EAs under construction II | Page 9 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

trading system under construction 4 replies

TradingRightByListen ingAndSpeaking (Trblas) Under Construction! 5 replies

Eas under construction III 24 replies

EAs under construction 68 replies

Am i under under-capitalized? 6 replies

Attachments: EAs under construction II
Exit Attachments

EAs under construction II

  • #161
  • Jul 14, 2010 4:36pm Jul 14, 2010 4:36pm
Quoting jiva34
You see how we are doing this....
Thank you for your work. Your time and effort..
If you post the charts showing where the losses were I can look at them and maybe suggest some changes.
What I see right away just from the data:

GBPUSD (Great Britain Pound vs US...
Hi Jiva,

just in short: I am doing all testing on Alpari uk live and it is a standard routine for me to download the 1m from history centre. I am aware that modelling quality is sucks and usually I am not relying on anything which is not 90%. But what shall I do, this ea always has the same modelling quality which is n/a in alpari uk live. It must have something to do with the brokers.....in my view

The bad trades are 8th and 29th of September in 2009 and 22nd of January and 9th of April in 2010
  • #162
  • Edited 5:03pm Jul 14, 2010 4:50pm | Edited 5:03pm
Quoting Haishang
Hi Jiva,

just in short: I am doing all testing on Alpari uk live and it is a standard routine for me to download the 1m from history centre. I am aware that modelling quality is sucks and usually I am not relying on anything which is not 90%. But what shall I do, this ea always has the same modelling quality which is n/a in alpari uk live. It must have something to do with the brokers.....in my view

The bad trades are 8th and 29th of September in 2009 and 22nd of January and 9th of April in 2010
we're working on resolving the modeling quality issue.:nerd:
For now we can just ignore it and keep running tests.
FXDD is giving 90% and the graphs are very similar to Alpari. So no big devergence there.
I ran the settings again with some more tweaks.
FXDD:
1 bad trade shown below. I will run it again with another tweak.

SymbolGBPUSD (Great Britain Pound vs. United States Dollar) Period1 Hour (H1) 2009.07.01 00:00 - 2010.07.12 23:00 (2009.07.01 - 2010.07.13) ModelEvery tick (the most precise method based on all available least timeframes) ParametersiMagicNumber=532753; GMTOffset=3; TradeStartHour=5; TradeEndHour=9; Lots=0.1; TakeProfit1=20; TrailingStop1=80; TakeProfit2=50; TrailingStop2=100; StopLoss=150; PendingPips=40; ExpireCandles=10; DelayOrder2=20; CheckSlope=false; DSS_name="Robby DSS Bressert Colored with alert v1.1"; DSS_EMA=8; DSS_stoch=13; Use_T3MA=false; T3MA_name="T3MA"; T3MA_Periods=400; T3MA_VolumeFactor=0.7; Channel_EMAPeriod=38; Channel_EMAShift=0; ShowComment=true; EXP_MA_1="-------- MA settings --------"; EXP_MA_2="-- Method 0:SMA, 1:EMA, 2:SMMA, 3:LWMA"; EXP_MA_3="-- Timeframe 0:current or minutes"; Use_MA1=false; MA1_Timeframe=0; MA1_Period=0; MA1_Method=1; MA1_Shift=0; Use_MA2=false; MA2_Timeframe=0; MA2_Period=0; MA2_Method=1; MA2_Shift=0; Use_MA3=false; MA3_Timeframe=0; MA3_Period=0; MA3_Method=1; MA3_Shift=0; Use_MA4=false; MA4_Timeframe=0; MA4_Period=0; MA4_Method=1; MA4_Shift=0; Use_MA5=false; MA5_Timeframe=0; MA5_Period=0; MA5_Method=0; MA5_Shift=0;
Bars in test6273Ticks modelled10624353Modelling quality90.00% Mismatched charts errors2




Initial deposit10000.00



Total net profit839.97Gross profit1140.39Gross loss-300.41 Profit factor3.80Expected payoff15.00

