Automatic trading system DAX 1H mini 1€

• 02/14/2017 at 1:14 AM #25026
  manel

  ATR system looks pretty good I’d say. At least there is some decent growth in the first 5 years as opposed to none with most other systems. Draw is is only 10% so not high at all. 20% of trades and profits are zero bar which is again in line with expectations.

  Comparison with system 3 below as well. They are pretty much identical results for the first 6yrs and then ATR starts outperforming gradually over the longer term.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/14/2017 at 1:32 AM #25030
  manel

  And here are the Ibex results – ATR vs System 1. See earlier post for first round results for this index. But in this revised ATR system the risk has certainly been reduced. The draw is now 37% instead of 54% and the no of consecutive losses has dropped to 6 from 9. It makes 30% less trades for the same total gain i.e is more efficient. Only slight negative is that the biggest single loss is 3x greater. But overall the system has definitely made a positive difference to the performance and risk taken.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/14/2017 at 8:56 AM #25045
  Raul Vg

  I have tried to include as a condition of purchase that ATR is above 20. And on the other hand the condition that ATR is below 20, to see if in the DAX it is significant that this indicator signals us or not. I see that both this below and above 20, the system is a winner, I do not think that in this case ATR is a determining factor in which the operation wins or loses. Adding condition c111 or c1111

  screenshot 1

  If d1 And t1 and c1 and c11 and c12 and c13 and c111 THEN

  1	If   d1 And t1 and c1  and c11 and c12 and c13 and c111 THEN

  screenshot 2

  If d1 And t1 and c1 and c11 and c12 and c13 and c1111 THEN

  1	If   d1 And t1 and c1  and c11 and c12 and c13 and c1111 THEN

  

  Raul Vg

  Participant

  Topics: 4

  Replies: 177

  Been thanked: 83 times
  02/14/2017 at 8:28 PM #25086
  manel

  Interesting. So I guess it’s as we thought, low vol conditions are not great for producing trade entry conditions and/or profits as the results show.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/14/2017 at 8:34 PM #25089
  manel

  No trades in Live or backtesting for the DAX or CAC systems today by the way. Should hopefully get a trade next Tuesday that gives us some live data to compare.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/14/2017 at 8:57 PM #25090
  Raul Vg

  According to the conditions it has to operate from Tuesday at 9am until Thursday at 6pm. That is, tonight or tomorrow could also operate. Yes, it is more beneficial with great volume, but I see that with little volume does not come out negative, there are no losses, I see that these last years has increased the volume as we said. But I do not see that there is a negative result in any case. I will try with some other condition to see if I can refine it, but I certainly see good expectations for this robot in both dax and cac. You could look to see 60 put take to see if they change much the results and get a little more in each operation

  

  Raul Vg

  Participant

  Topics: 4

  Replies: 177

  Been thanked: 83 times
  02/14/2017 at 9:21 PM #25098
  manel

  Yes, I agree, that’s my conclusion as well. The way it is works is pretty good for now (hence my decision to take it into live). The main thing is that in unprofitable or slow periods it is not losing any big money and that is what our testing over the last few days has shown. So I’m happy with that. This means like I said earlier, you just have to be patient and implement this over a long time horizon.  We could try alternate periods but I don’t want to overfit. Maybe we could look at that at a later stage when the system is performing well.

   

  Re. your comment – Are you sure about more entry trades possible tom/thurs ? I interpret the code as saying that entry is only on Day 2 and exit can be anytime up until Day 4/5 5pm.

  If d1 And t1 and c1 and c11 and c12 and c13 THEN if not onmarket then IF c1 THEN IF PositionPerf(1) < 0 THEN OrderSize = OrderSize*2 if ordersize<1 then ordersize=1 ENDIF ELSIF PositionPerf(1) > 0 THEN OrderSize = 1 if ordersize<1 then ordersize=1 ENDIF endif buy ordersize+n Contract At Market Endif If d2 and t2 then sell at market endif endif endif

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

  If   d1 And t1 and c1 and c11 and c12 and c13 THEN if not onmarket then IF c1 THEN IF PositionPerf(1) < 0 THEN OrderSize = OrderSize*2 if ordersize<1 then ordersize=1 ENDIF ELSIF PositionPerf(1) > 0 THEN OrderSize = 1 if ordersize<1 then ordersize=1 ENDIF endif buy ordersize+n Contract At Market Endif     If d2 and t2 then sell at market endif endif endif

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/14/2017 at 9:27 PM #25099
  Raul Vg

  Yes, sorry, you’re right, I think I’ve confused myself with code, I’ll try without limiting the trading day to see it coming out

  

  Raul Vg

  Participant

  Topics: 4

  Replies: 177

  Been thanked: 83 times
  02/14/2017 at 9:39 PM #25100
  manel

  Ok, no problem. I did some testing early on to change the days to other days of the week to see what the effect was but apart from a couple of cases in some instruments, for some reason Tuesday looked like being the best day for this system.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/14/2017 at 10:31 PM #25103
  Nicolas

  So you are still using the same profit and loss (fixed values, not dynamic ones) and you added ATR as a condition to enter market?

