because volume trends downward.
Price trend. Price trends downward into the start of a broadening bottom. Even if price rises a week or two before the chart pattern begins (overshoot), ignore it. The pattern is still a bottom. This arbitrary designation makes intuitive sense: A bottom should appear at the end of a downtrend, not when price is climbing to the moon and not if price spiked upward just before the pattern started.
Trendlines. Two trendlines drawn across the minor highs and lows outline the pattern. The top trendline should slope up; the bottom one should slope down. The diverging trendlines distinguish the broadening bottom from other types of chart patterns, such as the right‐angled broadening formation (which has one horizontal trendline) and the broadening wedge (both trendlines slope in the same direction).
Touches. To prevent confusion, I changed the guidelines to require at least five touches, three of one trendline and two of the other. Fewer touches increase the likelihood of misidentification (but it still might be a broadening bottom). Play it safe and look for at least five trendline touches.
Each trendline touch should be at or near a minor high (top trendline) or minor low (bottom trendline). What is a minor high or low? A minor high happens when price trends up, then drops back down, leaving a clearly defined peak. A minor low is just the same except flipped upside down: Price moves lower, and then heads back up, leaving a clearly defined valley.
In Figure 8.1, odd numbers tag the minor highs and the even numbers are the minor lows. Let me stress that the minor highs and lows need not be alternating, as shown in the figure. Just as long as you can count at least five touches, then that's fine. If price cuts through a trendline, then don't count that as a touch.
Notice that at the start of this pattern (on the bottom), price cuts through the lower trendline, but it doesn't count as a touch.
Whitespace. Figure 8.2 shows a problem with identification. This is from 3M (MMM) in August 2019. On the left side (“Bad”), the stock appears to make a broadening bottom. It has two touches of the top trendline and three on the bottom, as required. The problem is that white hole in the pattern, which I highlight at A. Price does not cross the pattern enough to fill the space.
Compare the left side with the right side (“Better”). It's the same picture except I show a down‐sloping channel. It's not perfect because of the spike at B, but it'll do. This is a better interpretation of a viable chart pattern compared to the left side. Do not cut off a turn like I've shown on the left side and call it a broadening bottom.
Volume. There is nothing magical or important about volume. I used linear regression from the start of the chart pattern to its end and found that volume trends upward most of the time in this chart pattern.
Breakout direction. The breakout price can be difficult to identify in a broadening chart pattern as the pattern develops. I look for the place where price pierces the up or down trendline or makes an extended move, hugging the trendline.
Figure 8.2 Don't accept patterns with too much whitespace. This happens when you cut off a turn and call it a broadening bottom.
If price closes outside a trendline boundary, then the penetration point becomes the breakout price. If price moves up (for example) and follows along the top trendline without piercing it, then I backtrack to the prior minor high and draw a horizontal line forward in time until price closes above the horizontal line. When that happens, that is the breakout point (providing the pattern obeys the other identification guidelines first). Use the same logic for a downward breakout.
Let me give you an example. Consider the broadening bottom shown in Figure 8.3. Where's the breakout from the broadening bottom?
Figure 8.3 Where's the breakout?
If you wait for price to close above the top trendline, then you'd have to wait until C. Imagine if the top trendline were steeper. You might never see a breakout.
If the pattern meets all of the identification guidelines (especially trendline touches) and price begins sliding upward (along the top trendline or downward along the bottom one), like you see from B to D, then go back to the prior minor high (upward breakout) or prior minor low (downward breakout) and use that price as the breakout price. In this case, point B is the prior minor high in the chart pattern and the breakout is point D. If you worry about buying into the broadening bottom too late, then skip the trade. Chart patterns are not like attending a party 15 minutes late. Promptness pays.
Focus on Failures
Figure 8.4 shows a typical broadening bottom failure. Price trended down from the March high. On 19 March (E), the company announced the pricing of a secondary public offering of nearly 8 million shares of common stock. The stock price took a hit and shares tumbled that day and the next, just before the start of the broadening bottom.
The broadening bottom formed innocently enough with price swinging from low to high (A). Price touched the top trendline three times and the bottom trendline three times, as one would expect in a well‐behaved broadening bottom (meaning at least 5 touches).
Figure 8.4 This broadening bottom breaks out downward, reverses, and busts the downward breakout.
At B, the stock broke out downward from the broadening pattern when price closed below the lower trendline. Because price trended downward going into the pattern, a trader might expect a downward breakout, too (so the pattern acted as a continuation of the downward price trend). Indeed, they might expect price to drop back to just above the December 2018 low (F).
However, the stock surprised traders when it stalled at C. It was even more of a shock when the stock began to stair‐step higher and closed above the top of the pattern at D. At D, the stock busted the downward breakout. After that, the stock was an airline taking off and flying into the clouds.
The behavior of the broadening pattern shown in the figure represents what I call a 5% failure. Price breaks out lower but fails to continue moving in the breakout direction by more than 5% before heading back up. The reverse is also true for upward 5% failures: Price climbs after an upward breakout by no more than 5% before tumbling.
Statistics
Table 8.2 shows general statistics for the broadening bottom chart pattern.
Number found. I dug up 1,238 patterns in 667 stocks starting from August 1991 to September 2019 but removed the bear market ones because they were too few when sorted by breakout
