Phoenix optimization - page 14
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Charting Phoenix 5.7.2
Thank you to Daraknor, Azbo, Bert, and others for their feedback and general contribution to our common cause. Just a few comments on optimization and development first.
I appreciate that Phoenix 6 will likely be a quantum leap ahead of version 5xx, but I wonder then what is the purpose of optimizing older versions. Many people are still running 5.6 and 5.7.0, and I have even been discouraged from testing 5.7.2.
Moreover, there are now 2 different versions of 5.7.2W, and it is not clear which one should be used in testing or optimizing. Preliminary tests show that backtesting with a TP near infinity is problematic. It appears that open trades prevent others from opening, and then all the remaining open trades close on Grace at the end of the backtest period. It is difficult to draw any conclusions from this.
Given the uncertainties in backtesting and optimizing in general, I have tried to move in the direction of visual evaluation using indicators and comparison to the method proposed by Malone (EASY). The attached chart illustrates the technique, which involves displaying EASY and Phoenix Signal indicators all in the same field.
Legend: yellow lines in chart are Price Action Channel; black lines are Envelope with Percent=0.05 and Period=13; aqua and Pink lines are Signal 2 in Phoenix with period and bars at 5; Signal 3 is the OSMA window at the bottom; Traders Dynamic Index is for EASY indicators.
Interpretation:
Trade signals are generated in EASY when closing price above or below Price Action Channel and the green line in TDI is above (below) 50 as well as the red and yellow lines.
Trade signals are generated in Phoenix when closing price is above or below the Envelope, and the aqua is above (or below) the pink line and the OSMA crosses the zero line.
By inspection, one can see that it is possible to place Phoenix signals very much in tandem with EASY signals by tuning inputs. As it happens, the tuned inputs tend to perform well in backtesting, although I have not quite completed the process yet.
My main question: is this a waste of time given that Phoenix 6 is just around the corner? Also, I would be very interested in any comments or suggestions for improving the visual method. Finally, how can I plot an indicator for Signal 4 in Phoenix?
Autumn
Before answering on the indicators you intent to build that seems to be necessary, I let you know that learning is NEVER a waste of time. Never...
Then... go on : you are doing a tremendous work shaking the coconut tree of Phoenix even sometimes I don't agree with you...
What will be Phoenix 6 ?
Nobody knows for the moment...
Perhaps the things you try to "built" now will be a great help to optimize P6
See you
NB I will reply later on your indicators
Phoenix 6 is a new engine around the same signals. You can use the same signal optimizations in Phoenix 6 as you can in Phoenix 5.7. A few signals are added as options (filtering, exit strategy, etc) but the core "when do I enter" data can be directly used. We have 4 types of signals:
enter signal (start a trade)
exit signal (when to exit)
filter signal (don't trade during conditions)
confirm signal (increase/decrease lot size based on conditions)
In Phoenix 5.7 Signal 5 is a time based filter, but I don't like that they are all lumped together. We can probably save a ton of CPU cycles by checking the filter signals first, and the enter signals second.
I might take DVBuySell and a few other parameters and make them filter or confirm signals rather than enter signals.
I'm finding I really miss object overloading from Python, C#, etc that would allow me to change which code executes based on how many and what type of variables are sent to a function. I'm doing a C language hack that is somewhat similar, but I don't have function functions or other tricks.
I'm going on a trip to visit relatives for 2.5 days, I'm handing off the code to the development list until I return.
Phoenix 6 is a new engine around the same signals. You can use the same signal optimizations in Phoenix 6 as you can in Phoenix 5.7. A few signals are added as options (filtering, exit strategy, etc) but the core "when do I enter" data can be directly used. We have 4 types of signals:
enter signal (start a trade)
exit signal (when to exit)
filter signal (don't trade during conditions)
confirm signal (increase/decrease lot size based on conditions)
In Phoenix 5.7 Signal 5 is a time based filter, but I don't like that they are all lumped together. We can probably save a ton of CPU cycles by checking the filter signals first, and the enter signals second.
