Registered User Joined: 2/28/2005 Posts: 825
|
Whew! This made today kinda exciting. After reading & re-reading this entire thread several times over I changed the scan for 120 days and ended up with 34 ascending triangle possibilities.
Thanks for letting me "peak in" and get educated today.
art
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
That is the best part of these boards.... they teach well beyond the moment the discussion happens.
- Craig Here to Help!
|
|
Registered User Joined: 1/8/2005 Posts: 118
|
How would I scan for stocks breaking out of this ascending triangle?
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
lsaloga,
The Finding Continuation Triangles And Wedges With Telechart topic explores breakouts in relationship to triangle formations to some extent. The topic contains a rather lengthy and detailed discussion that will probably take some time to explore fully. You may also wish to review the Spotting trend changes using linear regression channel sorts and EasyScan Design for Breakouts and Bouncers topics for more ideas on identifying breakouts.
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 1/8/2005 Posts: 118
|
Thanks Bruce, I'll check these out
|
|
Registered User Joined: 2/11/2013 Posts: 78
|
Bruce:
I found this old thread via search. I'm using a similar, but not identical approach and want to run it by an expert.
Goal: To find Ascending Triangle Patterns formed within the last 20 days.
My approach:
(MAXH10.10 / MAXH10) * 100 < 102 AND
(MAXH10.10 / MAXH10) * 100 > 98 AND
(MINL10.10 / MINL10) * 100 > (MAXH10.10 / MAXH10) * 100
As you can see, I'm using MAX and MIN instead of the AVG mentioned above in the thread. I did so because if I'm drawing the lines to outline a triangle, I'm drawing it through Highs and Lows...not through the avg. high/lows. To make it an ascending triangle, the low slope must just be higher than the high slope...although obviously I could add sensitivity controls.
My scan seems to work better than the approach above. Is ther eany flaws in my method that I'm missing?
Thanks in advance.
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
Using maximums and minumums of the highs and lows instead of the averages will return different results. It is subjective, but they could be seen improved results as well. I do not see anything logically wrong with your approach.
If I were writing this on my own, I would probably dispense with division whenever possible. So it would look more like:
.98 * MAXH10 < MAXH10.10 AND MAXH10.10 < 1.02 * MAXH10 AND MINL10.10 * MAXH10 > MAXH10.10 * MINL10
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 7/30/2014 Posts: 2
|
Hello,
I am trying to set up the scan for acending triangle. I dont see the MAX value to set in the tool.
I set my first pcf condition
High's slope 20 day (VALUES: 90 to 110)
I could not set the second PCF condition, since i could not type or make entry MAX. It does not allow me type MAX or i dont find anything max.
Low's slope 20 day (VALUES: 105 to MAX)
Do i also add the below to along with above two pcf
Add this to the scan:
AVGC35/AVGC35.35*100
i wuld appriciate your replu
Thanks
Muthiah
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
If something requires setting the rangefinder to a Value to MAX in TC2000 v7, you could instead choose Greater Than when setting up the EasyScan Condition in TC2000 v12.5.
The same basic idea is true of setting Min to Value in TC2000 v7. This would be done by choosing Less Than when setting up the EasyScan Condition in TC2000 v12.5.
To add the other Condition Formula to the EasyScan, you would save it with a name. Then choose it by that name after selecting Add Condition in the Edit tab of the EasyScan.
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 1/18/2017 Posts: 1
|
I have tried the above formulas and they are not kicking out what I am looking for. I am trying to scan for ascending triangles in x periods. I have seen many stocks break out that have flat highs and rising lows for 5 to 8 days. Examples include TSRO in early February and CYBE in early January. Any thoughts? Thanks
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
The following specific checks for rising lows and dropping or flat highs over 5, 6, 7 and 8 bars.
MINL5 > MINL5.1 AND MAXH5 <= MAXH5.1 AND MINL6 > MINL6.1 AND MAXH6 <= MAXH6.1 AND MINL7 > MINL7.1 AND MAXH7 <= MAXH7.1 AND MINL8 > MINL8.1 AND MAXH8 <= MAXH8.1
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 9/18/2017 Posts: 20
|
Is there a way this can be done but sourced to a bollinger band %B 20 of volume?
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
There are a lot of formulas in this topic. Some could be converted to be based on a a Bollinger Band %B 20 of volume, but some could not (anything which requires a differentiation between highs and lows in order to work such as the original response from Craig_S). What specific formula(s) do you want to convert (I am not going to convert them all)?
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 9/18/2017 Posts: 20
|
I wanted to convert AVGL20/AVGL20.20*100 so that I could source it to something else more easily like you did for me before!
Thanks in advance bruce
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
The following assumes the width setting is set to 1.00, but that can be changed ba changing the two instances of / 1 in the formula to something else (v17 formulas follow).
AVG((C - AVGC20) / 2 / 1 / STDDEV20 + .5, 20) / AVG((C20 - AVGC20.20) / 2 / 1 / STDDEV20.20 + .5, 20) * 100
That said, the formula as is would be problematic in my mind.
It can cross through zero and that can create problems with dividing the two values.
The other issue is that if you are OK with the division by zero and negative value issues created by the above, then you would probably be better off getting rid of the instances of / 2 / 1 and + .5 in the formulas in order to center the %B on 0 instead of .5 (but this will make the division by zero and negative value situation worse if it is something you do care about).
AVG((C - AVGC20) / STDDEV20, 20) / AVG((C20 - AVGC20.20) / STDDEV20.20, 20) * 100
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 9/18/2017 Posts: 20
|
With that formula I couldnt swap all of the "C"s for the formula i wanted to input.. I was only able to swap the C's that stood alone. For example the avgc20 piece I cannot switch out the C..
Thanks
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
You mean like this?
AVG((C) - AVG(C, 20) / 2 / 1 / SQR(ABS(SUM((C) ^ 2, 20) - 20 * AVG(C, 20) ^ 2) / 20) + .5, 20) / AVG((C20) - AVG(C20, 20) / 2 / 1 / SQR(ABS(SUM((C20) ^ 2, 20) - 20 * AVG(C, 20) ^ 2) / 20) + .5, 20) * 100
Or this?
AVG((C) - AVG(C, 20) / SQR(ABS(SUM((C) ^ 2, 20) - 20 * AVG(C, 20) ^ 2) / 20), 20) / AVG((C20) - AVG(C20, 20) / SQR(ABS(SUM((C20) ^ 2, 20) - 20 * AVG(C, 20) ^ 2) / 20), 20) * 100
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 9/18/2017 Posts: 20
|
Looks about right! youre a blessing to the TC community bruce!
|
|
Registered User Joined: 9/18/2017 Posts: 20
|
unfortunately though, I am getting un mismatched parenthesis error..
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
Please try the formulas again.
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Guest-1 |