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

GridDancer / GridMaster
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=4709
Page 38 of 41
Author:  acostafulano [ Fri Nov 18, 2016 3:41 pm ]
Post subject:  GridDancer / GridMaster

What are you guys using for stop loss, (in terms of US$ or Pips)???

Im using an average of USD 180 EquitySL (which would equate to around 1800 pips) that hasnt been hit yet, but I'm not sure if this is too high or low to be profitable in the long run.

Any suggestions?

Diego
Author:  renexxxx [ Fri Nov 18, 2016 7:35 pm ]
Post subject:  GridDancer / GridMaster

Not sure. I've disabled the EquitySL and I'm using the Max Currency Exposure filter in order not to get into a massive drawdown. If/when the filter kicks in, the dancers are basically waiting for the tide to turn. That's the theory at least, time will tell if this survives everything the market throws at it. Key is not to be greedy and not to over expose.

Here is my update for the week:
Capture.JPG
At some point it showed a scary 3% drawdown, but up to now, it has always been able to recover.

Have a good weekend!
Author:  kwchau [ Sat Nov 19, 2016 12:47 am ]
Post subject:  GridDancer / GridMaster

renexxxx » Sat Nov 19, 2016 3:35 am wrote: Not sure. I've disabled the EquitySL and I'm using the Max Currency Exposure filter in order not to get into a massive drawdown. If/when the filter kicks in, the dancers are basically waiting for the tide to turn. That's the theory at least, time will tell if this survives everything the market throws at it. Key is not to be greedy and not to over expose.
!
Here is my update on 18 pairs. It is doing well but not as good as Rene because I have had more loss pairs in the basket. The GBP pairs, except GBPJPY which is very profitable, caused the drawdown and lower profit. So I think in a live account, we should avoid the GBP pairs including GBPUSD, except GBPJPY.

The drawdown yesterday was caused by GBPNZD but dancer recover very well.

Next week, I will start a new test only on profitable pairs from the 18-pairs test and then I can compare results with Rene's standard pair set.
Author:  renexxxx [ Sat Nov 19, 2016 9:16 am ]
Post subject:  GridDancer / GridMaster

There are new versions of GridMaster (version 1.9) and HGIDancer (versioni 1.4) in Post 1 of this thread.

These updates are all about the automatic sizing of various parameters, in relation to the account's available equity.

For GridMasterv1.9 we now have the parameters:
  • BasketProfit / Equity: (Default: 0.0001). If the account equity is $10,000, BasketProfi6t will be calculated as $10,000 * <the value of this parameter>
  • MinProfit / Equity: (Default: 0.00005). Similarly, for a $10,000 account, MinProfit will be $10,000 * <the value of this parameter>
  • EquityTP / Equity: (Default: 0.0005). You get the idea.
  • EquitySL / Equity: (Default: -1.0). Same here. A negative value means: disable this function.
So, if the equity increases (and hopefully it does), the BasketProfit, MinProfit, EquityTP and EquitySL do too.

For HGIDancerv1.4 we have two new/changed parameters:
  • Maximum Currency Exposure / Equity: (Default: 0.00002) The default value makes the MaxCurrencyExposure 0.20 for an account of size $10,000, which seems right.
  • LotEquityRatio: (Default: 0.000005 (five zeroes behind the decimal point). This gives a base lot size of 0.05 for a $10,000 account. The calculated lot size will always be rounded to the nearest multiple of MINLOT.
In an attempt to make the dollar contribution of each pair to the overall basket profit the same, each HGIDancer works out a lotSizeMultiplier. The base lot size (as obtained from the LotEquityRatio * Equity) is then multiplied by this lotSizeMultiplier. So, in effect each HGIDancer will be placing trades with different lot sizes. How, does the HGIDancer work out this lotSizeMultiplier? Every minute, it looks around (on the dance floor) to see which other HGIDancers are trading and what symbol each of them is trading. Then it works out for each traded symbol what its typical daily range is (with the ATR) and multiplies this number with the symbols' TICKVALUE to quantify what the $-value would be if that pair would move its typlcal daily range. These numbers are added and the average is taken. Each Dancer's lotSizeMultiplier can then be calculated as the ratio of its own pairs 'Daily $-value' versus the average 'Daily $-value'. Hope this makes sense -- if not, you have to take my word for it. As the Dancer's do this little exercise every minute, it doesn't matter whether you are adding or removing Dancers to the terminal: the lotSizeMultiplier will be re-evaluated if you do.

