Trumpcard: a candlestick closing price breakout EA by SteveHopwood | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

optimizing my reverse candlestick closing strategy 14 replies

Candlestick Breakout 11 replies

Trumpcard's Mountain of Gold 11 replies

SteveHopwood EAs 4 replies

Trumpcard's Journal 47 replies

Attachments: Trumpcard: a candlestick closing price breakout EA by SteveHopwood
Exit Attachments

Trumpcard: a candlestick closing price breakout EA by SteveHopwood

  • #41
  • Oct 11, 2011 9:56am Oct 11, 2011 9:56am
Quoting zeropoint
Dear Trumpcard,

Actually, I interpret that we cancel the pending only when the low of the trigger bar is breached, for a long trade, and vice versa for a short trade.

Thanks!
I was actually wondering earlier whether to alter the rules to allow for this scenario but at the moment just gonna focus on the original setup.
  • #42
  • Oct 11, 2011 10:15am Oct 11, 2011 10:15am
In the below pic, we can see Bar 1 is the trigger and so the sellstop would be placed 1 pip below the low.

Bar 2 is an inside bar and so the pending order remains.

Bar 3 is an inside bar within Bar 2, ie an inside bar within an inside bar, and so the pending order continues to remain.

The pending order will be cancelled if Bar 4 does not breach the trigger bar's low by 1 pip. We can see Bar 4 was not another inside bar AND it did not activate the pending order so the pending order is cancelled and we wait for the next trigger bar to form.
Attached Image (click to enlarge)
Click to Enlarge Name: 2 inside bars.JPG Size: 64 KB
  • #43
  • Oct 12, 2011 9:46am Oct 12, 2011 9:46am
Another feedback. The statement:
Inserted Code
 
if (Low[cc] < Low[1] ) return(false);
in the WasCloseLow() function should be:

Inserted Code
 
if (Close[cc]<Close[1]) return(false);

Preferably, to be more inline with Dan's rule, it should be:
Inserted Code
 
if ((Close[cc]-0.0001))<Close[1]) return(false);

The 0.0001 is specific to EU.

A similar thing could be done for WasCloseHigh(), except in the other direction.
  • #44
  • Oct 12, 2011 5:50pm Oct 12, 2011 5:50pm
thanks again zeropoint. Would the code need to allow for for trading E/J and perhaps other yen pairs?
  • #45
  • Oct 12, 2011 7:32pm Oct 12, 2011 7:32pm
Quoting Trumpcard
thanks again zeropoint. Would the code need to allow for for trading E/J and perhaps other yen pairs?
Sorry, I did not include more details in case I clutter the thread too much. You can do this to handle other pairs:
Inserted Code
bool WasCloseLow(int multiplier)
{
   //Returns true if the previous close was lower than that of the previous LookBackCandles, else returns false
   int multiplier;
   if(Digits == 2 || Digits == 4) multiplier = 1;
   if(Digits == 3 || Digits == 5) multiplier = 10;
   if(Digits == 6) multiplier = 100;
   if(Digits == 7) multiplier = 1000;
   double buf=1*multiplier*Point; // Trigger candle has at least 1 pip higher close.
   int consult = LookBackCandles + 1;
   for (int cc = 2; cc <= consult; cc++)
   {
      if ((Close[cc]-buf) < Close[1] ) return(false);
   }//for (int cc = 1; cc <= consult; cc++)
   //Got this far, so is a low
   return(true);
}//End bool WasCloseHigh()
  • #46
  • Oct 13, 2011 5:49pm Oct 13, 2011 5:49pm
Quoting zeropoint
Another feedback. The statement:
Inserted Code
 
if (Low[cc] < Low[1] ) return(false);
in the WasCloseLow() function should be:

Inserted Code
 
if (Close[cc]<Close[1]) return(false);
Preferably, to be more inline with Dan's rule, it should be:
Inserted Code
 
if ((Close[cc]-0.0001))<Close[1]) return(false);
The 0.0001 is specific to EU.

A similar thing could be done for WasCloseHigh(), except in the other direction.
Nice one. Thanks. I am always mixing up my High/Open Close/Low mixed up.

Don't need the extra pip thingy; taken care of elsewhere.

Update in post 1.

Cheers zeropoint

  • #47
  • Oct 14, 2011 4:37am Oct 14, 2011 4:37am
EJ 13th Oct 2011
Attached Image (click to enlarge)
Click to Enlarge Name: ej 15 min 13-10-2011.JPG Size: 75 KB
  • #48
  • Oct 14, 2011 6:11am Oct 14, 2011 6:11am
Hi there, I've been running TC since Monday and it is doing very nicely. First half day was OOTB on 15 M Tf and 7 losing trades an the trot.

So I've set TP at 100, BE to 10, BE profit 5, TS 15 and SL 30. It's running on 1M EU and AU.

Since Monday afternoon I've had 63 trades with 45 wins and 18 losses.

Using 0.01 lots that makes £14.36 so far, not bad !!

Anyone else with good results ?

Just one thing though is the screen gets a bit messy with lots of old SL and trade lines which can get a bit confusing.

Any way of getting them deleted automatically when the relevant trades are closed ?

Cheers
  • #49
  • Oct 14, 2011 10:40am Oct 14, 2011 10:40am
For today, I can report 14 wins, 2 loss, using 10/10, on EU and EJ. EJ loaded just 2 hours ago...
I don't say
  • #50
  • Oct 14, 2011 11:41am Oct 14, 2011 11:41am
Quoting dragosd1
For today, I can report 14 wins, 2 loss, using 10/10, on EU and EJ. EJ loaded just 2 hours ago...
What TF?
  • #51
  • Oct 14, 2011 12:07pm Oct 14, 2011 12:07pm
Quoting acostafulano
What TF?
M15
I don't say
  • #52
  • Oct 14, 2011 11:40pm Oct 14, 2011 11:40pm
Quoting dragosd1
For today, I can report 14 wins, 2 loss, using 10/10, on EU and EJ. EJ loaded just 2 hours ago...
Great! was this 24 hr trading or GMT 9am to GMT 8pm?
  • #53
  • Oct 14, 2011 11:47pm Oct 14, 2011 11:47pm
Quoting Rascacio
Hi there, I've been running TC since Monday and it is doing very nicely. First half day was OOTB on 15 M Tf and 7 losing trades an the trot.

