stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

The Great HGB Foward Test
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=4044		 Page 25 of 27
Author:  JRforex [ Sun Feb 08, 2015 11:06 pm ]
Post subject:  The Great HGB Foward Test
WhoKnows » Sat Feb 07, 2015 7:42 am wrote:
The worst instances took trades on GBPJPY (net -500 pips) on 03/02 that the rest did not.

I don't have an 'original' chart of that day, but these GBPJPY trades don't make sense.
On a H4 chart that time would have been a Range period and Range trading should have been on false.
Also the exit of these trades is strange as these trades should have been closed well before the time stated at FX (as there would have been opposite waves and opposite Green arrow).

Were these bad instances really run with the correct set file??
I was wondering the same thing, but according to the screenshot taken by HGB, those GBPJPY trades were RAD trades, and there were no range wavies when those pendings were sent.

Order open:
11782389 open ChartScreenShot.png
Order close:
11782399 close ChartScreenShot.png
Author:  WhoKnows [ Mon Feb 09, 2015 11:10 am ]
Post subject:  The Great HGB Foward Test
trendinator » Sun Feb 08, 2015 1:05 pm wrote:
WhoKnows » Sat Feb 07, 2015 11:42 pm wrote:
Unfortunately you can't see the trade comments on FX
you can, just hover over the speech bubble icon on the right in the history list. or download the CSV via 'export' in the top right above the chart

Thanks ... there is a lot more in this FX stuff than I knew. Also thanks for the CSV files.
I saw that you already raised the issue of lots of trades opened and closed in the worst performing instances in another thread. Strange that happens only in part of the instances.

Also a question:
I saw somewhere that the Global Prime DEMO accounts only allowed for 30 trades being open.
Is this still the case, and could this influence the test? (if some instances already have different orders open they might not be able to open new trades).
Author:  milanese [ Mon Feb 09, 2015 11:23 am ]
Post subject:  The Great HGB Foward Test
WhoKnows » Mon Feb 09, 2015 11:10 am wrote:
trendinator » Sun Feb 08, 2015 1:05 pm wrote:
WhoKnows » Sat Feb 07, 2015 11:42 pm wrote:
..

Also a question:
I saw somewhere that the Global Prime DEMO accounts only allowed for 30 trades being open.
Is this still the case, and could this influence the test? (if some instances already have different orders open they might not be able to open new trades).
this was longtime ago with the old demo-accounts (before 07/2013), since than GlobalPrime has NO limit how much orders you can open on the demo..

Cheers :)

Tommaso
Author:  trendinator [ Mon Feb 09, 2015 11:34 pm ]
Post subject:  The Great HGB Foward Test
JRforex » Mon Feb 09, 2015 9:06 am wrote:
I was wondering the same thing, but according to the screenshot taken by HGB, those GBPJPY trades were RAD trades, and there were no range wavies when those pendings were sent.

Order open:



Order close:
looks like a rad trade that turned bad and was closed on the opposite rad arrow. starting to think having a massive stop loss is not such a good idea. perhaps it would work better to have a tight TP and SL.
Author:  WhoKnows [ Tue Feb 10, 2015 1:40 am ]
Post subject:  The Great HGB Foward Test
looks like a rad trade that turned bad and was closed on the opposite rad arrow. starting to think having a massive stop loss is not such a good idea. perhaps it would work better to have a tight TP and SL.
If you look at the pictures, what is worrying me more is why HGB did not close on the Yellow and BLUE waves, well before the opposite RAD. That would have saved most of the drawdown.

I have seen similar situations before: see e.g. my post
http://www.stevehopwoodforex.com/phpBB3 ... 72#p112872

My coding skills are rudimentary. Can somebody please have a look at the logic of the closure of trades.
Especially on the placement of brackets at e.g. line 3991 and if I interpret it correctly that closure is only on Yellow Wave signal.
Don't shoot me please. I am trying to learn. I did play with the code to see what happens, but it is difficult to wait for situations to arise, so you can evaluate your changes.
Author:  JRforex [ Tue Feb 10, 2015 3:14 am ]
Post subject:  The Great HGB Foward Test
WhoKnows » Mon Feb 09, 2015 7:40 pm wrote:
If you look at the pictures, what is worrying me more is why HGB did not close on the Yellow and BLUE waves, well before the opposite RAD. That would have saved most of the drawdown.