Absolute drawdown95.00Maximal drawdown364.21 (3.32%)Relative drawdown3.32% (364.21)
Total trades56Short positions (won %)28 (92.86%)Long positions (won %)28 (96.43%)
Profit trades (% of total)53 (94.64%)Loss trades (% of total)3 (5.36%) Largestprofit trade50.20loss trade-150.16 Averageprofit trade21.52loss trade-100.14 Maximumconsecutive wins (profit in money)43 (890.49)consecutive losses (loss in money)2 (-300.11) Maximalconsecutive profit (count of wins)890.49 (43)consecutive loss (count of losses)-300.11 (2) Averageconsecutive wins18consecutive losses2
Attached Image(s) (click to enlarge)
Click to Enlarge Name: v5gbpusd1hpe40tp20ch38.gif Size: 8 KB
Click to Enlarge Name: bad2.gif Size: 38 KB
  • #163
  • Jul 14, 2010 4:59pm Jul 14, 2010 4:59pm
I'm set up live account for EURCHF 4h. Looks like a short setting up for Hong Kong open.
Attached Image (click to enlarge)
Click to Enlarge Name: hong kong.gif Size: 30 KB
  • #164
  • Jul 14, 2010 5:22pm Jul 14, 2010 5:22pm
Quoting jiva34
I'm set up live account for EURCHF 4h. Looks like a short setting up for Hong Kong open.
Thanks a lot for sharing this great ea, Jiva! Good luck tonight...I am not as far as putting anything on testing but hopefully will be end of the week.....we are running Alpari uk live and Alpari uk demo.....just fyi....hopefully will be able to find some things....
  • #165
  • Jul 14, 2010 6:45pm Jul 14, 2010 6:45pm
Quoting Haishang
Thanks a lot for sharing this great ea, Jiva! Good luck tonight...I am not as far as putting anything on testing but hopefully will be end of the week.....we are running Alpari uk live and Alpari uk demo.....just fyi....hopefully will be able to find some things....
I slaved many hours on the GBPUSD 1h to find something that worked at the London open, nothing I was thrilled about.
Now you have hit on something first try!!!
I am very glad as the saying goes, "that you are on my side"....
The amazing thing about this ea is that the basic stragey is very very simple and sound. Based on price actions which always the best indicator there is. A breach of a hi lo channel.
But the ea can be tweaked and adjusted ad infinitum to almost any market condition and any pair. There are great functions available and they are optional so they can be used or not used as the situation dictate.

It's always a trade off, reduced risk means paying up for it by reducing profit trades too.
But it is well worth it, risk is very expensive and runs havoc on the nervous systems too.

so about the versatility of functions that can be applied..
Interesting that Taiyakixz and myself have gone back a forth a bit about using moving averages. Mostly I have argued they do not help. But I see in some cases they really can make a difference by filtering out bad entries even if they do filter out some good entries too. It is definitely worth it.

So here's is what I have come up using Haishang's idea... and playing a bit with the settings. Nice!!!
Thank you Haishang.

FXDD

SymbolGBPUSD (Great Britain Pound vs. United States Dollar) Period1 Hour (H1) 2009.07.01 00:00 - 2010.07.12 23:00 (2009.07.01 - 2010.07.13) ModelEvery tick (the most precise method based on all available least timeframes) ParametersiMagicNumber=532753; GMTOffset=3; TradeStartHour=5; TradeEndHour=9; Lots=0.1; TakeProfit1=20; TrailingStop1=80; TakeProfit2=50; TrailingStop2=100; StopLoss=150; PendingPips=42; ExpireCandles=10; DelayOrder2=0; CheckSlope=false; DSS_name="Robby DSS Bressert Colored with alert v1.1"; DSS_EMA=8; DSS_stoch=13; Use_T3MA=false; T3MA_name="T3MA"; T3MA_Periods=0; T3MA_VolumeFactor=0.7; Channel_EMAPeriod=38; Channel_EMAShift=0; ShowComment=true; EXP_MA_1="-------- MA settings --------"; EXP_MA_2="-- Method 0:SMA, 1:EMA, 2:SMMA, 3:LWMA"; EXP_MA_3="-- Timeframe 0:current or minutes"; Use_MA1=true; MA1_Timeframe=0; MA1_Period=200; MA1_Method=0; MA1_Shift=0; Use_MA2=false; MA2_Timeframe=0; MA2_Period=0; MA2_Method=1; MA2_Shift=0; Use_MA3=false; MA3_Timeframe=0; MA3_Period=0; MA3_Method=1; MA3_Shift=0; Use_MA4=false; MA4_Timeframe=0; MA4_Period=0; MA4_Method=1; MA4_Shift=0; Use_MA5=false; MA5_Timeframe=0; MA5_Period=0; MA5_Method=0; MA5_Shift=0;
Bars in test6273Ticks modelled10624353Modelling quality90.00% Mismatched charts errors2