It would make sense to set the Max Currency Exposure / Equity and LotEquityRatio the same for each HGIDancer, although you have the flexibility not to.

Have fun, I know I will.
Author:  kwchau [ Sat Nov 19, 2016 11:23 am ]
Post subject:  GridDancer / GridMaster

renexxxx » Sat Nov 19, 2016 5:16 pm wrote:
For GridMasterv1.9 we now have the parameters:
  • BasketProfit / Equity: (Default: 0.0001). If the account equity is $10,000, BasketProfi6t will be calculated as $10,000 * <the value of this parameter>
  • MinProfit / Equity: (Default: 0.00005). Similarly, for a $10,000 account, MinProfit will be $10,000 * <the value of this parameter>
  • EquityTP / Equity: (Default: 0.0005). You get the idea.
  • EquitySL / Equity: (Default: -1.0). Same here. A negative value means: disable this function.
Hi Rene, just wonder should the default values for BasketProfit, MinProfit and EquityTP be 0.001, 0.0005 and 0.005 respectively for a $10,000 account. Please accept my apology if I am wrong !

Cheeers !
kwchau
Author:  renexxxx [ Sat Nov 19, 2016 7:47 pm ]
Post subject:  GridDancer / GridMaster

kwchau » Sat Nov 19, 2016 9:23 pm wrote: Hi Rene, just wonder should the default values for BasketProfit, MinProfit and EquityTP be 0.001, 0.0005 and 0.005 respectively for a $10,000 account.
You are absolutely right! I added too many zeroes in my enthusiasm. Just re-download GridMasterv1.9 from Post 1.

Great spot, kwchau, and thanks for letting me know :clap: :clap: :clap:
Author:  kwchau [ Sun Nov 20, 2016 12:03 am ]
Post subject:  GridDancer / GridMaster

renexxxx » Sun Nov 20, 2016 3:47 am wrote: You are absolutely right! I added too many zeroes in my enthusiasm. Just re-download GridMasterv1.9 from Post 1.
You have good reason to be enthusiastic. GridDancer is getting better and better with your perseverance.
Thank you very much for sharing it with us :good:

cheers,
kwchau
Author:  vsmakkal [ Sun Nov 20, 2016 7:29 am ]
Post subject:  GridDancer / GridMaster

Thanks Rene, for the updated EA. Sounds like a well thought out upgrade.. Thanks Mate,

Sara
Author:  traderduke [ Sun Nov 20, 2016 2:21 pm ]
Post subject:  GridDancer / GridMaster

Thanks Rene , it just keeps getting better :clap:
Ray
Author:  acostafulano [ Tue Nov 22, 2016 6:52 pm ]
Post subject:  GridDancer / GridMaster

Hello Rene,

I've noticed something that has left me confused.

Just to note: My demos of HGI GM are doing great, but I'm seeing some behaviour that I would like to check with you, just to make sure that I'm understanding correctly what is going on with the trading of this little awesome beast you have created

Please look at my current demos:

https://www.myfxbook.com/portfolio/1g-h ... -1/1853941
https://www.myfxbook.com/portfolio/3g-h ... -2/1853943
hgi gm.zip
If you go through the history of closed trades, you will notice that it has been over 24hrs since the time of last closure, and if you scroll back through the last closures that the EA has done, you will notice that basically all of them correspond to the respective "EquityTP ($8 and $12 respectively), and not the "MINPROFIT" $1 and $2 respectively that the EA is supposed to be doing.

Additionally, if you go to the open trades you will notice as well that there are many combinations of winning+losing trades that could be being closed out for the "minprofit" of $1 and $2 respectively.
1.PNG
note: Im trading HGI Dancer 1.1 and GM 1.8

Now, I may be understanding the logic completely wrong, but is the bot not supposed to be doing very tiny closures along the way in order to keep DD controlled (as the Balance/Equity charts posted by lots of guys here correctly depict?)

Even though my demos are running great, I would still understand exactly how the closure logic works in order to be more comfortable when going live eventually.

Thanks and sorry if this is a misguided post

Diego
All times are UTC Page 38 of 41