Awesome Oscillator Crossover

• 10/13/2023 at 4:30 PM #222394
  OllieB

  Hi All,

  I am trying to code a screener which returns an awesome oscillator crossover.  This means the Ao goes from below 0 to above 0, or from above 0 to below 0.

  The screener code is attached.  However, it is not behaving as expected.  For some of the values returned the chart (with an indictor coded up in the same way) does show to AO going from negative to positive, but in other cases i get stocks returned where the AO has been positive for some time.  I have (tried) to attach a screenshot which shows this.  The top section is price (which can be ignored) but the lower section shows the AO and its been negative for approx 30 time periods (weeks in this case).

  Can anyone shed any clues as to this strange behaviour please?  The stock given as an example of unexpected behaviour is FBIO.

  Many thanks.

  Awesome Oscillator Crossover

  // sample screener code TIMEFRAME(WEEKLY) // Awesome oscillator has crossed the 0 line AO = (Average[AO1](MedianPrice) - Average[AO2](MedianPrice)) cAOup = ((AO[1] < 0) AND (AO > 0)) cAOdn = ((AO[1] > 0) AND (AO < 0)) // Set vals for screener if (cAOup) then c1WB = 1 elsif (cAOdn) then c1WB = -1 endif SCREENER [cAOup OR cAOdn](c1WB as "WB")

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

  // sample screener code TIMEFRAME(WEEKLY)   // Awesome oscillator has crossed the 0 line AO    = (Average[AO1](MedianPrice) - Average[AO2](MedianPrice)) cAOup = ((AO[1] < 0) AND (AO > 0)) cAOdn = ((AO[1] > 0) AND (AO < 0))   // Set vals for screener if (cAOup) then   c1WB = 1 elsif (cAOdn) then   c1WB = -1 endif   SCREENER [cAOup OR cAOdn](c1WB as "WB")

   

  

  OllieB

  Participant

  Topics: 10

  Replies: 24

  Been thanked: 4 times
  10/13/2023 at 5:43 PM #222397
  JS

  Hi Ollie,

  Why unexpected behavior?

  You determine the distance between two averages and in the case of FBIO there is a huge downtrend going on, so the “fast” average remains below the “slow” average…

  This can take a very long time indeed…

  

  JS

  Participant

  Topics: 22

  Replies: 1423

  Been thanked: 557 times
  10/13/2023 at 9:19 PM #222410
  OllieB

  Hi JS,

  Thanks for taking the time to reply.

  Unexpected as what i’m looking for is, in a downtrend, the AO to be under the zero line for the current time period but the previous time period the AO is above the zero line.  I am looking for the AO to cross from positive to negative.

  This stock shows current and previous period both below the zero line.

  Thanks,

  

  OllieB

  Participant

  Topics: 10

  Replies: 24

  Been thanked: 4 times
  10/13/2023 at 10:00 PM #222411
  JS

  Hi,

  Do you mean something like that…?

  

  JS

  Participant

  Topics: 22

  Replies: 1423

  Been thanked: 557 times
  10/14/2023 at 9:19 AM #222423
  OllieB

  Hi,

  Sort of.

  I don’t think i was clear in my first post.  The FBIO stock shows up as a result of a screener.  I have used the same code from the screener to code the indicator.  This shows that the indicator has not crossed down in the last period, its been down for ages, and therefore should not be appearing in the screener as a recent crossdown.

  Your chart, with new indicator, shows the point at which the AO crossed down?  If so, then that could work, yes!  That could be coded for a screener and only return stocks which are crossing up/down.

  Thanks.

  

  OllieB

  Participant

  Topics: 10

  Replies: 24

  Been thanked: 4 times
  10/14/2023 at 9:34 AM #222425
  JS

  Hi Ollie,

  I see what you mean now…

  What values do you use for AO1 and AO2…?

  

  JS

  Participant

  Topics: 22

  Replies: 1423

  Been thanked: 557 times
  10/14/2023 at 10:41 AM #222426
  OllieB

  Thanks.

  I -read my initial post and it wasn’t very clear.  Seemed perfectly clear at the initial time of writing!  🙂

  I am using 2 and 33, slightly arbitrary but those were the values for that screenshot.

  Appreciate your help!

