EMA Breakout Trading robot

[traveliron]

Hi Steven

I was intrigued to learn what you'd been up to whilst your Beast was taking a break, although I see on your MyFxBook that it has started again.

Could I just say this is probably the most logically layed out and succint presentation of a new system I have had the pleasure to read. Bound to fail! Seriously though it sounds as though you've had some success with this.

What you really want is a coder to respond and sadly I'm not one. But I do have space for a spare demo on the home PC so I'd like to test your idea manually.

Under your "Buy Trigger" paragraph you have two lines that are exactly the same. Should one be slightly different or was it just simply typed twice. I thought you might be after 3 options: just the 3EMA, just the 8EMA or, as in your current system, both the 3EMA and 8EMA. Actually having thought about it I don't think the 8 can "breakout" before the 3, so just the 8EMA might not be sensible.

I think you've covered all that would be useful in an EA and I can't think of much else to add.

It might be quite nice to have an adjustable trade trigger margin above and below the Asian range to experiment with. Maybe even allowing negative margins for within the range; you never know. We wouldn't have to use them but it would be nice to have the option for testing.

A few traders seem keen to use multiples of ATR in setting their TP and SL. I can see ATR being quite variable as volumes crank up at this hour of the trading day, so maybe not that useful.

Best of luck - Chris

[Bunker]

Hi Forexpippy,

Thank you for submitting this idea for possible automation. This is exactly what I have been doing for the last couple of weeks and I had great success with this system (demo). The only difference is that I am using squalou’s indi (sqDynamicBreakoutBox) to draw the box. What I like about that version is the possibility to add a “BoxBuffer” of x pips (I use 5). I avoided a few fake BOs because of that buffer. I also use the “Alert” function.

I wait for the 8ema to cross the buffer line before making a trade (I agree with traveliron, when the 8ema cross the line, the 3ema is already far ahead).

Sometime, I don’t even use the box (if there is a strong move of the 2 emas going north or south).

I know a bit of coding (being playing with Steve’s EA for over a year) but obviously not enough to even call myself a junior coder.

I will definitely keep an eye on that thread, so thanks again for initiating it

Bunker

[SteveHopwood]

Thanks for posting this forexpippy.

You have my attention, and are about to gain the attention of the rest of our community.

I added the indi to the attached chart - chart picked at random. The results appear to speak for themselves. The boxes are drawn from 4-7 am London time where I live, and stop at 23.00.

I am busy for a couple of days, so more next week. In the meantime, if anyone else wants to take over this project, then sing out.



Edit: Erwin, the inline drawing thingy is clearly working perfectly. Cheers.

[musky]

Still anything moving average based no matter how you spin it is still late to the party.

[SteveHopwood]

Still anything moving average based no matter how you spin it is still late to the party.

True, but by the same token, everyone is late to the party, apart from the hosts. It matters not a jot what you are trying to do, if you wait for some sort of signal, then you are late to the party.

You want the market to move 1 pip in the direction of the trade you want to take? Wait for that pip and you are already late for the party.

Here, you wait for the market to move out of the box and you are already late for the party. All you are picking up from successfully identifying the move are the scraps. Then there are the fake-outs that really bugger things up. Maybe an EA cross can help avoid them? Or Hanover? Both? A few of whip's ritually sacrificed chickens? You never know what might help. Actually, we do know; a bosum buddy in one of the Big Banks, but not many of us have those.

There are a variety of trade entries suggested in post 1 here. They will all result in the user being late to the party.

Every single robot I have ever coded has been late to the party.

Ok, so I exaggerate, but you see what I mean? Trading decisions have to be based on something. Base them on nothing and you are guessing. Base them on something and you are late to the party.

Unless you are Epsilon Megastar.

[SteveHopwood]

Just to let you all know that "forexpippy EA" is mentally under construction and should make its first appearance in time for market opening on Monday.

[gaheitman]

Hey Steve,

A quick question, since the BoxHigh is based on the Bid price, shouldn't the test for a buy use the Bid? The actual order entry would still use the current Ask.

Also, I think we will have to include the logic to pick the high and low for the boxes in the EA. As it stands now, we can't try to optimize the box parameters in strategy tester. I'll take a crack at it if no one else wants to try their hand.

George

I will pop this in here for now. I have to go out soon and do not have time to start a new topic - I will do this later. With your permission, I will move your first post here into a new thread as a sticky so that people can find it easily.

Coders, would you care to run your slide rule over the code? The bot takes its hi-lo from the box lines drawn on the screen by the indi. TrendDetectionModule() looks at all the choices made by the user and sets RsiTrend etc to the appropriate direction, then consults them as part of the trade decision process in LookForTradingOpportunities().

More later.

[SteveHopwood]

Hey Steve,

A quick question, since the BoxHigh is based on the Bid price, shouldn't the test for a buy use the Bid? The actual order entry would still use the current Ask.

Also, I think we will have to include the logic to pick the high and low for the boxes in the EA. As it stands now, we can't try to optimize the box parameters in strategy tester. I'll take a crack at it if no one else wants to try their hand.

George

I will pop this in here for now. I have to go out soon and do not have time to start a new topic - I will do this later. With your permission, I will move your first post here into a new thread as a sticky so that people can find it easily.

Coders, would you care to run your slide rule over the code? The bot takes its hi-lo from the box lines drawn on the screen by the indi. TrendDetectionModule() looks at all the choices made by the user and sets RsiTrend etc to the appropriate direction, then consults them as part of the trade decision process in LookForTradingOpportunities().

More later.

Please do George, that would be great. I have to go out soon and will not be back until much later on, so this would be a big help.

I have left the new thread up and running at http://www.stevehopwoodforex.com/phpBB3 ... 1023#p1023 and will send out a general 'new toy' notification later.

[SteveHopwood]

George, I forgot the first part of your question, sorry.

I am never sure whether to trigger a chart line-based buy on the Ask or the Bid. Behind the scenes, Khalid and I are having the same discussion about Luxor. Over there, we are trying the Ask and having that line showing on the chart as well.

All thoughts welcome, guys.

[gaheitman]

I trigger everything off of Bid since I can't see Ask on my charts other than in real-time. It's that simple for me.

Oddly enough, years ago before I was man enough to get up at 2am to trade the European session I wrote an EA (ok, it was more of an initial trade entry assistant ) that would place buy/sell orders based on hi/lo of the Asian session. I spent an hour or more one morning trying to figure out how my buy got triggered without price passing the upper bound of the box. My problem was actually due to higher spread at the time of entry than my bid+spread pending order expected, but it still illustrates the issue. I should have coded it for a market order when bid crosses the boundary like you've done. Well, like you'll change it to.

George

George, I forgot the first part of your question, sorry.

I am never sure whether to trigger a chart line-based buy on the Ask or the Bid. Behind the scenes, Khalid and I are having the same discussion about Luxor. Over there, we are trying the Ask and having that line showing on the chart as well.

All thoughts welcome, guys.