Gold Chart Patterns Cheat Sheet: 12 Patterns on XAUUSD With Stats Caveats

How to use this cheat sheet

This page is a reference, not a course. Each of the 12 patterns gets the same five fields: shape, trigger, target, invalidation and gold notes. If you want the fuller treatment of how patterns form and the psychology behind them, read chart patterns on XAUUSD first, then come back here when you are at the screen and need a quick check.

Two definitions used throughout:

• Trigger means the candle close that confirms the pattern, usually on the timeframe the pattern formed on. A wick through a level is not a trigger.
• Measured target means the pattern’s height projected from the breakout point. It is a planning aid. Always compare it with nearby support and resistance and round numbers, and take the nearer one seriously.

The 12 patterns at a glance

Gold adds its own flavour to each. The sections below pick out what changes on XAUUSD and show three of the patterns in detail.

Reversal patterns on gold

1 and 2: head and shoulders, regular and inverse

The head and shoulders is the pattern beginners find first and over-trade most. On gold, three points matter:

• The neckline is rarely flat. Draw it through the two troughs even if it slopes. A rising neckline on a top makes the pattern weaker; a falling one makes the break come sooner.
• The right shoulder is where gold likes to fake. A push above the right shoulder high that fails inside an hour is common and is often a liquidity sweep rather than an invalidation. Use a close, not a wick, as your invalidation.
• Volume is not available on spot gold in a reliable way, so the textbook “volume should fall on the right shoulder” rule cannot be checked. Use the futures volume on a COMEX chart if you want a proxy, or skip the volume rule entirely.

The inverse version mirrors everything. On gold, inverse head and shoulders patterns after a sharp sell-off into a round number (for example a drop that bottoms around 3,900) tend to be the ones traders remember, because the subsequent rally is fast.

3, 4 and 5: double and triple tops and bottoms

Gold’s second test of a level very often overshoots the first by a few dollars before reversing. A double bottom where the second low is $3 to $6 below the first is not a failed pattern; it is the normal gold version. The trigger is still the close above the neckline.

Triple tops and bottoms are rare and are frequently just a range in disguise. If the third test holds, treat the level as a rectangle edge and trade the range rules instead.

Continuation patterns on gold

6, 7 and 8: flags and pennants

Flags are gold’s most frequent intraday pattern, especially after the London open or an 08:30 New York data release. Points that matter:

• The pole should be impulsive. Three to eight strong candles with small wicks. A slow grind is not a pole.
• The flag should be shallow. A retracement of more than about half the pole is better treated as a pullback to a supply or demand zone, not a flag.
• Timing matters. A flag that forms into the 10:00 New York hour or into the 17:00 daily break has less fuel than one that forms right after the London open.

Pennants are flags with converging edges. Trade them the same way, but notice that pennant breaks on gold are more prone to a first fake in the wrong direction.

9, 10 and 11: triangles

Ascending triangles on gold often resolve upward, but the flat top is a stop magnet. Expect a wick above it and a close back inside at least once before the real break. The cleanest triggers come when the break happens in the first two-thirds of the triangle, before the apex, with a retest holding the flat edge.

Symmetrical triangles give no directional bias on their own. Use the higher-timeframe market structure to decide which side you are willing to trade and ignore the other.

12: rectangles

A rectangle is a range, and gold ranges a lot, particularly in Asia. The Asian range breakout lesson is really a rectangle lesson with a time element. The measured target (range height from the broken edge) is one of the few pattern targets that gold hits with some regularity, because ranges are usually sized by the session’s volatility.

About the statistics

You will see claims like “head and shoulders works 83% of the time” or “bull flags hit target 67% of the time.” Treat them with suspicion.

• No verified statistic exists for pattern success on XAUUSD. The widely quoted figures come from studies of US stocks over periods ending many years ago, with hand-picked patterns identified after the fact.
• Pattern definitions are subjective. Two traders drawing necklines on the same chart will mark different triggers and get different results.
• “Success” is defined generously. Many studies count a pattern as successful if price moves 5% in the expected direction at any point, which says nothing about whether a real trade with a real stop survived.
• Costs are usually excluded. On a leveraged gold account, spread and slippage around a breakout are material.

If you want a number you can trust, produce it yourself. Backtesting a gold strategy with free data walks through the method, and the trading math and expectancy lesson shows why a win rate on its own tells you nothing without the average win and loss.

A simple way to trade any pattern

Whatever the pattern, the same process applies:

1. Identify the pattern on H1 or H4 and draw its edges with closes, not wicks, where possible.
2. Check the higher timeframe. A bearish pattern inside a strong D1 uptrend needs a stronger trigger than one at the end of a weak rally.
3. Wait for the trigger close. Then decide: enter on the close, or wait for the retest of the broken edge. The breakout and retest lesson covers both.
4. Place the stop at the invalidation level plus a buffer for spread, and size the position from that distance with the position size calculator.
5. Set the measured target and compare it with the nearest significant level. If a round number or the previous day’s high sits well short of the target, plan a partial there.
6. Check the clock. If tier-1 US data lands before the target is realistic, decide in advance whether you hold, reduce or close. The gold market events page lists the dates.

Worked example (hypothetical)

Common mistakes

• Seeing patterns everywhere. Fix: require a clear pole for flags, three obvious peaks for a head and shoulders, and skip anything you have to squint at.
• Entering before the trigger close. Fix: a wick through the neckline is not a break. Wait for the close on the pattern’s timeframe.
• Ignoring where the target lands. Fix: compare the measured target with the nearest major level; take the nearer one seriously.
• Treating published win rates as facts. Fix: assume no verified statistic exists until you have produced one on your own data.
• Trading reversal patterns against a strong D1 trend. Fix: demand a retest and a lower-timeframe structure break before fading a trend.
• Using the same stop for every pattern. Fix: the invalidation level sets the stop, and the stop sets the size, never the other way round.

Limitations and honest caveats

• Patterns are visible to everyone, so their edges attract stops and sweeps.
• Hindsight makes patterns look cleaner than they are in real time. Practise marking them live on a demo account before risking money.
• Gold’s volatility regime changes. A pattern height that was a good target in a $40 ADR market may be far too ambitious when ADR is $75.

Pro tips

1. Count the attempts. The third test of a pattern edge is more likely to break than the first; the first is more likely to fake.
2. Mark the pattern’s midpoint. For rectangles and triangles, price holding on the far side of the midpoint after a break is constructive.
3. Pair patterns with candles. A bearish engulfing candle at a right shoulder is a stronger right shoulder.
4. Anchor to the session. A pattern that breaks during the London-New York overlap has more follow-through potential than one that breaks at 15:00 New York time.
5. Keep a pattern log. Screenshot every pattern you trade and every one you skip, then review monthly. Your own statistics will beat any you read online.

Quick checklist

1. Can I name the pattern and draw its edges without forcing it?
2. Does the higher-timeframe trend agree, or am I fading it with extra confirmation?
3. Has the trigger candle closed on the pattern’s own timeframe?
4. Where exactly is the invalidation, and what size does that stop distance give me?
5. What is the measured target, and what level sits short of it?
6. Is there tier-1 news before the target is realistic?
7. Have I logged the pattern, win or lose, for my own statistics?

Frequently asked questions