Double Top and Double Bottom: How to Identify and Trade Them

A double top is the simplest reversal pattern there is: price tries a level twice and fails twice. That simplicity is why it appears everywhere, and why most of the ones traders take are worthless.

Double Top and Double Bottom at a glance

What it is and why it works

A double top is two peaks at approximately the same price with a pullback between them. Price rallies into a level, sells off, rallies back and stalls at or very near the same place, then turns down again. The low of the pullback between the two peaks is the neckline, here it is a single horizontal level rather than a sloping line, and a close below it is the standard confirmation. The double bottom is the mirror image at the end of a decline: two lows at roughly the same price, with the high between them as the neckline.

The two peaks do not need to be exactly equal, and insisting they are is a beginner error. What matters is that the second attempt fails to make meaningful progress beyond the first. A second peak a little higher than the first is still a double top, and is often a better one, because the marginal new high sweeps the stops of everyone who sold the first peak before rolling over. A second peak far above the first is not a double top at all; it is a continuation.

The story is straightforward supply and demand. Sellers defended a price. Buyers regrouped and tried again with whatever demand was left. The level held a second time, which tells you the supply sitting there is larger than the demand available to absorb it. The break of the neckline is the moment the buyers who bought the pullback give up. That is all the pattern is claiming, and it is a reasonable claim, but only where the level had significance in the first place.

Which is the honest caveat. Price touching the same area twice is one of the most common things a chart does. Inside a range it happens continuously and means nothing at all, because the whole definition of a range is that price keeps failing at the same two edges. A double top is only informative when it forms at the end of an extended advance, at a level visible on a higher timeframe, where the second failure genuinely represents demand running out rather than price bouncing around inside a band. As with all chart patterns, the evidence base for the shape as a mechanical predictor is thin; the evidence for level-based supply and demand is what you are really leaning on.

How to trade it, step by step

1. Find the level before you find the pattern. Open the daily chart and mark the prices that have mattered, prior swing highs, old range boundaries, areas price has reacted to more than once. Then drop to your trading timeframe. A double top that forms at one of those prices is worth attention; one that forms at an arbitrary price is not, and you cannot tell the difference if you start from the shape.
2. Confirm there is a trend into the level. The pattern needs an advance to reverse. Check that price has been making higher highs and higher lows on the way in. If the last two weeks look like a horizontal band, you are looking at range behaviour and the correct trade, if any, is a range trade, not a reversal.
3. Check the separation between the two peaks. A genuine double top has a real pullback between the attempts, typically retracing a decent fraction of the prior advance and taking multiple bars to do it. Two highs three bars apart with a shallow dip between them is one consolidation, not two failed attempts. On a 4-hour chart, expect the two peaks to be at least several bars apart and usually more.
4. Accept peaks that are close, not identical. Treat the two highs as equal if they are within a small fraction of the recent average bar range of each other. A second peak that pokes slightly above the first and reverses is still valid and often stronger, because it has cleared the obvious stops before failing. A second peak that closes well above the first has broken the level, and the pattern is void.
5. Draw the neckline horizontally through the low between the peaks. Use the extreme of that pullback, and draw it once. For a double bottom, use the high between the two lows. This is your confirmation line and your invalidation reference, so it should not move because you want a trade.
6. Watch the second peak for evidence of failure, not just for a price. Strong upper wicks, candles closing near their lows, a slowing rally, or a clear rejection bar are what a level being defended actually looks like. Where you have meaningful volume data, lighter volume on the second attempt than the first supports the reading. This step separates traders who wait for the market to show them something from traders who guess at a top.
7. Enter on a close below the neckline, or on the retest of it. The break entry is a sell on the confirming close of the neckline break. The retest entry waits for price to return to the broken neckline from underneath and be rejected. Retests are common on this pattern because it is so widely watched. Choose your approach in advance and stick to it.
8. Put the stop above the higher of the two peaks. The pattern says the level held. If price trades decisively above the highest of the two peaks, it did not hold, and the reason for the trade is gone. A stop inside the pattern (halfway up, or just above the neckline) is not protecting against the pattern being wrong, it is protecting against normal noise, which is a different and worse job.
9. Project the measured move, then look for real levels near it. Measure from the peaks down to the neckline and project that distance below the break. Treat the result as a rough reference for whether the trade is worth taking rather than as a target the market owes you. If there is an obvious support shelf just above the projection, that shelf is the better exit.

