Triple Top pattern

What is a Triple Top?

A triple top is a reversal pattern built from three failed attempts to break the same price area. Price rallies to a high, pulls back, rallies to about the same high, pulls back again, and makes a third attempt that also stalls. The lows of the two pullbacks mark the support zone, and a break below it completes the pattern.

It is essentially a double top that refused to break down. After the second peak, the second pullback held, buyers tried a third time, and the resistance held again. Each failure tells the same story more firmly: someone is willing to supply shares or coins at that level, and buyers have now spent three rallies failing to absorb it. Because it takes longer to form than a double top, more participants notice the level, and more of them hold positions that depend on the support below holding.

Anatomy of the Triple Top

The three highs rarely line up exactly; a band of a few percent is normal. The pullbacks are where traders disagree. If A and B are at similar prices, the neckline is a clear horizontal line and the pattern resembles a range. If they differ, you have a choice: draw the neckline at the higher low (the first support price has to break), at the lower low (a later, more conservative trigger), or along a line joining both. PatternX uses the higher, shallower pullback.

How to identify a Triple Top

A step-by-step check for a triple top, following general practice (the exact numbers PatternX itself uses appear under detection below):

1. Prior advance. The pattern should cap an uptrend, not sit in the middle of a long sideways range.
2. Three highs in a tight band. Mark the three swing highs and check they sit close to their average.
3. Two real pullbacks. Each reaction should be deep enough to matter, not a small dip.
4. Reasonable spacing. The peaks should be spread over enough bars to be distinct tests, but not so far apart that they belong to different market phases.
5. Choose the neckline. Decide whether you will use the higher pullback low, the lower one or both, and stick with that choice.
6. Look at the third rally. A shorter, slower or quieter rally into the third peak suggests exhaustion.
7. Wait for the close below the neckline. Three highs without a break are still a range.
8. Plan the trade: a target one pattern height below the break, a stop inside or above the pattern, and a note of the other pullback low, which may act as support on the way down.

Confirmation

The confirmation is a close through the neckline, ideally on increased volume and without an immediate recovery. If you use the higher pullback low as the neckline, the lower pullback low is a second checkpoint below it; a decisive break of that level as well is stronger evidence that the range is being abandoned.

Patience matters especially with triple tops because they spend a long time looking like ranges. Before the break, the third high is just another test of resistance, and ranges break upward as well as down. A trader who shorts each top is betting on a resolution that may go the other way.

PatternX treats a fresh close below its neckline, measured on the latest bar (a bar that may still be in progress), as the breakout. It does not wait for the second pullback low to break.

Invalidation

Before the break, a sustained move above the three highs invalidates the triple top and turns the range into a potential upside breakout, often a sharp one as stops above the highs are triggered. Once price has broken down, the main failure signal is a recovery that climbs back above the neckline and stays there.

PatternX applies no pattern-based exit after a signal or trade opens: the pattern's stop level becomes the stop-loss. For a triple top that level is the neckline plus 50% of the pattern's height, halfway back into the range, unless your own stop setting is tighter. A close back above the neckline that stays below that level does not close the trade.

Volume

In a classic triple top, volume tends to decline from one peak to the next: each rally into resistance attracts fewer buyers, and the third is often the quietest. Volume then usually expands when the neckline gives way, as stops below the range are triggered and holders who waited through three failed rallies give up. Heavy volume on the third peak, by contrast, can mean the overhead supply is being absorbed ahead of an upside break.

PatternX does not test the falling-volume sequence. The only volume condition is the universal one used across all 20 detectors, which requires breakout-bar volume of at least 1.2× the 20-bar average. Volume also feeds the PatternX Score and, in PatternX AI, the volume-quality factor. See volume.

Common false signals

Triple tops have their own characteristic failures:

• Mistaking a range for a top. A sideways market produces repeated highs in a band; only a break of the lows turns that into a triple top.
• The fourth attempt. Many apparent triple tops end with price finally breaking above resistance on a fourth or fifth try.
• Neckline ambiguity. When the two pullbacks differ, a break of the higher low may stall at the lower one. Treat that lower level as a hurdle.
• Breaks into support. A higher-timeframe support level or moving average just below the neckline can halt the move before the target.
• Low-participation breaks. A thin break, especially in an illiquid instrument, is easily reversed and becomes a false breakout.

To reduce them, require a prior advance, wait for a fresh close below the neckline, check that both pullbacks were genuine, and plan for the lower pullback low as potential support.

Market context

Triple tops mean the most after an extended advance, at a level that is also resistance on a higher timeframe, and when the rallies into the second and third peaks show fading momentum. They mean less in markets that have been range-bound for a long time, where repeated highs are the norm, and a single catalyst such as earnings or a policy announcement can override them.

Because a triple top takes longer than a double top, the environment can change while it forms, so check that the conditions that made the first peak a top still apply at the third. PatternX does not verify a prior uptrend for this pattern: the reversal label is descriptive, and only rectangles check prior trend. Compare the bearish rectangle, which PatternX only flags after a prior decline.

