Chart Patterns That Actually Matter

By Silviu Vasilescu · Updated 1 August 2026 · 11 min read

There are catalogues listing sixty chart patterns. Almost all of them are variations on two ideas: price paused inside a trend and then resumed, or price ran out of buyers and turned. Learning the two mechanisms is worth more than memorising the sixty names, because the mechanism tells you when the shape means something and when it is a coincidence.

A pattern is a description of who has been trading and at what price. Nothing more. Every one below is presented with the market behaviour that creates it, because a pattern you can explain is one you can also reject — and rejecting the bad ones is where the money is.

1. Two kinds of pattern, one mechanism each

Continuation patterns form when a stock has moved a long way and pauses. The people who bought early have a profit and some of them take it; the people who missed the move are waiting for a pullback. Those two groups trade against each other for a while, which produces a sideways or slightly counter-trend drift on falling volume. When the profit-takers are done and the latecomers are still there, the original direction resumes.

Reversal patterns form when an advance stops attracting new buyers. Price reaches a level, fails to hold, tries again with less force, and each attempt leaves behind a group of people who bought near the high and are now underwater. Their eventual selling is the supply that turns the trend. This is why reversal patterns take time and involve repeated tests — a genuine change in control is a process, not an event.

Hold those two descriptions in mind and the catalogue collapses into something manageable. A flag, a pennant, a wedge and a rectangle are the same pause with different geometry. A double top and a head and shoulders are the same failure with a different number of attempts.

2. Continuation patterns

Flags and pennants. A sharp move, then a short, tight consolidation that drifts slightly against the move, then a resumption. A flag is roughly parallel; a pennant converges. Both should be brief relative to the move that preceded them and both should show volume drying up during the pause. That volume decline is the signal that supply is being withheld rather than distributed — without it you are looking at a stall, not a flag.

Bases and rectangles. A longer sideways range with a reasonably clear ceiling and floor. Time matters here: a range that has held for months has absorbed a great deal of supply inside a narrow price band, so the eventual break has more behind it than a break from a three-day pause. The most tradeable bases tighten toward the end — the swings inside the range get smaller, which says the disagreement between buyers and sellers is narrowing.

Ascending and descending triangles. One boundary flat, the other sloping toward it. An ascending triangle has a flat ceiling with rising lows: sellers are defending one price while buyers keep paying more for the dips, and each higher low is a small piece of evidence about who is more determined. Descending triangles are the mirror image.

Symmetrical triangles — both boundaries converging — are the weakest of the group, because a narrowing range with no directional bias tells you volatility is compressing without telling you which way it will release. They are common in the middle of indecisive markets and they break both ways. Trade them only with the direction of the larger trend, or not at all.

One rule covers all of them: a continuation pattern is only a continuation pattern if there is something to continue. A "flag" in a stock that has gone nowhere for six months is a rectangle in a range, and the trade you thought you were taking does not exist.

The pause and the resumption. What makes it tradeable is not the shape of the pause but what came before it and how volume behaved inside it.

3. Reversal patterns

Double tops and double bottoms. Price reaches a level, retreats, returns to the same level and fails again. The information is in the second attempt: it means the sellers who stopped the first advance are still there and were not absorbed. A double top is confirmed only when price breaks below the low between the two peaks — until then it is a stock trading at resistance, which is not the same thing and happens constantly.

Head and shoulders. Three attempts: a peak, a higher peak, then a lower peak. Read it as a sequence of weakening demand — the third rally could not reach where the second reached, and the last group of buyers is now above the market. The neckline joins the two intervening lows, and the pattern completes on a break below it, usually on expanding volume. The mirror image at a bottom is an inverse head and shoulders.

Rounding tops and bottoms. No sharp turn, just a gradual loss of momentum where the slope flattens, reverses and steepens the other way. These are less precise to trade because there is no clean level to break, but they are honest patterns and they show up on longer timeframes where large positions are accumulated or distributed over months.

Climax moves. A near-vertical advance on enormous volume, then a violent reversal. What happened is that the last available buyers all bought at once, and once they are in, there is nobody left to pay more. These are recognisable but genuinely difficult to trade: the same configuration precedes both a sharp top and a further leg up, and the volatility means stops have to be wide, which forces small positions.

