Back to Community
Please delete this post.

Sorry I changed my mind. Algo with some changes is too profitable to share :)

1 response

AQ,
This does a nice job of keeping steady growth.
Some description of this strategy might encourage comments.

Also it would be good to understand the rationale for some parameters.
I've not done principal component analysis on prices before and am not familiar the sklearn library.
Why is PCA done with apparently 0.9 dimensions instead of an integer value of 1?
Why use 600 days of history? Is this to improve the covariance estimate for a 50 stock portfolio?