Pippo: a Murrey Maths retrace auto-trading EA by SteveHopwood | Page 14 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Pippo: A Murrey Maths Retrace EA 188 replies

Maths (ratio) problem 16 replies

Do i need to be a maths genius? 12 replies

How important is a knowledge of maths? 96 replies

Stupid maths 19 replies

Attachments: Pippo: a Murrey Maths retrace auto-trading EA by SteveHopwood
Exit Attachments

Pippo: a Murrey Maths retrace auto-trading EA by SteveHopwood

  • #261
  • Jun 29, 2011 8:54am Jun 29, 2011 8:54am
Quoting macman
I suggest keeping a very close watch on this EA, if you are using it live, as my basic strategy test was margin called - as expected.
Ok, I have put some thought into this after seeing a death trade last week and eur/gbp today and how the pairs react at the next octave. When a pair reaches the octave beyond the yellow line another trade needs to occur in sequence and add to the basket. The last 3 death trades that have occured have reversed after hitting the third octave. The 2x multiplier in the hedge is too much if the trade reverses at the 3/8 octave as the loss from that will not even be close to breaking even when the basket is closed. Another bad thing could also occur from the hedging strategy: What if the hedge gets closed out and then makes another run against you? That's right, another hedge would open and thus the vicious process starts over. My recommendation is a full basket closure some 20-30 pips above the 3/8 octave. This would have easily solved the death trade issues recently.
Another idea is to be able to let our profits run a little if we chose to do so. Maybe we can use a reversi type function for this? Thoughts? Better ideas? I'm ready to hear them.
  • #262
  • Jun 29, 2011 9:39am Jun 29, 2011 9:39am
Quoting Marlin Man
Ok, I have put some thought into this after seeing a death trade last week and eur/gbp today and how the pairs react at the next octave. When a pair reaches the octave beyond the yellow line another trade needs to occur in sequence and add to the basket. The last 3 death trades that have occured have reversed after hitting the .
Another re entry octave line beyond the yellow line a 3/8 and a -3/8 would be my choice as well , it would have avoided 2 hedge trades that went wrong on my demo and 3 death trades that later reversed on my live account.

A basket stop loss option of 20-30 pips above 3/8 octaves would also help. Initially before the death trade option I was using the MPTM with a hidden stop of 35 pips .

I was thinking last night in the bath that if we could move the dashed line to a .7 fraction from a .5 fraction we can squeeze in a few more pips or maybe take profit at the next solid line ? I tried to do this myself and compiled succesfully but when I opened it on a chart the lines were all wrong, my programming skills are still in the early stages otherwise I would share the modifications .

Another thought add a trend detection like we have in Reversi .

In my live account I turn off Pippo 1 hr befre any significant news is about to happen. I trade 15mins so generally 1 hr is sufficient time . It makes a difference and avoids death trades .A news filter would be ideal here but thats a tall order .

Agree fully with Macman , you have to keep an eye on Pippo at this stage of its development .
  • #263
  • Jun 29, 2011 12:30pm Jun 29, 2011 12:30pm
I think I have sorted out the rapid trade-placement/deletion thingy. I managed to catch Pippo at it when I had time to look into what was happening. Update in post 1.

I don't promise, but it has stopped this behaviour on my demo.

  • #264
  • Edited 3:55pm Jun 29, 2011 12:42pm | Edited 3:55pm
Quoting SteveHopwood
I think I have sorted out the rapid trade-placement/deletion thingy. I managed to catch Pippo at it when I had time to look into what was happening. Update in post 1.

I don't promise, but it has stopped this behaviour on my demo.

Hi Steve,
I'm running a D1 Robot, a D1 manual (and of course I wouldn't on a manual setup) and a W1 Robot and I'm not seeing any of the trade problem.....I'm going to update the D1 Robot and see how it reacts....
Thanks for this Steve....

whip
  • #265
  • Jun 29, 2011 4:33pm Jun 29, 2011 4:33pm
I've had 3 or 4 different demos running with various setups on Pippo and all have had the same problem with death trades and dropping below the minimum margin for trading etc, except for 1 version which is running Pippo without any of the really volatile pairs since 13/06 and I haven't had any problems with death trades.
I think it has twice had to stop trading due to the margin restriction but quickly and easily traded out of that problem as well.
Last night, it went through like a breeze and banked plenty of pips.
I also haven't updated pippo ea on this demo (as it was doing so well) since the update in post #121.
Attached Image (click to enlarge)
Click to Enlarge Name: accountequityanaliser_110630.gif Size: 15 KB
Attached Image
  • #266
  • Jun 29, 2011 6:31pm Jun 29, 2011 6:31pm
Hi Steve,

