Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
Partial least squares experiment with transaction costs

I modified earlier algorithm to use only stocks that are priced over 100$ and attached are results - Which begs the question if 3 cents per share is realistic.

3 responses

I like the PLS idea - but I have a question, how did you come up with the context.symbols list? There is presumably some bias in this selection?

@Jack - I used a stock screener to filter stocks that are priced over 100$ and have a market cap > 10 billion.

Priced and market cap as of when? Can you link to earlier algorithm, or explain more about what the strategy behind this? This same group of stocks doesn't perform well when run from, say, 2006. What's the criteria for picking stocks?