stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Symphonie EA
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=435
Page 13 of 31
Author:  SteveHopwood [ Thu Mar 22, 2012 8:32 am ]
Post subject:  Re: Symphonie EA

slowkey wrote:My concern with this would be though in high momentum situations would we loose possible large gains or do conditions stay favorable. Of course if this is even possible what signal or signals do we use. Maybe Evaluator knows. This is me just thinking out loud Steve if this makes no sense in code or logic I understand. Also if we let this run for a few days we may find this a minor issue.
It is easy to add the facility to amend tp's that fall less than x pips on the wrong side of round and big numbers. We shall see if this is necessary.

:D
Author:  slowkey [ Thu Mar 22, 2012 8:38 am ]
Post subject:  Re: Symphonie EA

Well It nicely did go down another 30 pips for a total 60 pip gain on the two trades. Conditions changed and no further trade was placed. Very enjoyable to watch this happen. Very encouraging. Now really this night owl is to bed.
Author:  pt2004 [ Thu Mar 22, 2012 11:35 am ]
Post subject:  Re: Symphonie EA

Hi,

First time poster here... But long time lurker of symphonie.
just one doubt, shouldn't the EA reverse the position in a opposite signal??? Like we are in a sell but so far no TP, SL or BE hit and all conditions turn into Buy, shouldn't the EA close the sell and open a Buy, or have an option to do it???

Just a though .

Best regards and thank you Steve for the EA and Eval for the strategy .
Author:  SteveHopwood [ Thu Mar 22, 2012 11:52 am ]
Post subject:  Re: Symphonie EA

pt2004 wrote:Hi,

First time poster here... But long time lurker of symphonie.
just one doubt, shouldn't the EA reverse the position in a opposite signal??? Like we are in a sell but so far no TP, SL or BE hit and all conditions turn into Buy, shouldn't the EA close the sell and open a Buy, or have an option to do it???

Just a though .

Best regards and thank you Steve for the EA and Eval for the strategy .
I don't think any of the exit strategies wait until there is a full opposite-direction signal before closing a trade.

:D
Author:  pt2004 [ Thu Mar 22, 2012 12:06 pm ]
Post subject:  Re: Symphonie EA

Yes, you are right.
So the exit strategies are not working correctly , i had a open GU sell only using LTF, using exit St. 1, and the LTF trading condition was Buy, now it changed to none, but once i spot other situation like this i will take a print screen.

Best Regards
Author:  Lex [ Thu Mar 22, 2012 2:12 pm ]
Post subject:  Re: Symphonie EA

slowkey wrote:I have a nice trade on gbpjpy whose t/p is 3 pips below the round number. Problem is it can not get past the round number. Interesting problem for an EA.


Update .... well it retraced all the way back to entry. Now I would have closed this out manually. Any EA needs hands on help in these conditions but I want to let this thing alone and see what it can do on its own. There was a trade signal but not enough momentum to push down below a round number.

Now when it stalled out close to the round number the conditions were no longer favorable. It would be interesting if we could close trades based on a loss of signals when the market was greater than entry. The 60-5 did its job but this was not a high momentum move. The trade stopped short of its target which was 30 pips and conditions became quickly unfavorable.

My concern with this would be though in high momentum situations would we loose possible large gains or do conditions stay favorable. Of course if this is even possible what signal or signals do we use. Maybe Evaluator knows. This is me just thinking out loud Steve if this makes no sense in code or logic I understand. Also if we let this run for a few days we may find this a minor issue.
Hello,

I guess this is my first post here. Yay.

From what I understand of Eval's strategy, and depending on the exit strategy, the position, or half of it, should be closed "when the Cycle indicator shows an extreme spike (...) at the close of that candlestick".

So, if this happens before the "round number", the retrace to BE or even a loss would be avoided. I don't know if this condition is coded or working on the EA. For the time I've been testing it I'd say this isn't working. Of course, I may be wrong.