Initial deposit10000.00



Total net profit670.00Gross profit670.00Gross loss-0.00 Profit factor
Expected payoff20.94

Absolute drawdown64.00Maximal drawdown234.00 (2.25%)Relative drawdown2.25% (234.00)
Total trades32Short positions (won %)18 (100.00%)Long positions (won %)14 (100.00%)
Profit trades (% of total)32 (100.00%)Loss trades (% of total)0 (0.00%) Largestprofit trade50.00loss trade-0.00 Averageprofit trade20.94loss trade-0.00 Maximumconsecutive wins (profit in money)32 (670.00)consecutive losses (loss in money)0 (-0.00) Maximalconsecutive profit (count of wins)670.00 (32)consecutive loss (count of losses)-0.00 (0) Averageconsecutive wins32consecutive losses0
Attached Image (click to enlarge)
Click to Enlarge Name: v5gbpusd1hpe40tp20ch38 sma200.gif Size: 8 KB
  • #166
  • Edited 8:28pm Jul 14, 2010 8:17pm | Edited 8:28pm
just for the heck of it I changed the trading time from just Europe open to anytime. And the results show that trading only before Europe open is much better.
So it appears Haishang's idea has some good merit.

FXDD all day not just Europe open: just Europe open better results.

SymbolGBPUSD (Great Britain Pound vs. United States Dollar) Period1 Hour (H1) 2009.07.01 00:00 - 2010.07.12 23:00 (2009.07.01 - 2010.07.13) ModelEvery tick (the most precise method based on all available least timeframes) ParametersiMagicNumber=532753; GMTOffset=3; TradeStartHour=1; TradeEndHour=24; Lots=0.1; TakeProfit1=20; TrailingStop1=80; TakeProfit2=50; TrailingStop2=100; StopLoss=150; PendingPips=50; ExpireCandles=10; DelayOrder2=0; CheckSlope=false; DSS_name="Robby DSS Bressert Colored with alert v1.1"; DSS_EMA=8; DSS_stoch=13; Use_T3MA=false; T3MA_name="T3MA"; T3MA_Periods=0; T3MA_VolumeFactor=0.7; Channel_EMAPeriod=38; Channel_EMAShift=0; ShowComment=true; EXP_MA_1="-------- MA settings --------"; EXP_MA_2="-- Method 0:SMA, 1:EMA, 2:SMMA, 3:LWMA"; EXP_MA_3="-- Timeframe 0:current or minutes"; Use_MA1=true; MA1_Timeframe=0; MA1_Period=200; MA1_Method=0; MA1_Shift=0; Use_MA2=false; MA2_Timeframe=0; MA2_Period=0; MA2_Method=1; MA2_Shift=0; Use_MA3=false; MA3_Timeframe=0; MA3_Period=0; MA3_Method=1; MA3_Shift=0; Use_MA4=false; MA4_Timeframe=0; MA4_Period=0; MA4_Method=1; MA4_Shift=0; Use_MA5=false; MA5_Timeframe=0; MA5_Period=0; MA5_Method=0; MA5_Shift=0;
Bars in test6273Ticks modelled10624353Modelling quality90.00% Mismatched charts errors2



