Using Triangle Patterns to Determine Price Movement
Wayne Thorp recently spoke at the 2015 AAII Investor Conference. For information on how to subscribe to recordings of the presentations, go to www.aaii.com/conferenceaudio for more details.
Pick up any text on technical analysis and charting and you are likely to come across some discussion of price patterns and their potential use in trading. One pattern that appears quite frequently is the triangle.
In general terms, a triangle is formed by the convergence of two trendlines drawn on a price chart. One trendline connects the price peaks and one connects the price troughs of a security. Therefore, at least two points (peaks or troughs) are needed between which the trendline is drawn. Over time these peaks and troughs narrow and, depending on the price behavior, the triangle pattern may or may not be confirmed. Triangles come in several forms and can take on different names, but the most common are right triangles and symmetrical triangles.
For any triangle, there are two sides that are important—the upper and lower boundaries. With the right triangle, one side or boundary, either the top or bottom, is roughly horizontal while the other slopes either upward or downward. Whether the slope is upward or downward gives an indication as to the type of right triangle—ascending or descending.
An ascending triangle is identifiable by its horizontal top, which connects adjacent peaks, and its upward-rising bottom. Figure 1 shows an ascending triangle that formed for Minnesota Mining and Manufacturing (3M) over the period from December 1996 to February 1997. The upper boundary—the level of resistance—was formed around $86.20 and was tested several times during this three-month period. This level proved to be a strong resistance level, with intraday prices barely breaching it but never closing above it.
The ascending boundary line, which is formed by connecting the price troughs, is drawn through the lows of mid-December, late January, and late February. Each time the price did bottom out, it was at a level higher than the previous low. While the price closed below the line on January 24, it moved above the line the very next day, making the violation insignificant.
To better understand the mechanics of a triangle, think of the two “boundaries” as support and resistance lines. The upper boundary serves as a resistance line, prohibiting prices from moving above a certain level. At the resistance point, sellers take control of prices and prevent them from rising higher. Investors do not believe that the price will rise above that level. The lower boundary serves as a support line, where prices tend not to fall below.
While it is not crucial that the price actually touch the boundaries, the more times the price tests these boundaries (by coming close or touching them), the more credible they become. If the price does violate the boundary (a breakout) before the pattern has fully developed, these violations should be small and short-lived.
The decision of which points to use in drawing your trendlines is extremely subjective. Given the same price chart, a group of people could come up with several different patterns. Some people may draw their boundaries so that the price never violates them until the breakout, while others may allow intraday breaks. Short-term traders would tend to watch for intraday breaks, while longer-term investors may look for closing prices outside the pattern. The key is to develop a strategy with which you are comfortable.
Over time, these boundaries constrict prices to a tighter and tighter trading range until, typically, the price breaks through one of the trendlines and makes a “considerable” move either on the upside or downside. It is the anticipation of this breakout and subsequent price move that makes triangle patterns useful in trading.
With right triangles, it is the breaking of the horizontal boundary and subsequent “sustained” movement in that direction that confirms the pattern. If the sloping trendline is broken, the resulting move typically has less significance.
In the case of 3M, its closing price on February 24 was just above the horizontal boundary. Any time a triangle boundary is broken, especially when the closing price is outside the triangle, you should pay close attention to the subsequent movement. On the two trading days immediately following the close above the boundary, 3M rose 3.5 and 3 points, respectively.
In addition to the behavior prices exhibited during the formation of a triangle, trading volume plays a role in confirming both the formation and confirmation of triangles. During the formation of a triangle, trading volume tends to fall. At the beginning of the pattern, you can expect heavy or at least average volume and you can expect low volume prior to the price breakout. Patterns that have this volume confirmation tend to be more reliable than those that do not.
The bottom portion of Figure 1 shows the trading volume for 3M. Over the period the triangle pattern developed—mid-December 1996 to late February 1997—daily trading volume steadily dropped from roughly 17 million shares to just over five million a few days before the breakout occurred. The trendline illustrates this fall in volume, with only a few days in that time period registering trading volume above the line.
Shortly before the price broke above the upper boundary, there was an increase in volume. Once the price closed above the boundary, volume jumped to over 17 million shares for three consecutive trading days, until the demand subsided. At that point, prices leveled out and volume declined.
This increase in volume at the breakout solidifies the pattern. In the case of both ascending and symmetrical triangles, where the breakout is to the upside, the increase in volume is vital. If such a breakout is not supported by volume, you should not be surprised if the rally falters.
For descending triangles and symmetrical triangles with the breakout to the downside, an increase in volume is not as important. However, if volume does increase, the price movement is more credible.
Descending triangles look and behave the opposite of ascending triangles. The lower boundary is basically horizontal and connects the adjacent valleys, while the upper boundary is downward sloping.
Figure 2 is an example of a descending triangle for Dow Chemical that formed from October 1989 through June 1990. The lower, horizontal boundary formed at $60.50 and was tested on three separate occasions over the period. The descending upper boundary was formed at the first peak of $75.75 and you can see that the next two price peaks are lower then the ones that preceded them.
