Simple H1 GBPUSD EA | Page 257 | Forex Factory

Options

Go to
Permalinks
Print Thread

Related Instruments

GBP/USD

Similar Threads

SIMPLE EA for a simple method needed!! 2 replies

Trading is this Simple? Finally Explained simple and clearly! 46 replies

Simple 3 order GBPUSD EA 1,047 replies

Experiment with "Simple H1 GBPUSD EA" 4 replies

Simple 5m GBPUSD trade system 10 replies

Attachments: Simple H1 GBPUSD EA
Exit Attachments

Simple H1 GBPUSD EA

  • #5,121
  • Jun 19, 2010 3:55pm Jun 19, 2010 3:55pm
Quoting fcbrants
Heya Geoff,

I've been trying to figure out what us mere mortals can do to help the cause.... not that the cause appears to Need any help - I've been doing back flips & dancing on the ceiling since the day I found this thread, but just in case...

I read the MT4 docs (MT Terminal -> Help -> Help Topics -> Auto Trading Section -> Expert Optimization) & started an optimization session, Only varying the Slow SMA from 30 to 297, Step 1.

It looks like this process will keep my 3 GHz dual-core machine running at about 50% cpu utilization...
I love it Franko.
  • #5,122
  • Jun 19, 2010 3:56pm Jun 19, 2010 3:56pm
Quoting fcbrants

<Snip>

It looks like this process will keep my 3 GHz dual-core machine running at about 50% cpu utilization for some 25 hours.

<Snip>
Hmmm - MT terminal started off reporting that it would take 25 hours, but it just finished the task & reported 268 / 1,280 (268) and 1:14:01 / 4:39:30 - not exactly sure what that means...

Franko
  • #5,123
  • Jun 19, 2010 4:04pm Jun 19, 2010 4:04pm
Quoting fcbrants
Heya Geoff,

I've been trying to figure out what us mere mortals can do to help the cause.... not that the cause appears to Need any help - I've been doing back flips & dancing on the ceiling since the day I found this thread, but just in case...

I read the MT4 docs (MT Terminal -> Help -> Help Topics -> Auto Trading Section -> Expert Optimization) & started an optimization session, Only varying the Slow SMA from 30 to 297, Step 1.

It looks like this process will keep my 3 GHz dual-core machine running at about 50% cpu utilization for...
Hi Franko
The problem is that I have run so many optimisation passes of my own, tried
every standard indicator and even tried turning my laptop on its head but I
just can't get even the slightest improvement over the defaults that we are
using at the moment!

My only conclusion is that with the current strategy we are right on the optimum settings.

So where does that leave us, do we look for a different strategy or do we just
accept that what we have is as good as it's going to get?

I have used dozens of different EA's, spent so much money on "Robots" its
embarassing but I have never had one even half as reliable as this one!

I guess all I can ask of the great members of this thread is that they run the
optimiser on different broker accounts, I have only used Alpari UK for my own
back tests so I only know how the EA performs on other brokers through posts
on this thread.

I'm not sure how we can collate results for other brokers, there is no way of
saving data on the forum but I am willing to receive information by PM and build a
website to publish the results if you think it would be useful.
  • #5,124
  • Jun 19, 2010 4:05pm Jun 19, 2010 4:05pm
Quoting sanyhuis
v2 does 2 trades per night. 11#3 does 3 trades per night. The 3rd trade only kicks in when the market goes against the first two trades.
Run the comparison between V11#2 and v2.
You're right. This occured to me just before reading your post. So repeating my previous test with 2 order presets on V11 took about 5 seconds again. 14 trades. P&L: $296. Max DD 1.99%. Still better results over the short test period.
  • #5,125
  • Jun 19, 2010 4:17pm Jun 19, 2010 4:17pm
Quoting geoffb65
Hi Franko
The problem is that I have run so many optimisation passes of my own, tried
every standard indicator and even tried turning my laptop on its head but I
just can't get even the slightest improvement over the defaults that we are
using at the moment!

My only conclusion is that with the current strategy we are right on the optimum settings.

So where does that leave us, do we look for a different strategy or do we just
accept that what we have is as good as it's going to get?