  

  OllieB

  Participant

  Topics: 10

  Replies: 24

  Been thanked: 4 times
  10/14/2023 at 10:43 AM #222427
  JS

  Hi @OllieB

  I’ve tweaked your screener a bit and at least I can’t get FBIO as a result anymore…

  Results from screeners are sometimes inexplicable (you’re not alone) but I think it could be due to several factors…

  1. Bad / polluted data
  2. Rounding in penny stocks
  3. Dividend adjustments
  4. …

  Awesome Oscillator

  TIMEFRAME(Weekly) // Awesome oscillator has crossed the 0 line Avg20=Average[20](Close) Avg10=Average[10](Close) C1=Avg10 Crosses Over Avg20 C2=Avg10 Crosses Under Avg20 //cAOup = ((AO[1] < 0) AND (AO > 0)) //cAOdn = ((AO[1] > 0) AND (AO < 0)) // Set vals for screener if C1 then c1WB = 1 elsif C2 then c1WB = -1 endif SCREENER [C1 or C2](c1WB as "WB")

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

  TIMEFRAME(Weekly) // Awesome oscillator has crossed the 0 line Avg20=Average[20](Close) Avg10=Average[10](Close) C1=Avg10 Crosses Over Avg20 C2=Avg10 Crosses Under Avg20 //cAOup = ((AO[1] < 0) AND (AO > 0)) //cAOdn = ((AO[1] > 0) AND (AO < 0)) // Set vals for screener if C1 then c1WB = 1 elsif C2 then c1WB = -1 endif SCREENER [C1 or C2](c1WB as "WB")

  

  JS

  Participant

  Topics: 22

  Replies: 1423

  Been thanked: 557 times
  10/14/2023 at 4:37 PM #222454
  OllieB

  Hi,

  Thanks for the changes.  I’ve made some modifications to the screener to match the above, apart from changing “close” on the Average to “Median” to better reflect the Awesome oscillator.  I also plotted 2 MAs on the chart to help see where the cross overs are happening.

  48 of the 50 stocks returned from the screener looked good, i could see they matched the criteria.  However, 2 did not!!  I guess I’ll have to live with the 4% margin of error.

  You can see that in the screenshot, the MAs have not crossed, nor is the indicator showing a crossover, yet its returned in the screener!!

  Thanks.

  

  OllieB

  Participant

  Topics: 10

  Replies: 24

  Been thanked: 4 times
  10/14/2023 at 4:47 PM #222457
  robertogozzi

  Try this one:

  TIMEFRAME(Weekly) // Awesome oscillator has crossed the 0 line Avg20=Average[20](Close) Avg10=Average[10](Close) C1=Avg10 Crosses Over Avg20 C2=Avg10 Crosses Under Avg20 //cAOup = ((AO[1] < 0) AND (AO > 0)) //cAOdn = ((AO[1] > 0) AND (AO < 0)) c1WB = 0 // Set vals for screener if C1 then c1WB = 1 elsif C2 then c1WB = -1 endif SCREENER [c1WB AND range <> 0](c1WB as "WB")

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

  TIMEFRAME(Weekly) // Awesome oscillator has crossed the 0 line Avg20=Average[20](Close) Avg10=Average[10](Close) C1=Avg10 Crosses Over Avg20 C2=Avg10 Crosses Under Avg20 //cAOup = ((AO[1] < 0) AND (AO > 0)) //cAOdn = ((AO[1] > 0) AND (AO < 0)) c1WB = 0 // Set vals for screener if C1 then c1WB = 1 elsif C2 then c1WB = -1 endif SCREENER [c1WB AND range <> 0](c1WB as "WB")

   

  

  robertogozzi

  Moderator

  Topics: 69

  Replies: 16469

  Been thanked: 3643 times
  10/14/2023 at 4:47 PM #222458
  JS

  Hi Ollie,

  That’s right, there are always a few results that don’t meet the criteria… it remains strange.

  Otherwise, looks fine…

  

  JS

  Participant

  Topics: 22

  Replies: 1423

  Been thanked: 557 times
  10/14/2023 at 4:51 PM #222460
  OllieB

  A big thanks for all your help.

  Until next time….

  2 users thanked author for this post.

  robertogozzi, JS

  

  OllieB

  Participant

  Topics: 10

  Replies: 24

  Been thanked: 4 times
• Author
  Posts