trailing 23 EA open source

Hi everyone at Forex Factory. This was my old tread- http://www.forexfactory.com/showthre...t=74959&page=3

Anyway, I needed an EA so I hired a programmer. Cost me some cash but in the end it is worth it. In the positive nature of Forex Factory, I will open source my EA to anyone that wants it under a few conditions though.

1. No one is allowed to take it and sell it on ebay or wherever. Part of open source means that it stays open source.
2. I need testers. So if you decide to test, we need to create a way so it doesn't get too messy.
3. If you are a programmer and want to contribute first discuss with me what you want to change. Then, when you upload the modified EA, it should have the version number and a list of upgrades that were done to it.
4. If you just want to stay around and learn from this tread, than sweet.

In the up coming posts, I will open source and discuss what this EA does and some bugs that I have come up.

I'm looking forward to having a great time and building an awesome tread with your help!

Keep in touch

To do list: If you can help, private message me or just reply to the post
6/18 change arrays to iHigh and iLow so EA can be set on any chart and still use information from daily.

Here we go everyone- looks like this tread got moved to another forum. Oh well, hope everyone can find it.

This EA will do the following.

It will calculate the daily range from yesterday and use a 23% value of that range for the trailing stop and stop. Now, in order for it to place an order, the high or low from yesterday has to be crossed.

The formula should look like this. For the Buy= yesterday's high +2+ spread
and the Sell= yesterday's low-2

Now to for the trailing stop Buy= 23% of yesterday's range and the trailing for the trailing stop Sell= 23% + spread.

So what happens if yesterday's range is really low than 23 is the minimum value. This EA I have only does a straight 23 not a 23%. I consider this a bug.

This EA will only make 3 trades per currency pair per day. So it filters out any whipsaw action.

This EA will also not make a trade if the spread is over a certain amount. We can tailor this to each pair we use.

To use the EA, it needs to be set on a Daily chart.

Bugs:
My experience so far shows that this EA is good some days and bad others. There will often be a few large wins. But there will be often many stop outs at 23 pips. Another bug is that the spread should be included in the stop. Remember, to the EA and metatrader, the stop and trailing stop are 2 different set points.

So for example, the spread for GbpJpy is 9. So at 23- 9, is only 12 pips breathing room. The stop should be in this case 32 at the minimum. Also if your computer shuts down for whatever reason or Metatrader restarts, than counter restarts also.

Another problem is that it trades on the daily. This is a problem because you can not see exactly where the trade took place.

So basically thats it everyone. Lets improve this and get a tight family to maintain this EA. I see a huge potential for this EA. It needs some tweaking and some filters applied, but compared to some EA's I have seen, its doing pretty well. If anyone wants to dissect the code and confirm what I said or add anything else, please contribute.

-Mike

5 wins 4 losses on 8 currency pairs for a total of aprox. +100 pips. Last one hasn't closed yet.

I'm surprised nobody has any input so far. Later guys-

Wait- last position closed. 120 pips profit for Monday June 9th. This is on the IBFX demo account.

I've included a picture of my results. I need your help to make this a better and more reliable EA. I mentioned in the past post that I want a clean approach when posting results. Sorry for the earlier posts.

HAHA Yah, ok. Try telling that to the guys on ebay, and the fools that buy that stuff.

I wish you luck!

I guess you are right. The junk they sell on ebay is crazy. I put the open source comment to show my intentions of keeping it open source while the community contributes. Too bad your the first to contribute-

Just remember that neither you or me are fools that buy that stuff on ebay. Thanks for the luck-

perhaps you can run it through the backtester for us?

Sweet, just closed another win for 15 pips. No can do about back testing. I'm running Metatrader on Linux through Wine and one of the bugs is the backtesting portion doesn't really work. On the other hand, I've heard that the backtesting on mt4 is really bad so its not a huge loss for the Linux crew.

But if you would like to set up a demo account and run the EA in the background, even if its one currency, that would be awesome.

I think the bottom line is that we need to identify some trends. Remember, every win is actually a loss of 23 pip potential pips because of the trailing stop loss. I think if we can come up with some sound trade management, like for example at certain pip values the EA closes half of the position or something like. I still have to think about it but before any changes are made I think we should clean up the original code.

Have any programmers out there checked the integrity and flow of the code?

June 9th; +132 pips, 6 wins, 4 losses

This is a reply concerning open source and making profit from this EA. This is open source meaning anyone can look at the code. I'm not sure how to 'officially' open source it, but I believe I have to GPL it. I guess this is my official notice to the public that this will be open source now and in the future. I know that some people have put up their EA's and had the forex factory community contribute. At one point, the guy took the code, closed it and now sells it for $30/ month subscription. I will never do anything like that. I feel that if we make a stronger EA together that will make money, no need to sell it for money because we'll be making money.