I have used dozens of different EA's, spent so much money on...
Geoff, you're right. What we have is the best. Nothing else comes close. Even your manual system beats anything else out there. So leave well enough alone and optimize when needed, get the bugs when they appear and get on with things.
You've done a lot of work and it is paying off. To the point where we have extra time on our hands and don't know what else to do. It's as good as it gets and I for one am extremely thankful for it all.
  • #5,126
  • Jun 19, 2010 4:18pm Jun 19, 2010 4:18pm
Talking about Improving the profit as I see it as the only thing that can be achived as the EA works fine as it is.

I have noted that while the TP 1 is hit regularly TP 2 often hits the trailing stop, therefore is there a case for TP 1,2 to be set at the same Target say 20 pips in which case more profit would be achived or just to set 1 TP with a higher lot size....
  • #5,127
  • Jun 19, 2010 4:21pm Jun 19, 2010 4:21pm
Quoting geoffb65



I have only used Alpari UK for my own back tests so I only know how the EA performs on other brokers through posts on this thread.
You never know, even a blind pig finds an acorn every now & then...

My optimization results of the slow_sma (I think from 30 to 297, step 1) on ATC Brokers looks promising... using 60 instead of 197 looks like an improvement, at least to this blind pig


Franko
Attached File(s)
File Type: zip 2010-06-19-Vary-SMA.zip   50 KB | 284 downloads
  • #5,128
  • Jun 19, 2010 4:36pm Jun 19, 2010 4:36pm
Quoting fcbrants
You never know, even a blind pig finds an acorn every now & then...

My optimization results of the slow_sma (I think from 30 to 297, step 1) on ATC Brokers looks promising... using 60 instead of 197 looks like an improvement, at least to this blind pig


Franko
That is exactly what I am talking about, ATC are ENC brokers and I have not
had much feedback on ECN before.

I suppose we need a standard setup for each optimisation so we have a baseline to work from.

Good stuff Franko!
  • #5,129
  • Jun 19, 2010 5:07pm Jun 19, 2010 5:07pm
Quoting geoffb65
Good stuff Franko!
Oink!

I started running the optimization because I wanted to find a way to keep the EA from trading a day like 3/22/2010 to 3/24/2010, when a hiccup fooled it into trading. Giving the slow_sma a finer grain looked like a good place to start.

If we take the Sunday (3/28 for the 60 sma, 3/28 & 6/14 for the 197 sma) trades out (ie, close our eyes & pretend they're not there, not programatically), then the results are still nearly perfect, but it looks like the EA takes more trades, 120 trades (60) Vs. 108 trades (197).

Franko
  • #5,130
  • Edited Jun 20, 2010 7:25am Jun 19, 2010 5:44pm | Edited Jun 20, 2010 7:25am
Quoting fcbrants
You never know, even a blind pig finds an acorn every now & then...

My optimization results of the slow_sma (I think from 30 to 297, step 1) on ATC Brokers looks promising... using 60 instead of 197 looks like an improvement, at least to this blind pig
Franko
Pay attention on this:

Bars in test 2145 Ticks "modelled"

6793882 "Modelling" quality: n/a

Mismatched charts errors : 1727

This test is invalid..

#1

Modeling quality has to be at 90%

And

#2

No Mismatched errors should be present in the Report

Update your M1 Records in "History Center" and try again..

And I don't think you are serious to show statistics with

Relative drawdown 60.60% (83924.05)

60% drawdown?

Russian Roulette? Win or die?

Normally it should be 3-5%

So everyone just disregard those BS stats and move alone..

P.S

"using 60 instead of 197 looks like an improvement"

You are welcome to use 60 EMA instead of 197, also use 30% risk instead of 3%

Everything is possible
Viewer Discretion Advised: Shall we shag now or should we shag later? :-)
  • #5,131
  • Jun 19, 2010 7:41pm Jun 19, 2010 7:41pm
Quoting SkyzerFX
Pay attention on this:

Bars in test 2145 Ticks "modelled"

6793882 "Modelling" quality: n/a

Mismatched charts errors : 1727

This test is invalid..
Thanks Sky - I get the same # of chart errors on all of ATC's backtests, but only 53 errors on FXDD - what causes that? (Off to Google...)

The 45 slo_sma on FXDD yielded $386k profit (Yikes!!).

Would you mind taking a look at my FXDD results to see if you find any problems?

