Is Social Media a Proxy for Momentum & Does it Lead or Lag the Market?

Social media is a unique form of information dissemination.  To illustrate the unique signal represented from social media a partner of ours, Markit,  did a study to look at the correlation of social media based factors with traditional financial factors.  The resultant table below illustrates the unique information being presented with social media signals.  As you can see all correlations are near zero.

Correlations

Does social media lag or does it lead?

Many times social media discusses events as they unfold.  Some examples include:

Earnings Reports – $CLC CLARCOR Reports Record Third Quarter Diluted Earnings Per Share

New Products – Mobile Weekly Notes: Record iPhone Pre-Orders, Android One Platform, BlackBerry

Changes in Ratings – Dunkin Brands Group Receives “Overweight” Rating from Barclays

Technical Analysis – CL_F 87.40 point of interest.  Price Structure is short on all frames.  Looking for shorts now

Prior to an event (like those named above), social media actively discusses expectations and is likely to report the actual event first.  We find the ability of social media to beat traditional news sources is dependent on the size of market capitalization.  The largest market cap securities are actively followed by ALL media so social media’s lead-time in these cases is short.  As you move out of the top 100 market cap stocks, social media can beat traditional news outlets by minutes.  More importantly, social media discusses expectations for the potential impact on securities prices and consensus are formed.   Professional traders will discuss their opinions on the impact of these events.  In this regard, social media always beats the general public.  People are posting opinions immediately as events unfold.  When speaking about investments and projected movements of securities prices, we have found it’s not just what’s being said, but who is saying it and how many people are saying it.

Advertisements

Predictive Power of Social Media data in Sectors

Below is the out-of-sample cumulative open to close return chart for stocks with extreme sentiment scores prior to market open.  As you can see from the chart stocks with high sentiment scores (SMA S-Score > 2) subsequently outperform while stocks with low sentiment scores (S-Score <-2) subsequently underperform.  The benchmarks in this chart are the S&P 500 constituents open to close return and all equities with sentiment signals.

Universe

“Are some sectors better than others for social media?” is a question I regularly receive.  Like all predictive models some sectors are better than others.  Sectors with Business-to-Consumer focus are more predictive because there is more consumer commentary on Twitter in addition to the standard trading and analyst commentary.  Below are the cumulative return charts for consumer facing sectors.

Services

High sentiment stocks in the Services Sector have a subsequent open to close cumulative return of +250 percent since inception of SMA sentiment data.  Low sentiment stocks have a subsequent open-to-close return of -50% since inception of SMA sentiment data.

ConsumerGoods

Sentiment data has proven to be very predictive for Consumer Goods.  High sentiment stocks in the Consumer Goods Sector have a subsequent open to close cumulative return of +350 percent since inception of SMA sentiment data.  Low sentiment stocks have a subsequent open-to-close return of -50% since inception of SMA sentiment data.

SMA is the leader in sentiment for financial markets; we are continuously analyzing our data.  Contact me (joeg@socialmarketanalytics.com)  to learn more about SMA research on Sector and Industry performance.