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

Holy Graily Bob 'n Grid
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=4525
Page 88 of 99
Author:  ALoveSvpreme [ Thu May 19, 2016 4:09 am ]
Post subject:  Holy Graily Bob 'n Grid

Lifesys » Mon May 16, 2016 11:46 pm wrote:Always good to mind what we are really doing here.
Hello Lifesys and all

I was doing some thinking today about HGBnG and your trade manager. The hypothetical problem with the manager (not sure if it s problem since you've been trading successfully but...) when all positive trades are closed many of the hedged positions become unhedged and we are susceptible to a swift drawdown if the market decides to move further against the grid of open positions. Instead of closing all profitable positions, maybe we could close all profitable pairs. If a hedged pair is not in profit it would not be closed and left to sort itself out until it is closed on Friday or at the traders discretion.


& Another idea I had (this one is for Steve) is what if we used Stop Orders for WAVE andTREND trades, and RAD trades we use the Stop and limit or Non Hedged Stop Orders. Instead of using one or the other, we could use a mix of strategies for different signals.

Let me know what you think of these ideas.


Regards,

ALS
Author:  senzhi [ Thu May 19, 2016 4:33 am ]
Post subject:  Holy Graily Bob 'n Grid

Another idea I had (this one is for Steve) is what if we used Stop Orders for WAVE andTREND trades, and RAD trades we use the Stop and limit or Non Hedged Stop Orders. Instead of using one or the other, we could use a mix of strategies for different signals.
Firmly agree with the Non Hedged Stop Orders on this one. Just not so with Stop and Limit orders: I have quite a few RAD signals that show up either at or near top or bottom of the cycle.

Just like in the original HGB, pendings in the direction of the RAD signal proved to be either succesful or no loss at all. At least for me.

Serge
Author:  marcoFF [ Thu May 19, 2016 7:47 am ]
Post subject:  Holy Graily Bob 'n Grid

2016-05-19 09_43_15-2044141_ GlobalPrime-Demo - Conto dimostrativo (demo) - [XAGUSD,M30].jpg
I'll continue to leave a lot of money on the table.. :roll:
Author:  tomgu [ Mon May 23, 2016 1:34 am ]
Post subject:  Holy Graily Bob 'n Grid

Lifesys » Tue May 17, 2016 6:46 am wrote:Always good to mind what we are really doing here.
As a mathematician & coder it is a little irksome realising much trading is about dealing with mass human sentiment, often not logic. Hence best of intentions ---
Not sure why you would want a TM to close (H4 EA) HGBnG trades each day - makes no sense to me. Stated previous, big mistake is trying to pick which pair, which day.
Have used/looked at many 'Trade/Basket Managers' incl. Steve's brilliant MPTM & HGB BM, but for HGBnG wanted a simple "take off the cream" && 'keep me out of trouble' manager, hence - "TargetCloseAllTradesBE", as no other TM I know uses auto-incrementing targets. [Maybe we could add feature to MPTM] HGBnG already has many MPTM features anyway.

Traded Price Action manually for some years but test most SH EAs, I believe have a sound basis. With great respect for Bob (& Steve), last 6 months ran one 'control' HGBnG version whilst concurrently testing newer mods. on 16 pairs. When a newer version statistically performed better for >week on same pairs, the 'control' was replaced. As mentioned, to this point HGBnG v.Q on H4 (with code fixes), is far & away the best net performer. Using only H4 as H1 or less generates huge drawdown & stable.

Since 5 April 2016 my $x,000 live, real money, account has near doubled due to HGBnG v.Q + 'TargetCloseAllTradesBE-EA' to take off the cream & close everything down at end of each week - ie. zero(0) DrawDown at weekends.
Compounding current live account growth rate, just $1000 could become >$60,000 in 10 months with managed HGBnG.

At SH forum we do not give much credence to Empty4 B.testing but it can be useful to check stability. v.1Q gains well live & even b.tests at >300% p/a.

'Balance' is not real, net, all out, money in the bank, Profit. No point in having increasing balance when the DD is larger. Unsure why, but trying to replicate v.1Q with versions 2, all MA filters false; 'Offsetting' false; 'Stop_Orders_Only';, for me, the DrawDown is uncontrollable.
Since most have lost v.Q, I have attached copy I use on my live account. No guarantee it will perform same as management is paramount. Occasionally I do look & close some 'obviously' wrong direction whipsaw trades.
[3Q is the HGBnG generation, 260 is MagicNumber tag, tradetimes tweeked for Aust. GMT+10, PipFactor.mqh embedded, 'showButtons' option added as I find they clutter charts - sorry Steve]

regards
Paul
Hi, Paul
As you removed the trading hours, I am not sure when start on Monday. It close all 22pm Aust. GMT +10 on Friday night?
Author:  trader689 [ Mon May 23, 2016 3:19 am ]
Post subject:  Holy Graily Bob 'n Grid

im also a bit confused about v3q time inputs. there is no 'monday start hour' option like in other versions; as far as i can see i have mind set to 24h trading and it still has a message that it is outside trading hours
Author:  Logs [ Mon May 23, 2016 3:36 am ]
Post subject:  Holy Graily Bob 'n Grid

