What chart patterns are
A chart pattern is a shape that price forms over many candles as buyers and sellers push against each other. Where a candlestick pattern describes one to three candles, a chart pattern describes a whole battle over dozens of candles.
Patterns fall into two groups:
- Reversal patterns suggest a trend may be ending: head and shoulders, double and triple tops or bottoms, and wedges against the trend.
- Continuation patterns suggest a trend is pausing before it resumes: flags, pennants, rectangles and many triangles.
Traders use them for three practical reasons:
- They give a clear trigger (the breakout).
- They give a logical invalidation (the other side of the pattern).
- They give a reference target (the measured move).
None of that makes them predictive. A pattern is a structured way to plan a trade, not a forecast.
How chart patterns behave on gold (XAUUSD)
Gold has a few habits that every pattern trader should know.
- False breaks are common. Gold often pushes a few dollars beyond a neckline or triangle edge, triggers breakout orders and stops, then snaps back. This is the same behavior described in liquidity sweeps.
- News breaks patterns abruptly. A US CPI or jobs release at 8:30 New York time can complete, fail or invalidate a pattern in one candle. Patterns that are about to break into a major release are a coin toss you don’t need to take.
- Session rhythm shapes patterns. Asian-session ranges often form small rectangles or triangles. London frequently breaks them, and New York may extend or reverse the move. See the Asian range breakout guide.
- Round numbers anchor necklines. Double tops at 4,200 or a neckline near 4,100 are common, because many orders cluster around round prices.
- Dollar and yield sensitivity. A pattern can fail simply because the US dollar or real yields move hard. Know what is on the calendar and check the intermarket picture.
- Strong trends produce many continuation patterns. In persistent gold trends, flags and pennants appear repeatedly, while reversal patterns often fail until the trend genuinely tires.
How to identify the main patterns
Reversal patterns
| Pattern | How to identify it | Measured move |
|---|---|---|
| Head and shoulders (top) | Three peaks, middle highest. A neckline joins the two troughs between them | Height from head to neckline, projected down from the break |
| Inverse head and shoulders | Three troughs, middle lowest, neckline through the two peaks | Height projected up from the break |
| Double top | Two peaks at a similar price, separated by a clear trough. Confirmed on a close below the trough | Height from peaks to trough, projected down |
| Double bottom | Two troughs at a similar price, confirmed on a close above the middle peak | Height projected up |
| Triple top / bottom | Three tests of the same level, then a break of the intervening swing | Height projected from the break |
| Rising wedge | Price makes higher highs and higher lows inside two converging, upward-sloping lines. Often bearish | Back toward the wedge’s start |
| Falling wedge | Converging, downward-sloping lines. Often bullish | Back toward the wedge’s start |
Continuation patterns
| Pattern | How to identify it | Measured move |
|---|---|---|
| Bull / bear flag | A sharp move (the pole) followed by a small, orderly channel against it | Pole length, projected from the breakout |
| Pennant | A pole followed by a small symmetrical triangle | Pole length, projected from the breakout |
| Ascending triangle | Flat top, rising lows. Usually bullish | Triangle height at its widest, projected from the break |
| Descending triangle | Flat bottom, falling highs. Usually bearish | Height projected down |
| Symmetrical triangle | Falling highs and rising lows converging. Direction neutral | Height projected in the break direction |
| Rectangle | A horizontal range with at least two touches on each side | Range height projected from the break |
| Cup and handle | A rounded base (the cup), then a small pullback (the handle), then a break of the rim | Cup depth projected up from the rim |
Step-by-step identification
- Identify the prior trend. Reversal patterns need something to reverse. Continuation patterns need a trend to continue. Use market structure.
- Find the swing points. Mark the obvious highs and lows only. If you need to squint, it isn’t a pattern.
- Draw the boundaries. Necklines, triangle edges and flag channels. Use bodies or wicks consistently. See trendlines and channels.
- Measure the height. Record the pattern’s height before the breakout. It sets your target reference.
- Define the breakout rule. For example, a candle body close beyond the boundary by at least 0.25x ATR 14.
- Define invalidation. Usually beyond the last swing inside the pattern (the right shoulder, the last flag swing, the last triangle touch).
- Check the location and calendar. Is there a major level just beyond the breakout? Is high-impact data due?
A pattern isn’t complete until it breaks. Until then, a “head and shoulders” is just three highs. Many traders lose money pre-empting patterns that never confirm.
Variations and settings: from beginner to pro
Beginner approach. Focus on four patterns: double top/bottom, head and shoulders, bull/bear flag and rectangle. Trade them on H1 or H4, only after a candle close beyond the boundary, with a stop beyond the last internal swing.
Intermediate refinements.
- Symmetry filter. Shoulders that are roughly equal in time and height make cleaner patterns. Very lopsided patterns are open to interpretation.
- Flag depth filter. Many traders want a flag to retrace less than about 50% of the pole. Deeper pullbacks look more like reversals than pauses.