Initial deposit10000.00



Total net profit448.91Gross profit1649.84Gross loss-1200.94 Profit factor1.37Expected payoff5.10

Absolute drawdown858.31Maximal drawdown918.31 (9.13%)Relative drawdown9.13% (918.31)
Total trades88Short positions (won %)46 (91.30%)Long positions (won %)42 (85.71%)
Profit trades (% of total)78 (88.64%)Loss trades (% of total)10 (11.36%) Largestprofit trade50.00loss trade-150.16 Averageprofit trade21.15loss trade-120.09 Maximumconsecutive wins (profit in money)35 (859.84)consecutive losses (loss in money)2 (-300.31) Maximalconsecutive profit (count of wins)859.84 (35)consecutive loss (count of losses)-300.31 (2) Averageconsecutive wins13consecutive losses2
Attached Image (click to enlarge)
Click to Enlarge Name: v5gbpusd1hpe40tp20ch38 sma200 1to24.gif Size: 8 KB
  • #167
  • Jul 14, 2010 8:30pm Jul 14, 2010 8:30pm
Quoting jiva34
I'm set up live account for EURCHF 4h. Looks like a short setting up for Hong Kong open.
sell stop triggered.. it's only a trade if it fills...
Attached Image (click to enlarge)
Click to Enlarge Name: eufchf.gif Size: 20 KB
  • #168
  • Jul 15, 2010 9:36am Jul 15, 2010 9:36am
step by step...we are nearing

My settings are in the meantime the following:

t/p1 30
trailing s/l 1 30
t/p2 50
trailing s/l 2 50
slope false
s/l 150

p/l 1230 with 0.1 lots
trades total 78
winning trades 73

but i still have the issue that the t/p2 gets immediately modified to the entry level in many cases....How can I modify the setting for this occurance? I thought by adjusting the trailing s/l this could have been done but I assume it has something to do with the expiry candles or pending pips....I will try and find out, in the meantime if you have an easy answer, i would be so grateful as it would speed up my optimizing.

Attached as usual:

GBPUSD (Great Britain Pound vs US Dollar) Period1 Hour (H1) 2009.07.01 00:00 - 2010.07.13 23:00 (2009.07.01 - 2010.07.14) ModelEvery tick (the most precise method based on all available least timeframes) ParametersiMagicNumber=8435501; GMTOffset=2; TradeStartHour=5; TradeEndHour=9; Lots=0.1; TakeProfit1=30; TrailingStop1=30; TakeProfit2=50; TrailingStop2=30; StopLoss=150; PendingPips=40; ExpireCandles=10; DelayOrder2=20; CheckSlope=false; DSS_name="Robby DSS Bressert Colored with alert v1.1"; DSS_EMA=8; DSS_stoch=13; Use_T3MA=false; T3MA_name="T3MA"; T3MA_Periods=8; T3MA_VolumeFactor=0.7; Channel_EMAPeriod=38; Channel_EMAShift=0; ShowComment=false; EXP_MA_1="-------- MA settings --------"; EXP_MA_2="-- Method 0:SMA, 1:EMA, 2:SMMA, 3:LWMA"; EXP_MA_3="-- Timeframe 0:current or minutes"; Use_MA1=false; MA1_Timeframe=0; MA1_Period=5; MA1_Method=1; MA1_Shift=0; Use_MA2=false; MA2_Timeframe=0; MA2_Period=12; MA2_Method=1; MA2_Shift=0; Use_MA3=false; MA3_Timeframe=0; MA3_Period=20; MA3_Method=1; MA3_Shift=0; Use_MA4=false; MA4_Timeframe=0; MA4_Period=55; MA4_Method=1; MA4_Shift=0; Use_MA5=false; MA5_Timeframe=0; MA5_Period=197; MA5_Method=0; MA5_Shift=0;
Attached Image (click to enlarge)
Click to Enlarge Name: JIVA Dragon EA DSST3MA v5.0a GBPUSD 1h.gif Size: 6 KB
  • #169
  • Jul 15, 2010 10:12am Jul 15, 2010 10:12am
Quoting Haishang
step by step...we are nearing