Hi Trader689
'24h local time = 14', should fire up at 1400 your time.
I've set mine to 13 AEST and away she went ;)
Author:  Lifesys [ Mon May 23, 2016 3:40 am ]
Post subject:  Holy Graily Bob 'n Grid

Hi Tomgu
Brave discussing times here!!
As more use VPS services and DayLightSaving changes in countries, it does attract attention for an easy solution.
An excellent site http://wwp.greenwichmeantime.com/time-g ... rld-clock/ shows actual place times.
Great thing about companies like Global Prime is they can tell the time - all H4 candles are same span. [Those using non-NY close brokers, no-one can really help.]

In 'HGBnG' my set times would be -
Stop/Start 14 (+10 AEST) = 07 (+3 GPserver) = 05 (+1 BST Lond.) = 00 (-5+1 DST NY)
[for US East coast, as 'MondayStartHour' is an INT and 'if (h < MondayStartHour)' it should start at 00 Monday]
In 'TargetCloseAllTradesBE-EA' set times would be -
'CloseAll' 22 (+10 AEST) = 15 (+3 GPserver) = 13 (+1 BST Lond.) = 08 (-5+1 DST NY)

Of course you can change these to anything you like - I just find it convenient to trim 7hrs (bounces) off each end of FX week.

Not sure what you mean by "removed the trading hours".
What I do know is that certain versions of Empty4 (for some stupid reason) show the "// (coder) remarks" instead of Experts/Properties/Inputs panel Variable names. TM version attached is same but with double semicolons which (usually) fixes this absurd problem. You may need to do same with HGBnG & recompile. Just remember all times are LOCAL.

regards
Paul
Author:  TraderJoeForex [ Mon May 23, 2016 9:50 pm ]
Post subject:  Holy Graily Bob 'n Grid

Lifesys » Tue May 17, 2016 6:46 am wrote:Always good to mind what we are really doing here.
As a mathematician & coder it is a little irksome realising much trading is about dealing with mass human sentiment, often not logic. Hence best of intentions ---
Not sure why you would want a TM to close (H4 EA) HGBnG trades each day - makes no sense to me. Stated previous, big mistake is trying to pick which pair, which day.
Have used/looked at many 'Trade/Basket Managers' incl. Steve's brilliant MPTM & HGB BM, but for HGBnG wanted a simple "take off the cream" && 'keep me out of trouble' manager, hence - "TargetCloseAllTradesBE", as no other TM I know uses auto-incrementing targets. [Maybe we could add feature to MPTM] HGBnG already has many MPTM features anyway.

Traded Price Action manually for some years but test most SH EAs, I believe have a sound basis. With great respect for Bob (& Steve), last 6 months ran one 'control' HGBnG version whilst concurrently testing newer mods. on 16 pairs. When a newer version statistically performed better for >week on same pairs, the 'control' was replaced. As mentioned, to this point HGBnG v.Q on H4 (with code fixes), is far & away the best net performer. Using only H4 as H1 or less generates huge drawdown & stable.

Since 5 April 2016 my $x,000 live, real money, account has near doubled due to HGBnG v.Q + 'TargetCloseAllTradesBE-EA' to take off the cream & close everything down at end of each week - ie. zero(0) DrawDown at weekends.
Compounding current live account growth rate, just $1000 could become >$60,000 in 10 months with managed HGBnG.

At SH forum we do not give much credence to Empty4 B.testing but it can be useful to check stability. v.1Q gains well live & even b.tests at >300% p/a.

'Balance' is not real, net, all out, money in the bank, Profit. No point in having increasing balance when the DD is larger. Unsure why, but trying to replicate v.1Q with versions 2, all MA filters false; 'Offsetting' false; 'Stop_Orders_Only';, for me, the DrawDown is uncontrollable.
Since most have lost v.Q, I have attached copy I use on my live account. No guarantee it will perform same as management is paramount. Occasionally I do look & close some 'obviously' wrong direction whipsaw trades.
[3Q is the HGBnG generation, 260 is MagicNumber tag, tradetimes tweeked for Aust. GMT+10, PipFactor.mqh embedded, 'showButtons' option added as I find they clutter charts - sorry Steve]

regards
Paul

Thanks for confirming your positive results with this version Paul.

I have been playing with this and for anyone interested here is a 120 period Semifor CLOSE ALL set for H4 which I am using for an idea I got from the HGI + Sixths + bit of Semafor strategy posted here which I found interesting: http://www.stevehopwoodforex.com/phpBB3 ... =88&t=4720

I have adapted slightly using 30LWMA rather than 8EMA and RAD / Trend for trigger no Wavy since there is no Sixths just the 120 Semifor to close all at similar extremes and once price crosses the 30LWMA trade back to the opposite 120.