- Volume or volatility contraction. Gold spot has no central volume, but futures volume or a falling ATR 14 inside a triangle can show contraction before expansion. See ATR and Bollinger Bands.
- Neckline slope. A neckline that slopes against the pattern (downward in a top) tends to show more weakness than a flat one.
Pro refinements.
- Trade the retest of the boundary, not the break. The breakout and retest approach on a neckline usually gives a defined stop and better reward-to-risk.
- Read the internal structure. On the lower timeframe, the right shoulder of a head and shoulders is often a lower high and a change of character. That is the same idea in Smart Money Concepts language.
- Expect the sweep. On gold, the second top of a double top often runs slightly above the first. Pros wait for the reclaim below the first top rather than selling the exact equal high.
- Trade failures. A failed pattern is one of the more useful signals. If price breaks below a neckline and then closes back above the right shoulder, the shorts are trapped.
3 setups traders use
These are educational templates. Test them on your own data before using them.
Setup 1: Neckline break and retest
| Element | Rule |
|---|---|
| Context | A head and shoulders (or inverse) after an extended trend, at a higher-timeframe level |
| Trigger | A candle body closes beyond the neckline |
| Entry | Limit order on a retest of the neckline from the other side, or a lower-timeframe rejection at it |
| Stop (invalidation) | Beyond the right shoulder, or for a tighter version beyond the retest swing plus a buffer |
| Targets | First: the nearest support or resistance. Second: the measured move |
| Management | Partial profit at the first target. If price closes back through the neckline and holds, exit |
Setup 2: Flag continuation
| Element | Rule |
|---|---|
| Context | A clear trend with an impulsive pole (large bodies, ideally leaving a fair value gap) |
| Trigger | Price closes beyond the flag’s upper (bull) or lower (bear) boundary |
| Entry | On the breakout close, or a pullback to the flag boundary |
| Stop (invalidation) | Beyond the flag’s last swing low (bull) or high (bear) |
| Targets | The pole length projected from the breakout, checked against the next level |
| Management | If the breakout candle is followed by a close back inside the flag, reduce or exit |
Setup 3: Failed pattern reversal
| Element | Rule |
|---|---|
| Context | A widely visible pattern breaks out near a major level or right after a news spike |
| Trigger | Price closes back inside the pattern and beyond the opposite internal swing (for example, back above the right shoulder) |
| Entry | On that close, or on a pullback into the pattern |
| Stop (invalidation) | Beyond the false-breakout extreme |
| Targets | The opposite boundary of the pattern, then the measured move in the new direction |
| Management | Move quickly. Failed-pattern moves can be fast, and so can failures of the failure |
Breakout orders are especially vulnerable around 8:30 New York data, FOMC decisions and the Sunday open. Spreads widen and slippage can turn a planned 10-dollar risk into a much larger loss. Size for the worst realistic fill, not the ideal one.
Worked example (hypothetical)
Account: $25,000. Risk per trade: 0.5% ($125).
On a hypothetical H4 chart, gold has rallied from 4,000 to 4,140. It forms a left shoulder at 4,122, a head at 4,140 and a right shoulder at 4,121. The two troughs sit at about 4,101 and 4,100, so the neckline is flat at 4,100.
- Pattern height: 4,140 − 4,100 = 40 dollars
- Measured target: 4,100 − 40 = 4,060
An H4 candle closes at 4,092, below the neckline. The trader waits for a retest. Two candles later, price rallies to 4,099, fails to close above the neckline and prints a bearish engulfing on H1.
- Entry: 4,097 (sell on the H1 engulfing close)
- Stop: 4,108 (above the retest swing high of 4,101 plus a buffer, and well below the right shoulder)
- Stop distance: 11 dollars
- Position size: $125 ÷ $11 = about $11.36 per $1 move, so 0.11 lots
- Actual risk: 0.11 x 11 = $121
- First target: H4 support at 4,075 (22 dollars, 2R). Close half
- Second target: measured move at 4,060 (37 dollars, about 3.4R)
If an H4 candle closes back above 4,100 and holds, the breakdown has failed and the trade is exited. A stop at the right shoulder (4,121 plus a buffer) would have needed a much smaller position. This is an illustration of the reasoning, not a trade idea.
Combining chart patterns with other concepts
- Support and resistance: a double top at a weekly resistance zone is more meaningful than one in mid-range.
- Candlesticks: a shooting star or engulfing at the right shoulder or second top gives a precise trigger.
- Trendlines: triangles, wedges and flags are built from trendlines and channels, so the same drawing rules apply.
- Fibonacci: flags that hold the 38.2% to 50% retracement of the pole are generally viewed as healthier than deeper pullbacks. See Fibonacci on XAUUSD.
- Momentum divergence: a second top with lower RSI 14 is a common supporting clue. See RSI divergence.
- Wyckoff: many double bottoms and inverse head and shoulders are the same events Wyckoff describes as a selling climax, a spring and a test. See Wyckoff on gold.
- Liquidity: the false breakout of a pattern is often a liquidity sweep in disguise.
