Dax intraday scalping strategy

• 02/13/2017 at 10:48 AM #24941
  gforton

  Hello fellow traders,

  I have been monitoring this wonderful site for nearly one year now, admired the work of a number of experienced and creative minds.

  I would now like to share a piece of code with out that my wife, Dominique, and I have developed.

  We have tried the code live since January 6th on te DAX 1 minute and it has generated 260 Euros with a DAX mini 1 Euro CFD system, trading long only. I admit that I have aborted two trades going the wrong way manually.
  The code relies on the crossing of linear regression lines, augmented with a factor a or b to define the steepness or angle of the crossing (the steeper, the more bullish),
  combined with RSI and the Schaff Trend Cycle Index. We have also tried other indices for trent, such as the TDI, the ‘Bluster DAX’ and a lot of others, but the Schaff Trend Cycle Index in our opinion performs best.

  The code is designed to take whatever profit as soon as it get’s later than 17.45, provided that we are making any profit. This is to avoid higher spreads. This codes has been tested using a spread of 0,8, which is normal DAX spread during the office hours. We specifically don’t want to go overnight, although it as happened.

  All parameters have been optimized for the last 6 weeks, which have been rather annoying, given the horizontal progression of the DAX, but that’s life.

  We also have a ‘short’ version of this code, but it doesn’t work as good.

  We have also tried to combine our entry with CTFA’s grid trading and money managing strategy (congratulations on this one, by the way), but we have found that a simple take profit and stop loss renders the most profit. Perhaps you can do it.

  So, here it is. We hope that you will be able to improve this code and -who knows- combine with an efficient short version and/or money management strategy.

  Greetings from the ‘hell hole’ Belgium. “It’s tremendous”. Time zone is GMT +1

  Glen

  //------------------------------------------------------------------------- // Main code : DAX DOMI LONG 03/02/2017 //------------------------------------------------------------------------- DEFPARAM CumulateOrders = true // Definition of indicators purple = LinearRegression[7](close) blue = LinearRegression[10](close) yellow = LinearRegression[40] (close) a = 1 b = 0 indicator1 = RSI[31](close) indicator2 = time // Schaff Trend Cycle MCD = WilderAverage[20] - WilderAverage[50] ST =( MCD-LOWEST[3] (MCD) )/(HIGHEST[3] (MCD)-LOWEST[3] (MCD) )*100 STC = WilderAverage [round(3/2)] (ST) // Conditions to enter long positions c1 = (purple > (blue + a)) c2 = (close > (blue + b)) c3 = (close > close[1]) c4 = (blue > blue[5]) c5 = CountOfPosition <=3 c6 = (indicator1 < 70) c7 = (indicator1 > 49) c8 = (indicator2 > 080000) c9 = (indicator2 < 161500) c10 = STC > 27 c11= STC > STC [5] IF c1 AND c2 AND c3 AND c4 AND c5 AND c6 AND c7 AND c8 AND c9 AND c10 AND c11 THEN BUY 1 CONTRACT AT MARKET ENDIF IF LONGONMARKET AND (time > 174500) AND (PositionPerf > 0) THEN SELL AT MARKET ENDIF IF LONGONMARKET AND (yellow > (purple + 20)) THEN SELL AT MARKET ENDIF // Stops and targets SET TARGET pPROFIT 25 SET STOP pLOSS 135

  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

  //------------------------------------------------------------------------- // Main code : DAX DOMI LONG 03/02/2017 //------------------------------------------------------------------------- DEFPARAM CumulateOrders = true   // Definition of indicators   purple = LinearRegression[7](close) blue = LinearRegression[10](close) yellow = LinearRegression[40] (close)   a = 1 b = 0   indicator1 = RSI[31](close) indicator2 = time     // Schaff Trend Cycle MCD = WilderAverage[20] - WilderAverage[50] ST =( MCD-LOWEST[3] (MCD) )/(HIGHEST[3] (MCD)-LOWEST[3] (MCD) )*100 STC = WilderAverage [round(3/2)] (ST)   // Conditions to enter long positions   c1 = (purple > (blue + a)) c2 = (close > (blue + b)) c3 = (close > close[1]) c4 = (blue > blue[5]) c5 = CountOfPosition <=3 c6 = (indicator1 < 70) c7 = (indicator1 > 49) c8 = (indicator2 > 080000) c9 = (indicator2 < 161500) c10 = STC > 27 c11= STC > STC [5]         IF c1 AND c2  AND c3 AND c4 AND c5 AND c6 AND c7 AND c8 AND c9 AND c10 AND c11 THEN BUY 1 CONTRACT AT MARKET ENDIF     IF LONGONMARKET AND (time > 174500) AND (PositionPerf > 0) THEN SELL AT MARKET ENDIF   IF LONGONMARKET AND (yellow > (purple + 20)) THEN SELL AT MARKET ENDIF   // Stops and targets SET TARGET pPROFIT 25 SET STOP pLOSS 135

  1 user thanked author for this post.

