I’ve been studying the periodicity problem in more detail. After running all of the stocks selected and adjusting against the S&P, the optimized rotation schedule appears to be 66 days per stock – or roughly one trade a week. 179 days came in at a very close second.

I got this by looking at each stock’s return MINUS the S&P return for the same time period, in order to factor out random market movements.

In any case, it turns out that I WILL make a rotation tomorrow, if possible:

I’ll be selling CLH and buying CLF

CLF is Cliffs Natural Resources, in the STEEL industry.

The only thing that would cancel this plan would be a positive gap of CLF more than 1% greater than CLH. If that happens, I’ll check again at lunch and the end of the day. As long as they are within 1% of each other, I’ll make the trade.