DislikedHi Jor
Is this manual trading or with Steves EA? How long time have You been trading Your system?
Best wishes
ElovvIgnored
jor1234's PwR always-in auto trader by Steve Hopwood
- #22
- Sep 8, 2011 1:23pm Sep 8, 2011 1:23pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
The code is not quite right yet; mine has missed reversals as well.
I will look later.
I will look later.
Current positions EURUSD + 155 PIPS (100 LOCKED IN)
GOLD + 234 TICKS (139 LOCKED IN)
AUDUSD + 12 PIPS
CLOSED AUDUSD -39 PIPS (EA OPENED AN ORDER 12 PIPS LATER AFTER THE SIGNAL CAME IN, ACTUAL LOSS IN MANUAL WAS 26)
Again Steve once again, thanks for putting this together so quickly, as soon as we fix taking all the reversals then this will really take off
GOLD + 234 TICKS (139 LOCKED IN)
AUDUSD + 12 PIPS
CLOSED AUDUSD -39 PIPS (EA OPENED AN ORDER 12 PIPS LATER AFTER THE SIGNAL CAME IN, ACTUAL LOSS IN MANUAL WAS 26)
Again Steve once again, thanks for putting this together so quickly, as soon as we fix taking all the reversals then this will really take off
- #25
- Sep 8, 2011 2:48pm Sep 8, 2011 2:48pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Latest update in post 1.
As soon as I looked at the code, I saw what I had done wrong; were all logic errors so easily found.
As soon as I looked at the code, I saw what I had done wrong; were all logic errors so easily found.
Latest update in post 1.
As soon as I looked at the code, I saw what I had done wrong; were all logic errors so easily found.
. Im attaching the statement of my demo since I started running the EA yesterday afternoon. Just for the record there's a trade, the $ 975 loss in gold that was the result of a trade that as we discussed in the first few posts it should not have been opened as price it had moved away from the original entry point already when I loaded the bot onto the chart. trades after that one were in line and followed all the rules with the exception that orders were not being opened as soon as reversal happened but a bit later instead, but once again something that should be already fixed with the latest update . My apologies for the long read guys, but even with the entries bug it still performed tremendously and Im excited so many of you are following it
INITIAL BALANCE: $ 50 000.00
ENDING BALANCE: $ 52165.00
.: + 4.33% Since yesterday afternoon.
- #28
- Edited 4:51am Sep 9, 2011 4:39am | Edited 4:51am
- | Commercial User | Joined Apr 2007 | 8,331 Posts
OK guys, time for us to put our thinking caps on. What I have coded is a pretty accurate representation of jor's posted strategy. This has made a highly promising early start, so anything I add from here on can be an optional extra.
Just because jor has been successful trading his system manually does not necessarliy mean that the way he has been trading it is the best way. It may be, but time constraints prevent a system developer from experimenting too much.
We have all seen the problem of loading up the bot for the first time. Here are a couple of possible solutions:
Just because jor has been successful trading his system manually does not necessarliy mean that the way he has been trading it is the best way. It may be, but time constraints prevent a system developer from experimenting too much.
We have all seen the problem of loading up the bot for the first time. Here are a couple of possible solutions:
- Use a previous-candle PwR filter, that can be applied to trade entry and/or trade exit:
- for a buy, PwR must be smaller than that of the trigger candle's PwR value.
- for a sell, it must be bigger.
- Trade with the trend of a higher time frame. My bots usually include a shed-load of optional trend-detection functions that I can easily add here. I can add a higher tf PwR filter as well.
By making all the extra stuff available as an option, all the inveterate tinkerers at FF have the ability to play with this method to their hearts' content, or to stick with the 'always-in' philosophy of the original idea.
Any thoughts, gals and galesses?
- #32
- Sep 9, 2011 11:18am Sep 9, 2011 11:18am
I pm-ed Steve Hopwood and Jor1234 with the question if I could post a mod here in which I put in another pwr calculation on a higher timeframe, so the lower timeframe in which the trades are taken only takes positions in the direction of the higher timeframe pwr-trend.
