Interesting GBP/USD MA Crossover system

the idea right now is optimize every week over the past 4 months...but I'm going to try to make it optimize itself continuously. Or not so much optimize, but find correlations, such as: take the average difference between open and close over the last x periods, multiply that by y and you get z, and z becomes the period for one of the ma lines.

And have formulas like that for all the parameters. And it will probably be much more complicated by that.

I like the idea of the spike detector

i'll give a try on a test account

Thanks for sharing aicccia!
I'm testing your system now. For the moment I have still no trades (since 1st dec. and I'm in january), we will see.
Do you already try to backtest it since 1st dec? If yes when do you have trades?

PS: Why this thread is not in "trading system" instead of "Forex discussion" ?

Since 1st of december until now I had only one trade and in the journal I see that:

2007.02.17 11:05:20 2006.12.11 20:00 MoneyMaker01 GBPUSD,H1: close #5 sell 1.00 GBPUSD at 1.9533 at price 1.9574
2007.02.17 11:04:59 2006.12.11 00:00 MoneyMaker01 GBPUSD,H1: zero divide
2007.02.17 11:04:12 2006.12.07 00:00 MoneyMaker01 GBPUSD,H1: zero divide
2007.02.17 11:03:40 2006.12.05 17:00 MoneyMaker01 GBPUSD,H1: open #1 sell 1.00 GBPUSD at 1.9709 ok



The problem comes may be from the zero divide?

Any coments?

The EA doesn't seem to reopen a trade after the first one (in backtest). So each time a trade is closed I stoped the EA and restart it from the previous closing date. Results are good for the moment:
3 winning trades with an average result of 150 pips.

Could you share how do you determine all the entry parameters so we could test it on other currency pairs? (I would like to test it on GBP/JPY)

Many thanks!

could you tell me why it didn't open a trade there (see attach.)?

because of the spike settings.

but I've adjusted some other things so the next version would have taken that trade.

set your beginning balance to less then 500. I set mine to 200.

This is the next version. I've added:

1)An equity checker that will close the trade if you lose a certain percentage of the max equity. For example if things go well and you're up 200 pips, and then something happens and it drops to 130 pips and stays there for a few candles, and if you had your max equity to lose set at 75%, then the trade would be close. Keep in mind it measures all of the pip movements off a 3 EMA, so to keep spikes from messing this up.
2)Now the EA looks back up to 6 candles to see if there was a jump across the two MA lines.

thats it for now, but there might be a new one coming out tomorrow night.

keep in mind this EA does not get every move, it does not even get every big move, but when it does make a trade, there are several conditions saying "yes, yes, open now" so it's very unsual for things to go wrong in a big way. It is sensitive to ranging markets and it's possible to go 4-5 days without a trade if ranging is what it is detecting. The bottom line is the EA opens trades only in perfect conditions.

The EA was developed with the GBP/USD in min so I haven't tried it at all on other charts but if you want to optimize it for other currencies be my guest.
Make sure you check "genetic algorithm", have 90% quality data, and test at least 3-4 months. Optimize the parameters in two sets. The first set being: start stop
fastma 6 92
slowma 6 144
currency_momentum 6 92
trend_momentum 6 144
maxpipsfromEMA 1 50
trending 1 50
SpikeHeight 25 150

and the second set:

buy_maxhoursNeg 3 60
buy_maxpipsNeg 20 100
sell_maxhoursNeg 3 60
sell_maxpipsNeg 20 100
startcountpips 30 150
startcounthours 3 48
maxEquitytoLose .2 .8

when optimizing the first set, disable the stop settings and save equity settings but setting their values to false.

For the pairs that move very fast such as GBP/JPY, you might want to increase the numbers to stop optimizing to.

Close Trade #2
Febuary 15th, 11:00 AM
@ 1.9540
P/L= +28 pips

I'm just going to post the trades once a night instead of trying to post them right as they happen since they often happen very early in the morning or late at night. That last trade fell pretty abruptly, I saw it coming just from the trendlines but from the systems point of view, it wasn't close to be closed. So with the first version, I got just 28 pips out of it. But with the second version's equity saver, I would have gotten out at 1.9575 and earned 35 more pips, for a total of 63 pips. Though that was only about half of the wave, it was still a nice easy 63 pips in 24 hours.

Hi Aicccia, thanks for very promising EA. It`s really interesting system. Here are my results for 14.02 with 0.2 lots on backtest (MM2). And with 0.1 from jan. 2007.

idk...I don't like it. There should be three trades. I'll have to add some more things.

I have been forward testing the first EA posted with the input recommended, its not so good - keeps losing. It seems to open one trade after a losing trade gets stopped out.

make sure the equity save is set to false. I don't have the best values for it yet.

thx a lot for all your responses. I'll try this as soon as I can

hello,

i've downloaded the EA, and tried to backtest it for Jan 07 on GBP/USD 1 HR (and I have not changed any settings) but there are no trades ??

Could you please advise ?

Anthony

Hi Body

Please, why this is happened?

2007.02.19 08:45:59 2004.07.01 15:28 MoneyMaker_02 GBPUSD,H1: invalid price 1.81654997 for OrderSend function

I'm use MoneyMaker_02 with all configurations.

Thanks.

Try with risk percent at 0.001 (I think it's because the EA is developped for mini account)

I have still the problem that the EA doesn't want to reopen a trade after a losing one. On the picture I think that I would have a Sell after the first one (losing). Or maybe I missed something in the rules?

We are go for breakout!

Open Trade #3
Febuary 19th, 1:00 PM ET
BUY @1.9527