I might take DVBuySell and a few other parameters and make them filter or confirm signals rather than enter signals.
I'm finding I really miss object overloading from Python, C#, etc that would allow me to change which code executes based on how many and what type of variables are sent to a function. I'm doing a C language hack that is somewhat similar, but I don't have function functions or other tricks.
I'm going on a trip to visit relatives for 2.5 days, I'm handing off the code to the development list until I return.Nice trip Darak... and take care
Trailing Stops - and Money Management
Three updates to the TS strategy in 5.7.2.w
1. Make it kick in at one point in profit.
2. Make a TP an equal number of pips above current price as TS is below point. This is so that if your computer loses it's connection to your broker,the trade will close itself once it hits the TS or the TP.
3. Make sure that the SL or TS is not too tight to make the trade illegal.
Also parameterize the new Decrease Factor in the Money Management Section. No Settings File changes.
Code Here
Optimize EA a litle bit
Hello,
i set EA like others : ) , my profit factor was only 1.12 (Statement:link bellow) my question is: do i must change phoenix mode (TP SL...) or i must look mistakes in signals ? Eny idea or example for this small 500$ account
GeneralSettings====== General Settings
PhoenixMode=1
PhoenixMode,F=0 <
PhoenixMode,1=1 <
PhoenixMode,2=1 <--is this thing affect to results or its just for optimization
PhoenixMode,3=6 <
Lots=0.10000000
MaximumRisk=0.05000000
DecreaseFactor=0
MM=0
AccountIsMicro=1
PrefSettings=0
BreakEvenAfterPips=0
Mode1======= Phoenix Mode 1 (Classic) ==================
TakeProfit=84
StopLoss=84
TrailingStop=35
Mode2======= Phoenix Mode 2 (Second trade)==============
0
Mode3======= Phoenix Mode 3 (Three trades at once) =====
0
Signal1======= Signal 1
Percent=0.00320000
EnvelopePeriod=2
Signal2======= Signal 2
SMAPeriod=2
SMA2Bars=18
Signal3======= Signal 3
OSMAFast=5
OSMASlow=22
OSMASignal=2.00000000
http://freeweb.siol.net/mvilfan1/tds/500account.htm
It is good to see you working on optimizing! We noticed there were some flaws with the trailing stop system, so I updated it in 5.7.2. I wouldn't recommend 5.6.3 except for mode 1 no trailing stop, and it will still have some bugs that affect trades. 5.7.0 is stable, but the logic of trailing stop was weak/strange.
I noticed that Signal 4 was the primary signal used to avoid drawdown, it is your main 'filter' of bad trades. If you turn that back on, you might have better optimization results.
Example: a found profit 1.8 with 45% drawdown (2.22 -> 3.9) than i test (2.19 -> 3.9) and my account blow off
how can you trade for real with that ? You must allways optimize last week ?
No, you discard the results that don't show consistent wins. Generally when you optimize over a few months of data you have 30 almost identical results, 22 almost identical results, 17 almost identical results, and then groups of 5-8 results. Each of these is a "local maxima" in genetic algorithm speak. Usually the 30 and sometimes the 22 is curve fitting. I found some really good performers on data I didn't optimize from the third or 4th major set. In a previous example I posted, I categorized I think 20 sets of data. (My optimization was done on my dead computer.)
Im dont understand this clearly : you say 30 results from when to when
example from 1#1.1 -> today , 2# 1.2-> today , #3 1.3-> today
anoter question: do you optimize every signal for it self or you optimeze all signals at once (this take alot of time mybe weeks
)
Can you please look that link: nice move from 2.22 to 3.9 (profit factor 2.47)
but im sure that this setup will faild nextweek.
http://freeweb.siol.net/mvilfan1/tds/fx/250acc.htm
Do you trade for real with this EA ? How many months do you optimize 3 4 a year ,... hm i need some strategy for this
Is possible to have 10%-25% drawdown in long run