High and Tight Flag — The #1 Pattern in Bulkowski's Ranking

In Bulkowski's catalog, measured across tens of thousands of patterns, the top of the ranking isn't held by a head and shoulders or a cup and handle. It's held by a pattern most traders have never consciously played: the high and tight flag. The numbers from the Encyclopedia of Chart Patterns read like a typo: an average gain after breakout of +69% in a bull market, and of 307 studied cases, zero — literally zero — failed to gain at least 5% after breakout.

Before you rush to a scanner, the catch sits in the precondition. To even talk about a high and tight flag, price must first nearly double in about two months. That filters out everything except the most explosive markets — and that's exactly why the pattern is a gem: rare, but measurable. And since we're "no hype," we'll also show a newer study by Bulkowski himself that walks part of the legend back.

How to Identify a High and Tight Flag

1. The flagpole: a gain of at least 90% in under two months (roughly 42 daily candles). This is the constitutive condition — without price doubling, there's no pattern, just an ordinary flag, and ordinary flags have completely different, much weaker statistics. In Bulkowski's study the median rally was 102%, the average 111%, and the climb to the top took an average of 36 days.
2. The flag: a tight, shallow consolidation below the top. After the rally, price rests — ideally retracing 10-34% (measured from the flagpole's high) over about 10-29 days. The shape can vary — a flag, a small pennant, a tight rectangle, sometimes a shapeless drift — geometry is secondary to the proportions here.
3. A retracement shallower than half is the rule of thumb: a consolidation giving back most of the rally isn't "tight" — it's distribution.
4. Volume usually fades during the flag — the market goes quiet before the next move.
5. Context before the rally: a flat base. The newer study found that flags performing best were preceded by a flat or gently rising price. A rally launching from a violent low (like the 2008 bear-market bounces) produced the weakest results.
6. Confirmation: a breakout above the flagpole's high. The formation is only active once price clears the highest point of the rally. This isn't a formality — it's half of the edge (more on that shortly).

What the Numbers Say (Bulkowski — the Book and the Newer Study)

The book version (manually selected patterns, bull market): average gain +69%, 0/307 cases below the +5% threshold after breakout, rank #1 in the entire catalog. No other pattern came close to these numbers.

The later version — and this is where it gets instructive. Bulkowski repeated the study automatically: 1,018 stocks, 1995-2009, 2,588 patterns with no manual selection. Results:

The takeaways, one at a time:

The "+69%, zero failures" legend applies to manually selected patterns. When the computer took everything that met the numerical criteria, the average fell to ~27% and failures appeared. That doesn't invalidate the pattern — 27% average gain is still an excellent result — but it shows that part of the edge lived in the selector's eye, not in the pattern alone. In fact, in a newer time-based ranking Bulkowski himself placed it around 43rd out of 56.

Waiting for the breakout cuts risk roughly in half. Without confirmation, total failure is 33%; after breaking above the flagpole's high, it's 19%. It's one of the cleanest illustrations of "confirmation first" in the whole catalog.

The flag's proportions matter. Retracements of 10-34% and a width of 10-29 days are the best-performing zone. A flag giving back 60% of the rally, or one dragging on for a quarter, is a different animal.

A stop "below the flag" can be expensive. The flag's low sat an average of 26% below the breakout price — a mechanical stop there means a lot of risk on the position. Small consolation: in cases where price touched the flag's low and still went on to new highs, the loss on that test averaged around 10%.

Disclaimer: US stocks, daily timeframe. Crypto produces doublings in two months far more often than the stock market — but nobody has measured whether the edge's proportions carry over. Treat the criteria as a selection filter, and the numbers as a reference point.

How to Trade a High and Tight Flag

Step 1: measure the flagpole. A gain of at least 90% from low to high in ≤2 months, measured on extremes. "Almost doubling" over five months isn't this pattern. Also check what preceded the rally — a flat base is a plus, a violent drop right before the rally is a red flag.

Step 2: evaluate the flag. A retracement in the 10-34% zone, a duration of 10-29 days, fading volume. Deeper and longer consolidations statistically hurt the result.

Step 3: enter only after the breakout. An order above the flagpole's high, or an entry after a candle closes above it. Buying "into the flag" for a better price doubles the failure risk — that's not an opinion, it's a measurement (33% vs 19%).

Step 4: a stop below the flag's low — consciously. That's the natural invalidation level, but calculate how much it gives up in percentage terms: if it's too much, size down instead of dragging the stop into the noise right under the breakout.

Step 5: target from the measure rule, plus trailing. The classic rule for the high and tight flag: half the height of the flagpole, added from the flag's low. With a distribution this skewed (22% of cases gaining over 45%, 6% doubling), it makes more sense to take partial profit at the target and trail the rest — those tail outcomes are exactly what built this pattern's legend.

What not to trade: a "high and tight" without price doubling (the most common mistake — an ordinary nice-looking flag is a statistically average setup), entries before the breakout, and penny stocks pumped from $0.10 to $0.20 — Bulkowski deliberately excluded those from the study because they're manipulation, not demand.

How often does it show up? Rarely — and that's fine. In the automated study, 552 stocks out of more than a thousand generated 2,588 patterns combined over fourteen years — statistically a handful per stock per decade. If your scanner is spitting out high and tight flags every week, the market hasn't gotten generous — your criteria have drifted, most likely on the doubling condition.

Myth vs Measurement

The myth: "The high and tight flag is a failure-proof pattern — 307/307, you buy and double your account."

The measurement: 0/307 was true for manually verified patterns in one specific book study. The automated repeat on 2,588 cases showed 19% failures after breakout and an average of ~27% — excellent, but not magic. On top of that, the result has a skewed distribution: the median gain sits well below the average, which gets pulled up by rare rockets.

Where does the myth come from? A single number frozen in time. "Bulkowski's #1 pattern" is a true quote — but from the second edition of the book, before the update. The author himself made his own numbers more realistic; the internet quotes the prettier version. The broader lesson beyond this one pattern: every statistic has an expiration date and a methodology, and the gap between manual and automated selection can eat up half the edge.

No hype: the high and tight flag is probably the best momentum filter described in the pattern literature — it forces you to only play the strongest markets, in the direction of the trend, after confirmation. Even after the numbers get walked back, it's still a rare gem. Just remember the game is played on a fat-tailed distribution: most trades will be ordinary, and the year's result will be made by the one you don't cut too early.

FAQ

What is a high and tight flag? A bull flag on steroids: the precondition is a gain of roughly 90-100% in a maximum of two months, followed by a tight consolidation (a 10-34% retracement over about 10-29 days). In Bulkowski's book study: an average of +69% after breakout and 0/307 failures — pattern #1 in the catalog.

Why is it so rare? Because a price doubling in two months happens to only a handful of markets a year. That rarity is a feature, not a bug — the pattern is essentially a filter that selects for extreme momentum, not a shape to go hunting for on every chart.

Do its statistics still hold up? With a correction: the automated study on 2,588 patterns showed an average gain of ~27% and 19% failures after breakout (33% without waiting for the breakout). Key practices: only enter after the flagpole's high is broken, and prefer flags with a 10-34% retracement sitting on a flat base.

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