I don't know, I am testing it from about one week and I like it up now.....
I created an EA for MT5 myself about 2 years ago and it was quite professional, the problem was that being very complex there were things that got too complicated and the programmer I thought it was too much to see ...
the problem was that I earned a lot when the coins tried to correlate as usual, however there were situations in which this correlation took much longer than normal and was lost too much ... so I stopped testing.
in the end I don't know if I lost because there was a failure or because of the strategy
Anyway I know that there are two good strategies that can be complemented with a set of accessories or not, these are:
1) correlations
2) divergences
Any EA with these strategies should be promising in the beginning.
I wish we could do an EA that combined these two strategies, for example, that measured the correlation and used the divergence as a filter or vice versa.