Screen back in time for historical data

• 05/12/2017 at 9:52 AM #35298
  Mikael Andersson

  Hi,

  I was wondering if it is possible to screen back in time for historical data? Now when I start a screener I only get today’s screened stocks.

  Been looking through the threads but can’t really find/understand how to do this! Any help would be much appreciated.

  Best regards,
  Mikael

  

  Mikael Andersson

  Participant

  Topics: 2

  Replies: 10

  Been thanked: 2 times
  05/12/2017 at 10:47 AM #35304
  GraHal

  How about turning your screener into an indicator and then running the indicator on Backtest using the GRAPH function on whatever criteria it is that you want to see historically?

  Screener function is designed for now / real time.

  Indicator function can be used on past / historical data.

  Cheers
  GraHal

   

  1 user thanked author for this post.

  Mikael Andersson

  

  GraHal

  Participant

  Topics: 214

  Replies: 8826

  Been thanked: 2198 times
  05/12/2017 at 11:36 AM #35306
  Mikael Andersson

  Hi,

  Thankyou for your answer!

  I’m new to this scenario but wouldn’t it be to difficult to “screen” thousands of stocks based on indicator?

  Best regards,
  Mikael

  

  Mikael Andersson

  Participant

  Topics: 2

  Replies: 10

  Been thanked: 2 times
  05/12/2017 at 12:51 PM #35308
  GraHal

  Yes impossible! 🙂

  If your screener criteria related to, for example DLow and you wanted to see if this level was crossed over 5 days ago then why not use DLow(5) … would this approach give you what you need?

  I am intrigued  … please might you say why you want to see what stocks met your criteria at a certain time in the past?

  Thank You
  GraHal

  PS Easy way might be  … put your screener code on here and then we may suggest a solution?

  

  GraHal

  Participant

  Topics: 214

  Replies: 8826

  Been thanked: 2198 times
  05/12/2017 at 2:42 PM #35335
  Mikael Andersson

  I want to find out what indicators from different screeners would be most profitable against historical data. Finding the common factors for a succesful screener in the long run. I was hoping this could be done with screener or programmatically rather than manually verifying each stock against a set of different indicators.

  It may work if all indicators can be set back in time as you suggest. I will look more into this!

  Anyway, I have no idea what I’m talking about and it’s just an idea I have in my head right now! 🙂

   

  

  Mikael Andersson

  Participant

  Topics: 2

  Replies: 10

  Been thanked: 2 times
  05/12/2017 at 7:29 PM #35356
  GraHal

  Hahahaha nice one, made me laugh cos I do that …  good to be open to possibilities … not many folk around with an open mind in my experience!

  I don’t know fully what I’m talking about sometimes when I make comments on here, but one (bad) idea might spark another (good) idea?? 🙂

  You maybe could choose the instrument that you are most likely to trade eventually (Dax, EURUSD etc) and convert Screeners into Bots (can be easy …. even I can do it! :)) and then Backtest to see if profitable?

  I’ve done above before, but can’t recall any definite conclusion I came to … probably because I got ‘swamped’ with the effort involved due to so many Screeners on here?

  So in case above is a bit brief / cryptic … when screener criteria is met during Backtest (after convert to Bot) then a Trade executes … Long or Short as appropriate.

  GraHal

  1 user thanked author for this post.

  Mikael Andersson

  

  GraHal

  Participant

  Topics: 214

  Replies: 8826

  Been thanked: 2198 times
• Author
  Posts