I spotted this... the brackets in bold italic is in a different place when you check buys in comparison to when you check sells - which way is it supposed to be?:

Inserted Code
if (Band < 5 && [b][i]([/i][/b]UseZeljko && !BalancedPair(OP_BUY)[b][i])[/i][/b] && FilledBuys == 0) return;

Inserted Code
if (Band > 8 && [b][i]([/i][/b]UseZeljko && [b][i]([/i][/b]!BalancedPair(OP_SELL)  && FilledSells == 0[b][i]))[/i][/b] ) return;
  • #267
  • Jun 29, 2011 6:37pm Jun 29, 2011 6:37pm
In relation to the above post, this is how the code looked before I downloaded your latest update:

Inserted Code
if (Band < 5 && (UseZeljko && (!BalancedPair(OP_BUY) && OpenTrades == 0)) ) return;

Inserted Code
if (Band > 8 && (UseZeljko && (!BalancedPair(OP_SELL)  && OpenTrades == 0)) ) return;

So I'm guessing you just didn't apply the bracket placement change you did for the buy check, to the sell check?
  • #268
  • Jun 29, 2011 6:40pm Jun 29, 2011 6:40pm
seems like EA doing better if the Bands height is 60pips or above between each band.
I've been forwarding pairs on H4, and got dragged into a deep DD when the EA opens trade too close to one another.
I think mininum bands height filter will help avoid this, or I could switch to D1
  • #269
  • Jun 29, 2011 6:47pm Jun 29, 2011 6:47pm
Quoting the-game
In relation to the above post, this is how the code looked before I downloaded your latest update:

Inserted Code
if (Band < 5 && (UseZeljko && (!BalancedPair(OP_BUY) && OpenTrades == 0)) ) return;
Inserted Code
if (Band > 8 && (UseZeljko && (!BalancedPair(OP_SELL)  && OpenTrades == 0)) ) return;
So I'm guessing you just didn't apply the bracket placement change you did for the buy check, to the sell check?
Cheers. The first one is correct, so the sell check should be:
Inserted Code
   if (Band > 8 && (UseZeljko && !BalancedPair(OP_SELL)) && FilledSells == 0) return;

Fixed nearly-fixed update in post 1.

  • #270
  • Jun 29, 2011 7:01pm Jun 29, 2011 7:01pm
Actually, it still isn't right. As coded, it will abort trades at the next level if there is an open trade.

It needs to be:
Inserted Code
if (Band < 5 && (UseZeljko && !BalancedPair(OP_BUY) && FilledBuys == 0) ) return;
and
Inserted Code
if (Band > 8 && (UseZeljko && !BalancedPair(OP_SELL) && FilledSells == 0) ) return;
I think.

Fixed fixed fix in post 1. We will get there one day.

Possibly.

  • #271
  • Jun 29, 2011 7:13pm Jun 29, 2011 7:13pm
Quoting fxozgirl
I've had 3 or 4 different demos running with various setups on Pippo and all have had the same problem with death trades and dropping below the minimum margin for trading etc, except for 1 version which is running Pippo without any of the really volatile pairs since 13/06 and I haven't had any problems with death trades.
I think it has twice had to stop trading due to the margin restriction but quickly and easily traded out of that problem as well.
Last night, it went through like a breeze and banked plenty of pips.
I also haven't updated pippo...
Could you tell us the pairs you are trading? In my live acct I went back to 4 - USDCHF, GBPUSD, EURUSD, USDJPY.

Thanks in advance...
  • #272
  • Jun 29, 2011 7:20pm Jun 29, 2011 7:20pm
Thanks Steve.

I'm wondering on which pairs everyone is experiencing death trades?
Just trying to find a pattern of which pairs to stay away from when using Pippo

EDIT: I'm assuming it's the pairs that are "prone" to strong trends. Please discuss.


sevenjones posted this earlier in the thread:

Quoting sevenjones
My $10,000 demo account is up 1.5% in five days, trading the following pairs: aud/cad; eur/gbp; gbp/cad; aud/nzd; chf/jpy; cad/chf. So far, all of my trades have come on three of those pairs: chf/jpy; gbp/cad; eur/gbp. Drawdown has been minimal.

I like this EA so far. Hope it can keep dd to a minimum. My trading philosophy is to use mostly the low- and medium-volatility crosses and stay away from usd pairs.
Maybe he was on to something with avoiding USD cross pairs.
  • #273
  • Jun 29, 2011 8:01pm Jun 29, 2011 8:01pm
Hi Steve

My question/statement: -

When we hit the margin filter I would have expected any pending trades to be deleted, otherwise the margin position could potentially get worse when the price is hit and the trade is opened.
My H4 demo is down to 700% margin and there are still about 5 pending trades in the list.

