The Rectangles, Double and Triple Tops
The Triangular Price Formations, which we examined in Chapter 8, can be either Reversal or Consolidation Patterns. In the case of the Right-Angle Triangles, we know as soon as they have attained recognizable form in which direction the trend will (or should) proceed. With the Symmetrical Triangles, we have no way of knowing whether they point up or down until prices finally break away from them, although the odds are, as we have seen, the previous trend will be continued rather than reversed. In this respect, and in many others, our next class of technical formations, the Rectangles, resemble the Symmetrical Triangles. There are, in fact, so many points of similarity between them that we can forego any long and detailed discussion. (For illustrations in this chapter, see Figures 9.1 through 9.18.)
A Rectangle consists of a series of sideways price fluctuations, a “trading area,” as it is sometimes called, which can be bounded both top and bottom by horizontal lines. A glance at any one of the examples that illustrate these pages will show how it got its name. On rare occasions, you may discover a chart pattern whose upper and lower boundary lines are parallel but either slightly down-sloping or up-sloping. So long as their departure from the horizontal is trivial, they may be treated as Rectangles. You will also find, on occasion, patterns whose boundaries, while nearly horizontal, tend somewhat to converge. These may be considered Rectangles or Symmetrical Triangles; it does not matter which because the “prognosis” will be the same in either case.
If you will give a quick mental review also to the Head-and-Shoulders, the Complex, and the Rounding types of formations, you will see how, if you disregard the volume part of their charts, any one of these patterns might merge or grade into a Rectangle. As a matter of fact, however, you will seldom be left in doubt as to proper classification because the circumstances of trading, the type of buying and selling, which produce Rectangles are different, which is usually apparent.
We characterized the Symmetrical Triangle as a “picture of doubt.” The Rectangle might, with even greater propriety, be called a picture of conflict. Of course, any fairly compact price formation represents conflict in the supply–demand sense. A Head-and- Shoulders Top, for example, portrays a conflict between “strong” sellers and “weak” buyers with the outcome already clearly seen before the combat has ended. But a Rectangle defines a contest between two groups of approximately equal strength—between owners of the stock who wish to dispose of their shares at a certain price and others who wish to accumulate the stock at a certain lower figure. They bat the ball back and forth (up and down, that is) between them until ultimately, and usually quite suddenly, one team is exhausted (or changes its mind) and the other proceeds to knock the ball out of the lot. Nobody (often, not even the contestants themselves) can tell who is going to win until one line or the other is decisively broken.
We speak of two groups operating in the development of a rectangular trading area because, under present-day conditions, that is what is usually the fact behind the scenes. This, it should be noted, does not imply “manipulation” in any invidious sense. An investment trust or an estate or, in some cases, an individual heavy stockholder has good and sufficient reasons for selling at the top price (the “Supply Line” of the Rectangle) with no intent to mislead the public. Another investment trust or a group of insiders interested in the company may have equally good and, from their point of view, wise reasons for buying at the bottom price (“Demand Line”). Such are the forces at work in the market at the start of most Rectangular Chart Patterns, but if the “spread” between top and bottom lines is wide enough (say 8%–10% of the market value of the stock), the situation may quickly attract a following from quick-turn scalpers and the professional element
Thus, a syndicate holding a large block of U.S. Steel may decide to liquidate at 76, whereas another group decides to invest heavily in “Steel” at 69. The price of X will naturally fluctuate for a time between those two levels. Traders, seeing this, will try to ride the play, buying at 69 and selling at 76 (perhaps also selling short at 76 and covering at 69). Their operations will tend to accentuate or extend the Rectangle, although the number of shares involved in such parasitic trading is seldom great enough to affect the final outcome. As a matter of fact, this type of trading inside a Rectangle can be quite profitable at times, especially if protected by judicious stops (see Section II).