Size every one of those entries with the position size calculator and check the trade is worth taking with the risk/reward calculator before you commit.

The conditions it needs

The level is one the market already respects

Everything hinges on this. A double top at a price that has previously acted as resistance, or at the high of a prior range, or at an obvious round number that has been defended before, is describing real supply. The second failure confirms that the orders sitting there have not been consumed.

A double top at a price nothing has ever happened at is describing two random stalls. It has the same picture and none of the content. Mark your levels first and let them tell you where to look.

There is an extended move into the pattern

Reversal patterns need something to reverse, and the more extended and one-directional the move into the level, the more meaningful its failure. A market that has rallied for weeks and then fails twice at the same price has spent its buying; a market that has drifted sideways and touched the top of its band twice has done nothing unusual at all.

The practical test is simple: if you could not describe the preceding move as a trend to somebody else, the pattern is not a reversal pattern.

The second attempt is visibly weaker than the first

The information is in the quality of the second push, not in its price. Look for a slower approach, more overlapping candles, upper wicks that keep appearing, and closes that sit in the lower half of their bars. On instruments with genuine volume, a lighter second push is the classic tell.

When the second attempt arrives with large, decisive candles and simply stops dead, be more cautious, that is often a market gathering itself for a break rather than one running out of buyers.

The timeframe is high enough for the failure to mean something

On a 5-minute chart price fails at the same price several times an hour, mostly for reasons that have nothing to do with supply and demand and everything to do with the spread and short-term order flow. Those double tops carry almost no information.

Use the 1-hour chart as a floor for intraday work and the 4-hour or daily for swing trading. The higher the timeframe, the more each failed attempt represents a real decision by real size.

When it fails

• Trading it inside a range. This is by far the most common misuse. A range is a series of double tops and double bottoms, that is literally what the boundaries look like. Selling every double top at the top of a range is fine if you are deliberately range trading with targets at the other side, but calling it a reversal and expecting a trend is how traders bleed. Establish whether the market is trending or ranging before you interpret the shape.
• Shorting the second peak before confirmation. Traders see the second touch, decide the top is in, and sell immediately at the level with a tight stop above. Sometimes this works beautifully. More often the level breaks, because until the neckline goes there is no evidence the sellers have won; there is only a market at resistance, which is also what every breakout looks like just before it breaks.
• Pattern-hunting on low timeframes. Drop to M1 or M5 and you can find a double top somewhere on your watchlist every few minutes, in whichever direction suits your mood. Humans see patterns in noise as a matter of wiring, and short-timeframe price series are mostly noise. If you find yourself scanning for shapes rather than waiting at pre-marked levels, you have stopped analysing and started confirming what you already wanted to do.
• Requiring the two peaks to be identical. Rejecting valid patterns because the second high is a few ticks above the first, while accepting invalid ones because the two highs happen to line up in the middle of nowhere, is optimising the wrong variable. The tolerance on the peaks matters far less than the location.
• Ignoring the failed retest. If price breaks the neckline, returns to it, and then closes back above and holds, the pattern has failed. A failed double top traps everyone who sold the break and frequently produces a fast move to new highs. Treat a neckline reclaim as an exit, not as a chance to average down.
• Treating the measured move as a forecast. The peak-to-neckline projection is a convention that has been repeated so often it feels like a rule. It is not a prediction, and there is no mechanism that makes price travel that particular distance. Use it to size up the opportunity before entry, then manage against structure.

