Hi Zepp,
Rev0-5 is working for me. Pls check parameters again one more time.
Dennis/Frawan -> Let me know if it is working or not for you.
I’m still using version 4.
You can allow discretionary primers at any lotsize if you add say “AUTO PRIMER” to the comments and then check for that. As long as manual primers doesn’t have that exact text in the comments - no problem.
Re
D
Thanks for the tip Dennis. I have never used the Comments section. Will try that out. Seems an easy logical solution.
Just one thing, Comments (and Magic Number) can only be set at order placement. They can’t be modified later on.
Re
D
Noted, cant use magicnumbers here as we will be opening so many trades.
im having some problems with my 3 year backtest, metatrader closes all trades after about half a year,
it seems that im stopped out, over a 6 days period i loose about 7000$ in equity… it happends when im moving from the primer range, and back up in the direction of the normal trades…
Crisscross, does it keep setting primers when going the wrong way? because that would make an insane drawdown when going back up
sgude0 >> you are right, its not escalating trade sizes, but that is the same for both regular trades, and primers, so i still thinks a backtest is interesting, and then maybe in time crisscross can add a check, so that if balance is increased with x% & equity is over 50% then increase both trade size and primer size to the “next level”
Hi everybody,
I am trying to follow you and the development of this EA.
One question regarding backtesting: are you using “every tick” model? And GBPUSD M1?
Thank you,
Cristian
i have just edited my last post… found out why it shut down…
im using “every tick” model And GBPUSD Daily
here is the test until the stopout
its 29mb, so can take some time to load 
For me it worked fine (until all positions are closed at the end of course). This one is no primers, GBPUSD1, every tick.
Re
D
Same, as before, but started in 2008.
I’m not sure why the tester stops after 6 and 9 months, but I’m pretty sure the “close at stop” code means that strategy tester ended, as there’s no reason the trades should been stopped out otherwise.
Re
D
yes, and WITH primers its working fine until the pair is moving back up if we are far from the normal channel borders…
That does not surprise me. I’ll admit I’m having trouble grasping the finer points of Greenlanding, but I’m pretty sure primers should not be run automatically. I think it might be better if the primers functionality was turned on/off manually using variables. Basically, if you need primers you get a warning automatically, but they only run when you manually ask for them.
“Give me short (or long) primers of size X while rate is between x.xxx and y.yyyy” => execute!
I use trendlines on candles with stochastics, MACD and RSI (don’t ask about success rate please), but I would like to have the trades auto-placed and updated once I’ve manually decided to go for it. Adding primers just because we’re outside of the channel seems too random to me.
Re
D
yes, but the auto primers work great aslong as the trend is not turning and going a long way in the opposite direction, as you can see here from 1 jan to now
And you are right, primers should normaly be tweaked to the specific situation, but if we could get an working auto primer, that would be perfekt for situations where your not able to look at the EA all the time as an example, a vacation without a lot of wifi in the hotel 
Could be a problem with SL not being set, but more likely short primers are being set all the way up to ChannelBottom regardless of how bullish a turn might be. With a narrow channel like 1.6 to 1.5 that could be 2700 pips as I seem to remember some time spent at 1.37. A primer every 30 pips against 2700 pips would be… expensive?
Re
D
that is exactly what i think is happening
by the way, im still having some problems with the REV 5 version and primers… i have set the primer size to 0.04 but there is no 0,04 size trades opened at any time…
Guys,
The EA should not set primers if price is retracing back up to ChannelBottom.
Lets take eg of GBPUSD:
Channelbottom: 1.5300
PrimerStep: 30pips
1st primer: 1.5270
2nd primer: 1.5240
3rd primer: 1.5210
.
.
.
.
It continues so forth.
But please note that when 2nd primer is opened EA is automatically locking in 5 pips on 1st primer. Thus at one time only 1 primer is running without a SL. Previously opened primers are either at lockinpips or else in a trailing stop routine.
If price starts retracing back from lets say 1.4900 to 1.5300 (and the last primer opened was at 1.4900) then there will be no more primers placed during the retracement. If price breaks above ChannelBottom and again comes down below it then the EA will start placing Primers again.
If you see the results page then during “Close at stop” you will not find more than 4 primers active at any time and 3 of them will definitely be positive.
One of the problems I have seen during my backtests is that primers usually hit their SL of +5 pips or else hit the Trailing SL. very few primers go to hit their full TP. Hence I would advise that for every pair there has to be a realistic TP level set for primers. Even setting a 30pip TP for primers at a bigger lot will yield lot of money to equity.
