DOW M15 with Wilder average (Intraday / swing)

• 12/13/2019 at 9:33 AM #114746
  juju333

  Hi ! Here is my first very simple code, for DOW M15 (tested with 100 000 periods). 1 contract, 10 000 €, spread 2.5 pts, 1€ / order to represent overnight charges.

  Every comments (or test on 200000 periods) are welcome :-).

  DEFPARAM CumulateOrders = False // Cumul des positions désactivé noEntryBeforeTime = 000000 timeEnterBefore = time >= noEntryBeforeTime noEntryAfterTime = 230000 timeEnterAfter = time < noEntryAfterTime // Conditions pour ouvrir une position long indicator1 = WilderAverage[20](close)-WilderAverage[50](close) c1 = (indicator1 CROSSES OVER 0) indicator5 = MACD[7,14,5] (close) c5 = (indicator5 > 0) IF c1 AND c5 AND timeEnterBefore AND timeEnterAfter THEN BUY 1 CONTRACT AT MARKET ENDIF // Conditions pour ouvrir une position short indicator10 = WilderAverage[20](close)-WilderAverage[50](close) c10 = (indicator10 CROSSES UNDER 0) indicator13 = MACD[7,14,5] (close) c13 = (indicator13 < 0) IF NOT LONGONMARKET AND c10 AND c13 AND timeEnterBefore AND timeEnterAfter THEN SELLSHORT 1 CONTRACT AT MARKET ENDIF // Stops et objectifs SET STOP pLOSS 108 SET TARGET pPROFIT 525

  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

  DEFPARAM CumulateOrders = False // Cumul des positions désactivé   noEntryBeforeTime = 000000 timeEnterBefore = time >= noEntryBeforeTime noEntryAfterTime = 230000 timeEnterAfter = time < noEntryAfterTime   // Conditions pour ouvrir une position long indicator1 = WilderAverage[20](close)-WilderAverage[50](close) c1 = (indicator1 CROSSES OVER 0) indicator5 = MACD[7,14,5] (close) c5 = (indicator5 > 0)   IF c1 AND c5 AND timeEnterBefore AND timeEnterAfter THEN BUY 1 CONTRACT AT MARKET ENDIF   // Conditions pour ouvrir une position short indicator10 = WilderAverage[20](close)-WilderAverage[50](close) c10 = (indicator10 CROSSES UNDER 0) indicator13 = MACD[7,14,5] (close) c13 = (indicator13 < 0)   IF NOT LONGONMARKET AND c10 AND c13 AND timeEnterBefore AND timeEnterAfter THEN SELLSHORT 1 CONTRACT AT MARKET ENDIF   // Stops et objectifs SET STOP pLOSS 108 SET TARGET pPROFIT 525

   

  

  juju333

  Participant

  Topics: 3

  Replies: 26

  Been thanked: 12 times
  12/13/2019 at 10:25 AM #114804
  Nicolas

  Hi Juju,

  Thanks for sharing your idea.

  The strategy suffer from over optimization as you can see with 100k bars of out of sample testing. It is always a good idea to optimize with split periods between In Sample and Out Of Sample periods, to validate robustness of optimized variables.

  1 user thanked author for this post.

  juju333

  

  Nicolas

  Keymaster

  Topics: 83

  Replies: 20899

  Been thanked: 4868 times
  12/13/2019 at 11:46 AM #114812
  juju333

  Nicolas wrote:

  Hi Juju,

  Thanks for sharing your idea.

  The strategy suffer from over optimization as you can see with 100k bars of out of sample testing. It is always a good idea to optimize with split periods between In Sample and Out Of Sample periods, to validate robustness of optimized variables.

  Thank you Nicolas for the long term back testing ! 2019 alone seems a very good year, which is not the case the years before… Indeed, TP and SL are optimised only with my 100 approach, maybe someone could fin best values in the long term ? What would you advice for me with only a 100K approach ?

  

  juju333

  Participant

  Topics: 3

  Replies: 26

  Been thanked: 12 times
  12/13/2019 at 11:51 AM #114813
  juju333

  Finally, around 100% in 8 years (minus overnight charges, which might be high du to a very long time on market) : is this strategy worth optimizing it or is it quite common ? this 0 crossing of wilder average 20/50 is quitte efficient in my ‘manual’ trading setup and easy to automate. MACD prevents from taking too much wrong signals.

  

  juju333

  Participant

  Topics: 3

  Replies: 26

  Been thanked: 12 times
  12/16/2019 at 11:42 AM #114965
  Nicolas

  juju333 wrote:

  2019 alone seems a very good year, which is not the case the years before

  Of course, and surely because of optimization! If you optimize your variables, any strategy would look good for sure. That’s why I encourage you to explore concept of optimization into splitted periods aka “walk forward analysis”. There are plenty of topics discussing concept of robustness, but you can start with these videos:

  French videos:

  https://www.prorealcode.com/blog/video-tutorials/analyse-walk-forward-prorealtime/

  https://www.prorealcode.com/blog/video-tutorials/recapitulatif-sur-lutilisation-du-module-walk-sous-prorealtime/

  French topic: https://www.prorealcode.com/topic/optimisation-et-analyse-walk-forward-_-aides/

   

  

  Nicolas

  Keymaster

  Topics: 83

  Replies: 20899

  Been thanked: 4868 times
  12/17/2019 at 6:52 PM #115116
  juju333

  Thank you Nicolas, i’ll go and read all that

  

  juju333

  Participant

  Topics: 3

  Replies: 26

  Been thanked: 12 times
• Author
  Posts