, you would have to create a signal that would be of sufficient magnitude to swamp out the prevailing sentiment," Bollen contends. "That would be really tough to do without someone noticing."

A Flash in the Pan?

Meanwhile, a nagging question for cynics is how a model aiming to predict markets three or four days in advance will react to events such as last year's Flash Crash, in which stocks lost billions of dollars in value in a matter of minutes before a quick recovery. How will the algorithm measure and interpret public sentiment on days during which the markets swing wildly?

Bollen argues that while stocks could fall sharply on any given day, it takes much longer for the public mood to shift. And since the Derwent strategy is not as time-sensitive as those of high-frequency trading operations, its predictions are not as vulnerable to short-term market gyrations.

"Public mood is a little like a big ship - I don't think it's quite that easy to turn," Bollen says. "In terms of the Flash Crash, what you have is a feeding process that may have been caused by trading algorithms that operate at the level of less than a second. Ours is definitely not looking to do that."

Nonetheless, experts say that while Twitter is growing in popularity on Wall Street as an indicator, the trading community largely has cast a jaundiced eye toward the idea that the social network can be an accurate market forecaster. "Many traders will tell you they wouldn't stake their reputation on it and they don't believe in it," Progress' Bates says.

"I'm fascinated to see how this fund will do."

As the Senior Editor of Advanced Trading, Justin Grant plays a key role in steering the magazine's coverage of the latest issues affecting the buy-side trading community. Since joining Advanced Trading in 2010, Grant's news analysis has touched on everything from the latest ... View Full Bio

Previous
2 of 2
Next
Register for Wall Street & Technology Newsletters
Video
7 Unusual Behaviors That Indicate Security Breaches
7 Unusual Behaviors That Indicate Security Breaches
Breaches create outliers. Identifying anomalous activity can help keep firms in compliance and out of the headlines.