Caillou wrote:Hello to everyone,

This is my first post here and I want first to thanks Steve and all other members for such a great team.
I´ve been playing with Phantom EA´s for a long time, basically a grid strategy and sooner or later, always, a "death trade" appears. It is unavoidable. But this doesn´t mean that this EA will not be profitable in the long run.
I think we need 2 conditions for making money:

1.- The "death trade" must happen as little as possible. For that we have CZ boxes and we need taking trades for the less ranging pairs. Bob has made a lot of work here. A few months ago the most ranging pairs were (in his opinion): aud/jpy, aud/nzd, aud/usd, cad/jpy, chf/jpy, eur/gbp, gbp/cad, gbp/usd, nzd,chf, nzd/jpy, usd/cad, usd/jpy. We need trending pairs here.

2.- The "death trade" must be cutted when it happens, and it will happen, sooner or later. This wonderfull EA has already an option for that: "Rangeboundtradeclosure". Mabye 10 Buys/10 Sells is the limit, or 20/20, or 5/5......who knows, that is what we have to test.

Even with a few big losses, this EA can be profitable because we need this losses, otherwise our SL is our account.

Caillou wrote:Hello to everyone,

1.- The "death trade" must happen as little as possible. For that we have CZ boxes and we need taking trades for the less ranging pairs. Bob has made a lot of work here. A few months ago the most ranging pairs were (in his opinion): aud/jpy, aud/nzd, aud/usd, cad/jpy, chf/jpy, eur/gbp, gbp/cad, gbp/usd, nzd,chf, nzd/jpy, usd/cad, usd/jpy. We need trending pairs here.

2.- The "death trade" must be cutted when it happens, and it will happen, sooner or later. This wonderfull EA has already an option for that: "Rangeboundtradeclosure". Mabye 10 Buys/10 Sells is the limit, or 20/20, or 5/5......who knows, that is what we have to test.

Even with a few big losses, this EA can be profitable because we need this losses, otherwise our SL is our account.

acostafulano wrote:

Caillou wrote:Hello to everyone,

1.- The "death trade" must happen as little as possible. For that we have CZ boxes and we need taking trades for the less ranging pairs. Bob has made a lot of work here. A few months ago the most ranging pairs were (in his opinion): aud/jpy, aud/nzd, aud/usd, cad/jpy, chf/jpy, eur/gbp, gbp/cad, gbp/usd, nzd,chf, nzd/jpy, usd/cad, usd/jpy. We need trending pairs here.

2.- The "death trade" must be cutted when it happens, and it will happen, sooner or later. This wonderfull EA has already an option for that: "Rangeboundtradeclosure". Mabye 10 Buys/10 Sells is the limit, or 20/20, or 5/5......who knows, that is what we have to test.

Even with a few big losses, this EA can be profitable because we need this losses, otherwise our SL is our account.

I agree with you Caillou. We can't let this thing get to a grid of 18 open trades, PA in the middle of the grid and -900 pips in DD, that's simply not realistic (at least for some of us). I'd rather take the hit before it gets that bad and start over. Besides, during all that time in DD we can get fresh baskets and make up for the occassional hit. On top of this, if we choose the pairs wisely (creme de la creme) this will be an even more rare event.

What I feel comfortable with, for example, would be with a -250 pip per basket (or around $25 for 0.01 as I trade due to the size of my account (0.01 per 1000 USD).

For that, I was thinking of using a RangeboundTradeClosure of 10, (5 buys and 5 sells and close the grid at that point and start over), but looking at the code part of this function I noticed something:


bool IsBasketStopLossHit(int type)

/*
This is not a stop loss in the accepted sense.
This function returns true if OpenTrades + 1 would equal RangeBoundClosureTotal with an equal number
of buys and sells.
Called when a pending trade is about to fill
*/

and the code:

int half = RangeBoundClosureTotal / 2;

if (type == OP_BUY)
{
if (OpenSells < half) return (false);//Not at max yet
//Are at max, so will one more buy cause OpenTrades to = RangeBoundClosureTotal
if (OpenTrades + 1 >= RangeBoundClosureTotal) return(true);
}//if (type == OP_BUY)


I'm not sure how this works so I hope SH or some other coders can clarify this out:

Let me illustrate an example if I set RANGEBOUNDCLOSURE at 10.... A buy trade fills and then PA triggers 8 buys in a row. Imagine the TP is really high so it don't get hit and then the PA reverses all the way and starts triggering sell trades of the opposite side of the grid. My question is:

- Will this function close the basket at the 2nd sell triggered? (8+2= 10)?

or as the comment suggests

- Will the function close the basket only when "OpenTrades + 1 would equal RangeBoundClosureTotal with an equal number of buys and sells" (8+8= 16)?