Best timeframes and sessions
| Timeframe | Notes |
|---|---|
| Daily / weekly | Big, widely watched patterns. Measured moves can take weeks. Wide stops, small size |
| H4 | A good balance for swing traders. Most textbook patterns appear clearly |
| H1 | Good for flags, triangles and rectangles in the London and New York sessions |
| M15 and below | Mostly for entry timing within a higher-timeframe pattern |
Sessions (New York time, with GMT varying by daylight saving):
- Asian session (about 20:00–00:00): small rectangles and triangles often form.
- London (about 02:00–05:00): breakouts of overnight patterns, including many false breaks.
- New York AM (about 07:00–10:00): the strongest follow-through or the sharpest failures, particularly around 8:30 data.
Common mistakes
- Seeing patterns everywhere. Fix: only trade patterns that are obvious on the timeframe you’re using.
- Entering before the break. Fix: require a candle body close beyond the boundary.
- Ignoring the prior trend. Fix: no trend, no reversal pattern.
- Treating the measured move as guaranteed. Fix: take partial profits at the nearest opposing level.
- Stops inside the pattern. Fix: place invalidation beyond the last internal swing plus a buffer.
- Breakout orders into news. Fix: check the calendar and avoid pending orders into high-impact releases unless tested.
- Redrawing the pattern after it fails. Fix: once invalidated, it’s gone. Look at the failure as new information.
Limitations and honest caveats
- Subjectivity. Two traders can draw different necklines and triangle edges on the same chart.
- No verified universal success rates. Numbers commonly quoted for chart patterns come from varied markets, periods and definitions, and many are unverifiable. They don’t transfer reliably to gold. Test your own rules.
- Hindsight. Completed patterns are easy to spot. Live, they are ambiguous until the breakout.
- Measured moves are rough. They are a geometric habit, not a law. Price often stops short or overshoots.
- Macro overrides geometry. A surprise Fed decision or a sharp move in the dollar can ignore any pattern.
Pro tips
- Measure before the break. Write down the height and target in advance so you don’t adjust it emotionally later.
- Prefer patterns at the edges. Reversal patterns at higher-timeframe levels, and continuation patterns after a fresh break of structure, have a clearer story.
- Time matters. A flag that drags on for too long often turns into a range or a reversal. Some traders exit if the breakout doesn’t come within a set number of candles.
- Use the retest as a free test. If the retest of a broken neckline closes back inside, you learn the break failed with a small loss.
- Log failures separately. Track how often your patterns fail and what happened next. Failure follow-through is often the more useful data.
- Scale the stop to the pattern, then the size to the stop. A big pattern needs a wide stop and a small position. Never shrink the stop to keep the size.
Quick checklist
- Is there a clear prior trend (for reversals) or a clear trend to continue?
- Is the pattern obvious on this timeframe without squinting?
- Are the boundaries drawn with a consistent rule (bodies or wicks)?
- Have I measured the pattern height and noted the target in advance?
- Has a candle body closed beyond the boundary?
- Is my invalidation beyond the last internal swing plus a buffer?
- Is the nearest opposing level far enough away to make the trade worth it?
- Is there high-impact news within the next hour?
- Is my position size calculated from the stop distance?
- Do I know exactly what a failure looks like, and what I will do if it happens?
Frequently asked questions
What is the most reliable chart pattern for gold?
There’s no verified universal ranking of chart patterns for XAUUSD, and published success rates are rarely reproducible. Continuation patterns in a clear trend, such as flags, are often easier to define and manage. Test any pattern with written rules on your own data.
How do you calculate a chart pattern target?
The common method is the measured move: take the pattern’s height (for example, from the head to the neckline) and project it from the breakout point. Treat it as a reference area, not a promise, and check it against nearby support and resistance.
Should I trade the breakout or wait for a retest?
Breakout entries catch every move but suffer more false breaks. Retest entries usually give a tighter stop but miss strong moves that never return. On fast markets like gold, many traders split the difference with partial entries or require a candle close beyond the level.
What happens when a chart pattern fails?
A failed pattern traps traders who entered on the breakout, and their exits can push price sharply the other way. Many experienced traders watch for failures as setups in their own right, for example a head and shoulders that breaks back above the right shoulder.
Do chart patterns work on lower timeframes like M5?
They appear on every timeframe, but on M1 to M5 gold charts they’re often just noise and are easily distorted by news spikes. Patterns on H1, H4 and daily charts tend to be cleaner and more widely watched.
How long should a chart pattern take to form?
There’s no fixed rule. A useful guide is that a pattern should have enough swings to be obvious to other traders on that timeframe, typically at least 15 to 20 candles for reversal patterns. Very small patterns are less likely to be watched.
Sources and further reading
- CME Group — Gold futures overview — Contract specifications and trading hours for COMEX gold
- CFTC — Commitments of Traders — Weekly futures positioning, including gold
- Investor.gov — Stop order — How stop orders work, and their limits in fast markets
External sites are independent of XAUUSD Global. Check dates and data yourself.