Bullish Pennant pattern

What is a Bullish Pennant?

A bullish pennant, or bull pennant, is a short-term continuation pattern that combines a strong advance with a rapid squeeze in volatility. After the pole, a fast and near-vertical rally, price swings back and forth in a range that tightens with every swing: each high a little lower, each low a little higher. Drawn with two converging trendlines, it looks like a small pennant flying from the top of the pole.

The psychology is equilibrium under pressure. Profit-takers sell into the top of the range, but buyers step in at progressively higher prices, unwilling to wait for a deep pullback. As the range narrows, the market is in balance, yet that balance comes right after a burst of demand, which is why the conventional expectation is a resolution in the direction of the pole. The narrowing also means the eventual move can be abrupt, because orders cluster just outside an ever-tighter range. A pennant is, in effect, a miniature symmetrical triangle with a pole attached.

Anatomy of the Bullish Pennant

Thinking in halves is a practical way to judge convergence without drawing trendlines: if the back half of the consolidation sits entirely inside the front half and spans a smaller range, the pattern is narrowing. That is exactly what separates a pennant from a bull flag, which is a channel that need not narrow, and from a rectangle, whose edges stay parallel. Proportion matters too. The pennant should be small next to its pole, in both height and duration; a consolidation that rivals the pole in size is a different pattern.

How to identify a Bullish Pennant

1. Confirm the pole. A steep advance over a handful of bars, clearly faster than the trend before it.
2. Check where the pause begins. The pennant should form right where the pole ends.
3. Look for convergence. Lower highs and higher lows or, more simply, a back half that is tighter than the front half.
4. Keep it shallow. The pennant should hold in the upper part of the pole. A pullback that gives back more than about half of it is something else.
5. Keep it short. Pennants are brief compared with the trend that produced them. A long, slow triangle is a symmetrical triangle, not a pennant.
6. Check volume. It usually contracts as the range narrows.
7. Set the trigger above the upper trendline or above the most recent swing high in the pennant.
8. Plan the exits: a stop below the pennant's low and a target equal to the pole's height above the breakout.

PatternX's exact numbers for these checks are in the detection section further down.

Confirmation

A bullish pennant is confirmed by a close above the pennant, ideally with an expanding range and volume. Because the upper edge slopes down, there are two common triggers: a break of the upper trendline, which comes first, or a close above the most recent swing high inside the pennant, which comes later but filters more noise. PatternX uses a version of the second: a fresh close above the high of the pennant's second half, its narrower and more recent part. By rule, that level is below the first half's high, so PatternX can enter before price has reclaimed the top of the pennant or the pole. A long-standing rule of thumb also favors breakouts that arrive well before the apex. Keep in mind that PatternX evaluates the latest bar, which may still be forming.

Invalidation

The pennant fails when price closes below its lower edge, or when the pullback deepens past roughly half the pole and the advance is no longer just pausing. A breakout that immediately falls back inside the pennant is a warning, and a drift into the apex without a break suggests the buying impulse has faded. Under PatternX's rules, the stop is the pennant's lowest point. Because the second half's low must be above the first half's, that point is always in the first half, which typically makes it wider than a stop under the latest swing low. After a signal or trade opens, PatternX applies no further invalidation logic; that stop, or your tighter stop-loss setting, is the stop-loss.

Volume

Pennants have a characteristic volume rhythm: a burst during the pole, a steady decline as the swings contract, and a fresh expansion on the breakout. The contraction reflects a market waiting for information; the expansion shows the wait is over. A breakout on volume no higher than the pennant's average is easy to doubt, and one on surging volume is generally treated as more convincing.

PatternX tests only the breakout, and only in the universal way: the breakout bar's volume must be at least 1.2× its 20-bar average, the same test applied to all 20 patterns. Its pennant rules do check contraction, but of the price range (second half against first half), not of volume. If you want to see volume dry up inside the pennant, check the chart yourself before approving a signal.

Common false signals

Pennant breakouts fail for a few recurring reasons:

• Apex drift. When price wanders into the tip, the range is so narrow that any small move counts as a breakout, and many of those are noise.
• Fake-outs in both directions. Tight ranges invite stop runs above and below; a poke above the pennant that reverses can be followed by the real move down.
• Exhausted poles. A pole that is the last leg of a long, extended rally can be climactic buying rather than fresh demand.
• Thin breakouts that move price without drawing in participation.
• Mislabeled patterns. A slow, wide triangle after a modest move is a symmetrical triangle that can break either way, not a pennant.

A practical discipline is to require the breakout to clear a small buffer and to decide in advance how long you are willing to wait for it. See false breakouts for the general mechanics.

Market context

A bullish pennant is most convincing when the broader market supports the pole: a rising index or sector, price above its moving averages on the higher timeframe, and a catalyst that keeps buyers interested. Pennants often appear mid-trend, as the pause between two legs of a strong move, rather than at the very start of an advance. They are less dependable in low-volatility, range-bound markets, where narrow consolidations form constantly without a meaningful pole. Watch the volatility regime: a pennant is a compression, and a compression that forms when overall volatility is already very low may need an outside catalyst to resolve.

