Bull Flag Pattern — The Most Popular Continuation Setup Under the Microscope

The bull flag is probably the first pattern every trader learns: a steep "pole" of gains, a small downward-sloping "flag" of consolidation, and a breakout to the upside meant to deliver the second leg of the rally. Courses and trading blogs quote 65–75% success rates for it and call it "one of the most reliable continuation patterns."

Reality checks that story hard. In Thomas Bulkowski's database (Encyclopedia of Chart Patterns, hundreds of perfect trades on US stocks), flags have a 44% failure rate after an upside breakout — almost every second one doesn't even travel 5% in the breakout direction. Average move after breakout: 9%. The measure-rule target is hit in 46% of cases — meaning fewer than half of trades actually reach their goal.

So the world's most popular pattern turns in surprisingly weak numbers — unless you know which flags are even worth counting. The difference between "some random consolidation after a rally" and a flag that meets Bulkowski's criteria is the difference between a coin flip and a setup with a sound risk structure. This article breaks it down piece by piece.

How to Identify a Bull Flag

Identification criteria per Bulkowski — all of them, not a pick-and-choose list:

1. A flagpole is mandatory. Before the flag there must be an unusually steep, near-vertical up move lasting several days. Without a fast, straight-line rally there is no flag — just ordinary consolidation. The best flags have long, nearly vertical poles.
2. The flag is a small rectangle. Price moves between two parallel (or near-parallel) lines, usually sloping against the trend — that is, slightly downward. A horizontal variant is also acceptable.
3. Three weeks, max. A longer consolidation is already a rectangle or a channel — a different pattern with different statistics.
4. Volume shrinks during the flag — in 74% of cases on upside breakouts. Rising volume during the consolidation is a red flag for the classification.
5. A shallow retracement. The flag should give back roughly 30–50% of the flagpole. A deeper correction is no longer "a rest inside the trend" — it's a reversal candidate.
6. Confirmation: a close above the flag's upper line. The shape alone isn't a signal.

As a class, flags break out to the upside 60% of the time (measured in a bull market).

What the Numbers Say (Bulkowski, Encyclopedia of Chart Patterns)

Definitions:

• Break-even failure rate — how often price, after a confirmed breakout, doesn't even travel 5% in that direction.
• Average move — measured for flags on the short swing (to the end of the current price swing), not to a distant extreme like most patterns. That's why Bulkowski doesn't give flags an overall performance rank — they can't be honestly compared to patterns measured a different way.
• Target % — how often price reaches the measure-rule goal.

Results for flags in a bull market:

Interpreted without the sugar coating: nearly every second confirmed bull flag is a dud (move under 5%). When the move does happen, it's 9% on average — this isn't a "big second leg" pattern, it's a short swing-continuation pattern. The measure-rule target lands in fewer than half of cases, so any plan built on "always get the full flagpole" is statistically beaten before it starts.

Does that make the flag useless? No — it means it's a timing tool inside existing momentum, not a forecasting machine. Its value lies in the tight risk structure: a short stop below the flag against the chance of a several-percent push. Anyone who plays it as a fate-reversing signal is playing against their own numbers.

For contrast, it's worth knowing the exception that proves the selection rule: the high and tight flag — a special variant where price roughly doubles in about 2 months before the flag, and the flag itself is unusually tight. In Bulkowski's measurements it's the #1 pattern in the entire catalog: an average gain of 69% in a bull market, and not a single failure out of 307 cases to reach even 5%. We cover it separately in the article on the high and tight flag — the takeaway here is simple: it isn't "flag-ness" that creates the edge, it's extreme pole strength. An ordinary flag on an ordinary pole has ordinary, weak numbers.

How to Trade a Bull Flag (Measure Rule)

Entry. A close above the flag's upper line. A more cautious version: wait for a retest of the broken line and confirmation (a demand candle, a rejection) — this filters out some false breakouts at the cost of a worse entry price.

Stop. Below the low of the flag (safer) or below the upper line with a buffer (tighter, but stops parked right on the line get hunted regularly on the second test).