Markets this pattern shows up on most cleanly

• EUR/USD: Clean, heavily watched levels on the 4-hour and daily where failed retests are unusually well defined.
• Gold: Trends hard into round-number levels, which is exactly where genuine double tops form.
• NAS100: Strong directional moves into prior highs, though it also fakes through them often enough to demand confirmation.
• GBP/USD: Wide daily ranges and well-respected higher-timeframe levels make the two-touch failure readable.

For different levels of experience

If you are brand new

This is probably the first pattern you should learn, because it teaches the thing that matters most: price failing at a level twice is more informative than price failing once.

Start with a routine rather than a scan. Once a week, open the daily chart of two or three instruments and mark the horizontal prices where the market has clearly reacted before. During the week, only look for double tops and double bottoms when price arrives at those marks. Use the 4-hour chart. Wait for a candle to close beyond the neckline before you do anything at all. Put the stop above the higher peak, or below the lower low for a double bottom, and use the position size calculator so that the wider stop reduces your size rather than your account.

Two things will make this feel harder than it looks. You will see double tops everywhere once you know the shape, ignore the ones that are not at your marked levels. And you will be tempted to sell the second peak rather than wait for the neckline. Waiting costs you some entries; not waiting costs you money.

If your results are inconsistent

If your results with this pattern are inconsistent, the most likely reason is that you are treating range boundaries as reversal signals. A double top at the top of a two-week range is not a reversal, and a target based on the measured move will usually sit somewhere the range simply will not go. Sort your charts into trending and ranging before you interpret any pattern on them, and trade the shape differently in each.

The second common leak is entering at the level instead of after the break. Selling the second touch feels intelligent because the risk is small, and occasionally it is the trade of the month. But a market sitting at resistance is indistinguishable from a market about to break resistance until it moves. If you want that better entry, earn it with a rejection you can point to, a strong bar closing back below the level, rather than an assumption.

Finally, be strict about the second peak that pokes above the first. This is not an invalidation. A marginal new high that sweeps stops and immediately reverses is the strongest version of this pattern, because it has removed the sellers who were positioned early and trapped the buyers who chased the break. If your rules reject it, your rules are rejecting the best examples.

If you are experienced

The tradeable content here is not the M shape, it is the second test of a level and what happens to the liquidity around it. The most reliable variants are the ones where the second peak takes out the first by a small margin, sweeps the resting stops above it, and is rejected on the same bar or the next one: a stop run into supply. That gives you a defined invalidation at the sweep extreme and an immediate read on whether the level is being defended.

Watch what the market does with the neckline afterwards rather than committing to the pattern in advance. A break on expanding participation that then rejects the retest is a healthy sequence. A break that stalls immediately and grinds sideways under the neckline is usually a market that has run out of sellers as well as buyers, and it resolves back into the range more often than it continues. In liquidity terms, the question is whether the break found new sellers or merely triggered old stops.

Two practical points. First, on instruments where you have real volume, futures and index products, compare the effort behind each of the two peaks; a materially lighter second attempt into the same price is the substantive version of what the pattern claims. Second, be aware how crowded this shape is. Its very obviousness makes the neckline a stop cluster, which is both why the retest happens so often and why the initial break so often slips. Plan for the fill, not just the level.

Risk management for this strategy

The honest stop for a double top sits above the higher of the two peaks, and for a double bottom below the lower of the two lows. That distance is frequently larger than beginners expect, particularly if the pullback between the attempts was deep. Take the distance as given and shrink the position, never the reverse. Tightening the stop into the middle of the pattern is the single most common way traders convert a valid idea into a loss.

The trade-off between the break entry and the retest entry is really a risk decision. The retest gives you a smaller stop distance from a similar invalidation point, which is a better reward-to-risk on paper, but only on the subset of patterns that retest at all. Whichever you choose, run the numbers before entry with the risk-reward calculator and reject the trade if the measured move does not comfortably clear your risk. A pattern that is textbook-perfect but offers barely one-to-one is still a bad trade.

