Last Position and Performance fallacy

• 06/11/2016 at 2:00 AM #9229
  grizzly

  Hi Everyone,

  Could anyone help me get me started in creating some code ? I’ve looked at many functions (positionperf, countofposition) and can’t figure this one out.

  I’m looking to count previous positions. My thinking is to reject “gamblers fallacy”, i.e. Entries are not statistically independent when an overall system is following a trend.

  For example, if last my last 3 closed positions in profit were: Long , Long , Long => Count of 3 => The likelihood of a 4th Long resulting in gains is minimal, and therefore avoid taking a 4th Long (given prior conditions are met).

  Thanks in advance,

  Regards,

   

  

  grizzly

  Participant

  Topics: 5

  Replies: 57

  Been thanked: 16 times
  06/11/2016 at 5:35 AM #9230
  grizzly

  This will only show open positions and not closed.

  countoflongshares[1]>0 and countoflongshares[2]>0 and countoflongshares[3]>0

  1	countoflongshares[1]>0 and countoflongshares[2]>0 and countoflongshares[3]>0

   

  

  grizzly

  Participant

  Topics: 5

  Replies: 57

  Been thanked: 16 times
  06/11/2016 at 5:06 PM #9250
  Nicolas

  I think you can use a loop over the last n ‘positionperf’, to count if their profit were in the same row, and then modify a variable to tell your strategy to trade or not:

  count = 0 for n = 1 to 3 do if positionperf(n)>0 then count=count+1 endif next if count = 3 then allowtrade = 0 else allowtrade = 1 endif

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

  count = 0   for n = 1 to 3 do if positionperf(n)>0 then   count=count+1 endif next   if count = 3 then allowtrade = 0 else allowtrade = 1 endif

   

  2 users thanked author for this post.

  David, grizzly

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4774 times
• Author
  Posts