In times past, before the U.S. Securities and Exchange Commission (SEC) outlawed the practice, Rectangles were frequently created by the well-organized operations of a single “pool” or syndicate. Such a pool might undertake to accumulate a large block of stock in a certain company with a view to marking it up and taking profits when some piece of good news, of which they had inside knowledge, eventually became public. To acquire the desired “line,” they would find it necessary first to shake out shares held by other traders and uninformed investors. They might start their campaign by suddenly selling short a few hundred shares to quench any current demand and start a reaction. Then, on that reaction to the previously determined accumulation level, they would start to buy, scattering their orders carefully and avoiding any publicity. Their buying would, sooner or later, engender a rally, but then they would “plant” rumors around the boardrooms to the effect that such-and-such insiders were selling, or that a projected merger was being called off, or a dividend would have to be passed, and, if necessary, they would ostentatiously let out a few of their own recently purchased shares to give color to the rumor. The process might be repeated several times with the “pool” gradually securing more and more shares on balance until, finally, its intended line is completed or could not shake out more of the floating supply. Often, what was going on was fairly evident to the alert chartist back in the 1920s even before the operation was concluded, and perfectly evident, of course, as soon as prices broke out topside from their Rectangle
But such tactics are no longer permitted. “Wash sales” are strictly condemned; the constant policing of all exchange transactions and prompt investigation by the SEC of any suspicious news or activity in a stock effectually deters the blatant “pool” manipulations of previous years. This probably is the chief reason why Rectangles are nowhere near so common on the charts of the 1950s as they were in the 1920s. (EN: Not uncommon in the 2000s.)
Perhaps we can clear up various details of the Rectangle formation most quickly and easily by comparison with that most nearly related chart pattern, the Symmetrical Triangle, as follows:
- Volume—Follows the same rules as in the Triangles, gradually diminishing as the Rectangle lengthens. Any contrary development, unless it be a momentary news flurry, is
- Breakouts—Here also the same rules apply as with Triangles. Review volume requirements, margin of penetration, and so on
- False moves—Much less frequent from Rectangles than from Symmetrical A clearly defined Rectangle is, in fact, almost as reliable as a Head-and-Shoulders, although not as powerful in its implications.
- Premature breakouts—Slightly more frequent, perhaps, from Rectangles than from Triangles.
(Note: Both false moves and premature breakouts, in the sense in which we employ these terms, are indistinguishable at the time they occur from genuine breakouts. Following both false and premature breaks, prices return inside the pattern. But, in the case of a false move, the trend ultimately proceeds out of pattern in the opposite direction, while in the case of the premature move, the trend finally breaks out again and proceeds in the same direction.)
- Pullbacks—Return of prices to the boundary of the pattern, subsequent to its initial penetration (breakout), takes place more frequently with Rectangles than with Symmetrical Triangles. Our estimate would be that a Pullback or Throwback (the first is the common term for a rally after a downside breakout, and the second for a reaction following an upside breakout) occurs within three days to three weeks in about 40% of all
- Directional tendency—The Rectangle is more often a Consolidation Formation than a Reversal Formation, the ratio being about the same as with Symmetrical Triangles. As Reversal Patterns, Rectangles appear more frequently at Bottoms (either Major or Intermediate) than at Long, thin, dull Rectangles are not uncommon at Primary Bottoms, sometimes grading into the type of Flat-Bottomed Saucer or Dormancy described in Chapter 7.
- Measuring implications—A safe minimum measuring formula for the Rectangle is given by its Prices should go at least as far in points beyond the pattern as the difference in points between the top and bottom lines of the pattern itself, though they may go much farther. Generally speaking, the brief, wide-swinging forms, which appear nearly square in shape on the chart and in which turnover is active, are more dynamic than the longer and narrower manifestations. Moves out of the latter almost always hesitate or react at the “minimum” point before carrying on.
Relation of rectangle to Dow Line
The resemblance of this individual stock chart formation, which we have discussed under the name of Rectangle, to the Average formation known to Dow theorists as a “Line” has doubtless occurred to you. Obviously, their rationale and forecasting implications are much the same, but true Rectangles with sharply delimited Top (Supply) and Bottom (Demand) boundaries are truly characteristic only of trading in individual issues. Line formations in the Averages are seldom rigorously defined, with successive Minor Heights forming quite precisely at a certain horizontal tangent and successive Bottoms at a similarly precise horizontal level. If you will examine the separate charts of the issues composing an Average at a time when the Average is “making a Line,” you will surely find some of them showing an irregular uptrend, others showing an irregular downtrend, still others may be forming Triangles, and a few may be constructing Rectangles, or what not, but it is the algebraic sum of all these more or less divergent pictures that makes up the Average “Line.”
To be sure, there is some tendency on the part of active traders to sell (or buy) stocks when a certain Average reaches a certain figure, regardless of the status of individual issues involved. An investment counsel will occasionally advise his clients, for example, to “sell all speculative holdings when the Dow Industrials reach 500” (EN: or 5,000 or 15,000). But trading commitments based solely on general Average levels are so seldom followed consistently that they have little effect. (EN10: In the modern era, with the availability of index exchange-traded funds, this is no longer true.)
