Short Orders execution on Long only strategy

• 05/05/2020 at 10:41 AM #129985
  Bi

  Hi

  I am trying to build a long only strategy where I have the below trade management rules.

  1. StopLoss set at 1 ATR (14)
  2. First Target set at 1 ATR(14)
  3. Trailing stop starting at 2 ATR(14) and trailing step at 1.5 ATR(14)

  My StopLoss, First target and trailing seems to work but for some weird reason system is opening short positions even though I have no Short orders. The only sell orders I have exit ones, I tried add the code “Longonmarket” as well as “Countofpositions” still I am getting short orders

  Much appreciate if you could let me know where I am going wrong please?

  LongOnly Strategy

  // Definition of code parameters DEFPARAM CumulateOrders = TRUE // Cumulating positions deactivated // Conditions to enter long positions keltnersma = ExponentialAverage[20](close) keltnerpos1atr = keltnersma+ExponentialAverage[20](AverageTrueRange[1](close)*1) keltnerneg1atr = keltnersma+ExponentialAverage[20](AverageTrueRange[1](close)*-1) longindicator1 = Call Strategy01LongV1 longindicator2 = Call Strategy02LongV2 l1 = (longindicator1 = 1) l2 = (longindicator2 = 1) within1ATR = close >keltnerneg1atr and close<keltnerpos1atr //trade management lotsize = 2 //original position size Lot2Close =1 //close one lot while reaching first target stoploss = Averagetruerange[14](close)*SLFactor firsttarget= Averagetruerange[14](close)*PTFactor trailingstart = Averagetruerange[14](close)*2 trailingstep = Averagetruerange[14](close)*1.5 //trade entry and exits if l1 or l2 and within1ATR then Buy lotsize LOT AT MARKET SET STOP PLOSS stoploss endif //reset the stoploss value IF NOT ONMARKET THEN newSL=0 ENDIF //manage long positions IF LONGONMARKET THEN IF newSL=0 AND close-tradeprice(1)>=firsttarget THEN Sell lot2close LOT At Market newSL=tradeprice(1)+firsttarget ENDIF //first move (breakeven) IF newSL>0 AND close-newSL>=trailingstart THEN newSL = newSL+trailingstart ENDIF //next moves IF newSL>0 AND close-newSL>=trailingstep THEN newSL = newSL+trailingstep ENDIF ENDIF //stop order to exit the positions IF Countofposition>0 and LONGONMARKET AND newSL>0 THEN SELL AT newSL STOP ENDIF

  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

  // Definition of code parameters DEFPARAM CumulateOrders = TRUE // Cumulating positions deactivated   // Conditions to enter long positions   keltnersma = ExponentialAverage[20](close) keltnerpos1atr = keltnersma+ExponentialAverage[20](AverageTrueRange[1](close)*1) keltnerneg1atr = keltnersma+ExponentialAverage[20](AverageTrueRange[1](close)*-1)   longindicator1 = Call Strategy01LongV1 longindicator2 = Call Strategy02LongV2   l1 = (longindicator1 = 1) l2 = (longindicator2 = 1) within1ATR = close >keltnerneg1atr and close<keltnerpos1atr   //trade management lotsize = 2 //original position size Lot2Close =1 //close one lot while reaching first target stoploss = Averagetruerange[14](close)*SLFactor firsttarget= Averagetruerange[14](close)*PTFactor trailingstart = Averagetruerange[14](close)*2 trailingstep = Averagetruerange[14](close)*1.5     //trade entry and exits if l1 or l2 and within1ATR then Buy lotsize LOT AT MARKET SET STOP PLOSS stoploss endif   //reset the stoploss value IF NOT ONMARKET THEN newSL=0 ENDIF   //manage long positions IF LONGONMARKET THEN IF newSL=0 AND close-tradeprice(1)>=firsttarget THEN Sell lot2close LOT At Market newSL=tradeprice(1)+firsttarget ENDIF //first move (breakeven) IF newSL>0 AND close-newSL>=trailingstart THEN newSL = newSL+trailingstart ENDIF //next moves IF newSL>0 AND close-newSL>=trailingstep THEN newSL = newSL+trailingstep ENDIF ENDIF     //stop order to exit the positions IF Countofposition>0 and LONGONMARKET AND newSL>0 THEN SELL AT newSL STOP ENDIF

   

  (See attached strategy and code)

  

  Bi

  Participant

  Topics: 3

  Replies: 7

  Been thanked: 2 times
  05/05/2020 at 11:11 AM #129998
  GraHal

  I am trying to run it on my platform to see if I get the same … well weird??

  Is the Strategy01LongV1 and v2 in the Library on this website?

  

  GraHal

  Participant

  Topics: 204

  Replies: 8432

  Been thanked: 2048 times
  05/05/2020 at 11:19 AM #129999
  Bi

  Thank you, below the v1,v2 code. I believe rest of the conditions are in the strategy code itself.