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20828

  Been thanked: 4742 times
  02/15/2017 at 12:13 AM #25109
  Raul Vg

  to try…

   

  DEFPARAM CumulateOrders = true once ordersize=1 // Conditions MA1 = Average [1] MA2 = Average [2] c1 = Ma1>MA2 t1 = time >= 090000 t2 = time = 180000 d1 = dayofweek = 2 d2 = dayofweek = 4 if strategyprofit<=30000 then if ordersize>16 then ordersize=16 endif endif if strategyprofit>30000 then if ordersize>64 then ordersize=64 endif endif // buy indicator13 = MACD[12,26,9](close) c13 = (indicator13 < 30) indicator11 = ExponentialAverage[20](close) c11 = (close >= indicator11) indicator1 = MACDline[12,26,9](close) indicator2 = ExponentialAverage[12](indicator1) c12 = (indicator1 > indicator2) //sell indicator1111 = ExponentialAverage[20](close) c1111 = (close <= indicator1111) indicator111 = MACDline[12,26,9](close) indicator2111 = ExponentialAverage[9](indicator111) c1211 = (indicator111 < indicator2111) if not onmarket then If d1 And t1 and c1 and c11 and c12 and c13 THEN IF PositionPerf(1) < 0 THEN OrderSize = OrderSize*2 if ordersize<1 then ordersize=1 ENDIF ELSIF PositionPerf(1) > 0 THEN OrderSize = 1 if ordersize<1 then ordersize=1 ENDIF endif buy ordersize Contract At Market Endif If d1 And t1 and c1 and c1111 and c1211 THEN IF PositionPerf(1) < 0 THEN OrderSize = OrderSize*2 if ordersize<1 then ordersize=1 ENDIF ELSIF PositionPerf(1) > 0 THEN OrderSize = 1 if ordersize<1 then ordersize=1 ENDIF endif sellshort ordersize Contract At Market Endif endif If d2 and t2 then exitshort at market endif Set Stop Ploss 50 // from Trade Set Target PProfit 50 // from Trade

  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 82 83 84 85 86 87 88 89 90

  DEFPARAM CumulateOrders = true once ordersize=1   // Conditions   MA1 = Average [1] MA2 = Average [2]   c1 = Ma1>MA2   t1 = time >= 090000 t2 = time = 180000     d1 = dayofweek = 2 d2 = dayofweek = 4     if strategyprofit<=30000 then if ordersize>16 then ordersize=16 endif endif if strategyprofit>30000 then if ordersize>64 then ordersize=64 endif endif // buy indicator13 = MACD[12,26,9](close) c13 = (indicator13 < 30)   indicator11 = ExponentialAverage[20](close) c11 = (close >= indicator11)   indicator1 = MACDline[12,26,9](close) indicator2 = ExponentialAverage[12](indicator1) c12 = (indicator1 > indicator2)   //sell   indicator1111 = ExponentialAverage[20](close) c1111 = (close <= indicator1111)   indicator111 = MACDline[12,26,9](close) indicator2111 = ExponentialAverage[9](indicator111) c1211 = (indicator111 < indicator2111)   if not onmarket then If   d1 And t1 and c1 and c11 and c12 and c13 THEN   IF PositionPerf(1) < 0 THEN OrderSize = OrderSize*2 if ordersize<1 then ordersize=1 ENDIF ELSIF PositionPerf(1) > 0 THEN OrderSize = 1 if ordersize<1 then ordersize=1 ENDIF endif buy ordersize Contract At Market Endif   If   d1 And t1 and c1 and c1111 and c1211   THEN   IF PositionPerf(1) < 0 THEN OrderSize = OrderSize*2 if ordersize<1 then ordersize=1 ENDIF ELSIF PositionPerf(1) > 0 THEN OrderSize = 1 if ordersize<1 then ordersize=1 ENDIF endif sellshort ordersize Contract At Market Endif endif   If d2 and t2 then exitshort at market endif       Set Stop Ploss 50 // from Trade Set Target PProfit 50 // from Trade

   

  

  Raul Vg

  Participant

  Topics: 4

  Replies: 177

  Been thanked: 83 times
  02/15/2017 at 1:12 AM #25110
  manel

  @ Nicolas – Yes in respect to results shown in #25026 above, there is a fixed 50pt stop and limit and ATR has to be >20 for entry trade condition to be satisfied.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/15/2017 at 1:39 AM #25111
  manel

  Results are not good. System only works from mid 2012 but that too with a 54% draw. Prior to that it’s a 100% loss in 08.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/15/2017 at 1:43 AM #25114
  manel

  With regards to the ATR I suppose there may be some value in testing to see if linking the take profit/loss to the ATR produces better results. Maybe it could improve performance in the earlier years as well where the signal is correct but the take profits are too high.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  02/15/2017 at 8:36 AM #25123
  Joachim Nilsson

  Manel – I think the results are great! The only problem is that it only works in bulltrend. You have to be convinced that the market is in a bulltrend if you´re going to start it.

  

  Joachim Nilsson

  Participant

  Topics: 4

  Replies: 62

  Been thanked: 18 times
• Author
  Posts