Symphonie EA

[slowkey]

George of the two trades I had a chance to look at one came back and hit the 5 pip stop loss for 5 dollars. The other was a 38 dollar win. Since than I have had a lot of little losers and some winners. That on my new settings. Over all the 60-5 looks better with my new settings. I am going to take a look at it with the RSI filter later. Need to keep some of these trades from starting in low momentum ranging periods. The 240 with the rsi and first extreme indicator has a winning AUDUSD trade on it. Normally I would close some of these out manually but need to see what happens. It does not trade enough though to tell much and that is a problem. But it will be ahead. The 60 minute period might be better. The new settings have made this look better with the new extreme indicator.

[cuzgeorge]

The 60 minute period might be better. The new settings have made this look better with the new extreme indicator.

slowkey, this is great news, well done.
just go's to show that Eval has done his homework and there is substance in what he says.

okay, as far as timing and getting more trades, the best i found was dropping the time frames, but it sounds like you want to stick to the 60:5. okay, LHDT, had mentioned that you would get 75% more success rate following the 60:5 method. i dont know where he got that figure from, but it sounds right, no one can expect 100%.

Then , i found when you change the period setting on the RSI filter down to '9', it's very active and can trigger a trade quite fast. '21' is standard here and i'm still playing around at the 15 and 18 level. Can't confirm anything yet, but just got onto the puter and will check my results for the day. i also don't know what to do with 'applied price' in RSi, but will figure it out. any ideas here?

slowkey i also played with by changing the 100 period to 95, but again, still going to check. what i can tell you, a lot of people on the symphonie thread on FF would complain about getting in too late on a trade or exit too early, but Eval maintained that you should wait for all the signals and then move in. my 'homework' has found that Eval is right. he does'nt want this system to work exactly on a cycle, and has expressed himself very clearly when he wants the method to capture the 'meat' of a trade. very good advice.

trying to catch a movement right by the change of direction often gets people burnt and thats my motto, "I’ve never swung at a ball while it's still in the pitcher's glove” W. Buffet"

I have'nt got to trying Hanover yet, so i won't comment on anything there.

Also, AllAverages is still in the testing area, although i have set up a few charts with that settings tinkered with. I like AA because you have an option near the bottom of it's section to choose "candle shift", how many bars to look back for the signal. i used 1 for now, but if you look at the charts, 1 might be too quick and 4 too late, so still messing around here. i expected the exit strategies to kick in if the position was taken too early and went the wrong way, but it does'nt seem to be going my way here. likewise with slowkey, it seems i had much better results since yesterday, after i changed some period settings, will also confirm this with you later.

Nanningbob looks interesting and provides both a trigger and a trend detector. i've changed to lower TF's here and also used a few charts with it. I went lower TF's to entice more triggering, even though they might go the wrong way, but still i only went down to 60M and i think the exit strategy should take care of the rest. i must admit though, i don't fully understand Nanningbob yet and have yet to study Steve and Nanningbob's Big Bob EA. Its on a chart, but i have'nt got to the settings yet.



Okay, i think this is definately going to change the way i set up my trades. let me look at the results for the day and the settings, as there were far too many '0' profit results after they were in a good profit, dont know why TS or BE did'nt kick in. it might be the broker or the Empty4 , maybe even a good reboot will do the trick as i have'nt rebooted since sunday night. i might d/l the ea again and set up new charts as i might have messed up some of the settings and caused irreparable damage, hahaha.

okay, i'm blabbing again, so will check with you shortly,
regards
cuzgeorge

[cuzgeorge]

Latest update in post 1.

Assuming you guys have been using the previous updated version, then the most likely cause of the trade closure at a small loss is that I coded the exit strategies to apply to the LTF as well.

I have added a new input to the Exit Strategies inputs - ApplyStrategiesToLtf. If set to false (the default) S will only test the HTF for trade closure. See if this solves the problem.

This is a shockingly good idea, damn you're good. It helps to read the posts again as it always makes more sense, each time one reads them.
george

[cuzgeorge]

Hi Steve,

i remember this was posted by a member of your forum , LhDT in another forum, and i think it can be particularly usefull in this experiment. How can i emulater this, i can't get my head around this with the current options you have allowed. I have'nt had the time to fully study all the options, but to cut to the chase, here is what he had suggested. Dont know how relevant , or practical this could be , but heck, it makes a lot of sense.

Exit : Use stochastics cross (11,3,3, typical) opposite.

If you closed a trade when stochs crossed in opposite direction and you got a repaint few bars after (all indicators same direction again), reenter. Eval calls this a repaint play and it works pretty well !

Of course, for the trigger (M5+H1) you will have less opportunities but 75% better quality !

Another really important thing is trading session. Best used are London Open to London Close -1.