Reversal patterns fail more often than continuation patterns, for a structural reason worth remembering: they are bets against the most persistent feature of a chart. Trading one means accepting a lower hit rate in exchange for entering near an extreme, which is a legitimate trade-off but has to be made deliberately and reflected in how much you risk.

4. Breakouts, and why most fail

A breakout is price moving beyond the boundary of a pattern. The uncomfortable fact is that a large share of them do not follow through, and the reasons are mechanical rather than mysterious.

Stop orders cluster just beyond obvious levels, and clustered stops are liquidity that larger participants can trade against. Price pokes through the level, triggers the stops, fills the orders that were waiting there, and returns inside the range. Nothing about the pattern was wrong; the break simply had no demand behind it once the mechanical orders were done.

Three filters do most of the work of separating these:

• Volume on the break. A real break moves a lot of stock — meaningfully above the recent average. A break on ordinary volume has not consumed the supply at that level.
• The close, not the touch. Requiring a close beyond the level, rather than a print through it, eliminates most intraday probes at the cost of a slightly worse entry price. That trade is almost always worth making.
• Behaviour after the break. A genuine break tends to hold above the level on the retest. Price that immediately falls back inside the range has told you what the move was, and it costs far less to act on that than to argue with it.

Some traders skip the breakout entirely and buy the retest instead — waiting for price to break out, pull back to the old boundary and hold. That gets a tighter stop and a better price, at the cost of missing the moves that never come back. Both are defensible; what is not defensible is deciding which one you are doing after the break has already happened.

5. Context is the pattern

The same shape has completely different value depending on where it appears. A base breakout in a stock above its rising 200-day average, in a sector that is leading, while the broad market is advancing, is one trade. The identical shape in a downtrending stock during a market-wide decline is another, and it is not the same idea with slightly worse odds — it is a different bet with a different mechanism against it.

Before treating a pattern as tradeable, three questions cost nothing: What is the larger trend on the weekly chart? Is the sector participating? Is the broad market supporting this direction? Patterns that pass all three are a small fraction of the ones you will notice, which is the point — most of the work of a discretionary method is throwing candidates away.

This is also where pattern trading connects to screening: if your universe is already filtered for trend and relative strength, the context questions are mostly answered before you look at a single chart.

6. How to actually trade a pattern

Recognising a pattern is not a trade. A trade needs four prices written down before you place the order:

1. Trigger: the exact price and condition that puts you in — for example, a close above the range high on volume at least 1.5 times the fifty-day average.
2. Stop: the price that says the pattern failed. Usually back inside the range, below the last swing low, or a multiple of average true range below entry — beyond the noise, not inside it.
3. Size: computed from the distance between those two, so the trade risks the same as every other trade you take.
4. Exit when right: a measured target, a trailing stop, or a partial exit at a fixed multiple of risk with the remainder trailing. Decided in advance, because this is the decision that emotion damages most.

The traditional measured target for a rectangle or triangle is the height of the pattern projected from the breakout point, and for a head and shoulders it is the distance from head to neckline projected down. Treat these as rough expectations rather than predictions — useful for judging whether the reward justifies the stop before you enter, which is their real purpose.

7. Four ways pattern trading goes wrong

Seeing patterns everywhere. Human vision finds shapes in noise, and a chart with enough bars contains something resembling every pattern in the catalogue. The defence is written criteria — minimum duration, minimum prior move, volume conditions — applied before you get interested rather than after.

Anticipating completion. Buying a head and shoulders before the neckline breaks, or a base before the breakout, is trading a pattern that does not exist yet. Most incomplete patterns never complete. Waiting costs you a slightly worse entry and saves you the majority of the failures.

Redrawing after the fact. If a level breaks and you move the line to keep the pattern alive, you are no longer analysing — you are protecting a position. This is where a written stop earns its keep: it makes the question already answered.

Assuming a name confers reliability. A head and shoulders is a well-documented pattern, and it still fails a large fraction of the time. Patterns improve your odds slightly and give you a cheap place to be wrong. That is a genuine and useful contribution, and it is not the same as being right.

This guide is educational material about how markets and trading methods work. It is not financial advice and not a recommendation to buy or sell any security. Trading involves the risk of losing money, including more than you intended if you use leverage.