Rectangles from Right-Angle Triangles
In the preceding chapter, we referred to a type of partial “failure” in the development of a Right-Angle Triangle that necessitates reclassifying the Triangle as a Rectangle. Now that we have examined the latter pattern in detail, we need say little more about this phenomenon, except to note the odds still appear to be somewhat in favor of ultimate breakout in the direction originally implied by the incipient Triangle. The fact there is this slight presumption, however, certainly does not warrant disregard of an opposite breakout from the rectangular reconstruction.
Double and Triple Tops and Bottoms
To some of the old hands in the Street, our relegation of that good old byword, the Double Top, to a Minor Position in our array of Reversal Formations may seem almost sacrilegious. It is referred to by name perhaps more often than any other chart pattern by traders who possess a smattering of technical “lingo” but little organized knowledge of technical facts. True Double Tops and Double Bottoms are exceedingly rare; Triple Forms are even rarer. Additionally, the true patterns (as distinguished from chart pictures that might mistakenly be called such but are really assignable to some one of our other Reversal Formations) can seldom be positively detected until prices have gone quite a long way away from them, and can never be foretold or identified as soon as they occur from chart data alone.
But we are getting ahead of our story; we should first define what we are talking about. A Double Top is formed when a stock advances to a certain level with, usually, high volume at and approaching the Top figure, then retreats with diminishing activity, then comes up again to the same (or practically the same) top price as before with some pickup in turnover, but not as much as on the first peak, and then finally turns down a second time for a Major or Consequential Intermediate Decline. A Double Bottom is the same picture upside down; the Triple types make three Tops (or Bottoms) instead of two.
It is not difficult to skim through a book of several hundred monthly charts and pick out two or three examples of Major Double Tops and, perhaps, one or two Double Bottoms. One will find cases in which stocks made two successive Bull Market Peaks several years apart at almost identical levels. Such phenomena stand out, in distant retrospect, like the proverbial sore thumb, which undoubtedly accounts for the undue awe with which the amateur chartist regards them. He neglects, for the moment, to consider the fact a thousand other issues might have done the same thing but did not—that some of these even acted, for a time, as though they were going to Double Top, but then went on through and higher. Is there any practical utility for the trader or investor in the Double Top concept? Yes, there is, but it will be easier for us to formulate it if we first consider what is not a Double Top. Refer back for a moment to the Ascending Triangles and the Rectangles
when these start to evolve, the majority of the time their first step is the construction of two Tops at an identical level with an intervening recession, and with less volume on the second Top than on the first. In the ordinary course of events, a third Top will develop there, and ultimately, prices will break through and move on up to still higher levels. Thus, we see we must have some rule or criterion to distinguish a true Double Top Reversal Pattern from the Double Tops that do not imply Reversal when they appear as a part of a Consolidation Area in an uptrend.
No absolute and unqualified rule can be laid down to fit all cases involving stocks of different values and market habits, but one relative distinction quickly suggests itself when we study these different kinds of chart formations: if two Tops appear at the same level but quite close together in time and with only a Minor Reaction between them, chances are they are part of a Consolidation Area; or, if a Reversal of Trend is to ensue, there will first be more pattern development—more “work” done—around those top ranges. If, on the other hand, there is a long, dull, deep, and more or less rounding reaction after the initial peak has appeared, and then an evident lack of vitality when prices come up again to the previous high, we can at least be suspicious of a Double Top.
How deep is deep, and how long is long? Fair questions, to which, unfortunately, it is impossible to give simple, definite answers, but we can attempt approximations. Thus, if the two Tops are more than a month apart, they are not likely to belong to the same Consolidation or Congestion Formation. If, in addition, the reaction between the first and second high reduces prices by 20% of their top value, the odds swing toward a Double Top interpretation. But both of these criteria are arbitrary, and not without exception. There are cases in which the two peaks have occurred only two or three weeks apart, and others in which the “valley” between them descended only about 15%. Most true Double Tops, however, develop two or three months or more apart. Generally speaking, the time element is more critical than the depth of the reaction. The greater the time between the two highs, the less the need of any extensive decline of prices in the interim.
Given the conditions we have specified, namely, two Tops at approximately the same level but more than a month apart on the chart, with somewhat less activity on the second advance than on the first, and a rather dull or irregular and rounding type of recession between them, we can then be suspicious that a Double Top Reversal has actually evolved. Should a small Head-and-Shoulders or Descending Triangle start to develop at the second Top, as is frequently the case, we can be on guard, to the extent of protecting long commitments at once with a close stop or by switching to something else with a more promising chart picture.
