SeekingAlpha.com is a popular stock analysis website. It calls their best long and short ideas "Alpha Rich", claiming these ideas "frequently moves the stocks" (http://seekingalpha.com/articles?filters=sa-a-rich). I put these ideas to a test using Quantopian.
I manually extracted 20 "Alpha Rich" ideas and backtested them using minute-level data. The test spans roughly 2 weeks in Jun. The results:
- The return is -14.4% if we trade on the ideas after they are disclosed to the public (i.e., setting offset=0 in the code)
- The return is -0.4% if we trade on the ideas after they are disclosed to SeekingAlpha's paid pro users (24 hours before they are disclosed to the public, i.e., setting offset=-1 in the code)
- The return is 0.7% if we trade on the ideas 48 hours before they are disclosed to the public (i.e., setting offset=-2 in the code)
(The benchmark return is -1.6% in the same period.)
The results suggest that there could be some benefits to trade on "Alpha Rich" ideas if one has SeekingAlpha's pro membership, but it doesn't make sense to do so as a non-paid user. (I restricted holding time to 5 days. The conclusion could be different if one is seeking a medium to long term investment.)
As always, any ideas and comments are welcome!
