is it me or FAP20 when backtesing sucks badly?

lukx · #1 12-07-2008, 02:46 AM

I just backtested fap20 through whole November and only losses. Both long term FAP and scalper. 16 was good but 20 sucks badly. I guess I will stick to good old FAP

alan · #2 12-07-2008, 03:44 AM

Did you make sure you ran it on the proper time frame and what was your modeling quality?

I hope you do realize that you have to test the scalper and long term strategy on different time frame.

lukx · #3 12-07-2008, 04:53 AM

oh sorry my mistake I forgot to set GMT offset on fap20

BUt...
I run backtest on FAP 16 and 20 default settings exept moneymanagememnt off on 20 and lots set to 0.2 on both. (modeling every tick)

here are results:
FAP20
Strategy Tester Report
FapTurbo20
FXDD-MT4 Live Server (Build 220)

Symbol EURCHF (Euro vs. Swiss Franc)
Period 15 Minutes (M15) 2008.11.03 00:00 - 2008.11.28 22:45 (2008.11.01 - 2008.11.30)
Model Every tick (the most precise method based on all available least timeframes)
Parameters ____Scalper___________="--- Scalper EURCHF, EURGPB, GBPCHF or USDCAD M15 ---"; UseScalperStrategy=true; Scalper_Lots=0.2; Scalper_UseMM=false; Scalper_LotsRiskReductor=5; Scalper_MaxLots=10; Scalper_GMTOffset=2; Scalper_StealthMode=true; Scalper_UseCustomLevels=true; Scalper_RelaxHours=0; Scalper_SimpleHeightFilter=true; Scalper_TrendFilter=true; Scalper_TradeFriday=false; Scalper_OneTrade=0; Scalper_UseFilterMA=false; Scalper_PeriodFilterMA=100; Scalper_PriceFilterMA=0; Scalper_MethodFilterMA=0; Scalper_ExpertComment="FapTurboScalper"; ____FapTurbo__________="---------- fapturbo EUR/USD M1 ----------"; FapTurbo_Lots=0.1; FapTurbo_LotsRiskReductor=1; FapTurbo_MaxOrders=2; FapTurbo_MaxLots=10; FapTurbo_StopTime=9; FapTurbo_aaa=35; FapTurbo_bbb=48; FapTurbo_TakeProfit=140; FapTurbo_StopLoss=500; FapTurbo_TrailingStop=0; FapTurbo_DurationInHours=0; FapTurbo_CloseAfterXmonths=1; FapTurbo_PeriodMALarge=100; FapTurbo_PeriodMASmall=30; FapTurbo_PriceMA_0_6=0; FapTurbo_TypeMA_0_3=0; FapTurbo_ShiftMALarge=3; FapTurbo_ShiftMASmall=1; FapTurbo_LookForDays=2; FapTurbo_CriticalDays=45; FapTurbo_RelaxDays=0; FapTurbo_HedgingLotReducer=0; FapTurbo_Hedging=0; FapTurbo_TradeNFP=1; FapTurbo_TradeFriday=1; FapTurbo_Prudent=0; FapTurbo_SymAlligatorOnCritica=1; FapTurbo_FixedDirection=1; FapTurbo_ClsLsrOnMrktChnge=0; FapTurbo_AlwaysTrade=0; FapTurbo_LowLot=0.1; FapTurbo_TooGoodToBeTrue=50; FapTurbo_PrudentPeriod=30; FapTurbo_StartWorkTimeHour=0; FapTurbo_StartWorkTimeMin=0; FapTurbo_EndWorkTimeHour=0; FapTurbo_EndWorkTimeMin=0; FapTurbo_OneTrade=0; FapTurbo_SpanGator=0.5; FapTurbo_SlipPage=3; FapTurbo_ExpertComment="FapTurboEA"; ____Other_Parameters__="-----------------------------------------"; MagicNumber=12464336; ColorBuy=Blue; ColorSell=Red; WriteLog=true; WriteDebugLog=false; PrintLogOnChart=true; KEY="012345";