There was a an addition to the "extreme spike close", which consisted in exiting about 4 candles later, in case the spike was 'overrun' in the next few candles as it often happens. But I don't recall exactly where's Eval's post about this, so I can't say if it was meant to be a discretionary exit, therefore not 'codeable'.

I hope this can be useful. On the other hand, I may be talking about something very obvious already more than considered by you all, namely Steve. If that's the case, forget what I said.

And thank you all and Steve and Evaluator for this promising strategy and EA ;)
Author:  Lex [ Thu Mar 22, 2012 2:35 pm ]
Post subject:  Re: Symphonie EA

In fact it was very easy to find. Eval says, about Exit Strategy Two:
When the Cycle indicator shows an extreme spike (spike with colour) exit one half of the trade at the close of that candlestick. (One caveat.....once you get an extreme spike wait for 3 to 4 bars to see if the Extreme holds. If no repaint then exit at close of 4th bar/beginning of 5th.)
The second half of the order would continue to run until the trendline changes colour. Once the trendline changed colour one would exit at the close of that candlestick.
I suppose the exits, after the extreme and when the trendline changes color, are already coded in the EA. How else could we choose between the exit strategies?

So now I have a doubt: for these exit strategies to work, we should not input TP values, right? Leave them at zero?

Sorry for my idiocy
Author:  slowkey [ Thu Mar 22, 2012 2:38 pm ]
Post subject:  Re: Symphonie EA

Hello,

I guess this is my first post here. Yay.

From what I understand of Eval's strategy, and depending on the exit strategy, the position, or half of it, should be closed "when the Cycle indicator shows an extreme spike (...) at the close of that candlestick".

So, if this happens before the "round number", the retrace to BE or even a loss would be avoided. I don't know if this condition is coded or working on the EA. For the time I've been testing it I'd say this isn't working. Of course, I may be wrong.

There was a an addition to the "extreme spike close", which consisted in exiting about 4 candles later, in case the spike was 'overrun' in the next few candles as it often happens. But I don't recall exactly where's Eval's post about this, so I can't say if it was meant to be a discretionary exit, therefore not 'codeable'.

I hope this can be useful. On the other hand, I may be talking about something very obvious already more than considered by you all, namely Steve. If that's the case, forget what I said.

And thank you all and Steve and Evaluator for this promising strategy and EA ;)
I think you have made a good point. We have two Main signals based on 60-5 time frames. Should we by evaluators rules be exiting on the ltf going to none and candle close. The higher time frame can stay favorable for longer periods of time because there is less noise but this is a 5 minute trader. When the market is ranging we will get these retraces.
My strategy two Demo was closing its whole position it looks like but I think I have corrected that now. User error lol. Will see.
Author:  acostafulano [ Thu Mar 22, 2012 3:20 pm ]
Post subject:  Re: Symphonie EA

I have been running the EA since the last update.

I have 5 demos on, on all the pairs available in my platform:

- 60-5 with Exit strategy 1
- 60-5 with Exit strategy 2
- 60-5 with Exit strategy 3
- 60-5 with TP 15 SL 25
- 240-15 with TP 25 SL 35

My intention is to let this guys run indefinitely and try to find a pattern: Which times are the "trending" times for each pair?

The problem I've found is that sometimes (and due to be running this on so many pairs per platform) is the trade context busy errors 146... When more than one trade triggers at the beginning of any given candle, the Empty4 platform only lets 1 or 2 trades in and the I get some "context busy" for the rest of the pairs that also triggered... so trades that should've been triggered don't go through, only to (sometimes) trigger on the following candle and lose a bit of the initial movement, sometimes they miss the trigger and therefore the trade.

I know that it will be a bit more intensive on CPU, but for these kind of intensive hardcore testing, maybe a check of triggers at the beggining of each M1 candle could help avoid missing this trades. Anyone else encountering this problem? Any comments at all?
Author:  cuzgeorge [ Thu Mar 22, 2012 3:46 pm ]
Post subject:  Re: Symphonie EA

slowkey wrote:
Hello,

I guess this is my first post here. Yay.

