screening a strategy on each stock of us 500

• 03/26/2018 at 1:45 PM #66438
  Francesco78

  Hi, I was wondering if there is any way to screen a strategy on every single shares of the SP 500.

  Just to be clear. I want to see if a mean reverting strategy that works well on the SP 500 could be applied to a basket of stocks instead than to the whole index. It would be interesting to see if there is a particular basket that contribute to the “mean revertingness” of SP 500 and isolate it.

  Many thanks

  Francesco

  

  Francesco78

  Participant

  Topics: 43

  Replies: 242

  Been thanked: 77 times
  03/26/2018 at 1:48 PM #66439
  Nicolas

  We could simulate profit generated by a strategy of each individual stock of a shares list.. but it would be tricky, depends of the strategy itself! and it would only be possible on the last 254 bars (history limitation of ProScreener).

  If you could expose the strategy (or the probacktest code), I could lend an hand 🙂

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
  03/26/2018 at 2:18 PM #66444
  Francesco78

  Thank you Nicolas,

  the strategies I had in mind are the one that we published on the SP 500

  https://www.prorealcode.com/prorealtime-trading-strategies/sp-500-daily-rsi2-long-short-strategy/

  and

  https://www.prorealcode.com/prorealtime-trading-strategies/sp-500-reverting-strategy/

  Let me know if you think you can work something out of them and if I can do anything to help

  Thank you!

   

   

  

  Francesco78

  Participant

  Topics: 43

  Replies: 242

  Been thanked: 77 times
  03/26/2018 at 4:10 PM #66462
  Nicolas

  Quick example of the first one converted into a screener:

  a = 5 b = 20 cl = RSI[2]<a cs = RSI[2]>100-a if cl then long=1 short=0 entrylong=close endif if cs then long=0 short=1 entryshort=close endif if long and RSI[2]>100-(a+b) and close < open then long=0 gain=(close-entrylong)+gain //sell at market endif if short and RSI[2]<(a+b) then short=0 gain=(entryshort-close)+gain //exitshort at market endif screener[gain](gain)

  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 31

  a = 5 b = 20   cl = RSI[2]<a cs = RSI[2]>100-a   if cl then long=1 short=0 entrylong=close endif   if cs then long=0 short=1 entryshort=close endif   if long and RSI[2]>100-(a+b) and close < open then long=0 gain=(close-entrylong)+gain //sell at market endif   if short and RSI[2]<(a+b) then short=0 gain=(entryshort-close)+gain //exitshort at market endif   screener[gain](gain)

  Your a and b variables are optimized in ProBacktest, so I took dummy values for them. The gain is in points, not in money, depends of shares quantity you bought for each stock of course ..

  1 user thanked author for this post.

  Francesco78

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
  03/26/2018 at 4:12 PM #66463
  Nicolas

  Wait a minute, I think I made something wrong..

  1 user thanked author for this post.

  Francesco78

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
  03/26/2018 at 4:18 PM #66464
  Nicolas

  Hmmm, please review my code, I think it should be good now 🙂

  a = 5 b = 20 cl = RSI[2]<a cs = RSI[2]>100-a if cl and long=0 then long=1 short=0 entrylong=close endif if cs and short=0 then long=0 short=1 entryshort=close endif if long and RSI[2]>100-(a+b) and close < open then long=0 gain=(close-entrylong)+gain //sell at market endif if short and RSI[2]<(a+b) then short=0 gain=(entryshort-close)+gain //exitshort at market endif screener[gain](gain)

  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 31

  a = 5 b = 20   cl = RSI[2]<a cs = RSI[2]>100-a   if cl and long=0 then long=1 short=0 entrylong=close endif   if cs and short=0 then long=0 short=1 entryshort=close endif   if long and RSI[2]>100-(a+b) and close < open then long=0 gain=(close-entrylong)+gain //sell at market endif   if short and RSI[2]<(a+b) then short=0 gain=(entryshort-close)+gain //exitshort at market endif   screener[gain](gain)

   

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
  03/26/2018 at 4:34 PM #66465
  Francesco78

  Thank you Nicolas, the code is fine but I am not very confident with screeners. Let me understand, if I select “daily” screening the resultsa  are computed on the last 2 years right?

  

  Francesco78

  Participant

  Topics: 43

  Replies: 242

  Been thanked: 77 times
  03/26/2018 at 4:54 PM #66471
  Nicolas

  Yes, beginning 254 bars ago from now.

  1 user thanked author for this post.

  Francesco78

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
• Author
  Posts