Settings Files
Optimization of EURUSD
11-18-2006
I am attempting to do this now. It will take several days to even get a prototype going. I will keep you up to date here on the progress of my work.
Anyone with settings they create, please post them here.
I promised a settings file, but I have been too busy.
This was from Fikko, I'm attempting to put it in the appropriate place.Hi there,
I conducted a backtest on that setting for the period 2006-05-01 to 2006-11-17.
Please note the results are very different, with max drawdown 75.88% and the profit is just 71126.89!
Using BreakEvenAfterPips=35 will get a far better result, but the max. drawdown is still 50.19%.
Prehaps the backtest for any new settings should be up to present, not only up to a month ago.
Regards
Wilson
Wilson, I noticed that your settings were using 3 lots on an account size of $10,000 to start with, and later using 22 lots. What was your maximum risk?
I pasted the settings submitted from Fikko, but seeing the lot sizes made me a little scared. 
Since that is 6 times the default maximum risk it does make sense the lot sizes would be larger, but that also means (as your log shows) any consecutive losses are extremely painful. It still managed to turn 10k into 80K+ in 6 months after very large losses. Reducing the risk by 1/6th (to the normal amount) would reduce the drawdown *substantially*. I don't think I could ever run maxrisk=0.3 myself.
Would you mind running the tests again over the same data with a few changes? Set risk to 0.05, update to 5.6.4, and change the Consecutive signals to numbers between 1 and 200? (200 = 200 consecutive ticks of confirming signals)
When you're to it, could you please also reduce the initial account size to $5,000? We agreed that all the testing will use it to avoid confusion in the results and I think it is a good idea.
Thanks 
Wilson, I noticed that your settings were using 3 lots on an account size of $10,000 to start with, and later using 22 lots. What was your maximum risk?
I pasted the settings submitted from Fikko, but seeing the lot sizes made me a little scared. Since that is 6 times the default maximum risk it does make sense the lot sizes would be larger, but that also means (as your log shows) any consecutive losses are extremely painful. It still managed to turn 10k into 80K+ in 6 months after very large losses. Reducing the risk by 1/6th (to the normal amount) would reduce the drawdown *substantially*. I don't think I could ever run maxrisk=0.3 myself.
... and change the Consecutive signals to numbers between 1 and 200? (200 = 200 consecutive ticks of confirming signals)
This number may need to be revisit, since Dagnar found error in his code. Please, see msg nr 901 and 905: http://www.forexfactory.com/forexforum/showthread.php?t=8820&page=61&pp=15
Mario
Wilson, I noticed that your settings were using 3 lots on an account size of $10,000 to start with, and later using 22 lots. What was your maximum risk?
I pasted the settings submitted from Fikko, but seeing the lot sizes made me a little scared. Since that is 6 times the default maximum risk it does make sense the lot sizes would be larger, but that also means (as your log shows) any consecutive losses are extremely painful. It still managed to turn 10k into 80K+ in 6 months after very large losses. Reducing the risk by 1/6th (to the normal amount) would reduce the drawdown *substantially*. I don't think I could ever run maxrisk=0.3 myself.
Hi Daraknor,
I used maximurisk=0.3, I just followed the figure the above post.
OK, I am running optimization for that (1-200), it will take about 30 minutes to complete as my computer told me. Will post here the best setting.
Regards
Wilson
This number may need to be revisit, since Dagnar found error in his code. Please, see msg nr 901 and 905: http://www.forexfactory.com/forexforum/showthread.php?t=8820&page=61&pp=15 Mario
I saw that too. I made sure the description here was accurate with "ticks" of symbols, instead of "bars of symbols" which is what the change would do. Using 3 bars on M15 would mean the market needs to stay in position for 45 minutes before we execute a trade. There is also no scaling for very volatile markets. I post about this in the coding section. Thanks for keeping a look out, i am not sure which way we will go. (ticks or bars) I did send a PM to the author almost immediately after he posted the fix. No reply yet - I am leaving the code as is for now and documenting the feature as it works.
Hi Daraknor,
I used maximurisk=0.3, I just followed the figure the above post.
OK, I am running optimization for that (1-200), it will take about 30 minutes to complete as my computer told me. Will post here the best setting.
Regards
WilsonI'm glad you followed the directions Following directions told us the directions are scary. Thanks again Wilson
Tp=84 Sl=84
I'm glad you followed the directions
Following directions told us the directions are scary. Thanks again Wilson Hi, I'm back.
Period between 2006-05-01 to 2006-11-17
Settings marked with (*) are from optimization.
TP=84 SL=84
Net Profit: 11035.45 MD: 2039.49 (11.10%) Total Trade:69 Short won(62.5%) Long won(67.57%)
TP=84 SL=84 TrailingStop=50(*) Signal_count=2(*)
Net Profit:16022.59 MD: 2456.40 (9.91%) Total Trade:86 Short won(75%) Long won(78.57%)
Hi all ;-)
just to say i tried to optimize Phoenix 5.6.03 (thanks to fine .pdf provided here ;-) for usdjpy pair , and found a profit factor of 2.71, DD 22.36%, and max profit 26356$.
Phoenix was in mode 1 , and best setup was with take profit =45 and StopLoss = 80 (witch is not so far from outofthebox settings ;-).
Trailing stop and breakevenPips must stay at 0, else profit factor drop down.
Will try these settings during next week on demo account , just on this pair.
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Anyone with settings they create, please post them here.
I promised a settings file, but I have been too busy.
This was from Fikko, I'm attempting to put it in the appropriate place.
Take Profit = 84
Stop Loss = 84
Trailing Stop = 35
Tested with risk=0.3, Data May-October 2006
This will increase the profit from USD 114,433 to USD 345,913 (3 times)
Increase the winning probability from 83% to 89%
Profit factor increased from 1.54 to 6.17
Maximum Drawdown reduced from 53% to 22%
I try to apply the same setting for GBPJPY, but it doesn't work better than your setting.
So this is a very good result. You can try it.
If other people want to do testing on this new setting, you must go into the code of Phoneix, Line 122 (USDJPY) and change TakeProfit, StopLoss and TrailingStop.