that happens when your databse about one pair is incomplete, or when it doesnt make trades but all “dates” are there its because the bot has a problem.
first create a file handle
“int FileHandle_A = FileOpen(name, Opentype, Delimiter);”
delimiter is used on csv files
opentype = FILE_READ | FILEWRITE | FILE_CSV - check the fucntions on the help for more info.
and file name = name.
While its open you use “FileWrite(FileHandle_A, Var1, var2, “aaaa”,…);” <-- in case of CSV type. btw CSV = comma separated values, thats why you define a delimiter wich is optional.
REMEMBER YOU HAVE TO CLOSE THE FILE EVERYTIME THE “FILEOPEN” RESULT IS != -1!!
“FileClose(FileHandle_A);”
If you want to trade EUR/JPY with my new rules be careful. It’s not a proven thing yet, and it’s still in the testing phase. 
Good job Gif. Thanks a lot for working on this and posting it!
I’m still debugging my version. Since you posted your’s I’m going to hold off on releasing mine until I’m 100% sure it works right.
Thank you, if you need help building it tell me.
Also there seams that i forgot to add something on ea post… The current rule for opening orders on EA is 2 pending at same time or no trades that weekend. I will edit the v2 one so it places order on run time.
2 pendings = more precision.
1 order/time on normal time run = less precision, but will be suited to more agressive times (and bends abit broker order placing rules!)
May i know what is the latest setting of this strategy??
Many Thanks
Thanks!
Back again, and it looks like there’s some fantastic progress on the EA. This will be very good for me, as I live within the GMT period, and to trade past the bar is almost impossible with school on Monday mornings.
How does the backtesting upon the EA match up to that of the Excel file you’ve provided? I’d be happy to run it for the next daybreak for this trade.
Dear Phil838,
I have reading the thread, and I guess I missed Phil’s EA version. Could you please post again for download and info needed. Thanks so much. I thought I could find it but I must be overlooking it.
Dane
The settings are still the same as explained in the first post. Some people have adopted a few minor changes to the system, but I’m still trading it the same way as when I started.
We are working on some new rules for expanding the system, but it’s still in the planning/testing phase… For now the original rules are still the way to go!
My EA isn’t designed to be completely automated, so there really aren’t any backtesting results. It’s designed to enter the trades for you on Monday, and to handle resetting the stoploss to breakeven, but you still have to make the decision whether or not to enter the second trade as the week progresses.
That being said, if you do run it 24/7 and never turn it off it still makes money in the backtesting. It’s just not as much as it should make if used properly… 
I’ve got it where it works 100% correctly on IBFX with GBP/USD, but I’m still working on getting it fixed so it will work on any timezome/decimal system/currency pair.
I thought I would be done this week but encountered another major bug when it made a completely wrong trade on Tuesday. I’m not a professional programmer so the progress is slow. Plus the fact that this system only does 1-2 trades per week means I have to wait a whole week to test any changes.
I haven’t put it up yet, Dane. I said early this week it would be up but I found a problem with it, so I’m running a little behind. I think I’ve fixed the problem but I want to wait and see if it trades properly next week before I give it to the public.
Only 2 trades per week? … does it net in a lot of pips per trade? …
Depends on what you mean by “a lot.” We really don’t think so much in pips with this system, since the stoploss varies from week-to-week the pips always have a different value.
R-multiples or rate of return is a much better way to gauge a system’s success, but if you really want to know pips then this system has made 1490 pips from the first of the year to Aug 10th (sorry, I’m a little behind on my recordkeeping).
That comes to an average of 42 pips per week.
If you want some meaningful numbers the system has had a 21R return for the year, meaning if you risked 1% of your account per trade your account would have grown 21%.
Hi Phil,
Very promising strategy. Started reading it yesterday, just finished going through the thread. What caught my eye was this chart you attached about weekly high/lows.
I think a slight modification might help with this strategy. If I read the chart correctly, it shows that the weekly high or low tends to occur mostly on Mondays and Fridays. From what I understand, this means that it is highly probable that the market, lets say spikes down on Monday, stalls, and then chugs up throughout the week to end on a Friday high. Or it spikes up on Monday, and ends Friday on a weekly low. I hope you get what I mean. This means that Monday is a false breakout and risky. I propose we set this system to get us in after this false spike. 2 possible alternatives:
- Set the orders just before Monday Europe session to counter the false moves in Japan session, once London opens, they usually set the tone for the week for a currency pair like GBPUSD. (8 am GMT)
- After Monday ends (0 GMT Tuesday).
Oh, and to be conservative, I suggest taking 2 4H candles to decide the top and bottom of the range. Its your choice though.
Obviously, I have no idea if it will work (no backtesting done). Calling all the backtesting experts to investigate this idea further.
Good suggestions, kc0793.
I’m not basing this off any backtesting either, so who know which of us is right, but I don’t think it would help the system much in the long run.
The reason is that the system already has a way of dealing with the false moves. If one occurs you’ll simply lose the first trade and then win the second trade.
While it is true that you lose some of your profit on the first trade, that loss would be about the same amount of pips you’d lose when the market does pick a direction immediately. You have to remember that getting in later in the week means some of the movement will have already happened, and I see no difference in losing 50 pips to a bad trade vs. losing them to due to the market moving 50 pips before entering.
But like I said, that’s just as much a guess/opinion as your idea, and only backtesting will tell us for sure. I’ve got too much work piling up with EAs, writing, EURJPY backtesting, and making videos to test it right now, but I’m putting it on my to-do list.
i am not sure this analysis is that useful. we are using ATR/2 which already is reasonably safe assumption that the price will hit either above or below by the end of the week.
This sytem is successful only if the price knows where to go (particularly on early monday morning) and doesn’t spend too much time kicking out the stops.
Must be one of the best things someone said in this thread…
in fact the day itself estabilishing a route is a key fact for this sytem sucess.
the sucess would be 100% if the direction would keep going after the break… logicaly its doesnt more then does, but atr fix it nice because in the end, the tp when hit, rewards more then all losses.
I’ve done some more work on my EUR/JPY backtesting experiment.
From Oct. 2004 - Dec. 2005 (IBFX didn’t start showing Sunday data until Oct '04) EUR/JPY made a whopping 9.7% profit.
Even though 9.7 percent in 1.25 years isn’t so good I consider it a great success for the system. The reason is that market conditions were awful for the pair during that time. Most of the trades had less than a 2R ratio, and many of them were closer to 1R. If I were to see these conditions today I would stop trading this system, but I think it’s a good thing that the backtesting still made a slight profit during these times. It makes me think that even if things go in the toilet we’ll probably breakeven or make a little profit during the few months it takes for us to figure out the market is changing.
I did not test 2006, because the market conditions were the same as 2004/2005. I’m guessing it would probably be close to breakeven during that year too, but since I would not trade EUR/JPY during those conditions I didn’t bother to test it.
In 2007 we started to see market conditions similar to today. During 2007 the pair made 34.8%
I haven’t tested 2008 yet.
I’m going to trade both GU and EJ today, risking 1% on each trade. I’ve also opened up a live account on IBFX and will start testing my EA on that. I’ll be testing it on the live account, but only trading .01 lots, just in case there’s a problem.
If everything works well I’ll post the EA up after this week’s trades are complete.
Hi all,
I just got interested in forex a couple months ago; I’ve taken the school here and have been trading cowabunga successfully on my demo account.
I’ve decided to start trading the Sunday Breakout as well starting today. Can someone post the entries/SL/TP so I can make sure I did it right?
Thanks alot. And thanks to you Phil, for creating this thread. I’ve read alot of your posts and appreciate your insight.