Bit of fun for me but it looks promising...
Author:  TraderJoeForex [ Mon May 23, 2016 11:49 pm ]
Post subject:  Holy Graily Bob 'n Grid

TraderJoeForex » Mon May 23, 2016 9:50 pm wrote:
Thanks for confirming your positive results with this version Paul.

I have been playing with this and for anyone interested here is a 120 period Semafor CLOSE ALL set for H4 which I am using for an idea I got from the HGI + Sixths + bit of Semafor strategy posted here which I found interesting: http://www.stevehopwoodforex.com/phpBB3 ... =88&t=4720

I have adapted slightly using 30LWMA rather than 8EMA and RAD / Trend for trigger no Wavy since there is no Sixths just the 120 Semifor to close all at similar extremes and once price crosses the 30LWMA trade back to the opposite 120.

Bit of fun for me but it looks promising...

Here's an example visually :

120 period semafor printed high so EA would be waiting for price crossing below 30LWMA for setup and first RAD or TREND signal short to trigger stop and limit grid entry.

:jump:
Author:  Boulder [ Tue May 24, 2016 5:01 pm ]
Post subject:  Holy Graily Bob 'n Grid

As this is my first post on this forum I as well would like to take a few sentences here to express my fullest and sincerest gratitude. During the past few months, operating under the cloak of a 'silent lurker', SHF has become utterly invaluable to me and has allowed me to significantly enhance my own learning curve. The sheer knowledge, creativity and ingenuity to be found in here provided by people such as Steve Hopwood himself, Milanese, Renexxxx, Retireme, Lifesys, Wavegarrick, TraderJoeForex, JLCGarcia, Nanningbob, CaptainJack and many many other keep blowing my mind time and again. Therefore, people of SHF, thank you. Your selflessness and work truly have not gone unnoticed! :clap: :clap: :clap:

Now on to more relevant matters:

1. 130 errors

Like some of you, I as well encountered this persistent 130 error (invalid stops error) quite regularly with the latest version of HGBnG (V2c). Now logically this error will always pop up when the EA is either prevented from using the elements it needs to calculate the grid, TPs, SLs etc. or if it's just not feasible for the EA to send the orders through when e.g. the market is too close to the pending price. A potential solution for the latter is provided by Steve via SendImmediateMarketTrade. Personally my 130 errors were due to the former. I quickly discovered that it had to do with the default ATR indicator of Empty4. For some odd reason the daily ATR, regardless of the period, just wouldn't work for those pairs related to the error. ATR applied to every other timeframe (from M1 tot MN) did work however. A quick reinstall of an entirely new and fresh Empty4 terminal seemed to do the trick and now the HGBnG is happily doing its thing error-free :) I'm still clueless as to how I broke my ATR indicator though :?:

2. 132 errors

If you experience these errors, do not worry too much since this is related to the trading hours of Global Prime. More info to be found here https://www.globalprime.com.au/client-a ... ing-hours/. HGBnG will basically be halted for a few minutes during 23:59-00:02 (GP) platform time and normally should resume its operations thereafter.

3.
ALoveSvpreme wrote: I was doing some thinking today about HGBnG and your trade manager. The hypothetical problem with the manager (not sure if it s problem since you've been trading successfully but...) when all positive trades are closed many of the hedged positions become unhedged and we are susceptible to a swift drawdown if the market decides to move further against the grid of open positions. Instead of closing all profitable positions, maybe we could close all profitable pairs. If a hedged pair is not in profit it would not be closed and left to sort itself out until it is closed on Friday or at the traders discretion.
ALS these potential scenarios came to my mind as well at first but must say I had relatively little problems with it until this point after nearly 2 weeks of testing. Since we should not consider any EA as a set-and-forget one I always actively try to manage the demos. The moments I'd recommend to do the management are, as wisely suggested by Nanningbob, halfway during the London and US session. E.g. just a few moments ago, Lifesys' TargetCloseAllTradesBE-EA kicked in at 4% equity growth and closed all the winning orders that constituted this growth, leaving me with the following situation for GBPUSD:
first.jpg
In this case I'll just delete the pending sell orders of course. Now slightly more complex scenarios may indeed arise when e.g. the closest sell pending orders in the above picture would have been activated. These cases however do not occur too often since HGBnG is equipped with more than adequate offsetting mechanisms. Should the scenario occur, I'll practically always close out the entire position of the pair at a net profit and let HGBnG do its thing again from scratch. In this light, I do agree with you ALS that closing a single pair in net profit and letting a pair in net loss sort itself out might potentially enhance the performance (as opposed to the basket of positive trades). I for one just do not like having unhedged positions and find that timely, substantial capitalisations are much more rewarding to the mind and ease off one's endured stress levels. Regular management/intervention shall always be and remain paramount. Trading with consistency and discipline this way sure seems to enable a daily 1-2% compounded return on average (which is an incredible feat in and of itself..). In any case, HGBnG proves to be of sheer beauty. Thanks once again to breathe life in it people :)


Cheers,

Aram
All times are UTC Page 88 of 99