I just went to Viking’s eToro account to ask if a senior greenlander could overlook our discussion, and he had already mentioned on his wall that he will try to stop by later this week and fill us up with some more understanding. Yipii!!!
Re
Dennis
HI everyone. I would most certainly not see myself as a “Senior Greenlander”, but I do trade the Greenland myself and I’ve also written the FAQ for the strategy. It is still a work in process with Michael’s help of course. I’ve not read through the entire thread yet and would like to do so first before I give or make comments.
Thanks to everyone who’s been part of this so far, it looks really interesting and I’m sure it would be great if we can get this to work
hi there, just found this threat, and this look very interesting…
i just need to add that the primer explanation that you found of mine, is taken a little out of context…
i was explaining the basic about the primers to someone… that you use them to turn the drawdown into profit, and i then used the simpel explanaition to make it understandable… webzone told me that he would use 4-5 primers, but i think more small primers is a better choise, so back to this EA… amazing work…
is there a reason for only having 300 pips from top to bottom?
os that for slow primer implementing?
If you mean the ChannelTop and ChannelBottom variables, that’s probably just to provoke the primer functionality for testing purposes.
Re
Dennis
exactly, the ChannelTop and ChannelBottom variables… and yes… DOOOH… that would make alot of sense 
Hi Dennis,
How goes the backtesting?
I am waiting for Viking’s advice this weekend, then wiill see if anything else to amend in the EA. Else I plan to start with AUDUSD from next week.
What is your opinion?
I had some weird results where too many trades where clumped in a small area, basically a resetting of the algoritm. I haven’t been able to reproduce the effect manually, but I assume it happened as I was testing out different ranges, turning primers on/off, closing the chart etc. I’ve closed all trades in the account and started up fresh to see if it happens again. Will try to take notes of changes I do to variables (if I make any).
One thing I have noticed is that the range between opening new trades and deleting “far away” ones creates situations where a trade can be set, then cancelled repeatedly. Not sure if I could get into trouble with the broker if that continues, but maybe there should be a bigger distance between when to place new orders vs. when to delete them.
Re
D
i also see this in my demo account, i can count this happening up to 16 times in 1 min in my history
Guys,
I have noticed the same. Will try to identify and fix it soon.
damm, backtesting is sloooooooow…
is there any way to allocate more CPU and RAM to MT4 to speed up an backtest?
its not even close to maxing out my workstations resources :33:
Hi guys, thank you for all the attention and your efforts. I am more into creating the trading systems primarily. I am not an expert in EA´s . So this is why Zepp is now backtesting what you have done. I have seen many systems which enables automatic trading, however there were always flaws. Now Zepp is as I said the systems expert and he is hard to impress, but with the feed-back I have got from him so far, it´s the first time I ever hear him say that he is impressed. So looks like you have got something really good going here guys and that you are really excellent of what you are doing. When Zepp is finished testing it, using several currency crosses while doing so, we will discuss it further. But so far so good , so keep it up guys, you are doingbrilliantly so far. In regards to primers, it´s very sensitive , but the first small primer should be used when you have 40% of your equity left, which will give you plenty of time to change the equity and also evaluate the distance between the primers etc. But that is not the biggest issue as you can always take over manually in case you need to adjust the primers.
let me just add that i am no EA expert either! 
i have had programming back in school, but not for MT4… hahaha
but yes, im impressed, and yes, im trowing different scenarios at this EA
Guys,
for the above problem:
In the DeleteFarAwayPending order function, change the digit 4 to 5 in the code line:
" if((OrderOpenPrice()-Ask)>(5TPPoint))"
for all the cases.
This problem should go away and backtesting should become faster too.
One more thing that I was thinking:
If price goes below Channebottom (for eg lets take GBPUSD).
GBPUSD @ 1.5300.
Primers are placed automatically by EA. As price goes lower buylimits are placed (lets say a Buylimit is placed at 1.5400).
