The question of how useful social media can be to buy-side traders has been in the spotlight of late following Derwent Capital Markets' recent announcement that it shuttered its Twitter-based hedge fund last year after only a month of trading.
Rather than continue operating a hedge fund off a technology that analyzes tweets to predict which direction the market's heading, the firm decided it would be best served selling its platform as a tool for retail investors and traders.
In interviews with market participants of all stripes, Advanced Trading found that there was much skepticism surrounding Derwent's strategy, which nevertheless managed to earn a return of 1.86 percent during its lone month of trading. But even as social media-based sentiment analysis struggles to catch the buy side's imagination, experts say it still holds strong potential as a trading tool.
In an interview with Advanced Trading John Bates, the founder and chief technology officer of Progress Software, explains why platforms like Derwent's are likely to become a component of a buy-side trader's workflow five years from now.
Is the buy side at a point where they're just exploring this, or are firms now really beginning to add social media or news-based sentiment analysis this into their workflow?
Bates: I've seen a few examples in trading, but most traders would scorn this. I remember when this came out I got an email from one of our customers saying he'd rather put his private parts in a guillotine than trade on Twitter. I did take the pulse of what people thought about it and that was the most memorable one.
But I think you're going to see it. Just like when news analytics came out, people were trading on news for a while using kind of a fuzzy logic approach. We certainly had customers that were doing it. But then firms like Dow Jones and Reuters started putting tags in news and making it a sort of a structured market data product. There are a few firms who use that in their strategies. I don't think it's that widespread and there's even less people using Twitter as a trading signal, but it's happening.
Right now I think it's just one indicator that might be useful for things like finding the bottom of a market or determining whether people feel good or bad about the economy. There's different ways of using sentiment as an input into a strategy to help a human make a decision, but I'm not sure about it being the only input.
How useful will this technology be to the buy-side trader five years from now?
Bates: Social media is here to stay. No doubt it'll evolve into something else, but we have to take it seriously and we have to find the right ways to use it. Look at the Bloomberg and the benefit of the Bloomberg network. I could argue that Bloomberg was in the cloud years before everybody else, and that's social media years before everybody else. The benefit of the Bloomberg system isn't that it's most fantastic trading system. You're part of this network where you can communicate with all these people and that's somewhat more of a trusted network where you know people and whatnot. So that's one model that's been proven of how social media can work.
How does news-based sentiment analysis compare with social media-based analysis?
Bates: It's used more than social media, but it certainly isn't a dominant. It's a small percentage of firms which do that. Now I think the news is more trustworthy than social media because when you're doing it with news you're getting it from a trusted source. You're getting it from a Dow Jones, where assuming it's from a secure channel, you know that's coming from a trusted party. Now in social media like Twitter, I can tweet that Facebook's stock is going to go through the roof and I just made it up. It's not a trusted source.
That's why I think that taking the sentiment on a global basis is more useful because if you've got a million people, there will likely be some people who are making things up. But on the whole, you're getting a better signal-to-noise ratio if you roll up the all of the sentiment and use it as an economic indicator, and you're sort of balancing out the noise and the people who just make stuff up.
With the news it's more trustworthy and more event-based.