Mother of Dragons trading strategy…

Viewing 15 posts - 211 through 225 (of 523 total)

← 1 2 3 … 14 15 16 … 33 34 35 →

Author

Posts
05/03/2020 at 2:44 PM #129581 quote

eckaw
Participant

Veteran

Hi all, there’s some problems with the code I posted – the heuretics doesn’t work as it should. I will investigate and repost if better results are obtained.

swedshare thanked this post

05/03/2020 at 2:45 PM #129582 quote

eckaw
Participant

Veteran

nonetheless wrote:

Nice work with the ML application, are you sure there’s no crosstalk between them?

At the moment there isn’t as the Heuretics code isn’t properly implemented 😉

05/03/2020 at 2:45 PM #129583 quote

nonetheless
Participant

Master

These are the back test results I’m getting – position size 1, MM off.

@Moda, yours is not so far from the original. The results i posted last month were position size 2 to make it easier to compare with the DJ version.

DAX-5m-MoD-comparison.jpg

05/03/2020 at 2:55 PM #129591 quote

nonetheless
Participant

Master

eckaw wrote:

the Heuretics code isn’t properly implemented

That’s a shame, I tried yours with and without ML and it seemed to improve things. I tried a few ML versions of my own but couldn’t get consistent results so I gave up.

05/03/2020 at 9:34 PM #129644 quote

eckaw
Participant

Veteran

Attached is a version where Heuristics works as it should, and it doesn’t seem to interfere with Paul’s “dynamic live parameter adjustment” https://www.prorealcode.com/topic/dynamic-live-parameter-adjustment/

I’m not sure about the parameters, it looks like the value goes up to maxvalue = 5, which would have increased risk quite a bit. Maybe maxvalue should be lowered?

Results looks promising, will be running this in demo from tomorrow. I used spread 3.8 which is +1 of average spread, as the algo trades 24/7. What are your thoughts on the right spread to use for backtesting this?

Screenshot-2020-05-03-at-21.30.02.png Screenshot-2020-05-03-at-21.26.26.png

05/03/2020 at 9:34 PM #129647 quote

eckaw
Participant

Veteran

Forgot to attach the .itf! Here it is.

GraHal, nonetheless and Jabezz thanked this post

MoD-Dax-4.x-ML2.itf

05/03/2020 at 9:38 PM #129650 quote

Francesco
Participant

Veteran

I always use the maximum spread that an asset can reach during the day; i know that it’s a pretty extreme, but if a system can reach good results and shows a good robustness with those spreads, maybe it deserves to run in live.

eckaw thanked this post

05/03/2020 at 9:46 PM #129654 quote

eckaw
Participant

Veteran

Francesco wrote:

The ML version is very interesting, also reduces all the noise of the little profit trades. What if applied on the original trailing stop version and not on the less performing breakeven one?

Which version exactly is this @Francesco? I can try to apply it and see what happens.

05/03/2020 at 9:51 PM #129655 quote

Francesco
Participant

Veteran

Is the original nonentheless’s version on which Moda based his breakeven version. You’ll find it in the attachments of the topic: Dax 5m mother of dragons v2.
The differences with the breakeven versions are not huge, but surely better 🙂
Apart from that, could be interesting to see how ML works on trailing stops, maybe the performances will run up extremely, or maybe could be trash.. who knows

eckaw thanked this post

05/03/2020 at 10:37 PM #129683 quote

eckaw
Participant

Veteran

See attached – these are the results I get on the original DAX version with trailing stop. Less total win on Heuristics + “DLPA” (Paul’s Dynamic Live Parameter Adjustment code). Not sure it’s worth it? Parameters could possibly be improved. It wins more but has a bigger max loss.

Francesco and nonetheless thanked this post

Screenshot-2020-05-03-at-22.25.26.png Screenshot-2020-05-03-at-22.26.18.png Screenshot-2020-05-03-at-22.23.33.png

05/04/2020 at 8:13 AM #129704 quote

MrCool
Participant

Average

I still don’t get the same results, looks like it’s only in Januari it’s different. I got 3 losses in a row (see attached picture). Look like none got the same on his test as well.

5.png 6.png

05/04/2020 at 9:29 AM #129742 quote

Sebastian Arsjo
Participant

Junior

I get exactly same as you MRcool

Time settings difference between Sweden and England ?

05/04/2020 at 11:48 AM #129804 quote

MrCool
Participant

Average

Sebastian Arsjo wrote:

I get exactly same as you MRcool

Time settings difference between Sweden and England ?

It is, 1 hour but nonetheless and echaw got different results and they both in England right? Did you get any different results when you moved the time to English time Sebastian?

05/04/2020 at 3:40 PM #129868 quote

Sebastian Arsjo
Participant

Junior

No did not try to change, not even sure where i would do that in this new version 🙂

05/04/2020 at 6:00 PM #129906 quote

MrCool
Participant

Average

Sebastian Arsjo wrote:

No did not try to change, not even sure where i would do that in this new version 🙂

Shouldn’t make any difference tho, because there is no time variable in the code so something is wrong. And it’s only Eckaw that is getting this results so?
Author

Posts