Open-Source Algorithmic Trading

The way this (ostensibly) works is, you write an algorithm (MATLAB or R) that is successful on a historical dataset (looks like the last 12 months), they validate it, and apply their existing capital to it, with you earning a cut with (ostensibly) no capital outlay on your own part.

That’s a pretty cool model. The nut, of course, is the capital distribution, both quantity and timing.

Other thoughts:

  • Enable the workflow. There needs to be a hosted test box for experimentation and writing the scripts. That’s a product unto itself (with a constant stream of feedback) but worth it. API for that data.
  • Minimize the pain. Abstract the algorithm creation process (as part of the product above) to make it as clean as possible with constant testing and adjustment tools.


-Shlok
Sign up for my newsletter.

13. March 2011 by Shlok Vaidya
Categories: Thinking | Tags: | Leave a comment

Leave a Reply

Required fields are marked *