I have seen similar situations before: see e.g. my post
http://www.stevehopwoodforex.com/phpBB3 ... 72#p112872

My coding skills are rudimentary. Can somebody please have a look at the logic of the closure of trades.
Especially on the placement of brackets at e.g. line 3991 and if I interpret it correctly that closure is only on Yellow Wave signal.
Don't shoot me please. I am trying to learn. I did play with the code to see what happens, but it is difficult to wait for situations to arise, so you can evaluate your changes.
For RAD sells, HGB is coded so that those trades close on a Trend Up arrow, a Range Up arrow a RAD Up arrow and a TMA Range signal. The problem is that during that trade, the TMA signal was "TMA no signal", so the trade never closed, even when the wavies showed up. Can anyone provide some info about this? I thought that when a yellow wavy showed up, the TMA status changed to TMA range. :?

JR
Author:  trendinator [ Tue Feb 10, 2015 4:53 am ]
Post subject:  The Great HGB Foward Test
JRforex » Mon Feb 09, 2015 9:06 am wrote:
WhoKnows » Sat Feb 07, 2015 7:42 am wrote:
The worst instances took trades on GBPJPY (net -500 pips) on 03/02 that the rest did not.

I don't have an 'original' chart of that day, but these GBPJPY trades don't make sense.
On a H4 chart that time would have been a Range period and Range trading should have been on false.
Also the exit of these trades is strange as these trades should have been closed well before the time stated at FX (as there would have been opposite waves and opposite Green arrow).

Were these bad instances really run with the correct set file??
I was wondering the same thing, but according to the screenshot taken by HGB, those GBPJPY trades were RAD trades, and there were no range wavies when those pendings were sent.

Order open:



Order close:
ok i think ive figured out what happened. the massive candle (news spike) after the RAD caused the yellow wave to be painted in hindsight, as a candle or two late. so HGB totally missed it. also HGB has a logic flaw, it only closes on ranging waves (yellow) and not trend waves (blue). just whipped up a mod that im testing now to address this.
Author:  JRforex [ Tue Feb 10, 2015 5:10 am ]
Post subject:  The Great HGB Foward Test
trendinator » Mon Feb 09, 2015 10:53 pm wrote:
ok i think ive figured out what happened. the massive candle (news spike) after the RAD caused the yellow wave to be painted in hindsight, as a candle or two late. so HGB totally missed it. also HGB has a logic flaw, it only closes on ranging waves (yellow) and not trend waves (blue). just whipped up a mod that im testing now to address this.
You might also want to include a function to allow only one trade per bar, in order to avoid this:
12118051 open ChartScreenShot.png
The latest version of BoHGB includes a code snippet to avoid this, and it works OK.

JR.
Author:  trendinator [ Tue Feb 10, 2015 1:06 pm ]
Post subject:  The Great HGB Foward Test
JRforex » Tue Feb 10, 2015 3:10 pm wrote:
trendinator » Mon Feb 09, 2015 10:53 pm wrote:
ok i think ive figured out what happened. the massive candle (news spike) after the RAD caused the yellow wave to be painted in hindsight, as a candle or two late. so HGB totally missed it. also HGB has a logic flaw, it only closes on ranging waves (yellow) and not trend waves (blue). just whipped up a mod that im testing now to address this.
You might also want to include a function to allow only one trade per bar, in order to avoid this:



The latest version of BoHGB includes a code snippet to avoid this, and it works OK.

JR.
im pretty sure that is implemented in HGB 2t
Author:  nonlinear [ Tue Feb 17, 2015 9:58 pm ]
Post subject:  The Great HGB Foward Test
Have to say what is going on in this thread is pretty damn impressive. Nice work!
All times are UTC Page 25 of 27

Powered by phpBB® Forum Software © phpBB Limited