As of 11/20/2024
Indus: 43,408 +139.53 +0.3%
Trans: 17,002 -26.31 -0.2%
Utils: 1,055 +1.25 +0.1%
Nasdaq: 18,966 -21.33 -0.1%
S&P 500: 5,917 +0.13 +0.0%
|
YTD
+15.2%
+6.9%
+19.7%
+26.3%
+24.1%
|
 46,000 or 43,000 by 12/01/202418,000 or 16,600 by 12/01/20241,075 or 1,000 by 12/01/202420,000 or 18,400 by 12/01/20246,100 or 5,800 by 12/01/2024 |
As of 11/20/2024
Indus: 43,408 +139.53 +0.3%
Trans: 17,002 -26.31 -0.2%
Utils: 1,055 +1.25 +0.1%
Nasdaq: 18,966 -21.33 -0.1%
S&P 500: 5,917 +0.13 +0.0%
|
YTD
+15.2%
+6.9%
+19.7%
+26.3%
+24.1%
| |
|
46,000 or 43,000 by 12/01/202418,000 or 16,600 by 12/01/20241,075 or 1,000 by 12/01/202420,000 or 18,400 by 12/01/20246,100 or 5,800 by 12/01/2024 | ||
Initial release: August 29, 2017.
This article discusses a new trend change technique for investments (long term holdings). It reliably signals when you should sell a buy-and-hold stock.
Almost half (41%) of all long term channels (uptrends, really) I looked at started near the end of a bear market, but few (20%) ended because one approached.
When a channel shows a sell signal formed by three black candles (42% of channels have such a signal), the post-signal drop averages 43%. False sell signals occur 21% of the time.
After a sell signal occurs, it takes 3.7 years before a stock recovers to close above the top of the channel.
I was looking at charts on the monthly scale showing price channels, wondering how a stock behaves after a channel ends. Questions formed. Should I sell a long term holding? Should I even buy a stock showing one of the channels, knowing that sooner or later they end? Did a bear market end the channel? If not, did the channel's end precede the bear market and if so, by how long?
I started research to answer these questions when I found a reliable sell signal to mark the end of an up-sloping trend.
I manually found and catalogued almost 900 channels (trends, really), then reviewed my choices.
A picture shows best what I found.
The following chart is of 3D systems, on the monthly scale. I drew a channel at E, but I was more focused on rising price trends that followed a trendline, not on channels. If it happened to fall within the definition of a channel, then that was peachy. If not, that was fine, too. For me, though, it was easier to call them channels because my pattern finding software was already built to display them.
So I'll refer to all "trends" as channels in this article.
The trend change signal occurs after the channel is established. That means I set a minimum duration of two years for the channel. I discarded any channel shorter than that.
Then I looked at the channel end and found three black candles. Their appearance signals a trend change. A black candle appears on the monthly chart when the month's closing price is below the month's opening price.
I used my computer to log the 897 channels I found (manually) in 503 stocks from 1990 to August 2017. I looked for valleys in the stock to touch the up-sloping trendline at least three times.
The chart shows an example of such a channel at E. My focus was to find stocks which followed a trend for years, hence the monthly scale and the use of an up-sloping trendline.
The trend change signal begins with candle A. It's the highest one in the channel. It can be black or white.
If the candle is white, then look for three additional, consecutive candles, all of whom are black. If the highest candle is black, as shown in this chart, then look for the next two candles to be black. I'll show a clearer diagram of this in the next section.
If you find either pattern, then it reliably signals the end of the channel. Sell as soon as the three black candles appear.
For this chart, the inset shows the three black candles. Candle A in the inset matches candle A on the candle chart. Because candle A is black, I show two additional black candles, B and C, giving the 3-black candle sell signal.
Notice how a substantial decline takes the stock down to D.
The above figure makes the two signal variations clearer.
Find the channel's high price. If that candle is white, then look for the next three candles to be black. Otherwise, look for three consecutively black candles. The height of each candle and the configuration is not important. If you don't see either of these two configurations, then the channel will not have the setup (only 42% of channels show this setup).