Thx,
LhDT

PS : Next step will be to avoid ranging/choppy conditions. For this you can use ADX (don't trigger if <25).
PPS : For SL, I use ATR * 1.5

EDIT : Oh yes, don't bother using external indicator for trendLine, just use the build-in Empty4 CCI (iCCI). Up when >=0 and down when <0. Just to preserve CPU power.
__________________

Avoiding the small , choppy trade is vital from what i see, in all TF's as the market never always moves straight up or straight down. This way you move into the position at the time to catch the 'meat' Eval was talking about, and avoid all the false signals. This i mean of LhDT's comment about using ADX.
The indicators must line up, all of them. A position has to exit the moment a bar closes and and any of the exit conditions are met, as is with the latest version.

Thats when the 'repaint play' comes into the picture, if the market moves in the same direction again. Eval often mentions this that it's not over untill all the indicators show the opposite direction, especially his trendline. Not following this was vital to the symphonie system.

my new 5c into the formula, thoughts anyone please?
regards
cuzgeorge

[SteveHopwood]

Latest update in post 1.

Assuming you guys have been using the previous updated version, then the most likely cause of the trade closure at a small loss is that I coded the exit strategies to apply to the LTF as well.

I have added a new input to the Exit Strategies inputs - ApplyStrategiesToLtf. If set to false (the default) S will only test the HTF for trade closure. See if this solves the problem.

This is a shockingly good idea, damn you're good. It helps to read the posts again as it always makes more sense, each time one reads them.
george

Hehe. It often surprises me how many people appear to gloss over my posts - not you, this is just a general comment.

Why should anyone read my stuff? I am only the mechanic, after all.

[SteveHopwood]

Avoiding the small , choppy trade is vital from what i see, in all TF's as the market never always moves straight up or straight down. This way you move into the position at the time to catch the 'meat' Eval was talking about, and avoid all the false signals. This i mean of LhDT's comment about using ADX.
The indicators must line up, all of them. A position has to exit the moment a bar closes and and any of the exit conditions are met, as is with the latest version.

Thats when the 'repaint play' comes into the picture, if the market moves in the same direction again. Eval often mentions this that it's not over untill all the indicators show the opposite direction, especially his trendline. Not following this was vital to the symphonie system.

my new 5c into the formula, thoughts anyone please?
regards
cuzgeorge

George, remember this bit from post 1? "A contributor asked me to take a look at Symphonie shortly after I started this forum, but I declined because of my dislike of trading systems that require several pretty-coloured thingies to line up."

I often describe myself as, "Just the mechanic here" or somesuch - meaning that I am merely coding in response to other peoples' ideas. This is more true sometimes than at others.

Here, it is 100% true. I dislike Symphonie and cannot envisage trading it on a live money account. I dislike it because it requires several pretty-coloured thingies to line up before a trade is triggered. In my experience, by the time they all line up, the move is all but over.

That is OK here, because Eval is up front about not trying to catch the start of the move. I can even entertain dim hopes that my last para might one day be triumphantly deleted. Dim hopes, and I doubt it.

Relevant to here, what this means is: I do not keep up with developments in Eval's threads and so have no idea what is going on in them. If there is a thingy happening that I can code and that you feel is useful, then I will code it, but I need a blow by blow description of what that thingy is. I have no idea what your previous post refers to.

Cheers

[slowkey]

Steve, I sort have had the same thoughts all along about all the things that have to line up on the 60-5. I am back to the higher time frames and looking at a 240-1440 with rsi filter. This seems to quiet everything down a lot. I think if this is to be successful it will be on this higher time frame. I narrowed in on the setting and hope to have this running this evening on Demo. It is the only setup that ever looks good in the tester. Actually some pairs look fabulous up on these higher time frames but that does not mean anything till I can get them trading on a demo. I cannot make the 60-5 work. Although I have not given up entirely. That last little change you made with the LTF trigger switch helps a lot in the flexibility it gives us on these higher time frames.

[slowkey]

I think I am suffering from symphonie fatigue

Think I need a concerto or sonata. Anyway I have two 240 setups. I am going to take a break. I have my doubts about these working. They do not want to trade a lot. I have some Ideas I want to try to code my self. Just simple stuff for practice and motivation.

[cuzgeorge]

Hi Steve, Slowkey , all,
hope you are all having a good rest with your families and enjoying the holidays, God bless you all and keep you safe.

Steve,
i've got myself in a rutt checking out too many EA's so i'm going to slow things down a little. Actually a lot.

i've come to terms with the fact that the 'thingies' you mention are a great nuisance and can lead to some very frustrating losses, and bad tempers.

by nature i'm a bit of a gambler, so i'm giving this one a full whack! I have also accepted that there are no fireworks in forex trading and looking for the 'wild-west' here is not going to become a reality. i envisage several computers shooting off positions and buying and selling at a rapid pace. off-course i can't afford that, but that's the picture that comes to my head. I'm going to slow things down as i realise slowkey is headed in a better direction than me, based on his results compared to mine.