Looking at Figure 2, the Dow Chemical graph, you can see that the descending triangle pattern is confirmed when a definitive break of the lower boundary is made. The price stabilizes for a couple of days, at which point a common “post-breakout” pattern develops. This pattern—called a snapback—is formed when the price retests the first boundary after the breakout. In this case, Dow Chemical’s price moves back to the lower boundary before continuing downward. In order for the triangle pattern to be confirmed, the price should not re-enter the triangle during the snapback, except on an intraday basis. Re-entering the triangle could be a warning that the pattern will not be confirmed.
Symmetrical triangles tend to be the most common triangle pattern; they typically form during periods of consolidation or during a period of rest amid a trend. Unlike right triangles, where one of the boundaries is relatively horizontal, symmetrical triangles are made up of two sloping trendlines that converge. Prices will fluctuate between the two lines, with each move being smaller than the previous one.
An example of a symmetrical triangle is shown in Figure 3. The upper, downward-sloping boundary for Men’s Wearhouse is drawn through the major price peaks at $25.67, $24.67, and $23.33. Price troughs at $19.67, $20.17, and $20.50 form the upward-sloping, lower boundary.
Unlike the descending and ascending patterns, which give you an indication of the direction in which the price breakout will be, symmetrical triangles do not give you that “clue.” In order to trade a symmetrical triangle, you will have to wait until the breakout has occurred. For Men’s Wearhouse, the breakout was to the downside as the price violated the lower boundary at $20.40 and made a sharp decline thereafter.
Note also that there is the accompanying volume confirmation of the triangle pattern (see the bottom of Figure 3). In the middle of May, when the symmetrical triangle was just forming, volume topped out at 660,000 shares. Over time, as the price neared the breakout point, volume dropped to just 28,000 shares traded a day. In confirmation of the pattern, the day of the breakout volume rose to over 400,000 shares.
How can triangles be used as part of a trading system?
Many technical analysts believe that once a triangle pattern has been identified and confirmed, prices tend to behave predictably and you can reasonably predict the level to which prices will rise or fall once the breakout occurs.
A typical rule of thumb for calculating a price target is to base it on the height of the triangle at its widest point—the vertical distance between the two trendlines. As an example, refer again to Figure 1. The upper boundary is easy to determine since it is a horizontal line. In this case it is $86.20. For right triangles, the sloping boundary gives us a little more trouble (for 3M the lower boundary). The price we use is the first point used when drawing the trendline—$79.00 for 3M. Knowing the high and low prices for the triangle, you then subtract the values to find its height ($86.20 – $79.00 = $7.20). This range is then added to or subtracted from the price at which the breakout occurs. Therefore we can reasonably assume, if the pattern is confirmed for 3M, that the price will rise to at least $93.40 ($86.20 + $7.20), remembering, of course, that since this is an ascending triangle, we expect the prices to break the upper boundary and continue to rise. The target price for Dow Chemical using this rule of thumb is also shown in Figure 2. The technique is the same here, but since this is a descending triangle, we expect the price to fall below the lower boundary.
The measuring technique for symmetrical triangles is similar to that used for right triangles. We are still interested in finding the vertical height, but we need to use the first point used to draw both trendlines, since there is no horizontal boundary. Referring to Figure 3, the high point is where the price reversed itself to create the first peak, at $25.67. The low point is where the price first bottomed out to create the first trough, at $19.67. Again, you take the difference ($25.67 – $19.67 = $6.00) to estimate how much the price will move. However, with a symmetrical triangle you do not know the direction in which the breakout will occur. You typically have to wait until the breakout actually occurs because it is not in the nature of these triangles to provide you with a hint as to which way prices will go. Once the breakout happens, in this case, for Men’s Wearhouse to the downside at $20.40, we can then calculate the minimum price target by deducting the range from the breakout point ($20.40 – $6.00 = $14.40).
too good to be true?
While triangles, like many other price patterns, may provide you with an indication of an impending movement, technical analysis or any type of investment analysis is not an exact science. The example in Figure 4 proves this point. Here we see what appears to be a textbook example of a descending triangle forming for Consolidated Natural Gas. From late December 1987 until May 1988, prices tested but never fell below the $35.25 support level. Each time prices rose off this bottom, the subsequent high was always lower than the one that preceded it. This allowed for a downward-sloping upper boundary to be drawn, above which prices never rose. Over this same period, volume provided the classic confirmation, with most volume spikes being lower than the one prior. Volume above the trendline was short-term in nature.
Prices eventually did fall and close below the lower boundary on May 11. Since this is the first violation of the boundary, we should focus our attention on what happens next. The very next day the price rises and closes above the lower boundary—not an uncommon occurrence during the development of a triangle. However, as mentioned earlier, we should begin to question the pattern’s validity. The price flirts with the line for a few days until it, again, falls below it. This time, however, it continues to fall and, at this point, we are led to believe that the pattern has been confirmed and we can look to prices potentially falling to the target of $28.75 (using the same technique outlined earlier). However, within two weeks prices first stabilize and then reverse. Over the next couple of days, the price rises above both the lower and upper boundaries, thereby voiding the pattern. This shows how important it is to use other techniques in conjunction with technical analysis.
This discussion was meant to add some useful tools to your technical analysis toolbox. But it is important to remember that these techniques are not foolproof, and the perceived pattern may not develop according to your expectations.
In the end, no matter what type of trading strategy you follow over the short term, other forms of analysis, particularly a fundamental look at the position of the company, is crucial for long-term investment success.