On November 11 2013, a few minutes after 8 a.m. EST (Eastern Standard Time), news leaked out from a Canadian newspaper that Blackberry’s $4.7 billion buyout had collapsed. Wall Street wouldn’t find out for a full 180 seconds, when the newswires picked up the report in real time.
Investment clients at Dataminr, a New York City-based data analytics firm, had a leg up on the rest of the investing public. They received an email alert from Dataminr within seconds of the Blackberry news appearing in the Canadian newswire, and many of those clients – especially hedge funds – used the news to short the stock ahead of the rest of the investment community, who were late getting the news on Blackberry.
Another social media data analysis firm, Social Market Analytics (“SMA”), used its coverage of 400,000 Twitter accounts last August to tell its clients that positive chatter on Apple was percolating just before legendary trader Carl Icahn issued a Twitter statement stating he had purchased a huge chunk of Apple stock.
On both fronts, those early birds made a bundle on the alert and showed others that leveraging social media to get the fastest news impacting stock prices wasn’t just a theory, it was a reality.
"SSI - Social Sentiment Indicators”
As from the beginning of 2014, so-called “social sentiment indicators” - “SSI” are making big waves in stock market circles, as more evidence pours in showing that SSI really does give investors who leverage the technology an advantage over those who don’t.
According to a study from Markit, a financial data services provider, from December 2011 to April 2014 positive social media sentiment stocks have shown cumulative returns of 76% compared to -14% from negative sentiment stocks. Back in 2010, the Indiana Business School released a study in terms of which it reported Twitter data could predict the Dow Jones industrial average with 87.6% accuracy.
Tweets That Beat the Street
While Facebook offers some data mining opportunities, Twitter is the real hotspot for social indicator analytics. Twitter is a beehive of social media activity, with 645 million active users and 135,000 brand new users every day. Until 2012, however, the technology didn’t exist to splice, dice and slice Twitter feeds to discern fresh trading data. Once social sentiment indicator analysts began figuring out how to quantify all that streaming social media data - and offered the results to professional investors - they made good profits.
The technology, unsurprisingly, is highly sophisticated. The methodology relies on algorithms designed around key Twitter criteria – including averages, change, volume, volatility, dispersion of tweets and risk - to generate what company analysts refer to as “S-Scores," which are evaluations based on all of the above algorithms that mirror sentiment on a given stock, over a historical period of time (called a “lookback” period.)
These sentiment scores indicate whether the prevalent chatter is good or bad news for a given stock. With that information in hand, clients can act accordingly and trade the stock based on the sentiment score.
Too Much Data
That’s not to say social media indicators are easy pickings. SMA note that 90% of all the Twitter feeds that their analysts dissect are discarded - it’s the other 10% that reveal investment opportunities investors are clamoring for. Social media investment indicators do have a down side, however. It’s fairly easy for con artists to create Twitter feeds similar to publicly traded firms and to throw investors off the right track by tweeting false news about a company, like the buyout of an industry competitor or “foreshadowing” new product launches. Investment fraudsters purchase shares of the stock in advance and profit from Twitter watchers who fall for the bogus news.
The Bottom Line
No doubt, social media investment analysis is an ascending, but very much nascent, technology; so far it’s a winning one for early investing birds. However, the old data capturing adage, “junk in, junk out”, is certainly something to bear in mind. Given that Twitter is infested with spam "robot" accounts (those with the "sequential" account names, suspicious-looking tweets and extreme following/follower ratios) and Facebook estimates it has 83m fake accounts, it's highly likely that any social data haul will net spoof accounts.
So although a fascinating and perhaps useful concept it really is a case of “buyer beware” and basing an investment decision solely on social media indicators may not yield the desired results. #LOL. Not.