From what I understand of Eval's strategy, and depending on the exit strategy, the position, or half of it, should be closed "when the Cycle indicator shows an extreme spike (...) at the close of that candlestick".

So, if this happens before the "round number", the retrace to BE or even a loss would be avoided. I don't know if this condition is coded or working on the EA. For the time I've been testing it I'd say this isn't working. Of course, I may be wrong.

There was a an addition to the "extreme spike close", which consisted in exiting about 4 candles later, in case the spike was 'overrun' in the next few candles as it often happens. But I don't recall exactly where's Eval's post about this, so I can't say if it was meant to be a discretionary exit, therefore not 'codeable'.

I hope this can be useful. On the other hand, I may be talking about something very obvious already more than considered by you all, namely Steve. If that's the case, forget what I said.

And thank you all and Steve and Evaluator for this promising strategy and EA ;)
I think you have made a good point. We have two Main signals based on 60-5 time frames. Should we by evaluators rules be exiting on the ltf going to none and candle close. The higher time frame can stay favorable for longer periods of time because there is less noise but this is a 5 minute trader. When the market is ranging we will get these retraces.
My strategy two Demo was closing its whole position it looks like but I think I have corrected that now. User error lol. Will see.

Hi slowkey, all,

i agree with what you are posting about here. The concern i have with the 'wait for 3 or 4 bars' strategy is that there is no rule to this, and Eval expects a certain amount of 'forex savvy' thinking from the trader. I don't think that particular method is the one to follow for an EA. there are just some things that can't be done, for reasons like yesterday, and the day before. the market takes 8 hours of sideways wobbling, (breathing) and in 2 bars in 5M TF, 8minutes the whole days gains were gone. i get this from the Symphonie FF forum where the guys were complaining about it. Thats 2 bars guys, this would have had to be manually stopped if one could read the indis well. Its happening far too often lately.

What i found interesting, is the 5-60 rule that Eval is using. he watches the 5m TF and if the starts line up, he checks the 1H TF and if it is showing 2 or more of the same indis in his 5mTF, he makes his trade. i Believe it's his Emontion and Sentiment indicator in the 1H TF that he looks at as a minimum for executing this position. This is programmable, but in reality, i think this is about 2 or 3 trades a day. Eval could comment about it, i never tried that, i like to get in and out all the time, one of my weaknesses.

this leads to me to the progress on the latest EA and i can tell you , i'm losing, and it takes much less trades than before. I have NOT interfered even when i could see it's going the wrong way. This has caused me to wake up between the afternoon naps while watching the bars, and start fiddling with the settings Steve has allowed.Like you, i also noticed the EA not trading when it shoudl have and also not closing trades.

i'm thinking to try change the 60TF and put the 15TF, because we're not in it to make 300 pips from one position. thats very far from reality of a 5TF. i'm playing with 20-35 pips and equal SL. I've entered all options on different currencies and at equal and different settings for all the currencies (the main ones). The exit strategy is to close when the trend turns ( end of a candle in the 5TF, not 15.).I'm making notes but can't get back to see what happened a few hours ago. a movement starts from the lowest time frame, if it holds to 5TF, then it becomes and option, if the 15TF points in the same direction, then we jump in. Exit should probably be 1minute TF, but it can change or whipsaw, so we wait for confirmation from 5TF.

does this make sense or am i smoking my socks guys? :shock:

kind regards
george


Slowkey, i noticed you have a 'myfxbook' account and i think you have linked the EA there. Are you able to get info back on different time frames to see if there is a better time frame to use and possible suggest to Steve and Eval.? is this an option,? are we straying too far away from Eval's conditions? is it still symphonie? I believe Steve is going to try his best to get to as close as Eval can direct him to the ultimate symphonie system, but thinking logically, there are some actions in symphonie, that no code can produce.

this has become more obvious to me as i read the forums , and wondering if Eval can throw some light here and direct us for some of the testing. There's really not much more to add here.
All times are UTC Page 13 of 31