Bars in test 2253 Ticks modelled 327039 Modelling quality n/a
Mismatched charts errors 162

Initial deposit 10000.00
Total net profit 399.44 Gross profit 408.41 Gross loss -8.97
Profit factor 45.51 Expected payoff 7.99
Absolute drawdown 27.84 Maximal drawdown 72.01 (0.72%) Relative drawdown 0.72% (72.01)

Total trades 50 Short positions (won %) 24 (95.83%) Long positions (won %) 26 (92.31%)
Profit trades (% of total) 47 (94.00%) Loss trades (% of total) 3 (6.00%)
Largest profit trade 16.36 loss trade -4.91
Average profit trade 8.69 loss trade -2.99
Maximum consecutive wins (profit in money) 18 (142.49) consecutive losses (loss in money) 1 (-4.91)
Maximal consecutive profit (count of wins) 142.49 (18) consecutive loss (count of losses) -4.91 (1)
Average consecutive wins 12 consecutive losses 1

lukx · #4 12-07-2008, 04:54 AM

and fap16:

Strategy Tester Report
FAPTURBO16
FXDD-MT4 Live Server (Build 220)

Symbol EURCHF (Euro vs. Swiss Franc)
Period 15 Minutes (M15) 2008.11.03 00:00 - 2008.11.28 22:45 (2008.11.01 - 2008.11.30)
Model Every tick (the most precise method based on all available least timeframes)
Parameters ____Scalper___________="--- Scalper EURCHF, EURGPB, GBPCHF or USDCAD M15 ---"; UseScalperStrategy=true; Scalper_Lots=0.2; Scalper_UseMM=false; Scalper_LotsRiskReductor=1.5; Scalper_MaxLots=10; Scalper_GMTOffset=2; Scalper_StealthMode=true; Scalper_UseCustomLevels=true; Scalper_RelaxHours=0; Scalper_SimpleHeightFilter=true; Scalper_TrendFilter=true; Scalper_TradeFriday=false; Scalper_OneTrade=0; Scalper_UseFilterMA=false; Scalper_PeriodFilterMA=100; Scalper_PriceFilterMA=0; Scalper_MethodFilterMA=0; Scalper_ExpertComment="FapTurboScalper"; ____FapTurbo__________="---------- fapturbo EUR/USD M1 ----------"; FapTurbo_Lots=0; FapTurbo_LotsRiskReductor=1; FapTurbo_MaxOrders=5; FapTurbo_MaxLots=10; FapTurbo_StopTime=9; FapTurbo_aaa=35; FapTurbo_bbb=48; FapTurbo_TakeProfit=140; FapTurbo_StopLoss=500; FapTurbo_TrailingStop=0; FapTurbo_DurationInHours=0; FapTurbo_CloseAfterXmonths=1; FapTurbo_PeriodMALarge=100; FapTurbo_PeriodMASmall=30; FapTurbo_PriceMA_0_6=0; FapTurbo_TypeMA_0_3=0; FapTurbo_ShiftMALarge=3; FapTurbo_ShiftMASmall=1; FapTurbo_LookForDays=2; FapTurbo_CriticalDays=45; FapTurbo_RelaxDays=0; FapTurbo_HedgingLotReducer=0; FapTurbo_Hedging=0; FapTurbo_TradeNFP=1; FapTurbo_TradeFriday=1; FapTurbo_Prudent=0; FapTurbo_SymAlligatorOnCritica=1; FapTurbo_FixedDirection=1; FapTurbo_ClsLsrOnMrktChnge=0; FapTurbo_AlwaysTrade=0; FapTurbo_LowLot=0.1; FapTurbo_TooGoodToBeTrue=50; FapTurbo_PrudentPeriod=30; FapTurbo_StartWorkTimeHour=0; FapTurbo_StartWorkTimeMin=0; FapTurbo_EndWorkTimeHour=0; FapTurbo_EndWorkTimeMin=0; FapTurbo_OneTrade=0; FapTurbo_SpanGator=0.5; FapTurbo_SlipPage=3; FapTurbo_ExpertComment="FapTurboEA"; ____Other_Parameters__="-----------------------------------------"; MagicNumber=231313; ColorBuy=Blue; ColorSell=Red; WriteLog=true; WriteDebugLog=false; PrintLogOnChart=true; KEY="205233224";
Bars in test 2905 Ticks modelled 325211 Modelling quality 90.00%
Mismatched charts errors 0
Initial deposit 1000.00
Total net profit 405.99 Gross profit 414.96 Gross loss -8.97
Profit factor 46.24 Expected payoff 8.29
Absolute drawdown 16.38 Maximal drawdown 72.00 (5.16%) Relative drawdown 5.34% (66.17)
Total trades 49 Short positions (won %) 25 (96.00%) Long positions (won %) 24 (91.67%)
Profit trades (% of total) 46 (93.88%) Loss trades (% of total) 3 (6.12%)
Largest profit trade 16.36 loss trade -4.91
Average profit trade 9.02 loss trade -2.99
Maximum consecutive wins (profit in money) 18 (142.49) consecutive losses (loss in money) 1 (-4.91)
Maximal consecutive profit (count of wins) 142.49 (18) consecutive loss (count of losses) -4.91 (1)
Average consecutive wins 12 consecutive losses 1

