mechniv wrote: Yes, I suspected that Stoken's Max DD figure as reported in the blog post looked too good to be true.

I found the CAGR for every year from 1928 to 2024. Interesting. The vast majority of the time it was worse during the bull signal than during the bear signal. It makes sense. The strategy is literally waiting for a very long time before issuing a signal from a bottom, and waiting a very long time after a peak before selling.

Peak/Trough	Signal	Days	Gain/Loss
5/20/2015	8/20/2015	92	-4.64%
2/11/2016	6/2/2016	112	16.31%
9/21/2018	11/23/2018	63	-10.53%
12/26/2018	4/12/2019	107	24.03%

It is truly remarkable to me that anyone could possibly think this system would be market beating with such a low MaxDD. Prior to the signal in August 2015 the S&P 500 had already dropped -4.64%, and after the bottom prior to the signal to re-enter, it had already risen 16.31%. Clearly, you were better off not doing anything but just holding. The price on 8/20/15 was $2,035.73, and the price on 6/2/2016 was $2,105.26. This is a difference of 3.42%. The cost of missing the drop was not too much in this case, but I can assure you there are many other times when the cost is much, much more dear!