NYSE Enters Sentiment Analysis Market with SMA Alliance

The analytics goes through a rigorous filtering process, through which 90 percent is filtered out to ensure only relevant information is processed, according to SMA founder, Joe Gits.

A sentiment score, dependent on seven metrics, is then produced to signify the degree to which a particular stock has deviated from its base sentiment.

The rise of social media over the last few years has not gone unnoticed within the financial services industry, says Gits, whose previous company, Quantitative Analytics, was acquired by Thomson Reuters in 2007.
"The last year I was at Thomson Reuters, I didn't go to a meeting where someone didn't ask me about social media, how it was breaking news, and whether there was any way to quantify this data so that quantitative analysts could look at it," he says. "And there really wasn't, up until we launched."

Filtering Out the Noise
SMA is not alone in its provision of sentiment analysis, but Gits believes it possesses a number of facets that are unique, including the sophistication of its filtering process, which takes into account not only the number of mentions a stock receives, but the individual behind the mention.
Gits questions the value of raw sentiment, saying that the intrinsic "noise" of social media such as Twitter means that many tweets will lack value, at least as far as their utility for trading algorithms is concerned. Instead, he speaks of "indicative tweets" as those which truly possess value.
"The filtering is an absolutely key component because most of the Twitter feed is irrelevant for stock prices," he says. "And that's why I believe that a lot of other systems fall short. It's not necessarily that you want to get as much information as you can ─ it's how you filter that intelligently that adds the value."

Staying Ahead
Gits believes that identifying sentiment generated through social media can enable firms to get ahead of the game, especially now that social media has become one of the most regular providers of breaking news.
"There are a lot of examples where people are tweeting about particular securities 15 or 20 minutes before a lot of major news outlets pick it up, and obviously that's an eternity in our business," says Gits.

There are a lot of examples where people are tweeting about particular securities 15 or 20 minutes before a lot of major news outlets pick it up, and obviously that's an eternity in our business.

And it isn't just the speed of the data that can prove useful for trading firms. NYSE Technologies' Tom Watson, product manager of US derivatives market data, says there has been a noticeable improvement in the validity of the data produced via social media.

"The data has been changing," says Watson. "Twitter has been changing. There have been so many different users that have been coming onto the platform that the validity of the data has changed a little bit over the last couple of years. It's becoming more and more valid as it's been used more for business purposes over the last year."

NYSE Technologies will become the exclusive reseller for the SMA sentiment feed during this quarter.