I hope this clears up any misconceptions about my intent to distribute the code.

Hi derailed, thanks for sharing your system and the EA. I loaded into mt4 and attached to a GU chart but it says it only works on H1 time frame. I don't see what setting to change to make it work for the daily chart like your rules say.

Thanks again
dave

edit. OK now I see that it is a cross of the daily hi or low. All I had to do was read it!

You have to put it on the daily chart. How it works is that it will look at the last bar to make the calculations and rules. So if you put it on a 15 min chart, it will put in orders if price actions breaks through last 15 min high/ low. Same goes for the weekly or monthly.

I find this to be a bug. I want the EA to take its calculations from the daily but work on any chart. If we can fix this, we can then identify when the trades were placed and what was actually happening. So maybe all good trades happen in the morning and all the losses happen at night. Or the actual 24 hour period is better using a 4am to 4am instead of 12-12. Once we get enough people viewing this tread, I will update the very first post on what needs to be done for a final release of version 1.0.

I feel that if we master the basics and nail a fairly consistent EA, than we can play around with a statistics counter, nice colors, have it automatically calculate standard deviations, work on better position management and so on.

June 10th is almost over for me and today was another positive day with the trailing 23. +60 pips, 8 wins, 3 losses

Ignore the -2 pips loss on the EU, I was confirming if I can run on a smaller chart time frame. And it looks like you need to run this one on the daily. Can anyone help me modify the code so we can keep it on a 15 min chart but it still uses the daily chart for the calculations? This will help in finding micro trends and actually see what is going on when the trades are put in and closed.

June 11th was a bad day for this EA. Not much action.

June 11th: -83 pips, 2 wins, 4 losses

I don't know what went wrong since the charts are on the daily. If anyone can help me change the EA so I can set it on a 15 minute chart but it will still use daily data to make the trades, I could analyze where the trades go well and not so well. I'm impressed that the EA has been downloaded about 70 times but not so impressed since no one commented or posted on the tread.

You'll need to change the high[] and low[] arrays to iHigh and iLow functions.

The EA appears to be a bar breakout trader w/ a 23 fib trailing stop and a range filter. I've tested approaches like this before, they work very well on trends, not so well on ranges. If I was to improve it I'd look to see if we were also outside of a X period high/low. Like run a 10 day high and low and see if it's also above/below that too. In a way it's similar to the turtle strategy, they used the same entry approach on a 20 day high/low.

You might also want a fade catcher too. If stopped out, put a reversal entry point somewhere like mid-bar (right after the pivot or something). Anyway, just stuff to test. Had similar approaches work great one minute, then fail miserably. If you run a backtest be sure you run it over at least 5 years worth of data on at least a half-dozen pairs, you'll see how it behaves on whippy pairs.

Hi Rabid, thanks for the advice. I think if I'll have these running on 15 min charts I can dissect the trades more closely. Now, can anyone help me change the arrays to ihigh and ilow functions? I tried writing code for a simple pivot point plotter and couldn't do it. Yea, I suck at code writing.

June 12-17: 15 wins 24 losses with -260 pips
The market sucked. As you can see, all the loses are mostly on the 16th and 17th. Rabid's idea of having a stopped position reversing might be good but I can't really tell until I have the trades on 15 min. charts. I can't see what the market actually did when the position was stopped out.

Can anyone help change this EA so it uses ifunctions? I would really like to start watching the market to see if reversing stopped trades will work.

Hi derailed,

I really like the EA and looking forward to the results.

For your problem with the timefram in the EA change:
this two lines.

I just started to program in MQL
but I think you have to change the following lines
(Line 77+78) to
double High1 = iHigh(NULL, PERIOD_D1, Current + 1);
double Low1 = iLow(NULL, PERIOD_D1, Current + 1);

and all the High[1] to High1 and Low[1] to Low1
(Line 191 and 192)

that should change to have the EA calculate on daily even when placed on 15min chart.

Maybe someone could verify?

LiF

Yes this is correct, because High[] is a predefined array that only applies to the time frame on the chart it is attached to. So by using iHigh() you are looking at the array high[] inregards to the period or time frame D1. And the 1 in the shift of iHigh(Symbol(), 1440, 1) looks at the bar that just preceeded the current bar of D1 periods. So, iHigh(Symbol(),1440,0) is the current high for the day and iHigh(Symbol(),1440, 1) is the high for yesterday.

And the same logic applies to iLow(), iOpen(), iClose().

thanks smjones for confirming.

it shows i'm on the right way.

hope it helps derailed with the EA.

LiF

Lostinforex and Smjones- thanks for the advice. I'll try playing around. Of course, please feel free to mess around with the code as well