Now it works!
Thanks a lot...
If I can I would like to help you for improve it...
Between us all we must be successful and then we can all make profits!
- #121
- Apr 17, 2010 11:17am Apr 17, 2010 11:17am
- Joined Apr 2009 | Status: Trader | 2,035 Posts
Between us all we must be successful and then we can all make profits!
- #122
- Apr 17, 2010 11:48am Apr 17, 2010 11:48am
All help will be gratefully accepted!
Between us all we must be successful and then we can all make profits!
It looks really good. Probably I have some idea.
But before I would know why you decided stop loss at 150 and take profit 18 ??
Thanks
- #124
- Apr 17, 2010 12:05pm Apr 17, 2010 12:05pm
- Joined Apr 2009 | Status: Trader | 2,035 Posts
Ok, I already done some test.
It looks really good. Probably I have some idea.
But before I would know why you decided stop loss at 150 and take profit 18 ??
Thanks
Admitted, I used the optimiser on Control points only, because it takes too long to run it on ever tick.
- #125
- Apr 17, 2010 12:12pm Apr 17, 2010 12:12pm
- Joined Apr 2009 | Status: Trader | 2,035 Posts
One problem with backtesting an EA like this will be, there will be times included in the test that we would not normally be trading ie. Bank Holidays. This will greatly effect any test that is done, unless one is able to eliminate those days, (difficult to do running Strategy Tester), so the DD could actually be less than tested due to those periods. Just a consideration for everyone to examine the DD on their test and note when it occurs. Go back through the economic calendar and see if you would have been trading this during that time.
One...
With my other EA's I tend to not trade on Friday after NY open, I started doing this to avoid the crazy NFP moves but then decided to close every Friday afternoon!
- #127
- Apr 17, 2010 1:05pm Apr 17, 2010 1:05pm
- Joined Apr 2009 | Status: Trader | 2,035 Posts
Have you ever tried a trailing stop on the second lot?
Indeed there is a trailing stoploss of 18 on the second order.
The first order has a TP of 18 so when this is reached the trailing stoploss on the second order moves up to Break Even.
This makes the second order completely safe, almost a "risk free" entry!
Of course, there is a downside, if it goes pear-shaped we have 2 losses instead of 1.
- #128
- Apr 17, 2010 2:20pm Apr 17, 2010 2:20pm
Geoff,
Regarding something different to determine the trend, or maybe just verify the trend. Please have a look at the following thread here on FF. I have spent a long time testing the TrendMagic indictor and have tried to develop a system around it. It works very good on the 1HR. Your thoughts?
http://www.forexfactory.com/showthread.php?t=165948
SirWolf
Regarding something different to determine the trend, or maybe just verify the trend. Please have a look at the following thread here on FF. I have spent a long time testing the TrendMagic indictor and have tried to develop a system around it. It works very good on the 1HR. Your thoughts?
http://www.forexfactory.com/showthread.php?t=165948
SirWolf
Did you run this test on M15?
I have only tested it on H1.
Only 92% tut-tut!
That is quite impressive for such a simple strategy don't you think?
If only I could find a better way to determine the trend after NY close this EA would be absolutely fantastic!
What is your opinion kijkij?
Believe and Achieve!
- #129
- Apr 17, 2010 4:03pm Apr 17, 2010 4:03pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Attached is the first version of my multi-pair trading ea. Pop it on a demo tomorrow and shout out when bugs appear.
DO NOT USE THIS ON A LIVE ACCOUNT. THERE MAY BE BUGS.
I spend vast chunks of my life coding trading robots. I can spend time trading new ones, of answering questions about old ones. I cannot do both. I choose to do the former. With this in mind, I attach octomanoak's brilliant user guide to the robots in my Boris Schlossberg thread; it explains most of the inputs in the robots I code, regardless of the system they trade. As usual, I shall ignore questions whose answers can be found within this document.
To use this robot as a multi-pair trader, separate pairs in the PairsToTrade input with a comma, for example GBPUSD,EURUSD,GBPJPY. For the rest, users can generally go with the defaults.
The one point to note is this. I code any kind of pip value for 5 digit criminals. Adjustments are automatically made if your account is a 4 digit criminal account. This means that you have to multiply your normal inputs by 10 if your criminal is a 4 digit quoter.
This robot offers nothing better than the original in post one apart from the ability to trade a lot of different pairs on the same chart.
I hope the above makes sense guys. Not sure it will, because the whishkeysh hash been goodsh tonightsh.
