Great EA in backtest! - page 96
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
me, david and aragorn (1.85) but still looking for settings. brazilianTrader (1.88).
--------------
my results
eurusd
Total number of trades 46
Percentage profitable 69.6%
Total number of pips -10
usdjpy
Total number of trades 30
Percentage profitable 46.7%
Total number of pips -159
i've got an idea. i will prepare CT to place reverse trades on usdjpy. 160 pips are mine nex week! next month... well...I would appreciate if you do reverse code decision if you would please post.
Here is a screen shot for those that don't understand the process of this EA.
there will be good weeks, bad weeks,good months,bad months.
I have circled areas of flat or drawdowns.
every week won't be profitable.
setting risk to .1 I did mimimise draw down to 8% but returns on the year wasn't riches but was indeed a few thousand %....500 into 9000
the screen shot is only of the first couple of months of the back test befor it got to its max lots and while you could still make out the start peroid.
I have posted my settings and my used version all over the place and sugestions based on what I am doing. There is really nothing more I can contribute, acceptto enphasize every one must find what works for them on there own. MINE WILL NOT WORK FOR YOU. for what ever reason, broker, price filters, PC, server, connection, I dunno.
I must let this run for months. And I must get back to my teachings in my own thread on FFJ.
good luck to you all
Dave
David,
it is obvious that the EA won't trade good every week or even month. i know that. You have found Your settings, i havn't. i won't be satisfied until avarage 50-75 pips gain a week is achieved and not in backtests. in real. i do not trust backtests. i've made a little backtest for last week and guess what. usdjpy -40 pips only. not like in real -169. that is why i'll try to reverse trading logic and do some other experiments
i'm very happy that You contributed to this thread. appreciate Your ideas and Your work. good luck David.
I just back tested last 2 weeks and I got 20 trades.
Same as real # of trades. But only 4 losses
We already know back tests are not 100% accurate.
I am still ok with it
Trading Results
Hello all,
I have attached my trading statements for the past week for all of my three demo accounts.
All three accounts started on October 16th. This was the first week of trading.
Each of the statements includes the changed settings for all, or for an individual pair.
I have made one big mistake though! I didn't obey rule number #1 in forex trading:
STICK TO YOUR TRADING RULES!
I have changed the blocked trading hours sometimes around Wednesday because I thought
that the settings as posted on the CT website were better than fixed blocks of hours.
First, those hours are not accurate and sometimes incomplete, some negative trades are
a result of this. For example on 10/18 I had a big drawdown on $CHF (Demo1). The hour was
blocked but the news occured the hour after. On Demo3 I have the hours for the JPY pairs
completely wrong - what a shame.
I have to rethink my strategy for blocking hours in general. Maybe changing the code to
include half-hour blocks could be benefitial for the system. I don't know.
FXDD Demo1:
8 Pairs, 1H Timeframe:
GBPUSD: +025 pips (03 win, 01 loss)
EURUSD: -026 pips (01 win, 02 loss)
USDCHF: +017 pips (05 win, 01 loss)
USDJPY: -008 pips (04 win, 02 loss)
AUDUSD: no trades
USDCAD: +009 pips (01 win, 00 loss)
EURJPY: +029 pips (03 win, 00 loss)
GBPJPY: no trades
Total: +46 pips (17 win, 6 loss) 74% win ratio
FXDD Demo2:
6 Pairs, 30M Timeframe:
USDCHF: -041 pips (04 win, 04 loss)
USDJPY: -150 pips (11 win, 13 loss)
EURUSD: -004 pips (05 win, 02 loss)
USDCAD: +007 pips (06 win, 02 loss)
EURJPY: -070 pips (07 win, 07 loss)
AUdUSD: +006 pips (04 win, 01 loss)
Total: -252 pips (37 win, 29 loss) 56% win ratio
FXDD Demo3:
6 Pairs, 1H Timeframe:
USDCHF: -001 pips (02 win, 01 loss)
USDJPY: -011 pips (05 win, 03 loss)
EURUSD: -013 pips (00 win, 01 loss)
USDCAD: +009 pips (01 win, 00 loss)
EURJPY: -084 pips (04 win, 06 loss)
AUdUSD: no trades
Total: -100 pips (12 win, 11 loss) 52% win ratio
Account #1 seems to be the overall winner. Account #2 is going to be 86ed. A 30M doesn't work out for me.
Biggest winners:
EURJPY 1H +29 on Demo #1
GBPUSD 1H +25 on Demo #1
USDCHF 1H +17 on Demo #1
Biggest loosers:
EURUSD 1H -26 on Demo #1
EURJPY 1H -84 on Demo #3
If I could get rid of those loosers by blocking the rights hours of trading for a particular pair, I should be able
to bring overall results to a better level.
In addition I have to say that this is the first week of trading, and it doesn't mean anything.
AZBOfin
Dave I apologize for not responding earlier, I took a needed hiatus from beating my head against a wall...
Let me say that I will miss your input here, I hope you put the links to the other forums you are going to be active on so I can track you down when needed.
About my trading goal with this...
