Markit has launched a suite of social media indicators constructed to capture timely information gleaned from Twitter posts. The measures are based on analysis of the text content in daily Twitter posts. Tweets are filtered for financial trading relevance and scored for market sentiment content. Tweet scores are then aggregated for each stock to produce a sentiment measurement from which the indicators are derived.

Markit has partnered with Social Market Analytics (SMA) to provide signals of investment sentiment to help customers inform their trading strategies or investment process.

“What we think is unique is that SMA has built this product over the past two years specifically for the financial markets,” said Chris Hammond, director of the research signals team at Markit. “Some of the other products out there are look at sentiment of Tweets, but there are two key areas that SMA focuses on where the others are not.”

The first key area is the source of that Tweet. SMA has a ranking methodology which analyzes who’s tweeting about the stock and determines whether they are a valid source of information, whether they have a lot of followers and whether they have a history of posting indicative Tweets.

The second area that is unique is that SMA is looking at the Tweet sentiment in relation to the financial market. For example, SMA excludes Tweets about Apple where people are saying they like their iPhone; it is looking specifically for financial terms. “They have a natural language processing dictionary that is built specifically to analyze the sentiment relative to the financial markets,” Hammond said.

The social media indicators enhance the Markit Research Signals suite of investment factors which can be used to evaluate the expected performance of stocks-based sentiment indicators.

“We house a library of 400 factors that drive stock returns, including momentum, valuation, analyst sentiment, and quality,” Hammond said. “We have verified that the social media signals add value in that they have predictive power in the stock market, and we’ve also verified that the factors are different from other signals in our broad library. This lack of correlation is a really exciting aspect of the SMA data set.”

The research signals business that Hammond runs is part of Markit’s Indices business, led by Tim Sargent, which also includes Markit’s large economics business and their PMI surveys.

The library comes from a business that Market acquired in 2011 called QSG, which was founded by Sargent. “In 2001, we started with about 100 of these signals that were routinely used by hedge funds and investment managers, and we had an analyst team reviewing new data sets and new ideas and slowly but surely built up this library so after more than a decade we grew to about 400 signals,” said Sargent. “It’s really put us in a unique position to evaluate these new third party data sets.”

Markit’s factors span 12 categories, including measures of relative value, earnings momentum, earrings quality and price momentum. The factor analytics platform allows custom model building and strategy deployment for equity and fixed income.

“We’ve had great success in the corporate bond area, especially with our global iBoxx indices,” Sargent said. “A big part of the roadmap for teams here is to extend that value across more asset classes and develop more unique and innovative index offerings. Almost all of our work now is trying to identify things that aren’t already well understood or well- leveraged in the marketplace.”

Hammond’s team is working on a new set of economic factors related to Markit’s PMIs and sector performance. “You’ll continue to see us come out with those specialty areas that we think are unique opportunities to develop these return patterns that are of interest to investors,” Sargent said.

Featured image via Flickr/Garrett Heath under CC