This Buylimit will cause a selllimit to be placed at 1.5450.
But the initial strategy definition said that if price crossess bottom or top there should only be opposite orders placed until prices comes back within your channel.
So my question: should the EA place selllimit orders when Price is below ChannelBottom or vice versa when it crosses channeltop.
Think about it. I have prepared an EA for it. I attach herewith. this EA will not place selllimit orders if price is below Channelbottom and no Buylimit if price is above Channeltop.
I guess we can only know which is the best solution after baktesting through various situations.
@ZeppDK - list down the conditions that you want to test the EA on. We can split various currency pairs among each other and test them.
I have few conditions on my list:
- Extended one sided price action
- Lotsize variation.
- Primerlot size.
- Drawdowns.
Pls share other conditions that the EA should satisfactorily clear.
Cheers
Crisscross
Greenland strategy REV0-5.zip (2.41 KB)
PS: Rev 0-4 & Rev 0-5 are both working versions. Rev0-4 places sell/buylimit orders if price crosses below/above Channeltop/bottom.
While Rev0-5 does not.
Hmm, if you don’t place shorts under ChannelBottom you will basically revert to an elevator strategy in that range, which I don’t think is a problem if you set up a wide channel. However, as primers are now triggered by the borders, those two aspects might step on each other.
It’s my impression that having stragglers in either end of the scale is not a problem as the lots will grow with time and outstanding trades from long ago will hence be pretty irrelevant for the equity level as a whole.
Re
D
hey Crisscross,
you where right about the speed of rev. 4 in the backtest, as soon as i changed the 4 to 5, it was done in seconds instead of days
rev 5 does not seem to execute primers at all, as soon as we get outside the channel range there is no new trades executed…
so hard to tel wich primer setup works best
I was thinking about the same conditions for test as you, and especialy gbp/usd from january until now, and the usd/jpy pair…
because the gbp is sort of a normal primer situation, and the jpy would be the extreme primer case with big retractions…
and then ofcourse a 2-3 year test on both gbp/usd and eur/usd to see if it works on auto for such a long run with both normal and primer trading…
Could you post some charts / results of the backtest results when/if you have any? Would be interested to look at it, but not able to get this to work myself.
For me the primary purpose of the EA is to test that the system works; to actually trade it could be done more manually once results prove that the strategy can survive the markets in longer periods (which is really a key requirement given that the 500/750 system could take 2-3 years to complete).
I was thinking of one more thing. I will make the EA to check for open orders only if the Lotsize is equal to the variables [B]“Lots” [/B]or [B]“PrimerLots”[/B].
This will help us enter manual positions (discretionary) if we feel so.
Re: long time testing: Guys I will need help from both of you to give me the last 5 years worth data in CSV format. You can upload for prominent pairs like EUR/USD, GBPUSD, AUD/USD, NZDUSD, USD/CAD, EUR/JPY, GBP/JPY & EUR/GBP. If you have these then you can zip it in a file and post it here. I have tried to download but it takes a long time and I have failed many times to get it done. Your help will be most appreciated. To let you both know I am also keen on GOLD & SILVER.
Let me know your opinions.
i will post results as soon as i have collected some to compare, right now, im testing rev. 4 on gbp/usd from jan 2010 until now… that takes some time
Can the system really be tested against historical data when the EA doesn’t have automatic lot increases based on the account balance? The exponential growth factor wouldn’t be included as of right now from what I can see.
I really like the option for discretionary trading. I would also find it helpful if primers had some text in the comments. This would make them stand out, and they can be sorted in MT (which I think could come in handy from time to time).
Did you try to open a demo account at Alpari UK for the historical data? Four years of GBPUSD1 is over 220MB and that’s not something BabyPips appriciated in an attachement
Re
D
I will upload the next revision in some time which will make it possible to add manual primers. However you will have to keep the volume different from the value of variables “Lots” & “PrimerLots”.
No, didnt do that. Was too lazy. Will do it now :8: