Simple Strategy : TF 10S / Bollinger / RSI

• 05/26/2020 at 9:39 PM #133370
  Scooby

  Hi

  No I didn’t change the structure of the code.

  I Only change the bollinger bands, Stoch and SL/TP.

  I had duplicated the structure 2 or 3 times per algo to get more positions per day and a better results (I ope) for a long time.

  Following my current version for the DJI, you can adapt the values and tryout for the other instrument with the others parameters.

   

  Let me know if you find something to improve it 🙂

  2 users thanked author for this post.

  bertrandpinoy, OboeOpt

  

  Scooby

  Participant

  Topics: 9

  Replies: 52

  Been thanked: 26 times
  05/26/2020 at 10:21 PM #133374
  OboeOpt

  Thank you @scooby!

  Then I know where to start with the parameters! I don´t really understand what you mean with “duplicated the structure”, but I have to study your code more.

  I only have 100k to test on but I will do some curvefitting 😉

  (Looks like v1 performed better for the last 100k)

  

  OboeOpt

  Participant

  Topics: 11

  Replies: 101

  Been thanked: 7 times
  05/26/2020 at 10:43 PM #133377
  Scooby

  I duplicated the structure with for example one Bollinger at 20/20. A second one at 40/30 etc.. for each one I adapted the sl/TP and the stoch

  Like that I have more trades and a more linear curve.

  Yes it depend of the days 😉

  1 user thanked author for this post.

  OboeOpt

  

  Scooby

  Participant

  Topics: 9

  Replies: 52

  Been thanked: 26 times
  05/28/2020 at 10:00 PM #133663
  Scooby

  The current result in real for me. I know that there is not so much historic but it can be fine 🙂

  I know that noone think that 30 TF is a good thing, then I gonna work on the 1min TF ^^

  1 user thanked author for this post.

  swedshare

  

  Scooby

  Participant

  Topics: 9

  Replies: 52

  Been thanked: 26 times
  05/29/2020 at 10:20 AM #133759
  Francesco

  Ehy Scooby, it’s not that we don’t like the low timeframes systems… yes they are very difficult to manage, probably they’re more a dream then a reality, but the point is another one.

  I take a look at your codes and i noticed that you optimized some values of the system with precise numbers and without a logical sense, only and exclusively to make the previous 7 days profitable.

  Imagine that if doing this kind of overfitting could be very dangerous on larger timeframe systems, how could be this dangerous on a 7 days backtest? I invite you to pay attention by operating live with these systems

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
• Author
  Posts