A flag pattern is a short consolidation that slopes against the direction of a sharp prior move, forming a shape like a flag on a pole. The sharp move is the flagpole; the consolidation is the flag. Traders treat it as a continuation pattern โ the expectation is that the trend resumes in the direction of the pole rather than reversing.
A bull flag forms after a sharp advance and drifts slightly downward or sideways. A bear flag forms after a sharp decline and drifts slightly upward. In both cases the consolidation is orderly and shallow relative to the move that preceded it, which is the whole point: a violent move followed by a calm, shallow pause suggests profit-taking rather than a change of control.
This page covers the conditions that separate a flag from ordinary chop, how it differs from a pennant, why the tight stop it offers is where most sizing mistakes happen, and how flags fail.
What the shape implies
The flagpole is a period where one side was in complete control โ price moved a long way in a short time with little opposition. The flag is what happens next: some participants take profits, some fade the move, and price drifts back gently against the trend.
The information is in the contrast. If the counter-move were a genuine reversal, you would expect it to be as forceful as the move that preceded it. Instead it is shallow, slow and orderly. That asymmetry โ violent in one direction, gentle in the other โ is what traders read as the trend pausing rather than ending.
That reading is an interpretation, not a fact about the market. Plenty of flags resolve against the trend, which is why the invalidation level matters as much as the entry.
The conditions that qualify a flag
| Criterion | What to look for | Why it matters |
|---|---|---|
| Flagpole | A steep, near-vertical move over few candles | Without a strong pole there is no trend to continue |
| Flag slope | Counter to the pole, or sideways | A flag sloping with the trend is usually just continuation, not a pattern |
| Retracement depth | Shallow, commonly under half the pole | A deep retracement suggests a genuine reversal, not a pause |
| Flag shape | Two roughly parallel boundaries | Converging boundaries make it a pennant, not a flag |
| Duration | Short relative to the pole | A consolidation that runs long stops being a pause |
| Volume | Typically lighter through the flag than the pole | Falling participation during the drift, where volume data is reliable |
The volume row deserves a caveat: spot forex has no centralised volume, so tick volume is a proxy for activity rather than a measure of it. On futures, equities and indices the reading is more meaningful.
Flag, pennant and triangle
These three are commonly confused, and the distinction is in the boundaries of the consolidation.
| Pattern | Consolidation shape | Boundaries |
|---|---|---|
| Flag | Rectangular channel sloping against the trend | Roughly parallel |
| Pennant | Small triangle after a sharp move | Converging |
| Symmetrical triangle | Larger, often not preceded by a pole | Converging, longer duration |
A bullish pennant and a bull flag carry the same underlying idea โ a sharp move followed by a shallow pause โ and differ in whether the pause is a channel or a wedge. In practice, the trading logic is close to identical, and it is not worth agonising over which label applies to a particular consolidation.
Entry, stop and target
Entry. The conventional trigger is a close beyond the flag's boundary in the direction of the pole. More patient traders wait for a retest of the broken boundary, which gives a better price at the cost of missing the moves that never come back.
Stop. The stop belongs on the far side of the flag โ below the flag's low for a bull flag, above the flag's high for a bear flag โ with a buffer for spread and noise. That is where the premise of the pattern has failed: if price trades through the whole consolidation in the wrong direction, the move was not a pause.
Target. The convention is the measured move: take the height of the flagpole and project it from the breakout point. This is a convention traders use to set an initial objective, not a forecast. Price has no obligation to travel the length of the pole, and frequently does not. Treat it as one way of framing whether the trade offers enough reward to be worth the risk, then manage the position on what price actually does.
The sizing trap in a tight flag
This is the part that costs traders the most money on this pattern, and it is worth being blunt about.
A flag's appeal is that it offers a tight stop. The consolidation is shallow, so the distance from a breakout entry to the far side of the flag can be very small โ sometimes a fraction of the pole. A tight stop means that for a given percentage of risk, the position size is large.
That is fine when it is deliberate and dangerous when it is not. Two failure modes recur:
Sizing from habit rather than from the stop. A trader who normally uses a 40-pip stop takes a flag with a 12-pip stop and enters their usual position size. They have just tripled their risk on the trade without deciding to.
Forgetting that a tight stop is easier to hit. A 12-pip stop on an instrument that routinely moves 15 pips against you before resolving will be taken out regularly. Combining a large position with a stop inside the market's normal noise is how a run of small technical losses becomes a large drawdown.
The correction is mechanical: fix the stop from the chart, then compute the size from the stop and your risk percentage. Never the other way round. The lot size calculator does this for any balance and stop distance; for the pairs where flags are most often traded there are dedicated pages for EUR/USD and GBP/USD, and for gold โ where a "tight" stop is still several dollars โ the XAU/USD calculator.
Add spread and likely slippage to the stop distance before you size. On a 12-pip stop, a 1.5-pip spread is more than a tenth of the trade's risk.
How flags fail
The breakout reverses immediately. Price closes beyond the boundary, triggers entries, and comes straight back through the flag. This is the most common failure and the reason the stop sits on the far side of the consolidation rather than just beyond the boundary.
The flag deepens into a reversal. What looked like a shallow drift keeps going, retraces most of the pole, and the trend is over. A retracement past roughly half the pole is where most traders stop treating the structure as a flag at all.
The consolidation runs too long. A flag that persists for many candles stops being a pause and becomes a range. The longer it lasts, the less the pole tells you about what happens next.
There was no pole. The move labelled as a flagpole was ordinary trend movement rather than a sharp impulsive push. Without the contrast between violent and gentle, the pattern has no basis.
Reliability figures for flag patterns circulate widely and should be treated with suspicion โ they depend entirely on how the pattern is defined, which market, which timeframe and which entry rule. The only number that means anything for your trading is the one from your own trades.
Measuring it rather than trusting it
Record each flag trade with the pole height, the flag depth as a fraction of the pole, the stop distance, whether you entered on the break or the retest, and the result in R. That gives you the comparisons that matter: whether your retest entries beat your breakout entries, whether shallow flags outperform deep ones, whether the pattern works at all on your instruments and timeframe.
Fips's trading journal records trades in R and groups them by setup so those comparisons are a filter rather than a spreadsheet exercise, and account analysis shows the spread of outcomes rather than the average. Before committing money to a rule, backtesting it on historical data gives you the worst losing streak it has produced โ the number that determines whether your risk per trade survives a bad run.
Frequently asked questions
What is the difference between a flag and a pennant?
The shape of the consolidation. A flag is a channel with roughly parallel boundaries sloping against the trend; a pennant is a small triangle with converging boundaries. Both follow a sharp move and both are read as continuation patterns, so the trading approach is largely the same.
How far should a flag retrace?
Shallow is the point โ commonly under a third to a half of the flagpole. A retracement deeper than about half undermines the idea that the trend is merely pausing, and most traders stop treating the structure as a flag at that point.
Where exactly does the stop go?
On the far side of the flag, plus a buffer for spread and noise. Placing it just beyond the breakout boundary looks efficient but sits inside the consolidation, where price routinely trades before the move develops.
Is the measured move a reliable target?
It is a convention for setting an initial objective, not a prediction. Some trades exceed it, many fall short. Use it to judge whether the reward justifies the risk before entering, then manage the position on price rather than on the projection.
Do flags work on every timeframe?
The shape appears everywhere, but lower timeframes produce far more of them and a larger share are noise. A flag on a five-minute chart with a six-pip stop is dominated by spread and normal fluctuation. Anchoring intraday flags to a higher-timeframe trend filters much of that out.