The second scenario wouldn't work for me as I can't handle that much DD as a trader (I'll get too anxious hehe

Thanks

mantkinth wrote:Hmm hedge the hedges, but the death trade happens when its properly hedged lol... basically with Gerte when you run out of levels, and each side has the same amount of sells and buys is when your screwed hehe. Your loss will never increase but its a huge loss

Basically the worst case scenario can be calculated before starting up the grid.. Also if you reverse the grid, you are just adding to more possible losses, and not to mention the increased margin used.

I dont think there is an automatic way besides takign the loss (which Steve has programmed in already), unless you use your manual trading strategies to try to unbalance the sides and hope you can reduce the losses/or get out (aka martingaling and hoping you pick the right direction). But heck i doubt you can ever program something like into an ea which will do it automatically.

Going back to post 1, this ea can't be fully automated unless you take the hit when it happens.. its semi automatic as you need to save trades lol.

#Think about why the EA was made, and why we using Squalo's indicator... we are trading breakouts and we dont care which direction... problem happens when we have false breaks. If its a true breakout, we can lock in some profit and stop the death trade from ever happening.. its when we can't aka you enter a buy and it goes 1 pip before slamming through the sell level.. which goes for 20 pips, triggering the next level, then going all the way back past the first buy and even further hit the 1st buy level.. and back down to hit level 2 of sell, then back up.. as you add the lots and average down your positions as more levels are open the more you dig yourself into a hole

SWG123 wrote:I understand all that but what I meant to say (and I admit I didn't explain at all well ) when we get to level 10 long or short, then we are probably at an extreme high or low of an extremely wide range. This is where Nanningbob followers, for example, would be looking to apply countertrend trade strategies. I am suggesting likewise - does that make more sense?

SteveHopwood wrote:

SWG123 wrote:I understand all that but what I meant to say (and I admit I didn't explain at all well ) when we get to level 10 long or short, then we are probably at an extreme high or low of an extremely wide range. This is where Nanningbob followers, for example, would be looking to apply countertrend trade strategies. I am suggesting likewise - does that make more sense?

So guys, not one more word about the 'death trade'. Only an idiot imagines it can happen here.

SWG, would you care to expand on your idea?

acostafulano wrote:

SteveHopwood wrote:

SWG123 wrote:I understand all that but what I meant to say (and I admit I didn't explain at all well ) when we get to level 10 long or short, then we are probably at an extreme high or low of an extremely wide range. This is where Nanningbob followers, for example, would be looking to apply countertrend trade strategies. I am suggesting likewise - does that make more sense?

So guys, not one more word about the 'death trade'. Only an idiot imagines it can happen here.

SWG, would you care to expand on your idea?

I don't think it's a 'death trade' per-se. Eventually, it will get into profit because of the grid nature of the EA, but still, getting caught in an open grid of for example 9 buys and 9 sells (Right in the middle that DD can be nasty) and having to wait for the price to head a massive amount of pips in either direction to finally close that basket for a little profit (20, 30 pip or whatever it is) is something I would definitely try to avoid, simply by killing it early befor it gest nasty and starting fresh with a new breakout.

SteveHopwood wrote:
Plus, it has been mentioned occasionally in both threads, as I recall.

So, people, get a grip on the basics and stop wittering about them.

SteveHopwood wrote:

SWG123 wrote:I understand all that but what I meant to say (and I admit I didn't explain at all well ) when we get to level 10 long or short, then we are probably at an extreme high or low of an extremely wide range. This is where Nanningbob followers, for example, would be looking to apply countertrend trade strategies. I am suggesting likewise - does that make more sense?

It makes more sense that most of what I have read here tonight, with the mindlessly stupid wittering about the 'death trade' that shows conclusively that the witterer has no idea what this bot is about.

I would find if funny were I not returning from a day and a half away and having to catch up on a large number of posts and pm's. If only those here were of even a tenth of the quality of those in the Forex Robot thread. They are not. Most of them are rubbish.

So guys, not one more word about the 'death trade'. Only an idiot imagines it can happen here.

SWG, would you care to expand on your idea?