see above:
#154220
I mean on 1 minute timeframe 1 million bars, sorry didn’t explain it properly 🙂
why would you want to run it on 1min? it would need to be completely rebuilt on a backtest half the size.
DAX Hull SAR 2m Long only
Hello,
I make an test, and optimization with DAX 2m, and 200Kbars
I take the NAS 2m Hull SAR V5.3L from Nonetheless (take you again)
I make the optimization,from the 12min parameter, after from the 6min parameter, after, the 2min parameter, and SL and TP.
I don’t modify the trailling stop and trailling percent.
Here are the itf and picture
Thank you for your comment
Hey, thanks for that – nice thought, but not so pretty on 1m bars. I also tried a few things to make it work on the DAX but 2018 was always disastrous so I gave up.
A couple of things that aren’t helping is that you didn’t change the hours for Germany (the NAS runs on Wall St hours) so can probably be better optimized if 090000 – 173000
Spread should be 1.2
Also, you changed the Maxpositionsallowed to 1, which is the same as Cumulative orders = false … not sure if you meant to do that.
I test with Time 09 to 17, and is was worse. So i let 15h30 22h. Many times the Dax follow the US.
I test all ae with 1 lot, so it is easier to compare. Thanks for your comment.
rebParticipant
Master
Hi all
First of all, thanks for this strategy
@nonetheless, I try to use DJ 1m HULL-SAR v7.5 but it is quickly blocked by PRT.
The reason is : stop loss is nearer than the 10 points allowed by IG. The programm is stopped, even if there is an open position.
In which part of the code do I have to introduce this IG restriction (Stop>10pts)
Thanks in advance
Reb
You get that error because the minimum stop level changes at different times of the day. To fix it, tick the box that says ‘Readjust stops’ when you launch the strategy.
Please note that that system was optimized on a very short back test (v10.3) so I can’t really recommend it. None of the DJ versions work especially well over 1m bars.
rebParticipant
Master
Thanks,
I didn’t this tip on the 11th version
begeParticipant
Average
It would be nice to see a 1M backtest of DJ 1m HULL-SAR v7.5 since it looked promising on 100K. Any plans on a new DJ version optimised on 1M?
As you can see, it’s fairly disastrous in the OOS walk back. I have tried to rework this for the DJ but it’s not nearly as good as the NAS version.
rebParticipant
Master
I don’t have this report on the DJ version
More trades et more profitable
something wrong ?
No idea why yours is different to mine, but it’s still a disaster! ☹
NAS 2m 5.3L is the only one I’m running, attached is the forward testing since Dec 21 with MM
rebParticipant
Master
the single difference in the code, is the GMT+1 hour
In 2 min timeframe on DJ the algo is working a little bit better then in 1min. Attached two versions.
