Double Tops and Double Bottoms on the Gold Chart

What has to be true for the pattern to exist

Three components are required and the third is the one traders skip. First, a high that produced a visible reaction, meaning price turned away from it with intent rather than drifting. Second, a pullback deep enough to create a clear swing low, which becomes the neckline. Third, a second approach to that high which fails.

None of this is a pattern until the neckline breaks. A chart showing two similar highs with price sitting between them is a range, and ranges resolve upward about as often as downward. Shorting the second high before the neckline gives way is a bet on the range boundary holding, which is a different trade with a different stop and a far worse record.

On the timeframe question, the pattern needs enough candles to be meaningful. Two highs three candles apart on the five minute chart are noise. Two highs separated by a day or more on the hourly chart describe genuine repeated rejection.

The matched version and the sweep version

The textbook shape has both highs at almost the same price, within a few tens of cents on gold. This version is cleaner to draw and slightly less common, because a perfectly matched high leaves the stops above it untouched, and untouched stops tend to attract one more attempt.

The more frequent shape on gold has a second high slightly above the first. Price trades a dollar or two beyond the old high, triggers the orders resting there, and then fails to hold, closing back below the original level within the same session. Short term traders who shorted the first high are already out, which removes the supply that would otherwise fuel the decline, and the subsequent move down is often faster for it.

A marginally higher second high strengthens the pattern rather than invalidating it. What invalidates it is a second high that holds above the first on a closing basis for several candles, because that is a breakout, not a failure.

The neckline is the whole pattern

Draw the neckline at the lowest point of the pullback between the two highs, using the wick. That single line does three jobs: it confirms the pattern when it breaks, it provides the measurement for the target, and it becomes the level that should reject price on the retest.

The confirmation standard matters more than the shape. A wick through the neckline is not a break. What you want is a candle closing below it on the timeframe you identified the pattern on, ideally during London or the New York morning rather than in the thin hours, followed by an inability to reclaim it.

A sloping neckline is acceptable and common. If the low between the two highs is higher than the prior swing structure, the line tilts, and you use the actual low rather than a horizontal approximation. What is not acceptable is moving the neckline after the fact to accommodate a break that did not happen where you wanted it to.

Entry, stop and sizing in dollars

Two entries are defensible. The first is the close below the neckline, taken immediately, which captures the fast cases and accepts more false breaks. The second is the retest of the neckline from below, which on gold happens often enough to be worth waiting for and gives a much tighter invalidation.

The stop goes above the higher of the two highs with a buffer for spread, not at the neckline. A stop at the neckline will be taken out by the ordinary retest you were hoping to enter on. If the distance from the retest entry to above the highs is, say, 6 dollars, that is 600 dollars of risk on a full 100 ounce lot and 60 dollars at a tenth of a lot, which is how the sizing decision should be framed before entry rather than after.

If the pattern is large, and on the daily gold chart it often spans tens of dollars, the honest conclusion may be that the structural stop is unaffordable at your account size. Taking the trade with a tighter stop converts a reasonable idea into a coin flip.

Targets that are worth taking

The conventional measurement takes the vertical distance from the highs to the neckline and projects it downward from the break point. For a pattern 12 dollars tall, that implies roughly 12 dollars of decline below the neckline.

Treat it as a ceiling on ambition and check the path. If a strong daily support zone, the previous week low or a prominent round dollar level sits between the break and the measured objective, that obstacle is where the first exit belongs. Gold reacts at those levels frequently enough that running a target through them without taking something off is poor practice.

Partial exits also resolve the main tension in this pattern. Many double tops produce a quick move to the first obstacle and then a grinding recovery. Taking a portion early and leaving the rest with the stop moved to the break point lets both outcomes be acceptable, rather than needing the full measured move to justify the trade.

The mirror image at a low

Double bottoms work the same way with the roles reversed, and on gold they have a few distinct habits worth knowing. The sweep version is even more common at lows, because stops below an obvious support cluster are a convenient liquidity pool and sell side sweeps tend to be sharp.

The second difference is tempo. Rallies out of double bottoms on gold often start slowly, with price grinding above the neckline for several hours before the move develops, whereas declines from double tops are usually faster. A target that assumes symmetry in timing will have you exiting out of impatience.

The third is session dependence. Double bottoms that complete during the London open or on a New York data print tend to follow through; the same shape completing in the Asian session frequently stalls until London arrives and then gets retested.

Why many of these patterns fail

The honest position is that the pattern alone is not an edge. We have not measured one for double tops on gold and we are not going to present a figure we did not produce. The recurring failure modes are specific.

• Range misreading. Two highs inside a wider range are just the top of the range, and the neckline break is a move toward the range low, which holds.
• Trend context ignored. A double top in a strong uptrend is usually a pause, and the third attempt succeeds.
• Thin session confirmation. A neckline break in the hour after the daily settlement window is tested by almost nobody.
• News override. A scheduled release can invalidate the structure instantly, in either direction.

Our related work on the head and shoulders pattern on gold covers the same family of reversal structures and the same requirement for a confirmed neckline break.

How to use this on the chart

Build the pattern into a level based process rather than hunting for it. The patterns worth taking form at a place you already respected: a strong daily resistance zone, the previous week high, or a prior monthly extreme. Two highs in open space mean much less.

A workable sequence is to mark the zones before the session, note when price makes a second failed attempt at one of them, draw the neckline, and set an alert there instead of watching. When the alert fires, require a close beyond the line in an active session, then wait for the retest. Check the path to the measured objective for obstacles and size the position from the distance to above the highs.

Doing the marking in advance on the live XAUUSD chart is what keeps this disciplined, because a neckline drawn while price is already breaking it tends to end up wherever your position needs it to be.