What's your Favorite chart pattern?

Channel projection based on two lows and one high projecting to the next high.
Screen Shot 2021-06-30 at 10.23.11 PM.png
 
Mine is a minor gap up continuation.
Continuation is the operative word it'd seem.
What would be a couple of ways to recognize Continuation that a trader can watch for using a KC Continuation Trade at the Open?

Do you generally wait till 9:45 to jump on that KC Continuation Trade or if it (can it?) triggers right off the reel go grab a handful and take a bite?

As for minor gap up, how do you try to quantify that?
Does your Prior 2-Day High give all the indication of major vs. minor that an experienced eye would need?
 
Last edited:
I like a pattern where a detected cyclical inflection point is near a minimum followed by a predicted cyclical inflection point near a maximum.

Example for SAIC Science Applications International Corporation:
upload_2021-6-30_22-44-50.png

y axis is price. x axis is calendar days from 20210228 through 20210630 (123 calendar days). + signs are close prices (adjusted for splits and dividends and interpolated when the market was closed). smooth curve is parabola representing the trend (10 overlapping models). wavy curve is trend plus three cosine waves for the cyclical part (10 overlapping models).

A sample model is:
Code:
y = 91.2834777832031  -  0.141938492655754 * x  +  0.00139776105061173 * x * x
    +  2.9154794216156 * cos(twopi / 40.1655085347047 * x  +  3.46529150009155)
    +  2.08321166038513 * cos(twopi / 29.6157825190041 * x  +  2.67856240272522)
    +  1.53341674804688 * cos(twopi / 23.8787345374999 * x  +  1.54169023036957) ;
The trend is the parabola: 91.2834777832031 - 0.141938492655754 * x + 0.00139776105061173 * x * x.
The sum of the amplitudes is 2.9154794216156 + 2.08321166038513 + 1.53341674804688 == 6.53210783004761 which is the model's maximum absolute distance from the trend.
Software calculated the numeric values in the model (not including twopi) with genetic optimization for an attempted least squares fit.

The chart with the trend removed from this model (detrended) and the prices subtracted from the trend are:
upload_2021-6-30_22-45-5.png

This shows a cyclical inflection point on 20210630 (-5.40582674441694 away from the trend), and the next predicted cyclical inflection point on 20210715 (6.33096122298606 away from the trend).
These inflection points are close to the minimum and maximum for the combined cosine waves.
This predicts the next day (20210701) might be a good time to enter long and 20210715 might be a good time to exit.

The three cosine waves for this model are:
upload_2021-6-30_22-45-37.png

This visually shows the three cosine waves aren't too far from their minimums at the end of the actual data (20210630). And the three cosine waves are predicted to be close to their maximum on 20210715.
 
Last edited:
I like a pattern where a detected cyclical inflection point is near a minimum followed by a predicted cyclical inflection point near a maximum.

Example for SAIC Science Applications International Corporation:
View attachment 262200
y axis is price. x axis is calendar days from 20210228 through 20210630 (123 calendar days). + signs are close prices (interpolated when the market was closed). smooth curve is parabola representing the trend (10 overlapping models). wavy curve is trend plus three cosine waves for the cyclical part (10 overlapping models).

A sample model is:
Code:
y = 91.2834777832031  -  0.141938492655754 * x  +  0.00139776105061173 * x * x
    +  2.9154794216156 * cos(twopi / 40.1655085347047 * x  +  3.46529150009155)
    +  2.08321166038513 * cos(twopi / 29.6157825190041 * x  +  2.67856240272522)
    +  1.53341674804688 * cos(twopi / 23.8787345374999 * x  +  1.54169023036957) ;
The trend is the parabola: 91.2834777832031 - 0.141938492655754 * x + 0.00139776105061173 * x * x.
The sum of the amplitudes is 2.9154794216156 + 2.08321166038513 + 1.53341674804688 == 6.53210783004761 which is the model's maximum absolute distance from the trend.
Software calculated the numeric values in the model (not including twopi) with genetic optimization for an attempted least squares fit.

The chart with the trend removed from this model (detrended) and the prices subtracted from the trend are:
View attachment 262201
This shows a cyclical inflection point on 20210630 (-5.40582674441694 away from the trend), and the next predicted cyclical inflection point on 20210715 (6.33096122298606 away from the trend).
These inflection points are close to the minimum and maximum for the combined cosine waves.
This predicts the next day (20210701) might be a good time to enter long and 20210715 might be a good time to exit.

The three cosine waves for this model are:
View attachment 262202
This visually shows the three cosine waves aren't too far from their minimums at the end of the actual data (20210630). And the three cosine waves are predicted to be close to their maximum on 20210715.
Does this remind me of BioRhythms?
I'm trying to get this.
Is it an additive effect of areas "under" the curve in clumsy-speak? Thx.
 
Last edited:
Is it an additive effect of areas "under" the curve in clumsy-speak?

Yes, the cosine waves are added together to get the detrended distance from the parabolic trend. So they can be up to 6.53210783004761 above or below the parabola.
 
If you can find and run stats on a written set of instructions which prove that situation's probable outcome over a sizable population of trades, especially one that allows for a setup that gives you a stop so tight it squeaks, say maybe a pullback to a hammer at support, then maybe you can pullback some frogs to your pond.

That's the goal of Price TA... identifying correlations. What you have described is one of them.

KISS, as always. :)
 
Back
Top