Thanks!!

Franko (<--- Training Wheels still firmly attached...)
Attached File(s)
File Type: zip 2010-06-19-Vary-SMA.zip   153 KB | 278 downloads
  • #5,132
  • Jun 19, 2010 8:54pm Jun 19, 2010 8:54pm
And if you are going to run optimization test on the week end what is the spread that you are getting during the week end test..
It could be very high and would probably invalidate the test as well.
  • #5,133
  • Jun 19, 2010 8:58pm Jun 19, 2010 8:58pm
Quoting sanyhuis
Geoff, you're right. What we have is the best. Nothing else comes close. Even your manual system beats anything else out there. So leave well enough alone and optimize when needed, get the bugs when they appear and get on with things.
You've done a lot of work and it is paying off. To the point where we have extra time on our hands and don't know what else to do. It's as good as it gets and I for one am extremely thankful for it all.
Thank you very much. I couldn't agree with you any more. Anything else IMHO is off topic.
If you're scared get a dog.
  • #5,134
  • Jun 20, 2010 2:42am Jun 20, 2010 2:42am
Hi

I just want to confirm a point from the discussion about Sunday trading earlier. Am I correct to think that it's not advised to start the EA until Monday evening? My broker (Alpari UK) opens the feed at 2200 GMT tonight (GMT offset 2).

I didn't see any problems with losses on Sundays in backtesting, but of course can't go back and check as the spread at the Friday close was close to 30 pips!

Cheers
Charlie
  • #5,135
  • Jun 20, 2010 4:55am Jun 20, 2010 4:55am
Quoting FokkerCharli
Hi

I just want to confirm a point from the discussion about Sunday trading earlier. Am I correct to think that it's not advised to start the EA until Monday evening? My broker (Alpari UK) opens the feed at 2200 GMT tonight (GMT offset 2).

I didn't see any problems with losses on Sundays in backtesting, but of course can't go back and check as the spread at the Friday close was close to 30 pips!

Cheers
Charlie
I use Alpari UK and I will be trading tonight!
  • #5,136
  • Jun 20, 2010 4:55am Jun 20, 2010 4:55am
FWIW:

I'm with Alapri UK and since starting live trading with V5a at the end of April it did open a trade Sunday night/Monday Morning which ended up + pips

Have steadily upgraded as the new versions were made available and to date had no more Sunday night/Monday Morning trades
  • #5,137
  • Jun 20, 2010 6:41am Jun 20, 2010 6:41am
what exactly do the different presets (1 2 and 3) do?
thanks
  • #5,138
  • Jun 20, 2010 7:17am Jun 20, 2010 7:17am
Quoting sheppy
what exactly do the different presets (1 2 and 3) do?
thanks
As with 99.9% of questions asked here, the answers are in post #1: http://www.forexfactory.com/showthread.php?t=231650
  • #5,139
  • Edited 7:32am Jun 20, 2010 7:20am | Edited 7:32am
Quoting geoffb65
Simple H1 GBPUSD EA
running live Account for this EA now.. for back testing get amazing result... 100% (01/06/2010 - 18/06/2010), 1 fail -420 (27/03/2010-18/06/2010)

Thanks, Geoffb65.
  • #5,140
  • Edited 8:11am Jun 20, 2010 7:40am | Edited 8:11am
Quoting fcbrants

Would you mind taking a look at my FXDD results to see if you find any problems?

Thanks!!

Franko (<--- Training Wheels still firmly attached...)
You are welcome..

Ok, I saw those tests on FXDD

"Modelling quality 52.56% " That's too low, should be 90%

To run quality tests in MT4 go to "Tools"===>"History Center" find

GBPUSD double click on 1 min (M1)

and then click DOWNLOAD.. Cause you have not enough records in

your Database on either brokers..

When you running those test All you have to compare is Number of

LOSS TRADES, that's all it matter.. We want to minimize those loses.

People will get confused looking at those 400k win number in your test

get notice that Start Amount in those test $100 000

Maybe it's ok to start with few hundred bucks if blows,

then you lose just those few hundred bucks

Anyways I have more loss trades and fewer Total trades using 51 SMA instead of 197
Viewer Discretion Advised: Shall we shag now or should we shag later? :-)
1 trader viewing now