Power of News @ "Battle of the Quants"

Last week, I attended the "Battle of the Quants" conference that took place at the New York Marriott Marquis at Times Square. One of the discussions was on the topic of how to leverage news in quantitative strategies, which I obviously found to be an interesting discussion. Amongst the panelists were Armando Gonzalez from RavenPack, Rob Passarella from Dow Jones, Richard Brown from Thomson-Reuters, Christopher Ahlberg from Recorded Future, and Ben Weiss from StockTwits. The panel was moderated by Paul Hawtin from Derwent Capital Markets.

Hearing the discussion, I decided to dedicate this blog posting to social media and to some of the challenges facing investors who want to trade of information coming from the likes of Twitter, Facebook, or from general blog content.

While I believe that social media contains information that could potentially be leveraged in a professional trading or investment environment, there are some real challenges to making sense of it all. The wisdom of the crowds is a powerful force, but there are also definite limitations and concerns. Below, I have listed just a few of them:

• Needle in a Hay Stack: Even though social media may be useful in gauging the general mood or sentiment of the public or how people feel about their new iPad, it's by far more challenging to identify creditable market moving events such as merger or takeover rumors, the loss of business deals, bankruptcy fears, etc. Assuming that we have search and filtering technology to find such events, the question remains on whether the information is creditable enough to bet money on it - is someone anonymously trying to talk up a stock or do they have a real information edge?
• Do I Feel Like Blogging Today: In contrast to professional media, very few can make a living out of blogging and less so posting on Twitter. A person may be "online" today but be gone tomorrow making reporting consistency a real problem - how to know if you’re tracking the right sources at the right time?
• Historical Back-Test: For Quantitative Researchers, testing of hypotheses or backtesting of strategies are key when building investment or trading strategies. Often, building such archives are challenging based on social media not only due to the "do I feel like blogging" issue, but also since the landscape has changed dramatically in recent years. For instance, the number of people using Twitter has exploded just in the last two years - raising concerns about the usefulness of a backtest in earlier years. From my experience, how the market interprets certain news events can change dramatically over time. Therefore, one of the challenges trading of social media is to identify and understand these trigger points, which becomes difficult if the historical archives are short and inconsistent.
• Large Cap Bias: No doubt, even in professional media large cap companies get greater attention creating a large cap news bias. Still, I would argue that the bias is much greater in social media. You will probably find it difficult to get any meaningful news flow outside the top 100 largest companies in the world.
  

From a theoretical point-of-view, I can see the “attractiveness” in trading of events that might get broken early on some blog or through Twitter. My concern is though, that it will be difficult to feel confident enough to invest on a post from an anonymous blogger (typically with names like bigbucks1987, showmethemoney311, etc..) . There may be some blogs out there that could be interesting to track, and which might pass the "credibility filter" - but there are only a few. I think at best, one should focus on the "wisdom of the crowds" rather than the "wisdom of individuals" - then perhaps one could find an interesting strategy.