  GraHal

  

  gforton

  Participant

  Topics: 2

  Replies: 11

  Been thanked: 2 times
  02/13/2017 at 11:01 AM #24944
  Nicolas

  Hello Glen and Dominique !
  I know well Belgium, and living there is not like living in an “hellhole” 🙂
  You are right, it’s been almost a year since the website has been launched! We’ll celebrate our first year the 1st of March.
  I preferred moving your post from the library to the forum first, to discuss a bit about what have been optimized and what were the past performance. I think other fellow members will participate and I encourage them to do so.

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4774 times
  02/26/2017 at 5:07 PM #26433
  francoisph

  hello

  nice code…  I am looking for dax short code. The market is so high at the moment that I prefer shorting dax or dow jones to be honest.

  Main indicator that I like it is rsi pivot and macd.

  

  francoisph

  Participant

  Topics: 0

  Replies: 2

  Been thanked: 0 times
  02/26/2017 at 5:29 PM #26437
  GraHal

  I agree Franc, Glen can this Strategy go Short also please?

  Thanks
  GraHal

  

  GraHal

  Participant

  Topics: 214

  Replies: 8826

  Been thanked: 2198 times
  02/26/2017 at 10:59 PM #26468
  GraHal

  Glen maybe your Short version (you mention) performs better in a downtrend … as Dax appears to have started (until it goes up again! ;)). How did your Short version perform during down periods on Backtest?

  Cheers
  Graham

   

  

  GraHal

  Participant

  Topics: 214

  Replies: 8826

  Been thanked: 2198 times
  02/26/2017 at 11:05 PM #26469
  francoisph

  Didnt perform any trades at the moment. I am looking for algo code with low loss.

  

  francoisph

  Participant

  Topics: 0

  Replies: 2

  Been thanked: 0 times
  02/26/2017 at 11:05 PM #26470
  Derek

  Hello gforton,

  thank you for sharing your code.

  If you have been trading this strategy since new year and the variables are optimized for the last weeks have you “reoptimized” before posting or have the last six weeks been a out-of-sample testing period?

  

  Derek

  Participant

  Topics: 6

  Replies: 240

  Been thanked: 60 times
  05/08/2017 at 1:01 PM #34816
  GraHal

  Anybody got this working okay in a short version please?

  Many Thanks
  GraHal

  1 user thanked author for this post.

  reb

  

  GraHal

  Participant

  Topics: 214

  Replies: 8826

  Been thanked: 2198 times
  05/08/2017 at 7:56 PM #34856
  GraHal

  This looks promising, but needs to work in short mode also in my opinion?

  Alternatively the Long code attached could be improved as it loses in down periods.

  I have optimised as below and got results attached.

  Please might somebody be able to improve the Long Code and / or add Go Shorts?

  Many Thanks
  GraHal

  ///------------------------------------------------------------------------- // Main code : DAX DOMI LONG 03/02/2017 //------------------------------------------------------------------------- DEFPARAM CumulateOrders = false // Definition of indicators purple = LinearRegression[6](close)//7 blue = LinearRegression[8](close)//10 yellow = LinearRegression[40] (close)//40 a = 1 b = 0 indicator1 = RSI[35](close) indicator2 = time // Schaff Trend Cycle MCD = WilderAverage[20] - WilderAverage[50] ST =( MCD-LOWEST[3] (MCD) )/(HIGHEST[3] (MCD)-LOWEST[3] (MCD) )*100 STC = WilderAverage [round(3/2)] (ST) // Conditions to enter long positions c1 = (purple > (blue + a)) c2 = (close > (blue + b)) c3 = (close > close[1]) c4 = (blue > blue[5]) c5 = CountOfPosition <=3 c6 = (indicator1 < 70) c7 = (indicator1 > 49)//49 c8 = (indicator2 > 082000) c9 = (indicator2 < 163000) c10 = STC > 6//27 c11= STC > STC [5] IF c1 AND c2 AND c3 AND c4 AND c5 AND c6 AND c7 AND c8 AND c9 AND c10 AND c11 THEN BUY 1 CONTRACT AT MARKET ENDIF // IF LONGONMARKET AND (time > 194000) AND (PositionPerf > 0) THEN SELL AT MARKET ENDIF IF LONGONMARKET AND (yellow > (purple + 19)) THEN SELL AT MARKET ENDIF // Stops and targets SET TARGET pPROFIT 31 SET STOP pLOSS 72