Target — Bulkowski's measure rule. Measure the height from the start of the price swing (the low the flagpole started from) to the top of the pole, multiply by 46%, and add it to the flag's lower edge. Yes, the multiplier makes the popular "project the whole flagpole" version more realistic — the full-length move lands less than half the time, so treat it as the optimistic scenario, not the baseline.

Filters that improve results in the data:

• A tight flag beats a loose one. A flag where price meanders, pokes outside the lines and leaves "white gaps" performs worse than a compact, disciplined consolidation.
• Slope against the trend. The best combination: an uptrend plus a flag sloping downward.
• A flat base. A flag breaking out above a long flat base foreshadows a bigger move.
• Near the yearly low. The best results come from breakouts in the bottom third of the yearly range — flags of a "fresh" trend, not ones drawn after a year of gains.
• Half staff. The flag often appears at roughly the halfway point of the move (the move before ≈ the move after) — hence the "half-staff flag" nickname. It's a target-location heuristic, not a guarantee.

Position management. Since the average move is 9%, the plan has to be short: take partial profit at 1:1 relative to risk, and trail the rest with a stop dragged behind subsequent lows. The worst thing you can do with a flag is turn it into an investment position — the pattern promises a swing push, not a new trend. If price returns inside the flag after the breakout and closes below its low, the signal is broken: exit, don't negotiate. A broken bull flag often flows smoothly into a supply signal, as trapped buyers start giving up their positions.

Invalidation. A flag also stops existing once the consolidation runs past 3 weeks (reclassify as a rectangle or channel and recalculate the stats), or once the retracement deepens clearly below half of the flagpole — at that point you're looking at a potential reversal, not a rest in the trend.

Myth vs Measurement — Where Does That 70% Online Come From?

The gap between the "65–75% success rate" from courses and Bulkowski's 44% failure rate has three sources. First, nobody defines failure: if "success" is any upward move at all after the breakout, the win rate always comes out high — Bulkowski requires a minimum of 5% and measures systematically. Second, selective memory: the textbook flags shown on Twitter are the ones that worked; hundreds of failed flags never make it into the posts. Third, blurred definitions: without a steep flagpole and a 3-week limit, every pause after a rally gets called a "flag," and then you're measuring noise.

The honest conclusion: the bull flag works as an entry structure inside strong momentum — a short stop, a quick target, no romance with the position. It doesn't work as a standalone direction oracle. Selection (steep pole, tight flag, shrinking volume, shallow retracement) isn't cosmetic — it's the only thing separating the pattern from a random rectangle on the chart.

An example of the gap in practice: on crypto charts, "bull flags" get drawn on every pause after a green candle, often on the M15 with no trace of a pole. Then half of them "don't work," and the pattern gets blamed. In reality the pattern never occurred there — what occurred was a consolidation someone called a flag because they wanted to buy. If there's one sentence to remember from this article, let it be this: pole first, then flag, then breakout — and calculate the target with a multiplier, not a wish.

Disclaimer: Bulkowski's measurements cover US stocks on the daily timeframe, in a bull market. Nobody has rigorously replicated these numbers on crypto or lower timeframes — treat them as a reference point, not a 1:1 conversion.

FAQ

What is the real success rate of a bull flag? 44% failure rate after an upside breakout, an average move of +9%, target hit in 46% of cases (Bulkowski, US stocks, daily). The advertised 65–75% comes from material with no failure definition and no systematic measurement.

How is a bull flag different from a bullish pennant? A flag has parallel lines (a rectangle); a pennant has converging lines (a small triangle). Flags perform better: a 46% target rate vs 35% for pennants, and 44% failures vs 54%. Both need a steep flagpole and last a maximum of 3 weeks.

How do you set a target from a bull flag? The height from the start of the swing to the top of the flagpole × 46%, added to the bottom of the flag. Treat the full-flagpole projection as the optimistic case — it lands in fewer than half of trades.

FAQ

Rafał — Krypto Bez Ściemy

Trader and founder of Strefa Tradingu. He’s been breaking crypto down on YouTube for years — no hype, no signals, with a focus on market structure and risk management.

kryptobezsciemy.plInstagramYouTubeKlub Zero