The reply I got from Jor1234 is worth reading, so I quote it here:
Hey Jules, well first of all thanks for following the thread we started and thanks for looking for ways so we can improve the EA. I think any improvements we can come up with should be include in the EA that way we dont lose track of the original system, and we can build upon it. now in regards to the time frame I use to trade this in the 4 hour, but little by little I have been adapting and the 15 minute has produced the best results. The best example I can give you is when the swiss intervention happened this week the buy order came an hour and 15 minutes before the move happened and to see the results and what I mean just open a 15 minute chart with the indi on and you will see it. now in the other hand the 4 hour chart missed it completely. Let me know what do you think, and once again thanks for contributing to the thread
So I wont put the mod here, as per request of Jor1234. Maybe the original bot from post 1 can be equipped with a higher timeframe trade direction finding thingy. It can be either Steve's regular ones or the pwr-direction I mentioned.
But when I look at it now, I don't think an automate trade direction module is necessary.
The results as per now (I know, very early days) on the 14 "main" pairs:
The original EA on the 15 min. timeframe has a profit on the open trades of $210, with a closed loss of $16.
My mod on the 4hr timeframe, which only trades the pwr direction of the daily timeframe has a profit on the open trades of $153, with a closed loss of $26.
I know, it is absolutely not an objective test. The demos didn't start with the same pairs and on the same time. But my feeling says an always-in is better with this EA.
I also have the original EA on a daily timeframe with MPTM to lock in profits using the BE-setting and a trailing stop. I think adding the option of a trailing stop and BE-settings will be an improvement for the EA, like Jor1234 also mentioned in an earlier post. (off course we can use MPTM for that)
When you ask me which timeframe the EA gives the better results, 15 min or daily, than I agree with Jor1234. It is performing better on a 15min TF on my demos, but like said before, early days.
Well, enough to think about and to test. But as you can notice I am very excited about this EA and advise everybody to give it a try on demo!
The reply I got from Jor1234 is worth reading, so I quote it here:
Hey Jules, well first of all thanks for following the thread we started and thanks for looking for ways so we can improve the EA. I think any improvements we can come up with should be include in the EA that way we dont lose track of the original system, and we can build upon it
. now in regards to the time frame I use to trade this in the 4 hour, but little by little I have been adapting and the 15 minute has produced the best results. The best example I can give you is when the swiss intervention happened this week the buy order came an hour and 15 minutes before the move happened and to see the results and what I mean just open a 15 minute chart with the indi on and you will see it. now in the other hand the 4 hour chart missed it completely. Let me know what do you think, and once again thanks for contributing to the thread So I wont put the mod here, as per request of Jor1234. Maybe the original bot from post 1 can be equipped with a higher timeframe trade direction finding thingy. It can be either Steve's regular ones or the pwr-direction I mentioned.
But when I look at it now, I don't think an automate trade direction module is necessary.
The results as per now (I know, very early days) on the 14 "main" pairs:
The original EA on the 15 min. timeframe has a profit on the open trades of $210, with a closed loss of $16.
My mod on the 4hr timeframe, which only trades the pwr direction of the daily timeframe has a profit on the open trades of $153, with a closed loss of $26.
I know, it is absolutely not an objective test. The demos didn't start with the same pairs and on the same time. But my feeling says an always-in is better with this EA.
I also have the original EA on a daily timeframe with MPTM to lock in profits using the BE-setting and a trailing stop. I think adding the option of a trailing stop and BE-settings will be an improvement for the EA, like Jor1234 also mentioned in an earlier post. (off course we can use MPTM for that)
When you ask me which timeframe the EA gives the better results, 15 min or daily, than I agree with Jor1234. It is performing better on a 15min TF on my demos, but like said before, early days.
Well, enough to think about and to test. But as you can notice I am very excited about this EA and advise everybody to give it a try on demo!
Hey everyone I was trying to post earlier but I got a bit busy, apologies for that. I closed down every trade the EA had opened and I also all my orders in general because this is going to be my long weekend, last weekend I couldnt take it off so this weekend will be my official labor day long weekend holiday . with that being said I wont be around for the close today, nor the open on sunday and I dont like not being able to monitor those two so thats why I liquidated all positions both live and the EA we are testing.
