For how many years you test EA's before implementing it on your live account.

New comment
 
  • 10% (6)
  • 8% (5)
  • 15% (9)
  • 23% (14)
  • 43% (26)
Total voters: 60
 
This poll is about backtesting EA. So I have raised to understand for how many historic period generally we test EA.
 

Test it less than a year than go with it in real mode using a small capital. Once also real mode is confirmed (max 1 year is enough for this) you can invest bigger in the system if it proved positive.

 

Market is like a living organism. I created my oscillation EA and settings that worked 2-3 years ago may not work anymore (was very profitable on every chart until December 2012 - now after dec 2012 requires serious optimizing). In my  EA I included few currency presets. Some of them are for 6-year period, some for 2-year and some just for the last year. The shorter the term the more profitable it is, but the more unstable in theory it is. The longer the term - the less chance of breaking the pattern, but also less profitable it is. No strategy will work forever. You may re-optimize the settings from time to time to give people the illusion it is "perfect", but it will be just a huge BS. 

 
History is still history.... duration shouldn't be the main factor.
 
kchandek:
This poll is about backtesting EA. So I have raised to understand for how many historic period generally we test EA.
In my opinion 6 months back testing + demo account for  3 months to confirm your results.
New comment