   

  This is the V1 and V2 code

  //v1 atr = AverageTrueRange[14](close)*1/4 v1low = ABS(low - low[1])> atr v1lowcheck = low<low[1] v1close= close>close[1] if v1low and v1lowcheck and v1close then value = 1 else value = 0 endif return value //V2 atr = AverageTrueRange[14](close)*1/4 v1low = ABS(low[1] - low[2])> atr v1lowcheck = low[1] < low[2] notv1= close[1] < close[2] v2close1 = close > close[1] v2close2 = close > close[2] v2low = low>low[1] v2gap = abs(close[1]-open)<=atr if v1low and v1lowcheck and notv1 and v2close1 and v2close2 and v2low and v2gap then value = 1 else value = 0 endif return value

  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

  //v1 atr = AverageTrueRange[14](close)*1/4 v1low = ABS(low - low[1])> atr v1lowcheck = low<low[1] v1close= close>close[1]   if v1low and v1lowcheck and v1close then value = 1 else value = 0 endif return value   //V2 atr = AverageTrueRange[14](close)*1/4 v1low = ABS(low[1] - low[2])> atr v1lowcheck = low[1] < low[2] notv1= close[1] < close[2] v2close1 = close > close[1] v2close2 = close > close[2] v2low = low>low[1] v2gap = abs(close[1]-open)<=atr   if v1low and v1lowcheck and notv1 and v2close1 and v2close2 and v2low and v2gap then value = 1 else value = 0   endif return value

   

  1 user thanked author for this post.

  GraHal

  

  Bi

  Participant

  Topics: 3

  Replies: 7

  Been thanked: 2 times
  05/05/2020 at 11:33 AM #130001
  Francesco

  The problem is in the implementation of the lot closure when reaching first target; removing this condition from the trailing stop there’s no shorts

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  05/05/2020 at 11:41 AM #130003
  GraHal

  I confirm same … so you are not seeing things! Unless we are both deluded!? 🙂

  I get 1 short trade when running on DJI 1000 bars and 5 min TF and spread = 2.

  When I put spread = 4 I do not get a short trade on DJI 1000 bars and 5 min TF.

  That’s one of the weirdest thing I’ve ever come across since I’ve been using PRT??

   

   

  

  GraHal

  Participant

  Topics: 204

  Replies: 8432

  Been thanked: 2048 times
  05/05/2020 at 11:50 AM #130006
  GraHal

  I think it has something to do with the line of code below.

  When I change to ‘Buy at Market’, I get no shorts … yet!? 🙂

  Buy lotsize LOT AT MARKET

  1	Buy lotsize LOT AT MARKET

   

   

  

  GraHal

  Participant

  Topics: 204

  Replies: 8432

  Been thanked: 2048 times
  05/05/2020 at 12:15 PM #130007
  GraHal

  Somehow the code is buying 2 and selling 3 to exit??

  

  GraHal

  Participant

  Topics: 204

  Replies: 8432

  Been thanked: 2048 times
  05/05/2020 at 12:31 PM #130009
  GraHal

  Various values for ptfactor will give shorts or stop shorts also

  firsttarget= Averagetruerange[14](close)*PTFactor

  1	firsttarget= Averagetruerange[14](close)*PTFactor

   

  Seems there are several changes that will stop the Shorts, but how can a System go short when there is NO SellShort command in the code??

  Very worrying!!??

  

  GraHal

  Participant

  Topics: 204

  Replies: 8432

  Been thanked: 2048 times
  05/05/2020 at 12:54 PM #130015
  Bi

  thank you for looking into it. I do agree this is odd.

  Also I understand scaling out on a auto-trading is not possible but I am only try to backtest my strategy how I would manually trade it. Hence the first target close is important. But why market is shorting even with the below condition is beyond my understanding.

  IF Countofposition>0 and LONGONMARKET AND newSL>0 THEN

  1	IF Countofposition>0 and LONGONMARKET AND newSL>0 THEN

   

  

  Bi

  Participant

  Topics: 3

  Replies: 7

  Been thanked: 2 times
  05/05/2020 at 6:08 PM #130111
  Bi

  I think i fixed it, the issue was the calculation of breakeven and trailingstart were off and also I had to specify lot to close (remaining) at the end. Which I think is the reason for the short position.

  

  Bi

  Participant

  Topics: 3

  Replies: 7

  Been thanked: 2 times
  05/05/2020 at 6:39 PM #130116
  GraHal

  Yeah at certain conditions you are selling 3 after buying 2 (?) … still weird though as I would have expected a System Reject / Stop covered by one of IG’s ambiguous screen messages!?

  Maybe one day somebody might be able to make profitable use of this loophole!? 🙂

   

  

  GraHal

  Participant

  Topics: 204

  Replies: 8432

  Been thanked: 2048 times
  05/05/2020 at 6:47 PM #130119
  Bi

  haha yes! the shorts were actually making my system look profitable got a bit carried away with the results. Glad to have checked it.

  

  Bi

  Participant

  Topics: 3

  Replies: 7

  Been thanked: 2 times
  05/05/2020 at 6:49 PM #130120
  GraHal

  Re profitable … how about introducing (proper) shorts into the strategy?

  

  GraHal

  Participant

  Topics: 204

  Replies: 8432

  Been thanked: 2048 times
  05/05/2020 at 6:56 PM #130121
  Bi

  I do have shorts as well with the same logic reversed but wanted to keep the strategies Long only and Short only to review the results accurately. Also I am testing this on US shares where there is a long bias but have to test this out in other markets like FX or Commodities where trends are not one sided.

  Please feel free to improve it, if you do manage to please do let me know.

  

  Bi

  Participant

  Topics: 3

  Replies: 7

  Been thanked: 2 times
  05/05/2020 at 7:17 PM #130123
  GraHal

  Bi wrote:

  Please feel free to improve it

  I had a quick go this morning … I could see that it needed a filter to stop Longs when price was clearly in a big downtrend.

  I tried below and I used values of up to x = 10000 on a 5 min TF, but still it was taking Longs … I thought more weirdness!? 🙂

  Close > Average[x](Close)

  1	Close > Average[x](Close)

   

  

  GraHal

  Participant

  Topics: 204

  Replies: 8432

  Been thanked: 2048 times
• Author
  Posts