Trending on Twitter: Social Sentiment Analytics

February 20, 2014

If you’ve spent any time on Twitter, you know that tweets can be snarky, informative, nonsensical, hilarious, and sometimes provocative. This is the “sentiment” of the Twittersphere. What Twitter does for the reputation of Lady Gaga or Bill Gates also applies to companies and stocks. Knowing the fast-moving social sentiment of a particular investment opportunity could provide a trader with that extra edge. That’s why Bloomberg provides subscribers with social sentiment tools, which categorize opinion and provide charts and visualizations based on this data.

When Bloomberg integrated Twitter feeds into our platform via TWTR in April 2013, we also provided users the ability to monitor for unusual bursts of social media chatter about a company. Using Bloomberg Social Velocity (BSV) alerts, Bloomberg subscribers can see social media volume about, say, Ford, BP, or Samsung. To track a company, BSV scans tweets and StockTwits for so-called cashtags, social media’s way of indicating news about a stock, and any mentions of the company’s name.

With BSV, users don’t have to stare at a stream of tweets all day waiting for a spike in activity. BSV does the watching for them. When tweets start jumping, an alert is pushed out.

But knowing a topic is trending isn’t always enough for investors. The real question is: why is this topic important? Is the rise of interest in a company positive, negative, or neutral?

To answer that, Bloomberg also provides social sentiment analytics. Users can set up an alert to watch, for example, all of the tickers in the S&P 500 and be alerted to those with significantly positive or negative social sentiment. BSV monitors the flow of tweets about each stock and triggers an alert whenever sentiment crosses a particular threshold. From there, subscribers can see the overall volume of tweets and the mix of positive, negative and neutral comments, and can click their way to individual Twitter postings, which are available directly through the terminal.

Sophisticated tools are used to measure a tweet’s sentiment. The system relies on algorithms developed by a team of PhDs, product managers and R&D professionals incorporating statistical models and years of Twitter history applied to financial markets.

“Social signals provide another tool for investors to identify trends and events that may move or influence markets,” said Brian Rooney, Bloomberg’s Head of Product for News, Economics and Specialized Content. “Sentiment analysis helps derive meaning from the stream.”

To see how Bloomberg captures sentiment, take a look at Regeneron Pharmaceuticals Inc., one of the best-performing U.S. stocks of 2013. This month, Twitter chatter on Regeneron spiked several times, with positive sentiment. The below chart shows how after the first BSV alert was sent indicating positive sentiment on Feb. 10, the company’s share price rose more than $20 over the course of three days.

The blue line illustrates Regeneron Pharmaceuticals’ stock price, while the green, amber and red bars indicate social sentiment and volume.

Bloomberg’s social media tools are built completely in-house and are fully integrated with investment professionals’ workflow. Bloomberg’s 3,000+ software developers and infrastructure engineers around the world do everything from data gathering to creating sophisticated scoring algorithms to building the user interface.

So consider yourself alerted. Social sentiment tools are trending… and it’s positive.