Why different result on strategy tester?

To add comments, please log in or register
Erdenebayar Lamjav
1371
Erdenebayar Lamjav  

I tested my code last week, bellow first picture.

 

 And today I tested same timeframe, same period, but result has changed. I didn't change my source code. Why this happen? 

 

Sergey Golubev
Moderator
106721
Sergey Golubev  

MT4?

  • Different bars in test;
  • different time (from ... to ...);
  • different modelling quality (56.28% and 90.00%).

So, it should be different results in backtesting.
Martingale or similar system? If yes so it must be different ... 

Erdenebayar Lamjav
1371
Erdenebayar Lamjav  
Yes MT4, it's no martiangle, it's based on strong fast tick system
honest_knave
Moderator
2308
honest_knave  
Erdenebayar Lamjav:
Yes MT4, it's no martiangle, it's based on strong fast tick system

MT4 doesn't store individual ticks.

So it has to "make them up" in strategy tester.

The modelling might be close to the real thing, but they aren't actually real.

The effects are most pronounced on systems that are trading on a tick-by-tick basis (rather than using bar close, for example).

whroeder1
17928
whroeder1  
Also, the spread defaults to current, meaning that the spread used depends on exactly when you pressed the start button.
Keith Watford
Moderator
14472
Keith Watford  
Comments that do not relate to this topic, have been moved to "MT5 Strategy Tester".
Marco vd Heijden
Moderator
8419
Marco vd Heijden  


In the case of the OP it was an incomplete download of historic data as has been mentioned by Sergey.

Alain Verleyen
36572
Alain Verleyen  
Marco vd Heijden:


In the case of the OP it was an incomplete download of historic data as has been mentioned by Sergey.

There is always a reason, on 99.99% of the case I saw, the reason was a lack of knowledge or understanding or a typo...well, in summary a user issue.
To add comments, please log in or register