I am testing a EA using Perid M30 "ControlPoints-A very crude method based on the nearest less timeframe". It can produce 100% profit in one month. However when I test it with everytick mode or M5(control point), there will be big loss. Anybody can explain this?
Is there a mehod that can let the EA only trade the control point in real account? Guess we will profitable then?

control point is a crude approximation to the real thing. there is no such method in real trading.

For modelling, you have 3 choices:



I've done a lot of testing and have always used "every tick". I can understand the option to use open prices, if the strategy is based on that. But I honestly have no idea what the 3rd option (control points) can be used for.

So as a poster said, it doesn't reflect reality anyway, so do your testing with "every tick" and look for the best set-ups. Good and thorough testing take time. So conduct your testing scientifically and methodically and you will eventually get good results (if you system is good, of course ).

Happy trading,

__________________
Profitable trading starts here:
http://forexleader-forexleader.blogspot.com/