But we can find the best fit only by backtesting. Let me know if my logic is ok, but the EA performs otherwise.
Also Rev-5 does not set any selllimit/buylimit orders if price breaks below/above channel borders. This means that we are missing a lot of opportunities of price retracements and then continue along with trend.
Ranging pairs like AUD/USD have performed great. I will upload my test results later today.
that happends anyway… i can count 22 primers closed in loss in this test
https://dl.dropbox.com/u/93453374/Strategy%20Tester_%20Greenland%20strategy%20Rev%200-4.pdf
EDIT:
after looking at the order numbers, it looks more like the primers is just never closed, not even when in alot of profit.
because even though i have over 4000 orders total, i have primers with low order numbers like 73, 2510, 2507,2616, 2619 all closed when there is stopout… to take the most extreme one is the order 73, its opened at 1.57469 and the pair then drops to about 1.47 before it goes up again, so this primer should have been closed in profit
Hello all you wonderful people.
My name is Kent, and I have been a part of the Greenland/Hourglass system for quite a while now.
I have been helping Viking1961 with a lot of calculations during the time and we have made the system a bit more visual, so as many as possible would understand the basics.
It is wonderful to see that You are making an EA to trade the system. I have been looking around for a while to find something that could be made into a Greenland EA, and it seems that I have come to the right place here.
I have been back testing version 5. 3 years back now and it would have made above 100% income in that period. But that is with out any primers made.
It seems that the basics are right. But I might have some improvements ideas.
We have to keep in mind, that the system actual is not based on locked channel borders, but can be extended as needed. So it might be a good idea to make some calculations in % instead of pips. I’m not a shark in MQL5 so I can’t program my self. But I might be able to come with some inputs here so that we can make a nice working EA.
Here is my suggestions:
[B]Normal trades:[/B]
-
A normal trade shall count 0,5% - 1,5% of total capital (adjustable) Leverage x100
-
A new trade shall be opened in trend direction every 50 pips as standard (adjustable down to every 5 pips)
-
TP on normal trades shall be 50 pips as standard. (Adjustable)
-
When the rate has passed the opening point of with an (adjustable amount of pips, then a pending order on a normal trade against the trend is made. (This is to minimize the amount of red trades carried)
-
If a long or a short trade is open at a given rate (ex. 1.6000) then there will be no new trade on that rate, in that direction before the old trade is closed at TP
-
There is no Stop loss.
-
It should be possible to set a maximum size (units) for a normal trade (so when the income becomes high enough, it should flatten out the channel, to make a steady income, so that You can make withdrawals) (Primers shall overrule this function)
[B]Primers:[/B]
-
If the drawdown each pips raises to 0.06% of total Equity. Then primers are started
-
Primer size shall be more than total drawdown/3 , and less than total drawdown/2
-
Primer TP shall be 100 pips
-
When the rate drops or raise more 60 pips in green direction then SL for all primers should be kept 50 pips away from the opening point of the latest primer, so that you can only carry 2 primers at any time.
There might be more to it. But this is what I could think of right now.
I hope that You can use my input, to sort out some of the children illness that it might have at the moment.
I’ve set up a public profile on MT4pips so that we can all see the same results of demotrading. The data is available at mt4pips.com. At the moment I’m running version 4 with the following variables:
extern double Lots = 0.01;
extern double TP = 500;
extern bool UseSL = false;
extern double SL = 0;
extern bool UseBE = false;
extern bool UsePrimer = false;
extern double ChannelTop = 1.6000;
extern double ChannelBottom = 1.5000;
extern double PrimerLots = 0.01;
extern int PrimerStep = 300;
extern double PrimerTP = 1000;
extern double PrimerTSL = 250;
extern int Slip = 30;
Let me know if you have strong opinions about the setup.
Re
D
PS! Non-closed trades are shown under the Statement tab.
PPS! The opening balance is 3.000$ as 2.000 wasn’t an option at my broker.
Hi Zepp,
I did see your results. Many of the trades have closed on S/L. However the order 73 that you mention was abuy primer placed above 1.64 (@1.64495) but price never went forward much and after that dropped a long way. Hence the loss.
But there are other primers like order nos 3599, 3586 which should have had S/L’s but did not. I will check on that. I think it is because of the sell limits that are placed above the primers. They are confusing the EA. i will have to figure out exactly what is happening.
Can you test the same time range/same pair with Rev 0-5 and see if the same thing happens.