The higher TF's are going to be the new playground and results will trickle in on a weekly basis. Evaluator did say that you can expect 3 or maybe 4 positions in a day. Sure, i'll multiply that by a few charts and a few more demo accounts, and we're back in business.!! will post all the new details for the 'myfxbook' when they all set up so we can compare again with more volumes.

as for my previous post, where i confused everyone, i think. I was looking to avoid ranging markets and looking at what options we had. my lower time frames were to blame for this, and I almost suggested bringing in another thingy, but that's not the solution. I think i'll follow in slowkey's path and report from there, albeit a little different on the settings , but still along those lines.

This leads me to my first question, please don't shoot me down here, i know it's going to look simple to you, but please, read me through;

Code: Select all

extern int     TrendConfirmationCandleTF=15;//Candle for confirming the trend as a trade trigger

the above code refers to A TF for trend confrmation, are you sure it's a TF and not the bar count????
God-help me!!!

If you're still reading means i'm spared from the lions;

my second is the AllAverages indicator suggested in the code below. I have the indi in the correct folder and i have changed the settings to what i'd like to test. This will obviously filter through to the EA through the icustom command. In your opinion, am i doing more damage by doing this and messing with the symphonie system and your EA, or is it good practice to make changes like this. ? AM i supposed to keep it simple, or is it on good advice to tinker so much.?

Code: Select all

 return(iCustom(NULL, 0, "AllAverages_v2.1 cc", tf, price, period, mashift, method, buffer, shift));


}//End double GetAa(int tf, int price, int period, int mashift, int method, int buffer, int shift)


double GetAtr()

Code: Select all

bool IsTradeAlreadyHedged(int index)
{

   /*
   Trade will already be hedged if:
      - there is an opposite direction trade open
      - the opp trade has a lot size x HedgeLotsMultiplier

and finally, according to this code above, if for eg, i have a EURUSD position going long at a full point, and a USDCHF position going short for a full point, would the hedge still kick in for both pairs, or does the EA regard these positions as hedged already?

Now i must be baby-lamb meat for the crocs!

i ask this becuase i have not seen them in my tests, although i'm sure it did occur at some point but, i'd have to be in front of the screen to notice. my results wont point that out to me clearly, and i would'nt know how to check.!

Please spare me a thought Steve, i'm trying my best.
regards
cuzgeorge

[cuzgeorge]

Hi slowkey,
i went through your settings and i like the changes, i'm going to stick to larger TF's like you, the short are definately all over the place and not showing good resutls anyway. the 5:5 or 5:1 definately does not work and will attempt what you had mentioned before, maybe 240:60 and even higher.

for this, i'm going to try r-20, s-17, u-6.

my bar count i have set right down to 300 for cpu saving, and got rid of the irritating alerts as per Steve's post much earlier.
Emotion and Sentiment indi's i'm leaving standard as per Eval's spec.

i will be using a selection of SL and TP's , Stealth pips , BE's etc etc, but all much higher than before. i'd say between 50 and 500 pips per position.

i will only be using option 3, as i have no good results from option 1 or 2. there were obviously some outright wins, but you can see on the 'myfxbook' results, they really not the majority and never made up for the losses. I seldom caught TP, as per the comments section, of small amounts from 5 to maybe 20 pips.

ATR will be 'false' for all and i have changed rsi to 60-40. 240 will be my standard trend setting as per Steve's original, as 60M for this was no good.

Nanningbob v10 is going to be big with the new sets with mostly option (2) for smma.

i'm leaving slowkey up to you, as i think you'll have a much better understanding of this than me.

AA is also high up on the list, can't say what settings yet, but will post with results, and finally Hanover. I like this alot and it's a nice insurance policy i think, but i'm going to wait and see what Steve thinks of my previous post, and if i'm left with any dignity after his reply. 'i'm shakin in my boots'!

hope this send s you a good signal as to which way i'm going. i think our results should vary and depending if the markets pick up much for next week, i hope to have a good set of results to compare with all.

i'm going to set up multiple 'myfxbook' accounts, each with it's own set of most the pairs, and different set files for each account. this will make it much easier to see which set is more effective. I got a feeling we'll need about a month's testing, unless there is something drastically wrong with our settings, but this will become evident very quickly so we can change.

i'm also going to attempt the metals, also with a seperate account, with their own suitable settings.

my EXIT is going to be on the lower TF as per Steve's new option. i also like that alot and think it's going to be the winner next week.

stay well and enjoy the holidays
cuzgeorge