Cheers
DO NOT USE THIS ON A LIVE ACCOUNT. THERE MAY BE BUGS.
I spend vast chunks of my life coding trading robots. I can spend time trading new ones, of answering questions about old ones. I cannot do both. I choose to do the former. With this in mind, I attach octomanoak's brilliant user guide to the robots in my Boris Schlossberg thread; it explains most of the inputs in the robots I code, regardless of the system they trade. As usual, I shall ignore questions whose answers can be found within this document.
To use this robot as a multi-pair trader, separate pairs in the PairsToTrade input with a comma, for example GBPUSD,EURUSD,GBPJPY. For the rest, users can generally go with the defaults.
The one point to note is this. I code any kind of pip value for 5 digit criminals. Adjustments are automatically made if your account is a 4 digit criminal account. This means that you have to multiply your normal inputs by 10 if your criminal is a 4 digit quoter.
This robot offers nothing better than the original in post one apart from the ability to trade a lot of different pairs on the same chart.
I hope the above makes sense guys. Not sure it will, because the whishkeysh hash been goodsh tonightsh.
Cheers
- #130
- Apr 17, 2010 4:34pm Apr 17, 2010 4:34pm
- | Commercial User | Joined Jul 2009 | 5,375 Posts
Viewer Discretion Advised: Shall we shag now or should we shag later? :-)
- #132
- Apr 18, 2010 12:08am Apr 18, 2010 12:08am
- Joined Apr 2009 | Status: Trader | 2,035 Posts
Sorry geoff,
I don't have any robot and I never sell anyone. I am just following the thread as your EA have good potential to make money. Just looking for to improve it to remove that sharp drop downs.
Kashif Javed
Quote:
"Any one can remove the circled part(Very quick vertical drop down)?
But NOTE: You cannot use any eraser or scissor! http://cdn.forexfactory.com/images/s...m/big_grin.gif"
I'm sure some robot selling rogues use the scissor technique on their so called "live trading results".
- #133
- Apr 18, 2010 12:20am Apr 18, 2010 12:20am
- Joined Apr 2009 | Status: Trader | 2,035 Posts
Geoff,
Regarding something different to determine the trend, or maybe just verify the trend. Please have a look at the following thread here on FF. I have spent a long time testing the TrendMagic indictor and have tried to develop a system around it. It works very good on the 1HR. Your thoughts?
http://www.forexfactory.com/showthread.php?t=165948
SirWolf
I notice there is a .mq4 on your thread so I can take a new look at it.
Thank you for the information!
- #135
- Apr 18, 2010 12:39am Apr 18, 2010 12:39am
- Joined Apr 2009 | Status: Trader | 2,035 Posts
I'm not trying to ruin this thread, but, whole theory actually based on:
If US session direction point Down Asia Session must follow and vise versa..
Anybody actually check that, before spending valuable time?
Here (Chart zoomed out to show larger period) (GBPUSD H1):
Those Blue Vertical lines starts 1 hour after US session is close ( 10 hours covered/10 lines)
show me what we are talking about here? I see either 50/50 or just random
as usual.. Check it out:
did I ever say that I thought that you were trying to ruin this thread?
You usually make concise and valid comments however, looking back at the test results that you were kind enough to share with us, they do give a false impression.
The charts that you shared show a reasonable profit so, in view of your in depth analysis above, the charts that you published previously must be wrong and you were decieving us!
For any strategy to make a profit, no matter how small, there has got to be an "edge" otherwise you would see a loss or at best, break-even.
If your charts are genuine, this strategy does have a pronounced "edge"!
- #136
- Apr 18, 2010 12:48am Apr 18, 2010 12:48am
- Joined Apr 2009 | Status: Trader | 2,035 Posts
Attached is the first version of my multi-pair trading ea. Pop it on a demo tomorrow and shout out when bugs appear.
DO NOT USE THIS ON A LIVE ACCOUNT. THERE MAY BE BUGS.
I spend vast chunks of my life coding trading robots. I can spend time trading new ones, of answering questions about old ones. I cannot do both. I choose to do the former. With this in mind, I attach octomanoak's brilliant user guide to the robots in my Boris Schlossberg thread; it explains most of the inputs in the robots I code, regardless of the system they trade....
Gordon Bennet! You were burning the midnight whiskey, sorry, oil last night!
- #137
- Edited 3:40am Apr 18, 2010 3:23am | Edited 3:40am
- Joined Apr 2009 | Status: Trader | 2,035 Posts
I believe that anyone who has ever tried to design an EA will verify the truth of the following statement.
