Apologies if the formatting is a bit off in the table above, but you can see that holding SPY for the past 10 years has been hard to beat. I'm stating the obvious to members of this board, but when it comes to following a momentum system, simply using SPY (momentum based on market cap of US companies), has been difficult to beat - at least it sure has for me. Will that continue to be the case? It's hard to tell, but... certainly with all the index investing that is done these days, one could argue that SPY has a tailwind just from that alone.

I think it's worth considering that there are strong bull markets, and there are other times. In strong bull markets, the S&P is nearly impossible to beat and there is no real point in trying. At other times, the S&P can be much easier to beat, and it's that much more important to do so as the S&P isn't doing so well. We have had a huge bull market lately and the super-big-caps have dominated, so the index has been a very tough target, matching your experience. BUT...this too shall pass, I believe.

It could be worthwhile to think about ways to determine, in broad stroke fashion, whether or not you're in an ongoing strong bull market. Try fancy stuff only when that isn't clearly the case? You tend to see one situation or the other for years at a time.

As for trying to beating the S&P, even when it's doing well, the LargeCapCash screens are doing not so badly so far. I originally suggested, and track, Value Line versions, with a few variations. Here's one with a tiny bit of momentum, mainly "eliminate the worst" at the beginning. Created on data up to April 2020.
* Price / 52 week high top 50%
* ROE top 30%
* (Cash - Long Term Debt) top N

The original method was to run it with two month holds, top 40 hold-till-drop at 45. Run that way with 0.4% round trip friction, this has returned 24.8% in the first 4.25 years post discovery, beating the S&P by 5.87%/year. It has beat the S&P in 84% of rolling years in that post-discovery test. I very much doubt it will continue to lead by that much, and there will certainly be times that it lags the index for a while, but at least is isn't over cooked and seems to offer the *chance* of beating the index. I'm running something very similar to this with real money. One nice thing is that it scales well: almost all the picks are very large cap, and the turnover is very low, and the risk (single-stock exposure) at 2.5% of the portfolio is considerably lower than for the S&P at 7%. Since the final sort is on the absolute size of the (net) cash pile which happens only at large firms, the top picks from a recent month are the usual suspects: GOOG, MSFT, META, TSM, NVDA...

Jim