Strategy DayOpen Straddle for DAX

• 01/17/2020 at 12:43 PM #117067
  Francesco

  You’re absolutely right, if it weren’t for the tight backtest period I would have already implemented it in my portfolio.
  Let’s see if it’s possible to increase the time frame and obtain the same excellent results 😀

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  01/19/2020 at 5:55 PM #117216
  Francesco

  As said before, the sort timeframe is a problem for an ideal future live implementation of the strategy.
  So, i tried to make little fixes to the 1min code in order to fit in a 10min timeframe and have a larger backtest view.
  Results are not the best, but could be a starting point.

   

  1 user thanked author for this post.

  Paul

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  01/19/2020 at 8:25 PM #117246
  Paul

  @Francesco

  That timeframe is a lot better I think. What is a bit of issue with this strategy, is the way the entry is coded. The current way it is it checks first the long criteria, that’s why long has preference has the better results especially in a rising market.

  If you swap in the code below the long & short the strategy breaks down, because then it checks short criteria first. It’s because “elsif”

  if tradetime and tradeday then if pclong and mclong then orderpriceL=dayopen+((close/10000)*2)*pipsize buy positionsize contract at orderpriceL stop longtradecounter=longtradecounter + 1 tradecounter=tradecounter+1 elsif pcshort and mcshort then orderpriceS=dayopen-((close/10000)*2)*pipsize sellshort positionsize contract at orderpriceS stop shorttradecounter=shorttradecounter + 1 tradecounter=tradecounter+1 endif endif

  1 2 3 4 5 6 7 8 9 10 11 12 13

  if tradetime and tradeday then if pclong and mclong then orderpriceL=dayopen+((close/10000)*2)*pipsize buy positionsize contract at orderpriceL stop longtradecounter=longtradecounter + 1 tradecounter=tradecounter+1 elsif pcshort and mcshort then orderpriceS=dayopen-((close/10000)*2)*pipsize sellshort positionsize contract at orderpriceS stop shorttradecounter=shorttradecounter + 1 tradecounter=tradecounter+1 endif endif

  or you could code it like this below, which is in essence better; Then long above short, or visa versa give the same results. But then again results go down.

  if tradetime and tradeday then if pclong and mclong then orderpriceL=dayopen+((close/10000)*2)*pipsize buy positionsize contract at orderpriceL stop longtradecounter=longtradecounter + 1 tradecounter=tradecounter+1 endif if pcshort and mcshort then orderpriceS=dayopen-((close/10000)*2)*pipsize sellshort positionsize contract at orderpriceS stop shorttradecounter=shorttradecounter + 1 tradecounter=tradecounter+1 endif endif

  1 2 3 4 5 6 7 8 9 10 11 12 13 14

  if tradetime and tradeday then if pclong and mclong then orderpriceL=dayopen+((close/10000)*2)*pipsize buy positionsize contract at orderpriceL stop longtradecounter=longtradecounter + 1 tradecounter=tradecounter+1 endif if pcshort and mcshort then orderpriceS=dayopen-((close/10000)*2)*pipsize sellshort positionsize contract at orderpriceS stop shorttradecounter=shorttradecounter + 1 tradecounter=tradecounter+1 endif endif

   

  ps on the strategy you posted. if you remove

  if onmarket then if time >= closetime or (currentdayofweek=5 and time>=closetimefr) then sell at market exitshort at market endif endif

  1 2 3 4 5 6

  if onmarket then if time >= closetime or (currentdayofweek=5 and time>=closetimefr) then sell at market exitshort at market endif endif

  once closetime = 240000 // greater then 23.59 means it continues position overnight once closetimefr = 172900

  1 2	once closetime   = 240000 // greater then 23.59 means it continues position overnight once closetimefr = 172900

  then results go up.

  1 user thanked author for this post.

  Francesco

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  01/19/2020 at 8:48 PM #117249
  Francesco

  Good results with those fixes, but huge drawdown and not smooth curve 🙁

  Anyway, as i said it was a good starting point!

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  01/20/2020 at 6:43 PM #117344
  Paul

  the v6p in demo since 2 january

  2 users thanked author for this post.

  reb, Francesco

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  01/22/2020 at 9:41 AM #117493
  Wacko

  hello paul, well done for all your work. have you ever had canceled orders on the 1 min version because it happens 2 times out of 3. thank you in advance

  

  Wacko

  Participant

  Topics: 0

  Replies: 6

  Been thanked: 1 time
  01/22/2020 at 4:36 PM #117533
  Paul

  Hi Wacko. Yes it happens. it could be that the stop for entry is too near the current close to be placed and doesn’t meet the criteria of the minimum stop distance. In that case you can try to increase the 1.5/2 figures to something a bit higher.

  1 user thanked author for this post.

  Wacko

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  07/02/2020 at 12:35 PM #138077
  Franro

  Hi folks. I have been running daxv6p on live since 16th may. About 300€ profit.

   

  
  @Paul have you ran it since you posted in jan to 16th may without any improvements, if so whats the results?

  

  Franro

  Participant

  Topics: 25

  Replies: 219

  Been thanked: 23 times
  07/03/2020 at 12:40 AM #138198
  Paul

  no I haven’t ran it and didn’t look at the code for a long time!

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  07/03/2020 at 9:01 AM #138219
  Franro

  Well it has been profiting since may. Why did you stop it @Paul?

  

  Franro

  Participant

  Topics: 25

  Replies: 219

  Been thanked: 23 times
  07/03/2020 at 9:25 AM #138224
  Jaykay

  depending what version, i had several front testing,  deleted the data, (stupid) but it had hugh drawdown!!!

  

  Jaykay

  Participant

  Topics: 3

  Replies: 18

  Been thanked: 3 times
  07/03/2020 at 11:00 AM #138249
  Franro

   

  Well im talking about The v6p

  ---

  

  Franro

  Participant

  Topics: 25

  Replies: 219

  Been thanked: 23 times
  07/03/2020 at 11:22 AM #138253
  Paul

  it works mostly in uptrends, so it’s not good enough.

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  12/16/2020 at 1:40 PM #153972
  Indiana400

  Hi, i am currently testing the V5.3 on my demo account.  I have tried loading the D6p version and i get a warning message,

  Optimization limit.  Your back test exceeds the limit for repetitions for walk forward optimization.

  backtest.limit.optimisation.occurence,teasing

  How do i edit the variables ?  No lines are highlighted in red as not working…

  This looks like a great system by the way and i would really like to see it working for me.

  Lastly , what world time zone does it work best in?

  Many thanks

  

  Indiana400

  Participant

  Topics: 0

  Replies: 9

  Been thanked: 1 time
  12/16/2020 at 1:52 PM #153978
  nonetheless

  Open the optimization window, activate WF, reduce the number of repetitions, deactivate WF, close optimization window.

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
• Author
  Posts