As of 09/19/2019
Indus: 27,095 52.29 0.2%
Trans: 10,504 72.12 0.7%
Utils: 860 +3.49 +0.4%
Nasdaq: 8,183 +5.49 +0.1%
S&P 500: 3,007 +0.06 +0.0%

YTD
+16.1%
+14.5%
+20.6%
+23.3%
+19.9%

26,400 or 27,600 by 10/01/2019
10,200 or 11,100 by 10/01/2019
875 or 830 by 10/01/2019
8,000 or 8,750 by 10/01/2019
2,925 or 3,050 by 10/01/2019

As of 09/19/2019
Indus: 27,095 52.29 0.2%
Trans: 10,504 72.12 0.7%
Utils: 860 +3.49 +0.4%
Nasdaq: 8,183 +5.49 +0.1%
S&P 500: 3,007 +0.06 +0.0%

YTD
+16.1%
+14.5%
+20.6%
+23.3%
+19.9%
 
26,400 or 27,600 by 10/01/2019
10,200 or 11,100 by 10/01/2019
875 or 830 by 10/01/2019
8,000 or 8,750 by 10/01/2019
2,925 or 3,050 by 10/01/2019
 
Information on busted chart patterns is discussed in my book, Visual Guide to Chart Patterns. You can find information in the book in Chapter 22: "Busted Pattern Buy Setups" (starting on page 229) and in Chapter 25, "Busted Pattern Sell Signals" starting on page 271.
I show a picture of the book on the right.
If you click on the above link and then buy the book (or anything) while at Amazon.com, the referral will help support this site. Thanks.
$ $ $
Price can breakout of a descending triangle in any direction. When price moves less than 10%, reverses direction, and closes beyond the side opposite the breakout, it busts the chart pattern. For testing and safety, I used the top and bottom of the triangle as the price where a stock busts the triangle, not a trendline break.
As one might imagine, a descending triangle with a downward breakout that busts results in a higher gain. The rise averages 29% (median 24%).
Busted descending triangles occur 29% of the time in both bull (24%) and bear (5%) markets. In a bull market only, single busted descending triangles with downward breakouts occur 72% of the time. That means there are fewer double and triple busted triangles.
I show a chart of a single busted descending triangle in QLogic on the daily scale. The descending triangle is outlined with blue trendlines (which really look black on my 'puter). Price pierces the bottom trendline at A and confirms the chart pattern with an downward breakout. However, price does not drop far before it reverses. When price closes above the top of the triangle, which occurs at B (red line), it busts the downward breakout. If price continues to move higher by at least 10% above the red line, as in this case, then the chart pattern becomes a single busted descending triangle.
For a single bust, look for:
For the last point, 4, if price fails to move more than 10%, then it could be forming a double busted descending triangle.
The figure to the right shows an example of a double busted descending triangle in ColgatePalmolive (CL).
The descending triangle forms between the two blue trendlines. This one has price closing above the downsloping trendline for an upward breakout, at A. The upward move reverses and drops to B. When price closes below the bottom of the triangle, it busts the upward breakout for the first time. Price continues dropping, but not much  less than 10% below the bottom of the triangle.
Then price makes a straightline run up. When it closes above the highest peak in the descending triangle, it busts the chart pattern for the second time. That happens at C. Price continues rising (not all of the rise is shown) at least 10% above the red line (which highlights the highest peak in the pattern).
For a double bust, look for these elements.
If price fails to move at least 10% in the new direction, then it is a triple busted descending triangle.
Intel, pictured on the right, is a good example of a triple busted descending triangle.
The triangle is outlined in a thin blue line on the left of the chart. Price closes above the downsloping trendline at A, making an upward breakout. Price drifts lower to B, busting the triangle for the first time, before reversing direction. The drop below the blue line at B is less than 10%.
Price continues moving higher to C, where it closes above the red line (the top of the triangle). The move above the red line is less than 10% before price reverses, meaning price busted the triangle a second time.
Price drops to D. The drop below the lower red line is still less than 10% before price reverses. The triangle busts for the third time.
If price continues rising by at least 10% above the top red line, then the busting process would end at three busts. Otherwise, the stock can continue busting the triangle additional times by making tall swings above and below the chart pattern.
For a triple (or more) busted descending triangle, look for the following:
I found 1,615 descending triangles in 812 stocks dating back as far as July 1991 to September 2011. Few stocks covered the entire period. All of the descending triangles I found manually either using a historical search or real time (looking at my stocks each day). The real time additions prevented any lookahead bias since I am not privy to future price movements.
I then used software to measure performance and flag busted chart patterns.
Gauging performance uses the same method as I used to catalog nonbusted chart patterns. That is, the search for the new ultimate high or low proceeded as described in the glossary. Thus, the numbers reported in Busted Descending Triangles Results (next section) should be considered perfect trades. Do not expect to duplicate the results in actual trading. The numbers should be used only for comparison purposes to other chart patterns.
The following numbers are the results from perfect trades in bull markets, unless otherwise noted. Do not expect actual trading results to match those discussed below. Use the numbers only for comparison purposes with other chart patterns.
How often do descending triangles bust?
Upward Breakouts
Downward Breakouts
Descending triangles breakout downward more often than upward, so they tend to have more busts in that direction as well.
Of busted descending triangles, what is the frequency distribution?
Upward Breakouts
For upward breakouts, 43% of them will bust just once, meaning that over half (57%) will fail to drop more than 10% below the bottom of the triangle.
Downward Breakouts
If you trade busted descending triangles with downward breakouts, there is a 72% probability that it will bust just once. Thus, 28% fail to show price rising by more than 10% above the top of the descending triangle.
What is the average move for single busted descending triangles? As measured from the bottom (the lowest valley) or the top (the highest peak) of the chart pattern to the ultimate low or high, respectively, the move averaged:
Upward Breakouts
Downward Breakouts
The downward breakout comparison is not a fair comparison between all triangles and busted ones. Busted triangles measure from the highest peak in the triangle to the ultimate high. "All triangles" measure from a the low price on the day of breakout, where price pierces the downsloping trendline, to the ultimate high. In other words, the busted numbers are penalized. However, the busted numbers only reflect single busted patterns, not multiple busted ones. Again, the comparison is unfair.
* I decided to include busted triangles in the mix because a trader cannot know ahead of the breakout that the triangle will bust. To remove all of those that did bust is to remove most losing trades, which is unfair. Excluding busted triangles, upward breakouts gain 48% and downward breakouts lose 22%.What is the move for perfect trades after all busted descending triangles?
What is the failure rate of all busted descending triangles in bull markets? The answer appears in the below table.
Failure rate:  5%  10%  15%  20%  25%  30%  35%  50%  75%  >75% 