Yet, even all these signs together are not final and conclusive. The situation can still be saved, and often is. Let us take a look at what is, presumably, going on behind the scenes to create our chart picture up to this point. The first Top on relatively high volume was a normal incident and tells us little except that here, for the moment, demand met with sufficient supply to stop the advance and produce a reaction. That supply may have represented only traders’ profit-taking, in which event the trend is likely to push on up after a brief setback. But, when the reaction drifts off lower and lower until it has given up 15% and more of the stock’s peak market value, and flattens out without any prompt and vigorous rebound, it becomes evident that either the demand was pretty well played out on the last advance or the selling represented something more than short-term profit cashing. The questions then are these: did the first high give evidence of important distribution, and is there much more to meet at the same price range?
Nevertheless, as our chart picture shows, demand did finally come in and absorb enough of the floating supply to turn the trend around. When prices pushed up and began to run into selling again near the level of the first Top, that was to be expected on “psychological” grounds; many quick-turn operators naturally would take profits at the old high (perhaps with the intention of jumping right back in at a still higher price if the old high should be exceeded). Hence, a Minor Hesitation there was quite in order. But selling in sufficient quantity to produce another extensive reaction would be quite another matter. We have, by now, established a zone of Supply or Resistance at the peak levels and a zone of Support or Demand at the Bottom of the valley between. The final and decisive question now is this: will the “valley” Support reappear and stop the second decline?
The conclusive definition of a Double Top is given by a negative answer to that last question. If prices, on their recession from the second peak, drop through the Bottom level of the valley, a Reversal of Trend from up to down is signaled, which is usually a signal of major importance. Fully confirmed Double Tops seldom appear at turns in the Intermediate Trend; they are characteristically a Primary Reversal phenomenon. Hence, when you are sure you have one, do not scorn it. Even though prices may have already receded 20%, the chances are they have very much farther to go before they reach bottom.
As to measuring implications, the Double Top affords no formula comparable with what we have attributed to Head-and-Shoulders and Triangle Formations, but it is safe to assume the decline will continue at least as far below the valley level as the distance from peak to valley. It may not be so in one interrupted slide; on the contrary, considerable time may be required to carry out the full descent in a series of waves. Pullbacks to the “valley” price range, following the first breakthrough, are not uncommon. (Take into account the general rule that a Reversal Formation can be expected to produce no more than a retracement of the trend that preceded it.)
One more point: we have said the Tops need not form at precisely the same level. Use here the 3% rule we have previously laid down as a measuring stick for breakouts. A first Top at 50, for example, and a second at 51 1/2 would come within this limit. Curiously enough, the second peak often does exceed the first by a fraction. The important points are
(1) that buying cannot push prices up into the clear by a decisive margin, and (2) the Support below is subsequently broken.
In identifying a Double Bottom, we can apply all of the precepts we have formulated for the Double Top Pattern, but upside down. The differences between the two pictures are just what you might expect them to be, having in mind the characteristic differences between Head-and-Shoulders Tops and Bottoms, for example. Thus, the second Bottom is usually conspicuously dull (little trading volume) and is apt to be quite rounded, whereas the second Top in a Double Top is moderately active and nearly as sharp and “spiky” in contour as the first. The rally up from the second Bottom shows an increase in turnover, and volume should pick up to a marked degree as the valley level, or more properly, in this case, the height between the two Bottoms, is surpassed. Double Bottoms appear just about as frequently as do Double Tops at Primary Trend Reversals, and Double Bottoms also occur sometimes at the end of Intermediate Corrections in a Major Uptrend.
If you are familiar with some of the jargon of the Street, it has probably occurred to you that the second low of a Double Bottom is an example of the market action so often referred to as a “test.” In a sense, that is just what it is—a test or corroboration of the Support (i.e., demand) that stemmed the first decline at the same level. The success of that test is not proved, however—and this is a point to remember—until prices have demonstrated their ability to rise on increasing volume above the preceding high (the height of the rally between the two Bottoms). Until such time, there is always the possibility a second test (third bottom) may be necessary, or even a third, and that one of these will fail with prices then breaking on down into further decline. This thought leads us to our next type of Reversal Formation.