Be careful with the entry mechanics on the break itself. Necklines on double tops are among the most watched prices on a chart, so stops cluster there, and a market order into the break can fill well away from your intended level in fast conditions. If you are trading a widely followed instrument around a data release, that risk is materially larger.

Where Market Structure Pro fits

The judgement that decides whether a double top is worth trading is a context judgement: is this the end of a trend, or the top of a range? Traders get this wrong constantly, and they get it wrong precisely because the pattern looks identical in both cases.

Market Structure Pro exists to answer that question mechanically rather than emotionally. It fuses 27 tools into a single verdict (TRADE, TRANSITION or NO TRADE) with a confidence percentage, an A/B/C grade and a plain-English explanation of what is driving it. Its dedicated ranging filter is designed to say NO TRADE when conditions are choppy or directionless, which is exactly the situation in which a double top is a coincidence rather than a signal.

It also helps with the confirmation problem. MSP is non-repainting: state locks on the closed bar, so a neckline break that is only an intrabar wick does not produce a reading that quietly vanishes. That matters on this pattern more than most, because the neckline is a stop cluster and intrabar pokes through it are routine. MSP is decision support only (it places no trades, it is not a signal service, and it guarantees nothing) but on a pattern this easy to find and this easy to misread, an independent read on whether the market is actually trending is the difference between a setup and a shape.

Frequently asked questions

What is a double top pattern?

It is two peaks at approximately the same price with a pullback between them, forming at the end of an advance. The low of that pullback is the neckline, and a close below it is treated as confirmation that the level has held and the move has turned. The double bottom is the mirror image at the end of a decline, with two lows and a neckline drawn across the high between them.

Do the two peaks have to be exactly the same price?

No, and insisting on that is a common beginner error. Treat the peaks as equal if they are within a small fraction of the recent average bar range. A second peak that pokes slightly above the first and then reverses is still valid and is often the strongest version, because it has swept the obvious stops before failing.

How do you confirm a double top?

The standard confirmation is a candle closing below the neckline, the low between the two peaks, on the timeframe you found the pattern on. An intrabar spike through the level is not a break. Many traders then wait for price to return and be rejected at the neckline from underneath, accepting that some patterns never offer that retest.

Where do you place the stop on a double top?

Above the higher of the two peaks. That is the price which invalidates the idea that the level is holding, so it is the honest place for the stop. Placing it just above the neckline gives a tighter risk but sits inside normal retest noise and gets hit by moves that would have worked.

What is the target for a double top or double bottom?

The convention is to measure the distance from the peaks down to the neckline and project the same distance below the break point. It is a convention rather than a forecast: nothing obliges price to travel that far, and many valid patterns fall short. Use it to judge whether the trade is worth taking, then take profit near real support or resistance close to that projection.

Why do so many double tops fail?

Mostly because traders take them in the wrong place. Price failing twice at the same price is exactly what happens at the boundary of a range, where it carries no reversal information at all. Double tops are informative at the end of an extended trend, at a level the higher timeframe already respects, and much less so anywhere else.

Is a double top the same as an M pattern?

Yes. M pattern and W pattern are informal names for the double top and double bottom, taken from the shape the price makes. Some traders use the M and W labels for versions where the second peak or trough overshoots slightly, but there is no formal distinction and the trading approach is identical.

Can I trade double tops on a 5-minute chart?

You can find them there constantly, which is the problem. On very low timeframes the failures at a level are dominated by spread and short-term order flow rather than by supply and demand, so the pattern carries little information. Use the 1-hour chart as a realistic floor for intraday trading and the 4-hour or daily for swing trades.

Related reading

• Support and Resistance: The levels that decide whether a double top is a signal or a coincidence.
• Trends vs Ranges: How to tell an exhausted trend from a range boundary before you interpret the shape.
• Candlestick Patterns: The bar-by-bar rejection signals that make the second touch believable.
• Price Action Trading: Reading what price is doing at the level rather than trading its outline.
• Liquidity: Why the neckline is a stop cluster, and why the second peak so often overshoots.