Today even with the error of the late entry in GOLD for which I posted the screen demostrating it the EA produced a profit of 97 pips on my end. once again if I had the EA running on gold on time the results for the day should have been 242 pips instead, but hey it was my mistake not to have the EA on the chart so it is what it is as they say.
RESULTS FOR THE 2 DAYS WE HAVE BEEN TESTING
STARTING BALANCE: $ 50 000.00
ENDING BALANCE : $ 53 141.00
.: + 6.28%
I'm attaching the statement as well and to Steve, Jules and everyone who is working on making this better, I truly appreciate it
Have a great weekend.
Jorge
. with that being said I wont be around for the close today, nor the open on sunday and I dont like not being able to monitor those two so thats why I liquidated all positions both live and the EA we are testing.Today even with the error of the late entry in GOLD for which I posted the screen demostrating it the EA produced a profit of 97 pips on my end. once again if I had the EA running on gold on time the results for the day should have been 242 pips instead, but hey it was my mistake not to have the EA on the chart so it is what it is as they say.
RESULTS FOR THE 2 DAYS WE HAVE BEEN TESTING
STARTING BALANCE: $ 50 000.00
ENDING BALANCE : $ 53 141.00
.: + 6.28%
I'm attaching the statement as well and to Steve, Jules and everyone who is working on making this better, I truly appreciate it
Have a great weekend.
Jorge
Hi All
I have changed to Pwr calc in the code slightly to try and reduce oscillations around the 0 line to:
void GetPwr()
{
double ld_power1 = iBearsPower(NULL,0,144,PRICE_CLOSE,1);
double ld_power2 = iBullsPower(NULL,0,144,PRICE_CLOSE,1);
trend = none;
if (ld_power1 > 0 && ld_power2 > 0 ) trend = up;
if (ld_power1 < 0 && ld_power2 < 0 ) trend = down;
}//End void GetPwr()
i.e Both must be + or - to give a signal.
Am now positive on back-testing.
cheers
you just compare them both separately, while the original code is adding bulls to bears....
- #35
- Sep 9, 2011 4:20pm Sep 9, 2011 4:20pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
I agree with Steve, we should find better ways to make this EA even better, a trailing stop could be one of the things we can add to it. In my manual trading when it moves 30 pips my way I automatically move the SL to BE.
Quick recap yesterday as I explained I had to shut down trading for the day so I closed all the orders the EA had open but the EURUSD order that was open yesterday is now ghood for a bit over 300 pips.
CURRENT OPEN ORDERS: AUDUSD= + 128 PIPS
CRUDE OIL= + 192 TICKS...
In the normal run of events, a trailing/jumping stop will be hit and the trade will close.
Then at the start of a new candle, as coded, the bot will
- look at PwR
- see there is a 'trend' in place
- take a new trade
I think you have not used EA's much. EA's are pieces of software that do whatever they are programmed to do. As a human, having seen your trade closed by, say, a retrace to a trailing stop, you will wait until x events happen before re-entering the market. A robot can only act as I describe above.
I need to know what x is.
In the meantime, gobsmacking start by jor today. Your pairs minus crude up mega-pips. Live, I would have closed the trades, pocketed the profit and started again on Monday.
I can add all the usual mptm functionality quite easily.
In the normal run of events, a trailing/jumping stop will be hit and the trade will close.
Then at the start of a new candle, as coded, the bot will
- look at PwR
- see there is a 'trend' in place
- take a new trade
I think you have not used EA's much. EA's are pieces of software that do whatever they are programmed to do. As a human, having seen your trade closed by, say, a retrace to a trailing stop, you will wait until x events happen before re-entering...
so if buy is closed, wait till PwR turns down again.or.....
after (for example) buy order is closed, start counting, so when PwR turns down/red, mark a variable 1 (or true), and when it turns up again, allow buy trade again.
This way, after trailing stop, it has some room to retrace a bit, and start trading next PwR value (<>0)
have a nice weekend
- #38
- Sep 10, 2011 4:47pm Sep 10, 2011 4:47pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Latest update in post 1. Read the 'Subsequent edit' section for details.
- #40
- Sep 12, 2011 12:41am Sep 12, 2011 12:41am
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Hi Steve, can you add feature if trade close by trailing or BE, the robot do not trade until PWR indi change its color? thx very much....