Trading the 5 Min Bar

• 02/18/2018 at 7:56 PM #63057
  Aloysius

  Hi Grahal,

  I’m just searching for a winning strategy in live! and for the moment I found nothing… So here is the backtest in 5mn TF with your proposition of TP 5 and SL 15. I’ve tried an optimisation with lots of other variables but no configuration brings gain. I will test other possibilities and post here if i find something good. But I am very careful, because all the promising backtests I’ve seen have lead to losses in demo or real.

  

  Aloysius

  Participant

  Topics: 1

  Replies: 44

  Been thanked: 16 times
  02/18/2018 at 8:13 PM #63061
  GraHal

  Hi Aloysius the TP 5 and SL 15 I proposed were relevant only for the period I displayed / optimised over …  10,000 bars (and maybe going forward for weeks / months until there is an observed deviation from % winning trades and Gain to  Loss ratio etc following optimisation)

  Also if you backtested Gertrade version then you have backtested a 28 Sec TF version over 200,00 bars @ 5 Min TF?

  Attached results  of original Roberto 5 Min version over 100,000 bars @ £1 per point. I have only optimised TP, SL and the Trailing Stop Variables as below.

  I am considering that Algos on this Thread may need optimising every 3 months / 5000 bars.

  Cheers
  GraHal

  //------------------------------------------------------------------------- // Klinger-Rvi-Lsma DAX 5 min //------------------------------------------------------------------------- DEFPARAM CumulateOrders = False DEFPARAM FlatBefore = 090000 //no trades before 09:00:00 DEFPARAM FlatAfter = 213000 //no trades after 21:30:00 ONCE nLots = 1 //number of LOTs traded ONCE TP = 20 //23 pips Take Profit ONCE SL = 16 //16 pips Stop Loss RviVal, RviSignal = CALL "RVI by John Ehlers"[7] //7 KlingerVal, KlingerTrigger = CALL "Klinger oscillator"[25,44,55,1] //25,44,55,1 (ema) LeastSquareEMA = CALL "ELSMA - Least Square EMA"[11,4] //11,4 //*************************************************************************************** IF LongOnMarket THEN IF close < LeastSquareEMA THEN SELL AT MARKET //Exit LONGs when MACD reverses southwards ENDIF ENDIF IF ShortOnMarket THEN IF close > LeastSquareEMA THEN EXITSHORT AT MARKET //Exit SHORTs when MACD reverses northwards ENDIF ENDIF //*************************************************************************************** trailingstart = 20//10 trailing will start @trailinstart points profit trailingstep = 2 //13 trailing step to move the "stoploss" //reset the stoploss value IF NOT ONMARKET THEN newSL=0 ENDIF //manage long positions IF LONGONMARKET THEN //first move (breakeven) IF newSL=0 AND close-tradeprice(1)>=trailingstart*pipsize THEN newSL = tradeprice(1)+trailingstep*pipsize ENDIF //next moves IF newSL>0 AND close-newSL>=trailingstep*pipsize THEN newSL = newSL+trailingstep*pipsize ENDIF ENDIF //manage short positions IF SHORTONMARKET THEN //first move (breakeven) IF newSL=0 AND tradeprice(1)-close>=trailingstart*pipsize THEN newSL = tradeprice(1)-trailingstep*pipsize ENDIF //next moves IF newSL>0 AND newSL-close>=trailingstep*pipsize THEN newSL = newSL-trailingstep*pipsize ENDIF ENDIF //stop order to exit the positions IF newSL>0 THEN SELL AT newSL STOP EXITSHORT AT newSL STOP ENDIF //*************************************************************************************** // LONG trades //*************************************************************************************** a1 = close > open //BULLish bar a2 = KlingerTrigger CROSSES OVER KlingerVal //Klinger trigger is going north a3 = RviSignal CROSSES OVER RviVal //RVI long signal occurred IF a1 AND a2 AND a3 THEN BUY nLots CONTRACT AT MARKET ENDIF //*************************************************************************************** // SHORT trades //*************************************************************************************** b1 = close < open //BEARish bar b2 = KlingerTrigger CROSSES UNDER KlingerVal //Klinger trigger is going south b3 = RviSignal CROSSES UNDER RviVal //RVI short signal occurred IF b1 AND b2 AND b3 THEN SELLSHORT nLots CONTRACT AT MARKET ENDIF // SET TARGET PPROFIT TP SET STOP PLOSS SL

