@Patrick
I confirm backtest opening a trade, but live did not.
Regards, David
ALEModerator
Master
@patrick
the same for me, it did not open short position in real account
rgds
Ale
PereParticipant
Veteran
My backtest with V6 month/4 traded one short on 2/01/17 at 13:00, and soldshort on the same 2nd January at 21:00, and also traded one long on 2nd January at 21:00, still not closed.
But backtesting the versión V6B2, it does not trade the first short, only the long on 2nd January at 21:00, still not closed.
However, with versión V6, I get again the two trades, short and long, always in backtesting. I cannot explain this issue.
I have checked the setup with the last version V6. The only valid signal was a weekly/monthly crossover (11.485.8) at 2. Jan 17:00 triggered a long trade (11.586,8) at 21:00.
I can confirm a strange behavior in backtest mode. The code generate a sell signal at 2. Jan at 9:00 due signalline crossed under daily high and thats definetly wrong!
I can also confirm that no trade was triggered in my life account with V6.
That behavior must be a PRT error and I suppose that has something to do with the year change.
Hello,
That’s strange..I have the same behaviour..
Do you think that we have to ask PRT support??
best regards,
wp01Participant
Master
Don’t know if it helps but i have stopped it from automatic trading and activate it again from the BT. Than everything gets a date in 2017.
Don’t think that it makes any different but it is worth trying. Couldn’t find anything in the documentation from PRT.
PereParticipant
Veteran
Remember that there was also something strange on the first fake trade of 2016. Perhaps could it be done for the same reason…
For information, V6 SMI has opened a long position at 5.00pm today..
Had a go at updating Pathfinder HS 4h to V6, though I call it v5b5 until some kind soul has checked my alterations.
Regards, David
AlcoParticipant
Senior
Hi Dajvop,
Thanks for your update! I’m also looking to trade the HS, I see you only addapted the seasonal booster. But does this look like curve fitted?
Hi Alco,
Yes, that was what I suspected, but I needed a second opinion. It’s a fine line between optimization and curve fitting…
Hi guys,
I confirm strange behaviour on first days of 2017. I think the problem was that monlthlyHigh crosses over signalline (and so signalline crosses under monthlyHigh)
s1 = signalline CROSSES UNDER monthlyHigh
This happened because we moved 1 bar forward monthly calculation with
If Month[1] <> Month[2] then
If you revert to previous version, short position doesn’t start because monthlyHigh was refreshed on 2 jan at 5.00 and not at 9.00 (pathfinder skips operations before 9.00).
Remark that we changed Month[2] because didn’t detect crossing at 21.00 on last day of month.
I don’k know why backtest works different then real. Maybe on real montlyHigh is calculated well.
Hi Hakan,
I agree – it’s not so easy to keep track off the different versions and a kind of repository or cloud drive would be helpful. I prefer a solution that is connected to this forum.
@Nicolas: what is the best way to save all Pathfinder vesions with access rights for every member?
Thanks, Reiner
@Reiner: this is a question I asked me recently. Do these files need to be changed/deleted in the future?
@Nicolas: topic owner or a group of admins should be able to add, modify and delete and everyone should be able to see a list of bots and download them