lukx · #5 12-07-2008, 04:55 AM

so 16 was few $ better

alan · #6 12-07-2008, 07:37 AM

Hmmm...weird, there should not be any difference in teh quality of the backtest.

You fapturbo 20 test is worthless beacuse you have waay too many mistached chart errors and also you modeling quality is "n/a"...

Try to get it to 90% modeling quality like the fapturbo 16 test for a more meaningful comparison

lukx · #7 12-07-2008, 09:27 AM

Quote:

Originally Posted by alan

Try to get it to 90% modeling quality like the fapturbo 16 test for a more meaningful comparison

and how i can achieve that? I'm new to backtesting so sorry if my question is stupid

alan · #8 12-07-2008, 09:06 PM

Quote:

Originally Posted by lukx

and how i can achieve that? I'm new to backtesting so sorry if my question is stupid

No lukx, it's not a stupid question. Ok, first thing you gotta make sure is that you have accurate and complete historic price data. Go to Tools -> History Center - select the pair you wish to backtest and click on "Download"

As a shortcut you could also press F2

Once the download is complete you can go through the usual backtesting procedure, but remember to use the highest modeling quality possible.

Also I would HIGHLY recommend that you do NOT do your backtesting on a LIVE account because your broker's live price feed can pollute your history data. I typically setup a MT4 install JUST for backtesting and I make sure that I DO NOT connect to the broker at all or load up a chart. I just download historic price data and run the strategy tester.

azdave · #9 12-07-2008, 11:43 PM

For what it's worth, my v20 backtesting results on anything other than EUR/GBP were pretty dismal. For some reason, it seemed to close losers pretty quickly on that pair while it rode others out to pretty big losses on the other pairs. Until they get the money management thing fixed and it starts trading correctly (instead of .01 lots regardless of account balance), I'm just going to try and calculate the proper lot size for my account. We'll see how it goes I guess.

TwoFeathers · #10 12-12-2008, 03:15 PM

I too get worse results for FAP23 when backtesting, but unlike you, mine are much worse. From reading replies on this forum, I'm starting to think it's because I have an account leverage of 1:500. What I do know is that the lot sizes are different, using MM (or LRR) are different to that of FAP16.

I have backtested quite extensively and found that results, using Alpari, are similar for EURCHF on both live and demo accounts, significantly different for USDCAD between live and demo, and absolutely completely different on EURGBP between live and demo.