wp01Participant
Master
Ah, now i see what you mean. I already posted the OMX. Sorry forgot that.
I thought your question quite difficult, but it was actually easy 🙂
No, there was no difference with the one on page 73.
wp01Participant
Master
The corn september future has been added to the IG platform.
wp01Participant
Master
USD Basket:
Looks nice, but with only two trades in the past 7 years not very reliable.
wp01Participant
Master
I also checked Soyabeans Jul1. Following the roadmap it should be long, but the results are not ok.
My shortversion looks like a rollercoaster and not recommended. The R/W is 7,34 and that is not so good.
I also looked into the O-jay8 short version. That one looks better, but is optimized with a minimum of 15 contracts.
I assume O-jay8 checked that(?) and that is probably the minimum for traders in the UK who uses DFB. On the European mainland
we can not use that. For the non-DFB world is the minimum 20 contracts.
Hi Yes indeed, in the UK the minimum size for Soybeans is 15.
Had a go at PL, thoughts?
wp01Participant
Master
Hi parvus,
Thank you for your optimization of the Platinum EUR1 contract. Unfortunately the EUR1 contract does not exist anymore. You can find it in PRT, but not in IG
anymore. They replaced it with Spot Platinum mini (10 oz.). Unfortunately you can also not use that one because of the lack of enough history in the charts.
For the mini Platinum you should take the october ’97 future. This one expires september 22nd. so more than 2 months left.
Besides that, you made a long optimization. Nothing wrong with that of course when the results are fine, but the roadmap shows for July 1 and the following 5 months
a short (roadmap on page one), so i’m not very optimistic about a long version. If you want to try the Pl future, you can check O-Jay8’s version on page 74/75.
Thanks.
wp01Participant
Master
I noticed that when you optimize or import an itf file of any instrument it un-checks the “earliest date displayed” and selects a date in august 2010. If you want to
change that to “earliest date displayed” it un-checks the tick-by-tick mode. It looks like PRT did an update because i did not noticed it last week. It seems that the tick-by-tick mode is not applicable
anymore before 2010. Maybe this data wasn’t already there yet without noticing. Be aware of the different results with the different dates. Maybe it is already posted somewhere else in this forum.
That’s right, now the backtester automatically start the backtest at the Date where the tick history is present when you have checked the tick-by-tick mode.
wp01Participant
Master
Thank you for the confirmation Nicolas.
@wp01: Thanks for the guide. I gave it a go on DAX for july2 to practice. The bad trades looks very bad though, not sure how to make it better and safer.
wp01Participant
Master
@Oskar,
I also did the optimization and mine isn’t any better. Sometimes the results are crappy and there is nothing you can do about it.
The picture you showed is also without “tick by tick mode”. Actualy you should check that. It will change the setting to 2010 until now. The results are
even worse than. See my post above here on this page.
But did you also checked the results from Jul1? How i came to that result?
@wp01: Ok. You wrote that it’s good to have a high win-percent after the first two optimizations. I didn’t so I guess it was doomed with that one. Is it good to try with short instead maybe? Roadmap shows to buy though (not sure why but shorting isn’t on the roadmap?).
I suppose tick by tick mode don’t work before 2010 then. But to optimize we should have more history right? To optimize any system with only a few years and a few trades can’t be good. My guess is that tick by tick never worked before 2010 but the update fixed it so prt won’t continue to show error.
Will check the results for jul1 tonight, forgot that one. Thanks for your help 🙂
wp01Participant
Master
Sure you try a short one if you want to. Short isn’t on the roadmap because the overview is created early this year by Reiner. Pathfinder was a long-only swing strategy at that time. The adapted shortversion
was posted by him somewhere in May. But if one optimization isn’t working, i leave it and try something else. There are around 30+ commodities and indices.
The tick-by-tick seems not be working before 2010. With no other optimization i get results before 2010. It is better to have more history, but when it is not there, it is not there. Without tick-by-tick it is more
unreliable.