Candlestick Patterns on Gold (XAUUSD): Anatomy, Key Patterns, Context and Setups

What candlestick patterns are

A candlestick shows four prices for one period: the open, high, low and close (often shortened to OHLC). The thick part is the body, which spans the open and close. The thin lines above and below are the wicks (also called shadows), which show how far price traveled before being pushed back.

A candlestick pattern is a recognizable shape made by one, two or three candles. Traders use them because the shape tells a short story about who won the period:

• A long body means one side controlled the period from open to close.
• A long wick means price went somewhere and was rejected.
• A small body means neither side won. That is indecision.

Candles do not predict anything on their own. They describe what just happened. The skill is reading that description at the right place, in the right context.

How candlesticks behave on gold (XAUUSD)

Gold has habits that change how you should read its candles.

• Volatility is large and uneven. An H1 gold candle can range a few dollars in the Asian session and more than ten dollars during the New York morning. Always judge a candle relative to the recent ATR 14, not by its absolute size.
• News creates fake shapes. US data usually lands at 8:30 New York time (about 12:30 or 13:30 GMT depending on daylight saving). A spike at that moment can print a huge wick that looks like a perfect pin bar but is just a fast two-way reaction to the release.
• Stop runs make long wicks common. Gold often trades a few dollars beyond an obvious high or low, triggers stops, then reverses. That leaves long wicks at key levels, which is why hammers and shooting stars at liquidity pools are worth noticing.
• Session opens distort candles. The London open (about 03:00 New York time) and the New York open often produce sharp first candles. Sunday’s open at 18:00 New York time can gap, which creates odd-looking first candles of the week.
• Round numbers attract reactions. Patterns that form at whole numbers like 4,100 or 4,150 tend to get more attention than patterns at random prices.
• Candle close times vary by broker. Daily and H4 candles depend on your broker’s server time. Many traders use charts aligned to the 17:00 New York daily close so the daily candle matches gold’s trading day.

How to identify the key patterns

Use these definitions consistently. Write them in your trading plan so you don’t stretch them to fit what you want to see.

Single-candle patterns

Variations of the doji include the dragonfly doji (long lower wick, a hammer-like rejection) and the gravestone doji (long upper wick, a shooting-star-like rejection).

Two-candle patterns

On gold, because the market trades almost 24 hours, the next candle usually opens at or very near the prior close. Strict textbook rules that require gaps (written for stock markets) rarely apply. Most traders judge engulfing by bodies, not by gaps.

Three-candle patterns

Inside and outside bars

• An inside bar has its high and low within the prior candle’s range (the prior candle is called the mother bar). It shows contraction. Traders often trade a break of the mother bar’s range.
• An outside bar has a range that covers the prior candle’s high and low. When it closes strongly in one direction, it behaves like an engulfing pattern measured with wicks.

Steps to read any candle

1. Mark the context: higher-timeframe trend, range or transition. See market structure.
2. Mark the location: is price at a support or resistance zone, a prior high or low, or a round number?
3. Read the candle: body size versus recent candles, wick direction, close location.
4. Check the timing: did it form at a news release or a session open?
5. Wait for confirmation if your plan requires it: usually a break of the pattern’s high (for bullish) or low (for bearish).

Variations and settings: from beginner to pro

Beginner approach. Learn five patterns well: hammer, shooting star, bullish and bearish engulfing, and the inside bar. Trade them only at clearly marked support and resistance zones on H1 or H4, and only with confirmation.

Intermediate refinements.

• Use relative size filters. For example, an engulfing candle counts only if its body is larger than the average body of the last 10 candles, or its range exceeds 1x ATR 14.
• Require the wick of a hammer or shooting star to pierce a level and close back inside. That turns a generic shape into a failed breakout signal.
• Read the same pattern on two timeframes. An H4 hammer is often an M15 sequence of a sweep, a change of character and a pullback. The lower timeframe gives a tighter stop.

Pro refinements.

• Stop thinking in names. Pros read pressure: where price went, how fast it came back, and where it closed. A “pattern” is just a shortcut for that story.
• Watch the next candle more than the pattern. Strong follow-through within one or two candles is the market agreeing. A pattern that gets no follow-through is quietly failing.
• Fade failed patterns. If a bullish engulfing at support is immediately broken to the downside, the trapped buyers become fuel. A failed signal can be more useful than a successful one.
• Combine candles with session logic. A shooting star that forms when London sweeps the Asian high carries a clearer story than the same shape at a random hour. See London session gold trading.

3 setups traders use

These are educational templates, not signals. Test them before risking money.