In crypto markets

In crypto, pennant poles often come from squeezes: liquidations of short positions push price up in a few bars, then the market pauses to digest. On perpetual futures, open interest helps tell the story. A pole with rising open interest suggests new long positions, while one with falling open interest suggests shorts closing, which some traders regard as less durable. The PatternX Score includes small open-interest and funding-rate factors, though they carry little weight next to breakout strength and volume. Because crypto trades around the clock, a pennant can break in quiet hours; the 1.2× volume test is relative to the coin's own recent volume, not an absolute level, while the separate $3M 24-hour minimum screens out thin coins. See PatternX for crypto.

In stocks, ETFs and FX

In stocks, bullish pennants often follow news-driven moves: an earnings beat or an upgrade produces the pole, and the stock consolidates tightly before deciding whether to extend. On intraday charts, a pennant that forms during the midday lull may not resolve until activity returns later in the session, by which point it may have drifted toward its apex. Leveraged ETFs, which are part of PatternX's curated list, form pennants too, but they magnify both the pole and any failure. PatternX reaches these markets through your own broker connection: Interactive Brokers, Alpaca and Tradier cover US stocks and ETFs with automated execution, while manual-only brokers receive signals for you to act on. See stock chart pattern recognition.

Timeframes

PatternX scans for pennants on 1m, 3m, 5m, 15m, 30m, 1h, 4h and 1d charts, and the whole formation must fit in roughly the last 41 bars: a 5–15-bar pole that ended 6–25 bars before the latest bar. In clock time that is up to about 10 hours on a 15-minute chart, about 20 hours on a 30-minute chart and roughly a week on a 4-hour crypto chart. On a 1-minute chart, a pennant of a handful of bars is a few minutes of trading, where random wiggles can satisfy the halves rule; on 4h and daily charts the same rules describe multi-day consolidations. The next timeframe up adds to or subtracts from the score but never blocks a signal.

Example: the logic step by step

For example, with hypothetical numbers: a leveraged ETF near 40.00 rallies to 41.60 over 7 bars on the 30-minute chart, a pole of 1.60 (4%). It then consolidates for 10 bars. The first five swing between 41.75 and 40.95 (a range of 0.80); the last five between 41.50 and 41.15 (0.35, about 44% of the first). The second half's high is lower and its low higher, and the deepest pullback gives back 0.65 of the pole, about 41%. A new bar rises to 41.62 and is at 41.58 when evaluated, a fresh close above the second half's high of 41.50, on 1.4× its 20-bar average volume.

• Entry: about 41.58, still below the first half's high of 41.75.
• Stop: the pennant's low, 40.95, a risk of 0.63 (1.5%).
• Target: 41.58 + 1.60 = 43.18, a reward of 1.60 (3.8%), the measured move.
• Reward:risk: about 2.5, or about 2.2 once PatternX places the stop 0.2% below the pennant's low.

Both levels sit inside the default 10% take-profit and 5% stop-loss limits, so the pattern levels stand.

How PatternX detects the Bullish Pennant

PatternX's bullish pennant detector applies fixed rules to roughly the last 41 bars, identically in Classic and AI mode:

1. Pole: the strongest advance over 5 to 15 bars, at least 3%, ending 6 to 25 bars before the latest bar.
2. Pennant: the consolidation between the pole and the breakout bar, at least 4 bars, split into two halves.
3. Convergence: the second half's high must be below the first half's high, and its low above the first half's low.
4. Contraction: the second half's range must be no more than 75% of the first half's.
5. Retracement: the pennant may give back at most 55% of the pole, slightly more room than the 50% PatternX allows a flag.
6. Breakout: a fresh close above the second half's high on the latest bar, meaning the previous bar had not already closed above it.
7. Levels: stop = the pennant's low; target = entry + pole height.

The universal filters then apply: the breakout bar must close in the direction of the trade, with volume at least 1.2× the 20-bar average; ATR at least 0.05% of price; liquidity score at least 30; spread at most 0.25%; order-book imbalance against the trade no worse than −0.5; and, for crypto, at least $3M of 24-hour volume. More in how PatternX detects patterns.

How PatternX Classic evaluates it

PatternX Classic scores a detected pennant with the 13-factor PatternX Score. Geometry (capped at 85) and a fixed prior together make up 24% of the weight. The other 76% describes the market around the pennant: breakout strength and the volume surge, which reflect how decisively the squeeze resolved; trend alignment on EMA20/50 and agreement from the next timeframe up; and liquidity, spread, order-book imbalance, volatility (scored highest when ATR is 0.2–2% of price), funding, open interest and PatternX's live win rate on that market. The score becomes the confidence (Elite 80+, High 65+, Medium 50+, Weak below 50) and must meet your minimum, 65 by default and adjustable from 50 to 95. Levels are merged with your take-profit and stop-loss, reward:risk must be at least 1, the risk engine runs before any order, and unapproved signals expire after 15 minutes.

What PatternX AI adds

With PatternX AI on Pro or Elite, an accepted pennant passes through an 8-factor context evaluation: a fixed, transparent formula that never finds patterns. For a pennant breakout, trend strength (EMA12 versus EMA26) and volume quality (the last bar against its 20-bar average) speak to whether the advance still has fuel; the volatility regime factor scores ATR(14) as a share of price and favors readings around 1%; and confluence counts how many of trend, structure, volume and momentum agree. A context score under 48, or weak structure (below 35) combined with weak momentum (below 40), vetoes the trade; otherwise the confidence is re-weighted and a short template explanation is attached. Stops, targets and size do not change. How the AI layer works.