Opening a trade only when spreads are low enough.

New comment
 
Hi, guys, first ever post. I've been coding with the mql4 language for a few years now :) but my latest EA is worrying me. So far it is showing profit in a demo, but It starts over 200 trades per week!

Clearly slight changes in the spread value will have a massive inpact on the outcome of profit. What are the chances of a live account performing the same way, from your experience trading forex?

Also I've written some code that checks if the spread if under a predetermined variable value before placing the order, but the spread generally changes again by the time the trade actuality opens. Any ideas on how I could make this more efficient?

Appreciate any wisdom or advice you have to offer, thanks.
 
Trent Wedding :
Hi, guys, first ever post. I've been coding with the mql4 language for a few years now :) but my latest EA is worrying me. So far it is showing profit in a demo, but It starts over 200 trades per week!

Clearly slight changes in the spread value will have a massive inpact on the outcome of profit. What are the chances of a live account performing the same way, from your experience trading forex?

Also I've written some code that checks if the spread if under a predetermined variable value before placing the order, but the spread generally changes again by the time the trade actuality opens. Any ideas on how I could make this more efficient?

Appreciate any wisdom or advice you have to offer, thanks.

Rewrite your Expert Advisor in MQL5 - then you can test it in the mode "Every tick based on real data" - this is the most accurate testing mode, which is identical to the full historical data.

 
Vladimir Karputov:

Rewrite your Expert Advisor in MQL5 - then you can test it in the mode "Every tick based on real data" - this is the most accurate testing mode, which is identical to the full historical data.

Thanks for the repley. After years of testing my EAs in the back tester, only to have them fail on a demo I have litte faith left in any back testing.
Still, it's worth doing. Thanks for the advice :)
New comment
Reason: