Hello All,
I wanted to share an algo I've been paying around with in hopes to find some support from the community on how to improve the results. For some reason, I don't think my Pipeline and Ranking is optimal and could use some tips on how best to approach it with the Custom Factors used.
If you notice other improvements I could make to get the returns positive or more robust I'd love to implement them and throw this one into the Contest. Or at the very least, begin the conversation to rattle my brain and help me think outside the box.
I would really like to see how isolating each factor by industry or sector but could use some python/numpy experts help!
* Note, the algo currently shows the log error: WARN numpy/lib/nanfunctions.py:1147: RuntimeWarning: Degrees of freedom <= 0 for slice. and I'm unsure which line needs to be fixed.
Thanks