I am looking for more perhaps than this has to offer. I've been wrestling with it to try to wring out whatever it can give. That requires pushing the envelope and that is what I have done with it. I cranked up the risk, jacked up the reverse index, did whatever I thought I could to either make it take more trades or win more or lose less. I wasn't satisfied with only a couple trades a day and sometimes even less. I wasn't satisfied with 30% drawdowns which are unpredictable. I want more control than that. What frustrates me is how well it does perform in the backtest. What frustrates me is that you say you are getting similar performance from your live forward tests. I don't seem to getting so good performance from my live forwards.
I did some programming on this in the last couple days. Made two filters which perhaps are improvements. They are if nothing else two more switches to play with. I added an ADX filter on top of the one already there. And I added a DecisionValueFilter which taps into the logic of the EA and sets a limit on the minimum DecisionValue which it will allow to open. The idea being that a better quality signal results in a higher probability of winning. I'll post my upgraded version and you all can tell me what it does for you and what you think.
I may not have reached the full potential of what I could do with the settings already on the EA. I'm only beginning to understand them anyway...
I think I would tolerate 15% drawdowns if I could get that on a $350 initial deposit account. I've seen it do as low as 3% with larger initial deposits but I'm not working with a large account. I've got to find a way to make this build from the small account I have without wasting it.
The settings I am using are preset in this upgrade. I appreciate any suggestions you or others have to help me accomplish my goals. If I knew that I could get the performance live that I get from the backtester my anxieties would end.
A couple more things I have decided about this...
1- of the three current closing logics the pipsator is a piece of crap. I totally turned that one off, I tested all three of them independently and it always lost so I shut that one down. I found that leaving the other two on together is what does best and that is what this .htm report is of the two remaining closing strategies on.
2- I added another closing strategy which isn't exactly working right. My goal with it was to keep in trades which are with trends. When I did see it work in one test it made massive gains off single trades which it carried longer than the cyberia logic alone would have allowed. I moved something in the code and it stopped being as effective and I havn't tracked down how to get it working again yet. I'm just a novice developer. But I left it there because it's not hurting anything and I may or someone else might get it working again.
My decision value filter isn't working 100% either, not sure why it seems to allow trades it's told to block sometimes but it does.
The ADX filter seems to work perfectly but it doesn't make a big impact. It only blocks about 30 trades out of 400. Still I can assume those are or would be low probability trades so it's good to have them eliminated. I also left a piece of code at the bottom which is supposed print to the journal the time and Decision Value of valid orders for tracking purposes. I was trying to see what the DV's were of winers and losers to find what might be a good setting for the DVfilter.
I have seen that if I crank up the DV filter and use the ADX filter which was already on it...it can really reduce the number of overall trades And limit the number of losing or noisy trades...
it's just a couple more buttons to work with. One last thing I gave this version it's own magic number.
oops...one more last thing...
there is some code turned off in the DV filter which blocks values below zero. If that is on the other code which looks for a minimum user defined level doesn't seem to work, no idea why the two conflict. so i just tested both independently to see which one gave the greatest return and left it on and turned the other one off. The one that got turned off was the less than zero part. That might however be a safer way to fly. If you turn it all on you can get the drawdowns down to around 25 or 26%
I'll still look in to see how its going and make sugestions. I just got frusterated with every one asking for my settings and complaining to me when they don't work.
I posted several times as have you that every one has to come up with there own.
While your playing with the logic and code you can try to get it to use fractals in another way.
when a 4 hr fractal appears on the chart place a trade. instead of just using fractals to block a trade it should help initiate one. Once a 4 hr fractal appears indacating a short term top then it should sell at the top of the bar that triggered the fractal. or at the top of the next bar, using its own logic.
Fractals lag but a 4 hr fractal on a 1 hr chart should be able to get a 10 pip trade from it.
did that make sence?
I don't know anything about fractals yet. I still have much to learn.
I don't know anything about fractals yet. I still have much to learn.
here ya go....
http://www.investopedia.com/articles/trading/06/Fractals.asp
http://www.metaquotes.net/techanalysis/indicators/fractal
trading results
Hello all,
These are the results for my live account, first week of trading:
(Sorry, I don't have connection to my account right now, I'll post a statement at a later point)
Similar setup to my FXDD1 demo account with the difference of CCI and Pivot filters.
By turning off the filters, the EA traded 13 times on the live account compared to only 6 times on the demo account.
Again, my biggest losses occured during unblocked hours which should be blocked to begin with.
For the coming week I'll stick to Dave's settings and see what happends (thanks Dave).
Cyberia EA 1.85g on IBFX Live Account
-------------------------------------
Account started on October 16th, 2006
Trading Pair: USDJPY
EnableCCI=false, EnablePivot=false, ReverseIndex=3.82
ValuesPeriodCount=7, ValuesPeriodCountMax=7
StopLossIndex=2.5, AutoStopLossIndex=true
StaticStopLoss=15, EnableTrailingStop=false
10/16/06 to 10/20/06: USDJPY: 0 pips (10 wins, 5 losses)
Total: 0 pips (10 wins, 5 losses) 67% win ratio
AZBOfin