Predicting Breakout

As of 11/27/2023

Indus: 35,333 -56.68 -0.2%

Trans: 14,904 -190.22 -1.3%

Utils: 866 -0.47 -0.1%

Nasdaq: 14,241 -9.83 -0.1%

S&P 500: 4,550 -8.91 -0.2%

YTD

+6.6%

+11.3%

-10.5%

+36.1%

+18.5%

Targets Overview: 11/14/2023

35,450 or 34,050 by 12/01/2023

15,400 or 14,150 by 12/01/2023

900 or 825 by 12/01/2023

14,450 or 13,550 by 12/01/2023

4,600 or 4,350 by 12/01/2023

Chart Pattern Indicator:

on 11/17/23

As of 11/27/2023

Indus: 35,333 -56.68 -0.2%

Trans: 14,904 -190.22 -1.3%

Utils: 866 -0.47 -0.1%

Nasdaq: 14,241 -9.83 -0.1%

S&P 500: 4,550 -8.91 -0.2%

YTD

+6.6%

+11.3%

-10.5%

+36.1%

+18.5%

Targets Overview: 11/14/2023

35,450 or 34,050 by 12/01/2023

15,400 or 14,150 by 12/01/2023

900 or 825 by 12/01/2023

14,450 or 13,550 by 12/01/2023

4,600 or 4,350 by 12/01/2023

Chart Pattern Indicator:

on 11/17/23

Bulkowski on Predicting the Breakout

Predicting Breakout: Overview

Can we predict the breakout using the length of the trend start to the chart pattern versus the length of the chart pattern? Short answer: No. (sigh)

A chart of the time from the trend start to the start of a rectangle top versus the width of the chart pattern (the rectangle) shows there's no relationship between the rise leading to the rectangle and when a breakout will occur (that is, the length of the chart pattern).

Predicting Breakout: Method

I show a chart of MasterCard on the monthly scale because it illustrates what I was looking for.

I was interested in buying a security that looked like MA, but on the daily scale (so use your imagination). Is there a relationship of the length of the inbound price trend (the rise from A to the start of B) to the width of pattern B?

Experience says that there was, that the longer it took price to rise to the rectangle, the longer price would move sideways in the rectangle. Knowing the relationship between the two could help me better time the entry (buy just before an anticipated breakout upward).

To conduct the research, I used rectangle top chart patterns (shown here as B, the two horizontal red lines). I found 1,070 rectangle tops over the decades since I've been cataloging chart patterns.

I used the trend start, which is the lowest low before the pattern begins and just after price drops by at least 20%. See the glossary (trend start) for an exact definition, if you're interested. I scanned each rectangle and made sure to eliminate overshoot or undershoot which can interfere with where my computer thinks the trend begins. If overshoot or undershoot got in the way, I adjusted where the trend started manually, otherwise, I relied on the computer to calculate the start.

I only looked at the length of the trend start and the width of the rectangle, not whether the pattern appeared in a bull or bear market, or had an up or down breakout. Data started in October 1995 to April 2023 using 359 stocks taken from the database of securities I follow daily. Not all stocks covered the entire period. Market cap varies.

Predicting Breakout: Results

The figure shows the results of plotting the time from the trend start to the start of the rectangle on the horizontal axis versus the pattern's width on the vertical axis.

If there was a relationship between the two, say one day for each, you'd see the points lining up following a diagonal. The diagonal might be steep or shallow, depending on the ratio of the two widths.

Instead, we see narrow patterns that have short or long rises from the trend start.

Bulkowski on Predicting the Breakout