[bokonist]

In the first cut of your analysis (without rebalancing) the portfolio actually goes negative (I'm looking at the first graph in the notebook). An equity portfolio cannot go negative unless you are using leverage. Then from late 2012 to 2015 the portfolio rises by 400%. Looking at the companies in the portfolio, I don't see how they could have generated that kind of return. Is this algorithm using leverage? If not, I don't see how you can get those results. If the algorithm is using leverage, then comparing to a simple investment in the S&P is not valid.

[karenrubin]

You are absolutely right that the first version of my algo is using tons of leverage. I didn't realize what leverage was when I did this the first time.

The second version of the algorithm factors in leverage and there it hovers around 1.