Gamestudio Links
Zorro Links
Newest Posts
Blobsculptor tools and objects download here
by NeoDumont. 03/28/24 03:01
Issue with Multi-Core WFO Training
by aliswee. 03/24/24 20:20
Why Zorro supports up to 72 cores?
by Edgar_Herrera. 03/23/24 21:41
Zorro Trader GPT
by TipmyPip. 03/06/24 09:27
VSCode instead of SED
by 3run. 03/01/24 19:06
AUM Magazine
Latest Screens
The Bible Game
A psychological thriller game
SHADOW (2014)
DEAD TASTE
Who's Online Now
2 registered members (AndrewAMD, Imhotep), 567 guests, and 4 spiders.
Key: Admin, Global Mod, Mod
Newest Members
sakolin, rajesh7827, juergen_wue, NITRO_FOREVER, jack0roses
19043 Registered Users
Previous Thread
Next Thread
Print Thread
Rate Thread
Profit shape 9 years own Z1 #478310
09/29/19 19:48
09/29/19 19:48
Joined: Oct 2018
Posts: 90
Germanien
ags Offline OP
Junior Member
ags  Offline OP
Junior Member

Joined: Oct 2018
Posts: 90
Germanien
Hi there,

i'm developing my own Version of Z1 base on the black book and Ehlers. Got 4 diferent strategies on 16 assets. I train them with 10 periods WFO in 9 years and NumSampleCycles = 3, like the results and similar sample performance, but do not like the shape of the profit curve, excelent the first 3 years and almost flat the last 3 years.

I do not think it is because the markets (forex in this case) were more trending the first 3 years and now the market is predominantly in cycle mode...

Any idea what to look for or how to change the script to get a more steady ramp curve ?

Attached Files Z1_AGS_LIV_V1_AUDJPY.png
Re: Profit shape 9 years own Z1 [Re: ags] #478773
12/15/19 09:34
12/15/19 09:34
Joined: Oct 2018
Posts: 90
Germanien
ags Offline OP
Junior Member
ags  Offline OP
Junior Member

Joined: Oct 2018
Posts: 90
Germanien
Wow, more than 100 downloads of the chart but no comment... this seams to be a complex question crazy

Anyway, i read a lot the last weeks and developed the script further. I concentrated my efforts in the last 3 years time period, where the first script was very flat.
The key for me was in the development phase, and specifically the optimization range of each variable. In the first script i was using Zorro as a gun machine and shooting to evrithing...
Attached the much better results, 14 training cycles 86weeks / 46 days. Same 4 original algos with small changes and much thoughts on optimization range usage.

Attached the report and new equity chart, starting papertrade tomorrow 16.12.2019 to see if this is not overfitted...

A key for my changes in the way to optimize the variables and usage of the script was in the recommended book Trading Systems, from Jaekle/Tomasini. Now i'm starting with Ernest Chan, Quantitative Trading, let's see if i find something interesting in this one too.

Attached Files Z1_V3_EURUSD.png
Z1_V3.txt (175 downloads)
Re: Profit shape 9 years own Z1 [Re: ags] #478775
12/15/19 14:21
12/15/19 14:21
Joined: Feb 2015
Posts: 652
Milano, Italy
M
MatPed Offline
User
MatPed  Offline
User
M

Joined: Feb 2015
Posts: 652
Milano, Italy
Ciao, how the improved TS perform over all the original 9 years test period?

Re: Profit shape 9 years own Z1 [Re: MatPed] #478810
12/22/19 10:13
12/22/19 10:13
Joined: Oct 2018
Posts: 90
Germanien
ags Offline OP
Junior Member
ags  Offline OP
Junior Member

Joined: Oct 2018
Posts: 90
Germanien
Ciao, my V3 is trimmed looking at the markets 2017-2019, so is more suited for that period, and not surprisingly works not so good in the period before, but still catches good periods, specially 2016. But guess no one will stay confident in the system looking at 2015...

So the key question remains for me how to evaluate parameter-ranges to be used, extremely wide, or concentrated in a relative smaller period (?) 3 years seam OK for intraday bar period of 60 minutes system.

Attached the result of V3, trimmed for markets 2017-2019 and used for this test with other longer train-test cycles than originally designed 17-19. (179 weeks / 153 days Vs. 86 weeks / 46 days).
Will make another test using the longer period but with same train-test cycles to see if this improoves the period 2014-2017, but guess it won't change much.

Attached Files Z1_V3b11-19_EURUSD.png
Re: Profit shape 9 years own Z1 [Re: ags] #478812
12/22/19 12:06
12/22/19 12:06
Joined: Feb 2015
Posts: 652
Milano, Italy
M
MatPed Offline
User
MatPed  Offline
User
M

Joined: Feb 2015
Posts: 652
Milano, Italy
I do not change th ranges of optimization parameters over the training/testing period: not less than 5yr OOS. But that's only my approach.

Re: Profit shape 9 years own Z1 [Re: ags] #479500
04/02/20 18:58
04/02/20 18:58
Joined: Mar 2018
Posts: 62
hast29 Offline
Junior Member
hast29  Offline
Junior Member

Joined: Mar 2018
Posts: 62
Hi, I agree with MatPed, I think shortening the optimization OOS period may likely lead to overfitting and undermining the basic idea of the edge. For some reason though, the Z strategies are re-parametrized periodically, in terms of months, so I am still probably missing some parametrization fundaments in Zorro :-(.

Just my approach is to stay with the simplest original result, as you posted on 29.9.2019. Especially when you say it is consistent over many assets.
I trade it small with microlots and focus to adopt other strategies into the ensemble, to balance/compensate the flat periods and draw downs.


Moderated by  Petra 

Gamestudio download | chip programmers | Zorro platform | shop | Data Protection Policy

oP group Germany GmbH | Birkenstr. 25-27 | 63549 Ronneburg / Germany | info (at) opgroup.de

Powered by UBB.threads™ PHP Forum Software 7.7.1