As of 11/08/2024
Indus: 43,989 +259.65 +0.6%
Trans: 17,354 +143.48 +0.8%
Utils: 1,032 +20.02 +2.0%
Nasdaq: 19,287 +17.32 +0.1%
S&P 500: 5,996 +22.44 +0.4%
|
YTD
+16.7%
+9.2%
+17.0%
+28.5%
+25.7%
|
43,100 or 41,250 by 11/15/2024
16,800 or 15,700 by 11/15/2024
1,075 or 1,000 by 11/15/2024
19,000 or 17,600 by 11/15/2024
5,900 or 5,600 by 11/15/2024
|
As of 11/08/2024
Indus: 43,989 +259.65 +0.6%
Trans: 17,354 +143.48 +0.8%
Utils: 1,032 +20.02 +2.0%
Nasdaq: 19,287 +17.32 +0.1%
S&P 500: 5,996 +22.44 +0.4%
|
YTD
+16.7%
+9.2%
+17.0%
+28.5%
+25.7%
| |
43,100 or 41,250 by 11/15/2024
16,800 or 15,700 by 11/15/2024
1,075 or 1,000 by 11/15/2024
19,000 or 17,600 by 11/15/2024
5,900 or 5,600 by 11/15/2024
| ||
Initial release on 6/24/2021.
Multi-peaks are a new pattern I discovered/explored in June 2021, rising from my desire to find the reason why stocks tumble by more than 20%. What I found was that multiple peak patterns are likely not the cause (31% of regular multi-peaks and 44% of the variants that were wider than 91 days saw price drop 20% or more after the breakout). Hmm. Maybe those stats aren't as bad as I thought.
Maybe if you see a multi-peak, you should be worried.
Overall performance rank (1 is best): 10th out of 37 (where 1 is best)
Break even failure rate: 21%
Average decline: 16%
Pullback rate: 63%
Percentage meeting price target: 54%
The above numbers are based on 657 perfect trades. See the glossary for definitions. |
Multi-peak example |
I followed these guidelines when identifying the multi-peak pattern.
Characteristic | Discussion |
Price trend | Upward. The trend start (intermediate-term uptrend 3 to 6 months) must be below the bottom of the multi-peak. Ignore overshoot and ignore the pattern in an upward retrace of a downtrend. |
Shape | A flat top pattern with an irregular (or not) bottom. |
Peaks | Must have at least four peaks near the same price. No peak should soar above the others (as in the head of a head-and-shoulders top) |
Separation | No minimum, but major peaks I looked at were separated by at least a week. |
Confirmation | Price confirms the pattern and stages a downward breakout when it closes below the bottom of the pattern (the lowest low between the four peaks). Upward breakouts invalidate the pattern. |
Volume | Rising/falling or heavy breakout volume have no influence on performance. |
Patterns | I don't count multi-peak patterns if they were head-and-shoulders tops or included a pronounced 2B pattern. Rectangle tops were fine, though. A rectangle top can be a multi-peak pattern if it has four or more peaks, but a multi-peak pattern may not be a rectangle (because the bottom can be an irregular shape). Other patterns like ascending/descending triangles or broadening patterns were ignored. As long as the multi-peak obeyed these guidelines then I accepted the pattern. |
Variants | Some variants (2B type) performed better. I'll discuss them later in this article. |
Trading Tactic | Explanation |
Measure Rule
|
Measure rule | Reference the figure to the right. Take the height of the pattern from A to B (highest peak to lowest valley between the four tops) and subtract it from the price of the lowest valley (B) to get a target (C). Price reaches the target 54% of the time, so be conservative in your assessment. | |
Stop | Two percent of the patterns will see price rise above the top of the pattern on its way from the breakout to the ultimate low. Another 14% stop above the middle of the pattern but remain below the price of the pattern's top. The vast majority (70%) will return to stop somewhere above the bottom of the pattern but below the middle. | |
Price reversal | A trend change (20% or more drop) happens only 31% of the time. | |
Confirmation/breakout | The pattern confirms when price closes below the lowest valley between the four peaks. That's also called the breakout. |
This is an example of a nice looking multi-peak chart pattern. This one has four peaks, ABCD, which is the minimum required for the pattern. The four top out near the same price. The four peaks need to form a resistance barrier to an upward rise.
The pattern confirms as a valid one when price closes below the lowest valley between the two peaks. I show that in this example as red line E.
However, it might work better to trade confirmation of the double top and enter at F (a close below the blue line). That will get you in sooner than E, increasing profit and lowering the risk of a failed trade. I haven't investigated how often this works or how much profit increases.
At G, a pullback completes in late August or early September.
In this example, I show the four peaks as ABCD. The tops peak near the same price, but allow variations.
The pattern confirms as valid after the stock closes below blue line F. Notice that this pattern always forms as a top. By that I mean the move from E is upward but it need not be as long as that shown here.
The pattern needs something to reverse, though. Notice that the drop after F stops at support, circled.
This is the first of two variants I wanted to mention. Peak A is a 2B top appearing at the end of a multi-peak pattern.
On the first pass of my statistics, I included such patterns. On later passes and for the statistics on this page, I did not include it. However, excluding such patterns degraded the performance of the multi-peak. So that's why I mention it here.
In this version, the only difference between a valid multi-peak and this pattern is a 2B sitting near the right end of it. I show that as peak A. Peak A is the fifth top in the series in this example but you'll often see it as peak four. It shouldn't be too far above the other peaks but it should be obvious.
The multi-peak, 2B pattern confirms when price closes below the bottom of the lowest valley. I show that as blue line B.
I included many patterns in the following statistics, not just the 2B. For comparison, the statistics I used for the pattern (which exclude variants) are shown in parenthesis.
The average decline, failure rate, and measure rule target hit rate improve when variants are included. If you want to boost performance even more, look for patterns from first peak to last of at least 91 days (3 months). The average decline rises to 20% and failures drop to 12%.
In this variation, the multi-peak pattern is really a complex head-and-shoulders top. This one has two shoulders to the left and another two to the right of the head.
The shoulders are about equally spaced reflections about the head of their corresponding opposite shoulder. Think symmetry. I did not include this variation in the statistics.
-- Thomas Bulkowski
Support this site! Clicking any of the books (below) takes you to
Amazon.com If you buy ANYTHING while there, they pay for the referral.
Legal notice for paid links: "As an Amazon Associate I earn from qualifying purchases."
My Stock Market Books
|
My Novels
|