Number of descending triangles:  54  33  9  9  8  7  3  8  3  0 
Percentage:  40%  25%  7%  7%  6%  5%  2%  6%  2%  0% 
Cumulative:  40%  65%  72%  78%  84%  90%  92%  98%  100%  100% 
For example, there were 54 descending triangles with upward breakouts that failed to show price dropping at least 5% below the bottom of the descending triangle. Those 54 represent 40% of all busted descending triangles. On a cumulative basis (a running total), the 54 also represent 40% of all busted descending triangles with upward breakouts.
The median drop of all busted descending triangles is 6%. You can see that by interpolating between the 5% and 10% columns. Thus, half of all busted descending triangles will see price drop 6% in a bull market, providing they are traded perfectly.
Failure rate:  5%  10%  15%  20%  25%  30%  35%  50%  75%  >75% 

Number of descending triangles:  44  31  22  16  18  10  12  31  30  36 
Percentage:  18%  12%  9%  6%  7%  4%  5%  12%  12%  14% 
Cumulative:  18%  30%  39%  45%  52%  56%  61%  74%  86%  100% 
Read the table just as you did the prior one. Let's begin with the 10% column. Thirtyone triangles saw gains from over 5% to less than or equal to 10%. They represented 12% of busted triangles. On a cumulative basis, 30% of triangles saw price climb up to 10% after busting a downward breakout (as measured from the top of the pattern to the ultimate high).
How could you trade a busted descending triangle? The figure on the right gives an example.
A descending triangle appears in December 2010 in Family Dollar (FDO). Price breaks out downward from this pattern at A then gaps upward two days later. To bust the triangle, I require price to close above the top of the triangle and that occurs at B. Even if you placed a buy order at a penny above the top of the triangle, it would not have done any good except getting you in at or near the opening price the day the stock gapped higher. Assuming a fill at the open, the buy order completed at 30.20.
Price fades for several days until getting another boost and moving higher in a straightline run up.
Not shown, but the stock peaked in late May at 42.03, retraced to a low of 35.31 before another leg up to 51.81, eventually reaching a high of 56.92.
If you traded this one perfectly and sold in May, you could have made 39%. If you closed your eyes and weathered the two large drops along the way, the gain to 56.92 in May 2011 (oddly, the same day a year later) would be 88%
All of these tips apply to bull markets only!
For upward breakouts, when price drops after it busts, busted triangles with the breakout price below the 200day simple moving average tend to perform slightly better: 13% (52 samples) versus 11% (81 samples).
For downward breakouts, the rule is reversed. When the breakout price is above the 200day simple moving average, the busted triangle sees price rise by 33% (189 samples) versus 20% (only 51 samples) for those below the 200day SMA.
For a definition of the trend start, visit the glossary.
For downward breakouts, when the time between the trend start and the start of the descending triangle is less than or equal to the median 82 days, the stock rises an average of 33% (119 samples). When the time is more than the median, the rise averages 28% (118 samples)
The same applies to upward breakouts, but the numbers are different: 14% (64 samples) versus 9% (63 samples) for times of less than or equal to 119 days (the median for upward breakouts) versus longer periods, respectively.
When the trend start price is above the top of the busted descending triangle by less than the median 13% and the triangle breaks out downward, the busted pattern sees the stock rise by 34% (34 samples). If the trend start is above the busted triangle by more than 13%, the rise averages 19% (33 samples). The performance difference for upward breakouts shows no change, so don't worry about it.
Compare the trend start to the top of the triangle. If the high price at the trend start is below the top of the triangle and the triangle breaks out downward, then busted patterns see price climb an average of 33% (170 samples). When the trend start is above the triangle, the rise averages just 24% (67 samples).
As an entry setup, here are the rules for trading busted descending triangles.
For upward breakouts...
For downward breakouts...
 Thomas Bulkowski
See Also

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.
If Helen Keller is alone in a forest and falls, does she make a sound?