Extracting Alpha Information from Social Media Sentiment Data

In recent years, sentiments accumulated on social media have received considerable attention as predictive indicators for asset markets. Sentiment indicators are often appealing to investors since financial markets are known to be influenced not only by economic fundamentals but also by emotions.

• Twitter has drawn most attention as a source to derive research insights from social media. There are, however, many other social media sources with specific focus on financial markets which have not been widely adopted for data collection.
• Introducing reports, sentiment indicators with different weighting methodologies: unweighted raw message volume, exponential smoothing and average smoothing based on median. Using an “alpha”-factor for exponential smoothing.

StockPulse sentiments indicate weekly winners and losers

Two weekly adjusted stock buckets are put to the test by comparing their performance with the market over a time from 2013 until now. We allocate stocks carrying positive sentiment into one bucket and those with negative sentiment into another. The bucket formed by stocks with positive sentiment outperforms the market (Sharpe of 0.79 vs. DAX of 0.52), the one with negative sentiment stocks underperforms. A German standard portfolio consisting of 40 percent DAX futures and 60 percent Euro BUND futures can be improved significantly by adding a StockPulse model based on this insight. If you are interested in a detailed description of these findings please use the contact form below. —

[ninja_form id=4] —

Dataset Featured in this Post:

• German 30 Top Stocks

  Blue Chip Companies