My settings are in the meantime the following:

t/p1 30
trailing s/l 1 30
t/p2 50
trailing s/l 2 50
slope false
s/l 150

p/l 1230 with 0.1 lots
trades total 78
winning trades 73

but i still have the issue that the t/p2 gets immediately modified to the entry level in many cases....How can I modify the setting for this occurance? I thought by adjusting the trailing s/l this could have been done but I assume it has something to do with the expiry candles or pending pips....I will try and find out, in the meantime if...
1. Set one of the moving averages for trend confirmation. That will probably eliminate the 2 bloopers...Won't that be nice...
sma 200 can do the job most of the time...
Use_MA1=true; MA1_Timeframe=0; MA1_Period=200; MA1_Method=0


2. you do not need a delay 20 seconds on the second order, so set that to 0.. That is a special function I use for a unique situation that most traders won't ever encounter. That can adversely affect results..
DelayOrder2=0

Run the test again and let's see.. Then we can see about "fiddling" with the other parameters like stops and limits and pending amounts etc..
Good luck..
  • #170
  • Jul 15, 2010 10:35am Jul 15, 2010 10:35am
Here are the 2 alternatives:

First what they have in common:

T/P 1: 30
trailing S/L 1: 30
T/P 2: 50
Trailing S/L: 2 0
S/L: 150
Pending Pips: 40
expiry candles: 10
Delay 0rder: 0
check slope: false

The differences are the following
alternative 1: no ma
alternative 2: 197 ma


outcome:
alternative 1:
P/L 1555
trades 78
won trades 72
lost trades 6 (out of which 4 are forming basically the entire loss)

alternative 2:
P/L 1008
trades 46
won trades 42
lost trades 4


The remaining topic is that for order number 2 the 50 pips t/p is immediately modified to entry point level and hence the chance to make 50 pips is easily given away.....This makes me scrutch my head because I cannot find a solution for this baby.....
Attached Image(s) (click to enlarge)
Click to Enlarge Name: JIVA Dragon EA DSST3MA v5.0a GBPUSD 1h.gif Size: 6 KB
Click to Enlarge Name: JIVA Dragon EA DSST3MA v5.0b GBPUSD 1h.gif Size: 6 KB
  • #171
  • Jul 15, 2010 1:14pm Jul 15, 2010 1:14pm
Quoting Haishang
Here are the 2 alternatives:

First what they have in common:

T/P 1:...
Open the charts and look at the losing trades. They can tell you what caused the loss.
would more pending pips have helped avoid a big loss? If so, try adding just a few more pips to pending order. Make it 42 instead of 40. If you study the trade on the chart it can tell you by how much it missed not getting filled and simply expiring. It's these extreme cases that cause the problems. Obviously they can't be ignored.

minimizing risk always comes at a cost. Like insurance. I prefer to pay for insurance.
adding the moving average means some of the big bad trades can be avoided but undoubtedly it will be at the cost of missing a few good ones.
The dynamics are that of balancing..
http://t1.gstatic.com/images?q=tbn:Y...ng-act-001.jpg

It may be possible to catch some really big runs by not putting a limit on the second order. Sometimes the 2nd order will fall back to break even for 0 gain. But 0 gain means more or less 0 loss also.
By putting a limit on the second order it cannot catch a big move, and these big moves happen all the time esp in GBPUSD.
The ea is designed so the 2nd position has different ways of exiting.

1. before 1st position reaches its limit, in which case both positions will stop out. Bad trade..
a. taking out the preset fixed stop

2. after 1st position reaches it's limit
a. a trailing stop set by user.
or
b. falling back to breakeven for 0 loss/0 gain
c. for a long: closing below the channel
for a short: closing above the channel

So if no limit is set and 1st position is reached, the 2nd position could catch a big run without any stop set. Kind of all or nothing. Either it falls back to break even 0 loss/0 gain or it catches the big one and exits when a candle closes that breaches the channel. The channel itself acting as trailing stop...
A trailing stop could be set for the 2nd position that was fairly large (100?) to allow for retracing but to also not give back too much when the run peaks and retraces...
A move of 200 or 300 or more can be caught that way. So giving back a few 50 pips for a few breakevens might be worth it.

