Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
OLMAR with Fundamentals and Momentum Ranking

So, just as a side project that came from pure curiosity I wanted to see how Grant's OLMAR algorithim with Fundamentals would work with different sorting and ranking functions that Johnny Wu has been posting around Value + Momentum. This backtest worked pretty well but I've tried a few others and it doesn't really hold up that well. There's room for improvements in the algo and I'm sure if someone plays around with the fundamentals and filters they might hit on something good.

NOTE: This strategy doesn't follow any of the papers on the Value + Momentum, violates the overfitting rules and momentum kind of runs against the basis of OLMAR to a degree but it was just a fun side project because... why not?

CODE CREDIT:

  • Grant Kiehne's OLMAR with Fundamentals
  • Garyha's logging and recording features: Just look at one of his posts, his logging and reporting is awesome.
  • Based on Johnny Wu's Value + Momentum
1 response

You might also try using the actual OLMAR algo, instead of the version I cooked up using optimize.minimize, since you introduce no additional constraints. I think that the iterative solver is slower, and may not be as robust, as well (might not find a solution, when there actually is one). I don't have time now, but I'll try to provide an example, when I get the chance.

Another potential improvement would be to run the optimization every minute on a fixed portfolio, and average the weights. This might eliminate volatility and bias in picking a specific time of day to optimize and trade.