Backtester FAQ ?
You've got some with t/p set and other orders without t/p set. In fact, you've got the same order going through twice at the same second - one with the error and one without.
- Vooch
- Vooch
Wonderful isn't it? :)
I also notice boolean parameters are not an optimisation choice either. You'd think you'd be able to flip between true and false in an optimisation test.
Martin
I also notice boolean parameters are not an optimisation choice either. You'd think you'd be able to flip between true and false in an optimisation test.
Martin
> I also notice boolean parameters are not an optimisation choice either.
> You'd think you'd be able to flip between true and false in an optimisation test.
In some scripts, I've setup start=0,step=1,stop=1 tests using external variables. Then, loop through them to find out which is the best scenario. It works out fine for me.
> You'd think you'd be able to flip between true and false in an optimisation test.
In some scripts, I've setup start=0,step=1,stop=1 tests using external variables. Then, loop through them to find out which is the best scenario. It works out fine for me.
Cool, thanks for the tip, Vooch.
I'd still think MQ shouldbe making allowance for a true/false option in the short term as well as documenting how experts should be coded to make better use of the backtester - that's if they themselves have an understanding of real world experts.
Having said that, I discovered I had allowed insufficient buffer space for my custom indicator. Increasing this allowed for the indicator to cover back to a March 05 test period. Results are quite similar and appears to be not being used.
Have you used any custom indicators in any of your backtesting ? If so, any difficulties or 'gotchas' that other users (including me) should be aware of ?
Martin
I'd still think MQ shouldbe making allowance for a true/false option in the short term as well as documenting how experts should be coded to make better use of the backtester - that's if they themselves have an understanding of real world experts.
Having said that, I discovered I had allowed insufficient buffer space for my custom indicator. Increasing this allowed for the indicator to cover back to a March 05 test period. Results are quite similar and appears to be not being used.
Have you used any custom indicators in any of your backtesting ? If so, any difficulties or 'gotchas' that other users (including me) should be aware of ?
Martin
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
I'm really struggling with this Backtester, I'm afraid. I'm looking for perfect results - that's just silly given the lack of true tick data. I'd certainly hope for some consistancy throughout the results though.
I'm testing a strategy that simply does this :-
Places 2 x Buy Stops and 2 x Sell Stops based on a breakout range.
The first order has a set Profit Take and if reached, the second moves to a breakeven point and trails a custom moving indicator, moving the SL into locked in profit once the custom indicator moves past the breakeven point.
Looking at http://www.cooney.com.au/fx/StrategyTester.htm really confuses me. I've got a Modelling quality of 25% which appears crap - expert is running on an M5 chart / timeframe.
Modifys only seem to be working very late in the short test period - I've got M1 data back to Feb 2005.
There's a huge profit somehow realised as the last trade for Order no. 93 that's around 1.5 months old. That's a weird one :)
I'm running build 177 July 13 2005.
Does anyone have any tips for better setting up for backtesting ?
Cheers
Martin
Last