Setup 1: Rejection candle at a higher-timeframe level

Setup 2: Engulfing reversal after a liquidity sweep

This setup works on the same logic as a liquidity sweep. The candle is just the visible footprint of the sweep and reclaim.

Setup 3: Inside bar breakout in trend

Worked example (hypothetical)

Notice that the confirmation entry cost 3.40 dollars of extra stop distance versus entering on the hammer’s close. That is the trade-off of waiting. Use the position size calculator to run your own numbers.

Combining candlesticks with other concepts

Candles are a trigger, not a strategy. They work best inside a framework:

• Levels: support and resistance and supply and demand zones tell you where a candle matters.
• Structure: market structure tells you which direction a candle should agree with.
• Larger patterns: a shooting star at the right shoulder of a head and shoulders carries more weight than one in isolation.
• Trendlines: a hammer on the third touch of a trendline combines two pieces of evidence.
• Timeframes: use multi-timeframe analysis to read the same candle as a sequence on a lower chart.
• Momentum: a bearish engulfing at a new high with RSI divergence gives two independent hints of exhaustion.
• Fibonacci: a pin bar at the 61.8% retracement is a common confluence; see Fibonacci on XAUUSD.

Best timeframes and sessions

Sessions (New York time):

• Asian session (about 20:00–00:00): smaller candles. Inside bars and dojis are common and often mean nothing more than a quiet market.
• London (about 02:00–05:00 for the most active window): frequent sweeps of the Asian range, so rejection candles at those highs and lows are worth watching.
• New York AM (about 07:00–10:00): the largest candles, especially around 8:30 data. Patterns that form right at a release need extra caution.

See gold trading hours worldwide for conversions to your local time.

Common mistakes

• Trading shapes in the middle of nowhere. Fix: only trade patterns at levels you marked before the pattern formed.
• Ignoring the trend. Fix: take reversal patterns against the trend only at major higher-timeframe levels, and with smaller size.
• Judging size in dollars. Fix: compare body and wick size with ATR 14 and recent candles.
• Calling news spikes “pin bars”. Fix: wait at least one or two candles after a high-impact release before reading the shape.
• Using the pattern low as a tight stop. Fix: add a volatility buffer. Gold routinely retests wick extremes.
• Learning 40 patterns. Fix: master five, define them precisely, and test them.
• Mixing broker close times. Fix: use one chart feed, ideally aligned to the 17:00 New York close, for all pattern analysis.

Limitations and honest caveats

• No verified universal win rate exists for any candlestick pattern on gold. Figures you see quoted online are rarely reproducible. Your own tested sample is the only number that means anything for your rules.
• Patterns are timeframe-dependent. An H1 hammer can be an M15 downtrend with one sharp bounce. The same price action can look bullish on one chart and bearish on another.
• Candles ignore volume and context by design. Two identical hammers can come from very different order flow.
• Hindsight bias is strong. Patterns look obvious on a finished chart. Live, half-formed candles change shape until they close.
• Candle shapes depend on data. Different brokers’ feeds and server times produce slightly different OHLC values.

Candlesticks are a language for describing price, not an edge by themselves. Any edge comes from combining them with location, context and disciplined risk management.

Pro tips

1. Wait for the close. A candle that looks like a hammer at minute 50 can close as a marubozu down at minute 60.
2. Measure the wick against the level. The best rejection candles pierce a known level and close back inside. The pierce is the trap; the close is the signal.
3. Track the “second candle” rule. Log how often the candle after your pattern follows through. That single statistic tells you quickly whether a pattern works in your setup.
4. Respect the body of big candles. After a large marubozu, the 50% level of its body often acts as a reference for pullbacks.
5. Note failures. A bullish pattern that is fully broken within one or two candles is information. Consider it as a potential continuation setup in the other direction.
6. Avoid patterns into the 17:00 New York break. Liquidity thins before the daily close, and the last candles of the day can be unrepresentative.

Quick checklist

1. Is the higher-timeframe context clear (trend, range or transition)?
2. Did the pattern form at a level I marked before it appeared?
3. Does the pattern meet my written definition, with size judged against ATR?
4. Did it form away from a high-impact release or session-open spike?
5. Does the close location support the idea?
6. Has the confirmation rule in my plan been met?
7. Is my stop beyond the pattern extreme plus a buffer?
8. Is there enough room to the next opposing level for the trade to make sense?
9. Is my position size calculated from the stop distance and my risk %?
10. Will I log the result, including whether the next candle followed through?

Frequently asked questions