long-short when Moving average changed color

• 02/02/2017 at 2:20 PM #23607
  Sofitech

  @Dave.

  Results seems to be very good. And maybe too good to be true. How many 0 bars when the position close in this backtest ?
  Could you insert the code so we can try to understand if results are realistic.

  

  Sofitech

  Participant

  Topics: 56

  Replies: 150

  Been thanked: 10 times
  02/02/2017 at 5:20 PM #23642
  Dave

  @enzo, done simply with a 50 trailing stop. As humans we always try to make things complex!. 🙂

  
  @sofitech, as you can see in the image, no 0 bars.maybe it is too good to be true given my more than basic programming skills!. I’m mobile so can’t post the code yet for fear of reprimand from @nicolas  who has to keep correcting:-)… Maybe later from PC.

  Dave

  

  Dave

  Participant

  Topics: 3

  Replies: 29

  Been thanked: 4 times
  02/02/2017 at 8:26 PM #23688
  Joachim Nilsson

  Hey!

  Can someone upload the file? I cant get it to work if I copy the code?

  

  Joachim Nilsson

  Participant

  Topics: 4

  Replies: 62

  Been thanked: 18 times
  02/03/2017 at 8:32 AM #23727
  Dave

  Ok, here it is for you to try. Be kind enough to feedback how it performs longer term as I don’t have any historic data to work with. If it performs well then maybe I’ll get some fame if @Nicolas considers it for the library?. Any ideas for improvement are also welcome although I would probably need help to implement it!.

  Dave

  (Test was done on DAX daily – I read on here recently “why does everybody trade DAX” – Simple = Volatility!)

  DEFPARAM CumulateOrders = False // Cumulating positions deactivated //DEFPARAM FLATBEFORE = 080000 //DEFPARAM FLATAFTER = 180000 //takeprofit = 50 //takeprofit in points //stoploss = 100 //stoploss in points fastma = ExponentialAverage[75](close) slowma = ExponentialAverage[195](close) //fastma bullish and bearish conditions fastbullish = fastma>fastma[1] and fastma[1]<fastma[3] and fastma>slowma fastbearish = fastma<fastma[1] and fastma[1]>fastma[3] and fastma<slowma //conditions for long = fast ma ABOVE slow ma and turning in agreement with slow ma trend if not longonmarket and fastbullish then buy 10 share at market endif //conditions for short = fast ma BELOW slow ma and in agreement with slow ma trend if not shortonmarket and fastbearish then sellshort 10 share at market endif //set target pprofit takeprofit //set stop ploss stoploss set stop ptrailing 50

  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

  DEFPARAM CumulateOrders = False // Cumulating positions deactivated //DEFPARAM FLATBEFORE = 080000 //DEFPARAM FLATAFTER = 180000     //takeprofit = 50 //takeprofit in points //stoploss = 100 //stoploss in points fastma = ExponentialAverage[75](close) slowma = ExponentialAverage[195](close)   //fastma bullish and bearish conditions   fastbullish = fastma>fastma[1] and fastma[1]<fastma[3] and fastma>slowma fastbearish = fastma<fastma[1] and fastma[1]>fastma[3] and fastma<slowma   //conditions for long = fast ma ABOVE slow ma and turning in agreement with slow ma trend if not longonmarket and fastbullish  then buy 10 share at market endif   //conditions for short = fast ma BELOW slow ma and in agreement with slow ma trend if not shortonmarket and fastbearish then sellshort 10 share at market endif   //set target pprofit takeprofit //set stop ploss stoploss set stop ptrailing 50

   

  

  Dave

  Participant

  Topics: 3

  Replies: 29

  Been thanked: 4 times
  02/03/2017 at 10:51 AM #23751
  enzo_52

  HI GUYS, I DID NOT BACKTEST ABOUT STRATEGY, JUST I FOUND THE RIGHT SETTING ABOUT  PRINCIPAL TREND WITH THE CROSSES OF 2 EMA  (75 AND 195) (NOT GREEN AND RED ) THEN I WONTED ANOTHER MOVING AVERAGE FOR TO GO ONLY SHORT  WHEN PRINCIPAL TREND WAS DOWN  AND GO LONG WHEN PRINCIPAL TREND WAS UP.

  DAVE DID ANOTHER SYSTEM  THAT WORKS GREAT  ….

  I THINK FOR IMPROVED THE RESULT TO DO A POSITION SIZE TOO  ABOUT 2% OF CAPITAL

   

  

  enzo_52

  Participant

  Topics: 21

  Replies: 76

  Been thanked: 0 times
  02/03/2017 at 11:08 AM #23752
  enzo_52

  I TRAIED THE TS..BUT SEEM DOES NOT WORK..

  

  enzo_52

  Participant

  Topics: 21

  Replies: 76

  Been thanked: 0 times
  02/03/2017 at 1:23 PM #23775
  Joachim Nilsson

  Thanks Dave

  This is how it looks when I backtest it with v10.3 live. Dosen´t seem to work…

  

  Joachim Nilsson

  Participant

  Topics: 4

  Replies: 62

  Been thanked: 18 times
  02/03/2017 at 1:31 PM #23778
  Dave

  I’m on 10.2 whether that has an effect on it, I’m not sure?.As you can see the code is pretty simple so cannot explain it.

  Dave

  

  Dave

  Participant

  Topics: 3

  Replies: 29

  Been thanked: 4 times
  03/01/2017 at 8:45 AM #26758
  Dave

  @Enzo,sofitech,joachim.

  Now I’m on 10.3 I’ve had a look at the code and got it working. As predicted, the results are not favourable, on one test only making £300/10yrs! and most of the others negative. I think beacause the large ma’s are so flat it could not initiate  a trade without increasing lookback. Fast ma’s with little lag perform much better with the code as they need to reflect PA on whichever timeframe your using. Dave

  

  Dave

  Participant

  Topics: 3

  Replies: 29

  Been thanked: 4 times
• Author
  Posts