Pairs - I'm got everything open that the broker provides, some of the more exotic ones are providing great returns, e.g. Turkish Lira,

Thanks

Paul
  • #274
  • Edited 9:40pm Jun 29, 2011 9:28pm | Edited 9:40pm
Quoting batemap
Hi Steve

My question/statement: -

When we hit the margin filter I would have expected any pending trades to be deleted, otherwise the margin position could potentially get worse when the price is hit and the trade is opened.
My H4 demo is down to 700% margin and there are still about 5 pending trades in the list.

Pairs - I'm got everything open that the broker provides, some of the more exotic ones are providing great returns, e.g. Turkish Lira,

Thanks

Paul
I like the idea of all the pairs open (maybe not GbpJpy) but I think this EA to work properly, has to have enough cash in the account to avoid these Margin restrictions. It needs to open those next progressive trades to recover.

I have actually closed almost net zero open trades in other pairs to loosen up the margin which I shouldn't have to do.

I am now using a 0.05 starting lot size on a $10k account. That is probably too big as well but time will tell. I also want enough margin to be able to apply my own logic to some kind of recovery.

Right now everything is so quiet I have almost nothing open.

Dave R.
  • #275
  • Jun 29, 2011 10:03pm Jun 29, 2011 10:03pm
I have scoobs and ForexKiwi set to False and False. Just easier for testing...No need to worry about margin doing .01 on a demo.
  • #276
  • Jun 29, 2011 11:41pm Jun 29, 2011 11:41pm
Quoting droland
I like the idea of all the pairs open (maybe not GbpJpy) but I think this EA to work properly, has to have enough cash in the account to avoid these Margin restrictions. It needs to open those next progressive trades to recover.

I have actually closed almost net zero open trades in other pairs to loosen up the margin which I shouldn't have to do.

I am now using a 0.05 starting lot size on a $10k account. That is probably too big as well but time will tell. I also want enough margin to be able to apply my own logic to some kind of recovery.

Right...
Hi Dave

I agree with your statement 'It needs to open those next progressive trades to recover.' I've changed my Lots to 0.01, 0.02, 0.04 and 0.06 to try and prevent the margin restrictions.

So, from the filter point-of-view, should it consider all open and pending trades? This would be the conservative approach, assuming all the pending prices will be hit and the trades will be opened.

Paul
  • #277
  • Jun 29, 2011 11:53pm Jun 29, 2011 11:53pm
Quoting batemap
Hi Dave

I agree with your statement 'It needs to open those next progressive trades to recover.' I've changed my Lots to 0.01, 0.02, 0.04 and 0.06 to try and prevent the margin restrictions.

So, from the filter point-of-view, should it consider all open and pending trades? This would be the conservative approach, assuming all the pending prices will be hit and the trades will be opened.

Paul
Well, I am still studying this thing. I like catching whatever trades where I can but I think I will go in and turn off some of the pairs if I get too many open.

Like Steve says, you really can't set and forget but I do like what I see.

Dave R.
  • #278
  • Jun 30, 2011 12:48am Jun 30, 2011 12:48am
Sorry Steve.

Seems to be still doing its open/close thingy.
Attached File(s)
File Type: zip DetailedStatement.zip   51 KB | 314 downloads
  • #279
  • Jun 30, 2011 2:57am Jun 30, 2011 2:57am
The problem with the opening-closing thingy will be a simple logic thingy that I cannot solve.

Within void LookForTradingOpportunities() there are two lines of code that tell Pippo to check that a proposed pending trade will balance an existing filled trade. These are

Inserted Code
if (Band < 5 && (UseZeljko && !BalancedPair(OP_BUY) [color=Red]&& FilledBuys == 0[/color]) ) return;
and

Inserted Code
if (Band > 8 && (UseZeljko && !BalancedPair(OP_SELL) [color=Red]&& FilledSells == 0[/color]) ) return;

The highlighted bits are intended to allow Pippo to continue to send >L1 trades if there happen to be unbalanced trades open - either because of manual entry or two pendings filled at the same time.

Maybe this is an unnecessary hoop to jump through. Try deleting the highlighted conditionals and see if that stops the thingy.

  • #280
  • Edited 6:04pm Jun 30, 2011 12:46pm | Edited 6:04pm
I thought I would post my live account of Pippo. This started June 28 so just over one month.

Return was 12.5% for 6 days.

Dave R.

Note: Number of days trading corrected
Attached Image (click to enlarge)
Click to Enlarge Name: DetailedStatement.gif Size: 5 KB
1 trader viewing now