Here are the rules.
| 1. Number of stocks tested | 503 |
| 2. Number of samples found | 897 |
| 3. Period studied | 1/1//1990 to 8/29/2017 |
| 4. Dips within channel: 10-15% | 2.1 times |
| 5. Dips within channel: 15-20% | 1.4 times |
| 6. Dips within channel: 20-25% | 0.7 times |
| 7. Dips within channel: over 25% | 1.0 times |
| 8. How often does setup occur? | 42% |
| 9. False sell signals | 21% |
| 10. Average drop after sell signal | 43% |
| 11. Average drop after any 3 black candles | 27% |
| 12. Average time to close back above channel high | 3.7 years |
| 13. Percentage of channels which start in a bear market (avg: 5 months before end) | 23% |
| 14. Percentage of channels which end in a bear market | 16% |
| 15. Percentage of channels which start within a year of a bear market | 18% |
| 16. Percentage of channels which end within a year of a bear market | 4% |
The above table shows statistics I uncovered about this setup. Here are explanatory notes.
4-7: These lines show how often price retraces before the end of the channel. For example, I found that the average channel length was just over 5 years long. During that time, the average channel would experience a 10% to 15% drop twice (2.1 times). A drop exceeding 25% would happen once.
8. I looked at all channels and found 42% of them had a 3 black candle setup.
9. I also checked how often a 3 black candle setup did not signal the end of the channel. That happened 21% of the time. So if you use this setup, you'll be making a mistake by selling once ever five trades.
10. The average drop after a sell signal (measured from the closing price of the third black candle) is 43%. That's why it's so important to sell a holding that breaks out of a long-term channel.
To find the bottom of the dip, think of this as measuring the depth of a V. I found the two top ends (channel high followed by a close above the channel high) and then measured the depth between them.
11. This shows the drop after any 3-black candles, including those from channels. Dips after a channel (43%, line 10) are more severe than from all 3-black candles (27%).
The dip measures the largest drop from the closing price of the third black candle to the date the stock closed above the top of the highest candle in the 3-black candle pattern.
12. After the channel ends, price drops and takes an average of 3.7 years before it returns to close above the top of the channel.
13-16. I looked at channels and found where they appeared in relation to bear markets. Two bear markets spanned from 3/24/2000 to 10/10/2002 and 10/12/2007 to 3/6/2009 as measured by the Dow Jones Industrials.
I found that 23% of the channels started within a bear market (starting an average of 5 months before the end of the bear market), and 16% of them ended within one. I also found that 18% of channels started within one year after a bear market ended, but only 4% of channels ended within a year of the start of a bear market.
In other words, 41% (23% + 18%) of channels will start near the end of a bear market, but few channels will end because of one 20% = 16% + 4%)
I show a picture of Accenture (ACN) on the monthly scale.
Price moves up and down in waves, but follows an up-sloping trendline, shown here in red. The top of the channel is at B. It's the highest price and it also marks the end of the channel (I didn't show the top channel line, but it's not important anyway. What's important is that price climbs at a steady pace, following a trendline).
For the 3-black candle exit signal, notice that B is a white candle. So I would wait to see if the next three consecutive candles are black. They are, and they end at C. I would sell my holding at the opening price at candle D.
In this example, price makes a double top at peaks BA, confirmed when the stock closes below the valley between the two tops.
Notice that candle A also begins a 3-black candle signal. However, candle A is not the highest candle in the channel (B is higher), so this one should be ignored.
Similarly, look at candle G. As the channel develops, candle G will be the highest candle. It's white, so you'd look at the next three candles to see if they have black bodies. Which they do. This is a valid 3-black candle exit signal. It's also a dud because price makes a new high afterward.
As the above table says, these types of fake exit signals happen 21% of the time. But in this case, it's likely you would not have drawn the red trendline until after the third trendline touch (the next candle after the 3-black candles touches the trendline for the third time).
For statistics purposes, the exit signal occurs after the channel is in existence for two years, so it's a valid but fake exit signal. It was included in the statistics as a bad exit signal.
Finally, recognize that less than half of monthly channels will have a 3-black candle exit signal but when they do, they can save you a lot of money. In this example, if you sold the stock at the opening price the month after the signal, you would have sold your holding at 33.06 and the stock dropped to a low of 21.29, a potential savings of 36%.
-- 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
|
I parked my car in a tow-away zone. When I returned, the entire area was missing. -- Steven Wright
on