Triple Tops and Bottoms
Logically, if there are Double Tops, then we might expect that there will also be Triple Tops, which will develop in somewhat similar fashion. The fact is that Reversal Formations, which can only be classed as Triple Tops, do occur, but they are few and far between. Many patterns evolve at an important turn from up to down in the trend that contains three Top points, but most fall more readily into the category of Rectangles. For that matter, any Head-and-Shoulders Formation, particularly if it be rather “flat” with the head not extending much above the level of the two shoulders, might be called a sort of Triple Top.
The true Triple Top (as distinct, that is, from other types of three-peak formations) carries a recognizable family resemblance to the Double Top. Its Tops are widely spaced and with quite depth and usually rounding reactions between them. Volume is characteristically less on the second advance than on the first, and still less on the third, which often peters out with no appreciable pickup in activity. The three highs need not be spaced quite so far apart as the two that constitute a Double Top, and they need not be equally spaced. Thus, the second Top may occur only about three weeks after the first and the third six weeks or more after the second. Also, the intervening valleys need not bottom out at exactly the same level; the first may be shallower than the second and vice versa. Also, the three highs may not come at precisely the same price; our 3% tolerance rule is again useful here. Yet, despite all these permissible variations, there should be, and generally is, something suspiciously familiar about the overall picture, something that immediately suggests the possibility of a Triple Top to the experienced chartist.
The conclusive test, however, is a decline from the third Top that breaks prices down through the level of the valley floor (the lower one, if the two valleys form at different levels). Not until that has occurred can a Triple Top be regarded as confirmed and actually in effect; so long as demand persists at the valley price range, the trend can be turned up again. Only in those cases in which activity is conspicuously lacking on the third peak and then begins to show Bearish characteristics by accelerating on the ensuing decline is one justified in “jumping the gun.”
Note this formation qualifies as a Triple Bottom in every detail—spacing between Bottoms, extent in percent of intervening rallies, volume. Of course, its completion in October 1942 did not necessarily forecast that “NG” would climb to 33, as it ultimately did. But the fact that many other stocks were making sound Major Bottom Formations at higher price levels at the same time certainly warranted the conclusion that “NG” was on its way up, and that it was a bargain at 5.
Triple Bottoms are simply Triple Tops turned upside down, with the same qualifications noted when discussing Double Bottoms. The third low should always be attended by small volume, and the rise therefrom must show a decided increase in turnover and carry prices decisively above the Tops of the rallies that formed between the Bottoms. One is never justified in “jumping the gun” on a presumed Triple Bottom Formation unless nearly every other chart in the book is in an unmistakably Bullish position. The risk of premature buying is expressed in a saying one sometimes hears in the boardrooms to the effect of “a Triple Bottom is always broken.” This is not a true saying. Once a Triple Bottom has been established and confirmed by the necessary up-side breakout, it seldomly fails—it almost always produces an advance of distinctly worthwhile proportions. But an uncompleted “possible” Triple Bottom chart picture must be regarded as treacherous. Stick to the breakout rule and you will be safe.
Triple Tops are sometimes referred to as “W” Patterns because of their occasional resemblance to that capital letter on the chart. There is a sort of hybrid between the Double and Triple Top, in which the middle one of the three Tops does not attain the height of the first and third, and thus, even more strikingly resembles a “W.” For the same reason, Double Tops are sometimes called “M” Formations.
Because the elements in Double and Triple patterns are normally spaced well apart in time, they are often easier to detect and appreciate on a weekly chart than on a daily. Monthly graphs disclose numbers of widely spread Double and Triple Bottoms but, on the other hand, are too coarse to reveal many good Double and Triple Top Patterns.
In our foregoing discussion of the Triple Top, we referred to a sort of intuition that comes with experience and enables a technical analyst to recognize the potentialities for Reversal of a certain chart development, sometimes long before it has reached a conclusive stage. This is a not uncommon talent, but it is one that is seldom attained except through searching study and long experience (in which the latter usually involves a few expensive mistakes). The reader of this book need not despair of acquiring “chart sense” and without undue cost—if he will concentrate on his study, watch, check, and double-check every new development on his charts, and “keep score” on himself.
It has been said that chart interpretation is not a science but an art. It is not an exact science, to be sure, because it has no rules to which there are not exceptions. Its finer points defy expression in rule or precept. It requires judgment in appraisal of many factors, some of which may seem, at times, to conflict radically with others. But to call it an art, which implies the need for genius, or at least for a high degree of native talent, is certainly improper. Say, rather, that it demands skill, but a skill that can be acquired by anyone of ordinary intelligence.