  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

  ///------------------------------------------------------------------------- // Main code : DAX DOMI LONG 03/02/2017 //------------------------------------------------------------------------- DEFPARAM CumulateOrders = false   // Definition of indicators   purple = LinearRegression[6](close)//7 blue = LinearRegression[8](close)//10 yellow = LinearRegression[40] (close)//40   a = 1 b = 0   indicator1 = RSI[35](close) indicator2 = time     // Schaff Trend Cycle MCD = WilderAverage[20] - WilderAverage[50] ST =( MCD-LOWEST[3] (MCD) )/(HIGHEST[3] (MCD)-LOWEST[3] (MCD) )*100 STC = WilderAverage [round(3/2)] (ST)   // Conditions to enter long positions   c1 = (purple > (blue + a)) c2 = (close > (blue + b)) c3 = (close > close[1]) c4 = (blue > blue[5]) c5 = CountOfPosition <=3 c6 = (indicator1 < 70) c7 = (indicator1 > 49)//49 c8 = (indicator2 > 082000) c9 = (indicator2 < 163000) c10 = STC > 6//27 c11= STC > STC [5]         IF c1 AND c2  AND c3 AND c4 AND c5 AND c6 AND c7 AND c8 AND c9 AND c10 AND c11 THEN BUY 1 CONTRACT AT MARKET ENDIF //   IF LONGONMARKET AND (time > 194000) AND (PositionPerf > 0) THEN SELL AT MARKET ENDIF   IF LONGONMARKET AND (yellow > (purple + 19)) THEN SELL AT MARKET ENDIF   // Stops and targets SET TARGET pPROFIT 31 SET STOP pLOSS 72

   

   

  

  GraHal

  Participant

  Topics: 214

  Replies: 8826

  Been thanked: 2198 times
  05/08/2017 at 8:22 PM #34860
  GraHal

  As agreed on another Thread we need to state Times to a standard … times in above code is UTC +1.

  UTC / Universal Time Clock  is not a timezone it is a ‘time standard’ and it’s best if we all state what our times are in our codes in UTC mode.

  (UTC is not an abbreviation for Universal Time Clock but it is the easiest way to to remember! 🙂

  GraHal

  

  GraHal

  Participant

  Topics: 214

  Replies: 8826

  Been thanked: 2198 times
  09/21/2017 at 5:19 PM #46848
  halgo

  I did not compile the code in MetaEditor.
  gave some errors.
  Can you help me.

  

  halgo

  Participant

  Topics: 0

  Replies: 1

  Been thanked: 0 times
  09/21/2017 at 5:30 PM #46853
  Despair

  Jesus halgo, you are very wrong here. This forum is for prorealtime and not for metatrader. Nothing you find here will work with metatrader.

  

  Despair

  Blocked

  Topics: 97

  Replies: 1034

  Been thanked: 218 times
  10/25/2017 at 12:04 PM #50468
  Leo

  Hi gforton,

  Which variables are to be optimized?

   

  

  Leo

  Participant

  Topics: 17

  Replies: 357

  Been thanked: 144 times
  06/10/2020 at 4:16 PM #135462
  deletedaccount051022

  Hi gforton,

  I too trade DAX Index and the constituent stocks and really liked your strategy.  After having run some extensive back testing I started running your strategy on a live trading account, albeit with a small amount of risk capital.  Provided the DAX on a daily basis is in an uptrend the performance if very good.

  I was just wondering if you had been using this of late or worked on enhancing it?  Would be interested to swap ideas.

   

  Best regards,

  S

  

  deletedaccount051022

  Participant

  Topics: 26

  Replies: 139

  Been thanked: 99 times
  06/10/2020 at 8:50 PM #135506
  nonetheless

  Attached is a quick re-working of this for DAX 2min.

  It is a disaster. Has the distinction of perhaps the highest annual profit and worst robustness score I’ve yet to see. Possibly the most curve fit thing since Beyonce went shopping.

  Still, the basic code seems interesting and this could be a starting point if someone else wants to take it on.

  2 users thanked author for this post.

  Scooby, Francesco

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
• Author
  Posts