Classical Chart Patterns
Pauses that resume
Most of the time a trend is not moving. It is resting: a few weeks of sideways chop that looks, to a nervous holder, exactly like a top. Continuation patterns are how you tell a pause from an ending — and the tell is almost always the same, because a pause is a trend catching its breath while an ending is a trend changing hands.
During a sharp advance, early buyers are sitting on profits and want to ring the register; new buyers want in but not at the top. A consolidation is the negotiation between them. If the new buyers absorb the supply, the pattern resolves upward and the trend continues with a fresh set of holders whose cost basis is higher — which is why breakouts from long bases run. If the supply overwhelms them, the same shape resolves downward. The shape does not decide; it just holds the negotiation in a readable frame.
- SYMMETRICAL TRIANGLE — lower highs and higher lows squeezing into a point. Both sides are giving ground; it is the most neutral of the four and the trend it interrupts gets the benefit of the doubt.
- ASCENDING TRIANGLE — a flat ceiling with rising lows. Buyers keep paying more while a fixed wall of supply sits overhead; when that wall is eaten, the move is usually fast. Descending is the mirror.
- FLAG and PENNANT — a short, tight drift against the trend after a near-vertical move (the "pole"). These are the briefest patterns, days rather than weeks, and the sharpest resolutions.
- RECTANGLE — a clean horizontal range between parallel support and resistance. It is a continuation pattern only when the trend that preceded it was strong; otherwise it is simply a range, and you trade it as one.
A stock runs from 30 to 42 in eight sessions — a 12-point pole — then drifts sideways-to-lower between 39 and 41 for two weeks on shrinking volume. That drift is the flag. The conventional target takes the pole's length and adds it to the breakout point: a break at 41 projects roughly 41 + 12 = 53. Note what the measurement is really claiming: that the second leg of a move tends to resemble the first, because it is driven by the same crowd with the same conviction. When the flag takes much longer than the pole, or drifts far more than a third of it, that assumption is weakening and the target should be treated with suspicion.
Inside any healthy continuation pattern, volume dries up — fewer and fewer shares change hands as the range narrows, because the argument is running out of participants. Then the breakout comes on a visible surge. That sequence, quiet then loud, is what separates a genuine pattern from a chart that merely looks like one. A "triangle" whose volume stays heavy throughout is not consolidating; it is a fight, and fights are resolved by whoever is bigger, which you cannot see. A breakout on volume no larger than the average days before it is the single most common false break there is.
Murphy's practical guidance is that a triangle should resolve somewhere between half and three quarters of the way to its apex. A pattern that squeezes all the way into the point has lost its energy: the two sides ran out of disagreement rather than one side winning, and the breakout that follows is far more likely to be a false one that immediately reverses. If you are still waiting at the apex, the trade has already told you something — the setup expired, and the correct action is to remove the drawing rather than to wait harder.
Calling something a "continuation pattern" before it breaks is a prediction dressed as a label. The identical symmetrical triangle appears at market tops and resolves downward often enough that trading it purely on the assumption of continuation is a losing habit. The professional version: form the expectation from the prior trend, place the order on the side you expect, and place a stop on the other side of the pattern. Then you are paid when you are right and you are out cheaply when the label was wrong.
Reversal patterns and continuation patterns are not two families. They are one question — is the crowd still recruiting? — asked in two situations. Both are drawn with converging or parallel lines, both are confirmed by a break, both are measured by projecting their own height, and both live or die on whether volume shrank inside and expanded on the exit. Learn the volume signature and the shapes become almost incidental.
A market rallies hard, then trades in a narrowing triangle for six weeks. Volume during those six weeks is roughly flat — no drying up at all — and the eventual break comes on average volume. How much confidence does the measured target deserve? (Very little. Both volume conditions failed: the consolidation never ran out of participants, and the break was not endorsed by new ones. The shape is textbook and the evidence behind it is absent, which is the exact profile of a false breakout.)
Symmetrical triangles are neutral and defer to the prior trend; ascending and descending triangles name the side that is running out of patience; flags and pennants are brief pauses after near-vertical moves and are measured by the pole; rectangles continue only when what preceded them was strong. Inside a real pattern volume dries up; on a real break it expands. Resolve triangles before three quarters of the way to the apex, and always place the stop on the far side of the pattern so a mislabelled shape costs you a small, known amount.
Pick a chart with a clear consolidation in the last six months. Draw the pattern, then add the volume panel and answer two yes/no questions: did volume shrink inside the pattern, and did it expand on the break? Write the two answers and the outcome. Repeat on five charts. You are building the only filter that matters here, and you are building it from your own instrument rather than from a textbook's chosen examples.
Classical Chart Patterns