As someone who has written hundreds of EA's I believe the following to be true....
1. If an EA produces a profit of any description, no matter how small, it is worth pursuing because, in order to show the slightest profit it must first pay the spread and if it can do that, the underlying strategy must be profitable!
2. If the strategy only works 50% of the time an EA based on it will very definitely produce a loss, which makes it less effective than a coin-toss, simply because of the cost of the spread.
3. Successful manual trading strategies rarely convert to profitable EA's because the simple software that we use to create EA's is nowhere near as powerful or innovative as the human brain.
(There may be exceptions to this but here I relate my own experience.)
If anyone would like to add to this list I will be eternally grateful!
As someone who has written hundreds of EA's I believe the following to be true....
1. If an EA produces a profit of any description, no matter how small, it is worth pursuing because, in order to show the slightest profit it must first pay the spread and if it can do that, the underlying strategy must be profitable!
2. If the strategy only works 50% of the time an EA based on it will very definitely produce a loss, which makes it less effective than a coin-toss, simply because of the cost of the spread.
3. Successful manual trading strategies rarely convert to profitable EA's because the simple software that we use to create EA's is nowhere near as powerful or innovative as the human brain.
(There may be exceptions to this but here I relate my own experience.)
If anyone would like to add to this list I will be eternally grateful!
- #138
- Apr 18, 2010 4:05am Apr 18, 2010 4:05am
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Attached is the first version of my robot.
A few points to note:
A few points to note:
- there will inevitable be a few bugs. The robot is for demo only for now.
- it is a multi-pair robot, so users can trade as many pairs as they want. Add extra pairs by separating them with a comma in the PairsToTrade input i.e. GBPUSE,EURUSD,USDCHF
- the default PairsToTrade of just GU reflects this being a GU strategy. Users can experiment by adding pairs, but that is not what this thread is about.
- make sure the default lot size is acceptable to your criminal.
- Trading_Hour is input in your computer's local time, so you need to know what time Sydney or Tokyo open in your part of the world - no gmt offset stuff. The default is not a recommendation - it had to be something.
- any input requiring pips is coded for 5 digit wally-plonker-criminal accounts. The robot automatically adjusts if yours is a 4 digit criminal. If 4 digit account holders change the defaults, remember to multiply your normal setting by 10
- you have the choice of sending immediate market trades during your trading hour, or pending trades at PendingOrderPips distance from the market. Defaults to immediate.
- the robot sends two trades, one with an 18(0) pip tp, the other to be managed by the trailing stop function. Highly sophisticated management techniques are also available using MPTM - follow the link in my signature to find it.
My code is open-source and free. Users may pinch it, adapt it, use it in their own work. Just don't sell it, that's all. That could be mis-construed as a dig at Geoff. It is not; it is a genuine plea for honesty.
- #139
- Apr 18, 2010 4:36am Apr 18, 2010 4:36am
- Joined Apr 2009 | Status: Trader | 2,035 Posts
Attached is the first version of my robot.
A few points to note:
[list][*]there will inevitable be a few bugs. The robot is for demo only for now.[*]it is a multi-pair robot, so users can trade as many pairs as they want. Add extra pairs by separating them with a comma in the PairsToTrade input i.e. GBPUSE,EURUSD,USDCHF[*]the default PairsToTrade of just GU reflects this being a GU strategy. Users can experiment by adding pairs, but that is not what this thread is about.[*]make sure the default lot size is acceptable to your criminal.[*]Trading_Hour...
What a pleasure it is to witness such genuine altruism!
Footnote:
Steves reference to "a dig at Geoff" refers to the fact that I use EA's of my own design to manage trading accounts for a large number of clients.
I had previously made Steve aware of this fact and, I think, assured him that I would never use this thread as a means to obtain another EA for my business.
My aim here is to repay a debt that I owe various members of FF for their invaluable contribution to my forex education over the past couple of years.
Ending up with a working EA on this thread that everyone can profit from is my sole aim!
- #140
- Apr 18, 2010 4:58am Apr 18, 2010 4:58am
- | Commercial User | Joined Jul 2009 | 5,375 Posts
Oh, I never thought that you were trying to ruin this thread,
did I ever say that I thought that you were trying to ruin this thread?
For any strategy to make a profit, no matter how small, there has got to be an "edge" otherwise you would see a loss or at best, break-even.
If your charts are genuine, this strategy does have a pronounced "edge"!
to improve it. If we can squeeze 100 pips weekly out
of it, that be more than perfect.
Viewer Discretion Advised: Shall we shag now or should we shag later? :-)