I want to ask for assistance to see if something can be done to this EA to make it worthy of actually trading and profitable.

I have made several versions of it thus far with slight variations in each on the basic principle which is a macd crossover.

I see amazing things on small timeframes with lousy modeling quality. I wish someone would explain to me why these never translate into actual performance in either forward test or live account trading.

Even so, there may still be something of merit here if it can be filtered or something appropriate to a pair and timeframe.

Attached Files

	mcarty.mq4 (10.6 KB, 66 views)
	mcarty2.mq4 (11.6 KB, 51 views)
	mcarty3.mq4 (11.9 KB, 67 views)
	mcarty3a.mq4 (11.9 KB, 78 views)

Most promising results in 99% backtest are with EURUSD and USDJPY using mcarty3 and mcarty3a.

Attaching 2 best.
Hoping to do some optimizing, I agree that there is some promise here.

Attached Images

	3EURUSD1h(-208).gif (9.2 KB, 186 views)
	3aUSDJPY1h(-427).gif (9.6 KB, 190 views)

Attached Files

	3EURUSD1h(-208).htm (239.6 KB, 53 views)
	3aUSDJPY1h(-427).htm (210.9 KB, 17 views)

have you tried shorter timeframes like 1m and 5m?

do you have any suggestions for a filter parameter?

I'm on. I'm MACD worshiper! What do you want me to do? Just run forward test? Any specification? Which part need to be more focus? on tweaking settings? Entry signals? Exit signals? Reverse condition? Swing trade condition? Secure profit condition? Drawback Ratio? Risk management? Money management? Sharp ratio? I strongly belive that bunch of people running with a same settings on a same EA will end up reduncdunt. Anyway, I'm awaiting for instruction to start.

sorry for the slow reply, welcome aboard. Doesn't sound like you need directions. Sounds like you could give them. My as yet unanswered question is regarding the performance of this signal on the smaller timeframes like 5m and 1m.

I have backtests showing phenomenal performance for 6.5 years right up until about Oct. 2006 then it turns and loses as phenomenally as it was winning before. I don't understand why. I have no explanation besides some sort of data issue?

secondly I need help with filtering suggestions on ANY timeframe to make a version which actually makes more than it loses in forward tests.

Attached Images

	mccarty EURUSD 5m at .8.gif (5.3 KB, 169 views)
	mcarty3_5m.gif (5.3 KB, 177 views)

Attached Files

	mccarty EURUSD 5m at .8.zip (1.14 MB, 35 views)
	mcarty3_5m.zip (161.9 KB, 35 views)

Reading your backtest, might need some time to digest before i start testing them. I dont wanna waste time on hopeless system. I sincerly hope that I can get some good point in your EA. Anyway, thanks for sharing the effort.

all I ask is that you have a look, see what you can make of it, before you decide it's hopeless. Whatever you decide as long as you can tell me your reasoning it's not a waste of time. Nothing will defeat us like contempt prior to investigation. I sincerely hope you can get a good point in it too. Keep me posted.

when i work for my EA, the same thing happaned like you .
the exactly time turn result is 2006-9-12
i think that market changed since that day. change more trend in little frame that make our EA loss atter 2006-9-12