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81

  //------------------------------------------------------------------------- //                          Klinger-Rvi-Lsma DAX 5 min //------------------------------------------------------------------------- DEFPARAM CumulateOrders     = False DEFPARAM FlatBefore         = 090000                      //no trades before 09:00:00 DEFPARAM FlatAfter          = 213000                      //no trades after  21:30:00 ONCE nLots                  = 1                           //number of LOTs traded ONCE TP                     = 20                      //23    pips Take Profit ONCE SL                     = 16                          //16    pips Stop Loss   RviVal, RviSignal           = CALL "RVI by John Ehlers"[7]                  //7 KlingerVal, KlingerTrigger  = CALL "Klinger oscillator"[25,44,55,1]         //25,44,55,1 (ema) LeastSquareEMA              = CALL "ELSMA - Least Square EMA"[11,4]         //11,4 //*************************************************************************************** IF LongOnMarket THEN IF close < LeastSquareEMA THEN SELL AT MARKET                                      //Exit LONGs when MACD reverses southwards ENDIF ENDIF IF ShortOnMarket THEN IF close > LeastSquareEMA THEN EXITSHORT AT MARKET                                 //Exit SHORTs when MACD reverses northwards ENDIF ENDIF //*************************************************************************************** trailingstart = 20//10    trailing will start @trailinstart points profit trailingstep  = 2     //13    trailing step to move the "stoploss" //reset the stoploss value IF NOT ONMARKET THEN newSL=0 ENDIF //manage long positions IF LONGONMARKET THEN //first move (breakeven) IF newSL=0 AND close-tradeprice(1)>=trailingstart*pipsize THEN newSL = tradeprice(1)+trailingstep*pipsize ENDIF //next moves IF newSL>0 AND close-newSL>=trailingstep*pipsize THEN newSL = newSL+trailingstep*pipsize ENDIF ENDIF //manage short positions IF SHORTONMARKET THEN //first move (breakeven) IF newSL=0 AND tradeprice(1)-close>=trailingstart*pipsize THEN newSL = tradeprice(1)-trailingstep*pipsize ENDIF //next moves IF newSL>0 AND newSL-close>=trailingstep*pipsize THEN newSL = newSL-trailingstep*pipsize ENDIF ENDIF //stop order to exit the positions IF newSL>0 THEN SELL AT newSL STOP EXITSHORT AT newSL STOP ENDIF //*************************************************************************************** //                                       LONG trades //*************************************************************************************** a1 = close > open                                         //BULLish bar a2 = KlingerTrigger CROSSES OVER KlingerVal               //Klinger trigger is going north a3 = RviSignal CROSSES OVER RviVal                        //RVI long signal occurred IF a1 AND a2 AND a3 THEN BUY nLots CONTRACT AT MARKET ENDIF //*************************************************************************************** //                                      SHORT trades //*************************************************************************************** b1 = close < open                                         //BEARish bar b2 = KlingerTrigger CROSSES UNDER KlingerVal             //Klinger trigger is going south b3 = RviSignal CROSSES UNDER RviVal                       //RVI short signal occurred IF b1 AND b2 AND b3 THEN SELLSHORT nLots CONTRACT AT MARKET ENDIF // SET TARGET PPROFIT TP SET STOP PLOSS     SL

   

   

  2 users thanked author for this post.

  Aloysius, Gertrade

  

  GraHal

  Participant

  Topics: 212

  Replies: 8789

  Been thanked: 2189 times
  02/19/2018 at 12:01 AM #63075
  Gertrade

  @ Grahal,

  I realized the importance of history, I recognize that 20,000 units of history on the TF below the minute, it seems to me insufficient to validate a strategy. Right now, I’m doing tests on a TF 1 min, but I’m facing a major problem, how is it possible to make optimizations of variables of a complete strategy on 100,000 units? My PC is not powerful enough for that.

  

  Gertrade

  Participant

  Topics: 4

  Replies: 170

  Been thanked: 60 times
  02/19/2018 at 10:08 AM #63111
  Aloysius

  I Grahal,

  Yes I backtested Gertrade version for 28s, in 5mn TF, because you announced good results in this configuration. But your answer shows we have not the same approach of backtests: when a backtest is good for a little period it has been optimised for, and loosing at other periods, I don’t trust it and don’t think it will be good for some weeks. Anyway, here is the 200 000 bars backtest for Roberto code in 5mn, in the version you just posted: it looks promising, thank you.

  1 user thanked author for this post.

  GraHal

  

  Aloysius

  Participant

  Topics: 1

  Replies: 44

  Been thanked: 16 times
  02/19/2018 at 10:46 AM #63117
  GraHal

  @Gertrade you say … how is it possible to make optimizations of variables of a complete strategy on 100,000 units? My PC is not powerful enough for that … is your PC seeming to do nothing for ages then a quick spurt of results in the optimisation table then nothing then spurt … etc? If Yes then we all have that issue not matter how powerful our PCs … its due to optimise processing happening on the remote PRT Server.

  If No to above, then what symptoms are you getting that makes you think you PC is not powerful enough?

  GraHal
  PS  @Aloysius I do think same as you re not trusting if loose over longer Out of Sample / OOS period. Thank you for posting the 200,000 bar results.

   

  1 user thanked author for this post.

  Gertrade

  

  GraHal

  Participant

  Topics: 212

  Replies: 8789

  Been thanked: 2189 times
  02/19/2018 at 3:13 PM #63171
  GraHal

  @gertrade having said above re a spurt then a few results show in the Table then wait then another spurt etc … I have just backtested / optimised another Algo over 100,000 bars @ 5 Min TF and it completes in around 30 seconds with a lot less wait and a lot faster spurts.

  So speed of optimising / backtest does depend on how complex the code is and if there any ‘ Indicator Calls’ (the code I’m checking doesn’t have any Calls).

  GraHal

  1 user thanked author for this post.

  Gertrade

  

  GraHal

  Participant

  Topics: 212

  Replies: 8789

  Been thanked: 2189 times
  02/19/2018 at 11:08 PM #63227
  Gertrade

  GraHal wrote:

  … its due to optimise processing happening on the remote PRT Server

  That reassures me, I thought my PC was partly involved.

  I’m able to strategize on up to 120 trades.
  Beyond that, I get impatient, it becomes difficult to optimize each variable.

   

  

  Gertrade

  Participant

  Topics: 4

  Replies: 170

  Been thanked: 60 times
• Author
  Posts