
The 4N model results are the product of simple condition 1 signals while the 9N model results are the product of both condition 1 and 2 signals.
Both models get to the same place eventually, you just buy a lot more risk exposure with the 9N model. As a partial tradeoff for reduced risk exposure the 4N model does incur additional commission expenses (Schwab traders get a break when using SCHF in lieu of EEM) but that's a small price to pay (IMHO) for the added safety of a shorter term hold period.

Interestingly, the topography of both models shows the same attractive ski jump pattern with a gradual decline in performance from optimized N day values. What we don't want to see is a pattern that looks like Niagara Falls since that will indicate that the results are probably over-optimized, unique to a very narrow range of parameters, and cannot be sustained.
This is another reason to review performance potential of a pair by checking the slope of the equity curve (chart D above) to confirm there are no major hiccups along the path.
No comments:
Post a Comment