In crypto markets

Triple tops are common in crypto ranges that form under a widely watched level, such as a round number or a previous cycle high. In a 24/7 market the three peaks can span weekends and different regional sessions, so the participants at each test may differ. Stop runs above the range are frequent; a wick that briefly pokes above the other highs can still leave the three peaks within PatternX's 3% band around their mean.

Perpetual futures add positioning information: open interest building through repeated tests while funding stays positive can mean longs are crowding into a level that keeps failing, which can intensify a breakdown. PatternX's $3M 24-hour volume floor for crypto, plus its liquidity and spread filters, screen out the thinnest markets. See perpetual futures pattern trading.

In stocks, ETFs and FX

In equities, triple tops can develop over weeks under a level investors anchor to, such as a prior high, and the third test can coincide with an earnings date. Overnight gaps can open below the neckline and skip the trigger level entirely; the entry then sits further from the stop and closer to the target, which worsens reward:risk.

For ETFs, a triple top reflects resistance across the whole basket; leveraged ETFs magnify the swings between the peaks. In FX, triple tops are common at round numbers. Because PatternX's thresholds are percentages, such as its 2% minimum pullback depth, qualifying triple tops in major currency pairs are more likely on higher timeframes than on 1-minute or 5-minute charts. See stock chart pattern recognition.

Timeframes

Needing three peaks and two pullbacks, the triple top is slower than the double top on any timeframe: intraday versions can take a session or two, daily ones often months.

PatternX gives the triple top its own, longer window: the detector needs at least 70 bars of history and reads the latest 110 bars. The three peaks must span 16 to 90 bars from first to third, and the third peak must be no more than 30 bars old at the break. On a 15-minute chart that means a formation spanning 4 to 22.5 hours; on a 4-hour crypto chart, roughly 2.7 to 15 days; on a daily stock chart, about three weeks to four months. Larger formations are not seen. PatternX scans eight timeframes, from 1m to 1d.

Example: the logic step by step

For example, a hypothetical stock on a 1-hour chart:

1. All three peaks are within 0.5% of their mean; pullbacks A and B are about 4.4% and 3.5% below it.
2. Neckline = the shallower pullback, 78.10. Height = 80.93 − 78.10 = 2.83.
3. Target = 77.90 − 2.83 = 75.07.
4. Stop = 78.10 + 50% × 2.83 ≈ 79.52, or 79.68 with the 0.2% buffer.
5. Risk 1.78 to make 2.83: reward:risk ≈ 1.6.

Pullback A's low at 77.40 sits between entry and target, so the trade has to get through that earlier support. With the 10% take-profit and 5% stop-loss defaults, both pattern levels are nearer, so they apply.

How PatternX detects the Triple Top

PatternX's triple top detector is a fixed set of rules, run on every scan of every chart you track:

• It needs at least 70 bars of history and reads the latest 110 bars.
• Swing points use a window of three bars on each side, so a peak is only known three bars after it is made.
• It takes three consecutive swing highs spanning 16 to 90 bars from first to third, each within 3% of their mean.
• Both pullbacks between them must be at least 2% deep relative to that mean.
• The neckline is the shallower pullback, the higher of the two lows.
• The third peak must be no more than 30 bars old.
• Breakout: the latest close is below the neckline while the previous close was not (a fresh break); that bar may still be forming.
• Stop = neckline + 50% of the height; target = entry − the height (the mean of the peaks minus the neckline).

Before anything reaches you, the match must also clear the market filters shared by all 20 detectors: a breakout bar finishing below its open, with volume 1.2× or more of the 20-bar average; ATR of 0.05% of price or more; a liquidity score of 30 or more; a spread within 0.25%; no order-book imbalance against the short beyond −0.5; and, on crypto, $3M or more of 24-hour volume. See pattern detection methodology.

How PatternX Classic evaluates it

Every plan includes PatternX Classic, where a triple top match receives a PatternX Score from 0 to 100, computed deterministically from 13 weighted factors. Geometry (16%, capped at 85) and the pattern type's reliability prior (8%, a fixed input rather than anything measured from results) supply 24% of it. The higher-timeframe factor, worth 7%, reads 100 bars of the next timeframe up and can only move the score; it never rejects a signal.

The score sets the signal's confidence and its band: Elite at 80 or more, High at 65, Medium at 50, Weak below that. Your minimum confidence (65 by default) must be met. The pattern's levels are reconciled with your own settings, keeping the tighter stop and the nearer target; reward:risk under 1 is discarded; and Elite accounts can add optional minimum reward:risk, ATR-band and maximum-spread filters. Signals expire after 15 minutes, and every order passes the risk controls first.

What PatternX AI adds

PatternX AI, on Pro and Elite, re-examines a triple top once the deterministic engine has already accepted it, and never finds patterns itself. Of its eight fixed factors, two matter most here. The market-structure factor looks back only 60 candles, using the last three swing highs and lows there, so against the detector's 110-bar window it judges the recent part of the formation rather than all of it. The confluence factor rewards agreement between trend, structure, volume and momentum, which a break from a long, flat range may or may not show.

The outcome is a veto, when the context score is under 48 or structure is under 35 while momentum is under 40, or else a re-weighted confidence. Stops, targets, position size and risk checks stay as they were. More in Classic vs AI.