So I've set TP at 100, BE to 10, BE profit 5, TS 15 and SL 30. It's running on 1M EU and AU.

Since Monday afternoon I've had 63 trades with 45 wins and 18 losses.

Using 0.01 lots that makes £14.36 so far, not bad !!

Anyone else with good results ?

Just one thing though is the screen gets a bit messy with lots of old SL and trade lines which can get a bit confusing.

Any way of getting them...

hi Rascacio, were you running the bot 24 hours? That sounds like a lot of trades taken especially with a 100tp. Was there more than 1 trade running at a time on each pair?
  • #54
  • Edited 12:19am Oct 15, 2011 12:08am | Edited 12:19am
see pic
edit: the pic should read that a TB's CLOSE needs to be lower than ALL 3 previous bars' closes
Attached Image (click to enlarge)
Click to Enlarge Name: TB ID.JPG Size: 55 KB
  • #55
  • Oct 15, 2011 12:40am Oct 15, 2011 12:40am
for the first 2 weeks of October, EJ is showing 58 wins and 42 losses while EU has 72 wins and 53 losses.
  • #56
  • Oct 15, 2011 2:04am Oct 15, 2011 2:04am
Quoting Trumpcard
hi Rascacio, were you running the bot 24 hours? That sounds like a lot of trades taken especially with a 100tp. Was there more than 1 trade running at a time on each pair?
Hi Trumpcard, I was running the bot OOTB from Sunday to Wednesday then thought I'd give it a whirl 24Hr. No multiple trades but often as a trade closed it would open another straight away.

The 100 Tp was just to try and keep the trades open to catch any major moves. I messed about with the Ts an Be a few times but not really sure what's best yet as it's only been a week testing so far !!!!

I'll update my trade numbers later as I'm away from the computer for a few hours today.
  • #57
  • Edited 6:46am Oct 15, 2011 6:09am | Edited 6:46am
Quoting Trumpcard
Great! was this 24 hr trading or GMT 9am to GMT 8pm?
It was from 9:00 GMT until end of the day. 10/10, but EJ was loaded only in the last 3-4 hours of trading.
It was, at the market close, 15 wins and 4 loss. Pretty good.

Cummulative results 13.10-14.10, total 24 wins, 13 loss (on Thursday I was playing with 20/20, and was only on EU).

Settings are 10/10, look back candles 3, no trade management at all, hours to trade 9-24 GMT.
I don't say
  • #58
  • Oct 15, 2011 1:52pm Oct 15, 2011 1:52pm
Given the high win/loss rate of the bot, I've have took the liberty to add to it a Martingale sequence, from Steve's Spider. Hope you don't mind, Steve!

Also fixed TS issues, due to copy/paste . If you use TS feature, be advised, this version trails the price movement BEFORE the BE, as soon as profit is > 0. I prefer this option, to minimize the losses.Don't really know how this will interfere with BE option, I don't use it.

It's worthless to test this in backtester, due to huge spread will kill your equity fast. I think setting are pretty self-explanatory...

Do NOT use this on real account! It's for demo testing purposes only. So, have fun and post forward results, I have a feeling this could have legs.

P.S. I'm just learning MQL, so, code is not bulletproof! Steve is really busy with Hanover, so I hope everything is fine into this.At least it works in backtester, we will see Monday...Funny thing, as I'm writing this, backtester is running,I'm just verifying logical operations, but it shows profit after 1 month, with 80 spread, yes, 80! Hahaha....
Attached File(s)
File Type: mq4 Trumpcard auto trading robot by Steve Hopwood(with Mart).mq4   56 KB | 501 downloads
I don't say
  • #59
  • Oct 15, 2011 2:12pm Oct 15, 2011 2:12pm
Quoting dragosd1
Given the high win/loss rate of the bot, I've have took the liberty to add to it a Martingale sequence, from Steve's Spider. Hope you don't mind, Steve!

Also fixed TS issues, due to copy/paste . If you use TS feature, be advised, this version trails the price movement BEFORE the BE, as soon as profit is > 0. I prefer this option, to minimize the losses.Don't really know how this will interfere with BE option, I don't use it.

It's worthless to test this in backtester, due to huge spread will kill your equity fast. I think setting are pretty self-explanatory...

Do...
Just a word of caution, I've been testing this manually on E/U in Forex Tester 2, going back to 2008 and picking random days, weeks, months using the default M15 10/10 and also trying 20/20. So far my account has been eaten alive. Not conclusive I know but I've had more losses than wins so please don't anyone try using Martingale on a live account just yet!
  • #60
  • Oct 15, 2011 3:34pm Oct 15, 2011 3:34pm
Quoting Swg123
Just a word of caution, I've been testing this manually on E/U in Forex Tester 2, going back to 2008 and picking random days, weeks, months using the default M15 10/10 and also trying 20/20. So far my account has been eaten alive. Not conclusive I know but I've had more losses than wins so please don't anyone try using Martingale on a live account just yet!
As I just said! Two days of forward testing doesn't mean nothing!

On the other hand, Martingale can make the difference, if you select max lots 0.64, you can lose only if you have 7 losses in a row! But, again, just TEST it!
I don't say
0 traders viewing now