Did you follow that?
  • #172
  • Jul 15, 2010 4:25pm Jul 15, 2010 4:25pm
Quoting jiva34
Open the charts and look at the losing trades. They can tell you what caused the loss.
would more pending pips have helped avoid a big loss? If so, try adding just a few more pips to pending order. Make it 42 instead of 40. If you study the trade on the chart it can tell you by how much it missed not getting filled and simply expiring. It's these extreme cases that cause the problems. Obviously they can't be ignored.

minimizing risk always comes at a cost. Like insurance. I prefer to pay for insurance.
adding the moving average...
it was quite a long night for me and I am not sure if I can follow you, Jiva. I try my best but probably it will take until tomorrow that you get a qualified answer.....is that ok?
  • #173
  • Jul 15, 2010 4:26pm Jul 15, 2010 4:26pm
and sorry for the late reply....we should be more in tune in 2 weeks time when we are back in shanghai
  • #174
  • Jul 15, 2010 5:15pm Jul 15, 2010 5:15pm
Quoting Haishang
and sorry for the late reply....we should be more in tune in 2 weeks time when we are back in shanghai
  • #175
  • Jul 17, 2010 3:48pm Jul 17, 2010 3:48pm
Subscribing so we don't have to private

So only eurchf yes??
  • #176
  • Jul 17, 2010 4:41pm Jul 17, 2010 4:41pm
Quoting stifland
Subscribing so we don't have to private

So only eurchf yes??
Greetings..
Have been reading the Raygun site, Old dog...
Thanks for mentioning it..
We are using pending orders to try to avoid the consolidation areas..
And the buy at the top and sell at the bottom syndrome..
There are a couple new modifications for limiting losses and improving entries on the drawing table also.
Taiyakixz will be tied up for a few weeks. So they will remain on the table for a while.
When they come into effect it will greatly improve the results.
For now we continue to study how the trades are being placed given different pairs and settings.
This is not a ready made system. It is under construction and development.
When it is ready we will go to the trading systems forum.

http://www.acme.com/jef/construction/ani-gear.gif
  • #177
  • Jul 17, 2010 4:52pm Jul 17, 2010 4:52pm
Quoting stifland
Subscribing so we don't have to private

So only eurchf yes??
If you want to you could run the setting for EUCHF. I have run it (7/1/09 to 7/1/10) on AAAFX and FXDD and Alpari all with same results. 90% quality.
  • #178
  • Jul 17, 2010 5:23pm Jul 17, 2010 5:23pm
Quoting jiva34
If you want to you could run the setting for EUCHF. I have run it (7/1/09 to 7/1/10) on AAAFX and FXDD and Alpari all with same results. 90% quality.

I am running test now with settings in post 1 to see what i get on my broker.
And to verify, t3 and all MA's are on False?
I will post results. TY Jiva.
  • #179
  • Edited 7:02pm Jul 17, 2010 6:39pm | Edited 7:02pm
Quoting stifland
I am running test now with settings in post 1 to see what i get on my broker.
And to verify, t3 and all MA's are on False?
I will post results. TY Jiva.
sorry, it's full time job just keeping the first page updated. I need to change those setting, they were for v4.1 which didnt have moving average options..
for v5: channel 2, sma 400, make sure slope check is false...
which broker?
  • #180
  • Jul 17, 2010 9:39pm Jul 17, 2010 9:39pm
Quoting jiva34
sorry, it's full time job just keeping the first page updated. I need to change those setting, they were for v4.1 which didnt have moving average options..
for v5: channel 2, sma 400, make sure slope check is false...
which broker?

so only MA 1 at 400?

I am using those original settings on v5--still running, but I don't see any losses yet. do you post results just with screen capture?

Ikon_Royal
0 traders viewing now