Can you unpack this a bit more? Let's say you establish trend based on a 20/50/100/200 ema stack (no crosses - strong trend) - what's the entry point?
Post-retracement trend-resumption, maybe?
There are countless ways of identifying those: including the good, the bad and the ugly.
If you want to look at it in terms of moving averages, maybe something like "When the price dips below a 10-period MA and then crosses back above it again, with those other MA's not having crossed each other and the 100-EMA still rising?" That should be
at least a "reasonable" entry-point (for a long trade, of course)?
(How often it happens is another matter. Not necessarly "10-period" specifically, but just to illustrate the idea? That's just one very simple example - perhaps too simple.)
I'm wondering how to strengthen this system, especially by finding earlier trends!
"Strengthen" can mean different things to different people, can't it?
Some people mean "finding higher probability entries", i.e. entries with a higher win-rate (this isn't my approach and I respectfully suggest that it probably shouldn't be yours, either).
Some mean "strengthen the overall results" by trading more frequently,
even if that incurs a slight overall drop in win-rate (that would be more my approach).
You could do that, perhaps, by trading the same thing but on a faster time-frame so that you get more trades in?
My experience of doing that sort of thing is that the decrease in overall win-rate (which there can easily be) is
more than compensated for by the increase in trading frequency.
Key concept: the inverse correlation between overall win-rate and trading frequency is a
non-linear one (and there are reasons for that), and some people find that "faster" works in their favour, overall - and I'm one of them. "Just saying" ... (many will disagree, doubtless, but sometimes through prejudice rather than through statistical analysis).
PS: The pre/post-entry trade management things (including position-sizing, stop-loss placement and adjustment, and target(s)) are
always going to be more important, collectively, in terms of their overall influence on the results, than the exact entry-point! Don't fall into the trap of imagining that "if you can just find better entries, everything else will somehow take care of itself": it won't!