Ensuring statistical significance when optimising intra-day systems

• 01/03/2022 at 1:57 PM #184399
  borderlineJim

  If optimising over 200k bars: what amount of trades would be considered statistically significant?

  How many trades are required to ensure robustness testing is effective?

  Obviously a few hundred trades is not going a good indication of well a startegy performs in the future.  When traditional WFA is taken into consideration, using 5 iterations for example ( IS and OOS), this means even less trades are utilised for extracting optimial parameter values.  I’m struggling to find answers…

  Are there any ways to increase statistical significance that experienced traders use ( for example, testing on different markets, adjusting parameter ranges to check for ‘sensitivity’ etc etc)

  thanks

  Jim

  

  borderlineJim

  Participant

  Topics: 15

  Replies: 24

  Been thanked: 8 times
  01/03/2022 at 3:25 PM #184414
  nonetheless

  this is worth a look if you haven’t seen it already:

  https://www.prorealcode.com/topic/day-month-year-strategy-robustness-tester/

  1 user thanked author for this post.

  borderlineJim

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  01/07/2022 at 2:11 PM #184838
  borderlineJim

  ahhh interesting!

  May be just what I am looking for.  Thank you Sir!!

  

  borderlineJim

  Participant

  Topics: 15

  Replies: 24

  Been thanked: 8 times
  01/07/2022 at 2:57 PM #184843
  nonetheless

  this is the version that most people seem to use: #106992

  then you need the excel template to analyse the results: #107211

  be sure to read all the instructions as it’s a bit fussy, but worth it once you understand how it works and what it’s doing.

  1 user thanked author for this post.

  borderlineJim

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
• Author
  Posts