Double Bottom: How Equal Is Equal?
A double bottom is two lows at approximately the same price separated by a recovery, completed when price breaks above the high between them. The tolerance on approximately is never specified, so the pattern's frequency depends entirely on how strict the person looking decides to be.
Covered on this page: TradingView.
How it works
Price falls, makes a low, recovers, falls again and stops at approximately the same price. Drawn on a chart the shape resembles a W, which is where the alternative name comes from.
The pattern is not complete until price breaks above the high between the two lows. That is the confirmation, and until it happens there is no double bottom — only two lows and an expectation.
Saying “a double bottom is forming” is therefore a forecast, not a description. It is worth separating those two things every time, because the second low failing to hold is an entirely ordinary outcome and the pattern’s supporters rarely count it.
The unspecified tolerance
Nothing in the definition says how close the two lows must be. Within a tick? Within 1%? Within 3%? Every source says “roughly equal” and stops there.
That single omission decides how many double bottoms exist on any chart, and it does so after the fact, when you already know how the story turned out. A low 2% below the first is a double bottom if the recovery worked and a failed support test if it did not.
Even an index that rose for decades flips direction constantly: the S&P 500 closed up on only 53.1% of its trading days from 1950 to 2026. A series that turns that often contains an enormous number of local lows, and any two of them at a similar price qualify under a rule with no tolerance in it.
So the fix is a written tolerance, decided before you look. “The second low must be within 0.5% of the first, and at least 10 bars later” is not a better theory of markets — it is a constraint that stops the pattern being found wherever you want it.
In practice: the parts that are actually specified
The measured move is the vertical distance from the lows to the middle high, projected upward from the break. Arithmetic derived from the shape — its use is deciding in advance whether the distance to the target justifies the distance to the stop, not predicting where price will stop.
The second low is the most obvious price on the chart by the time the pattern is visible. Everybody who identified the formation put a stop just under it, which makes it a concentration of resting orders — the exact situation the liquidity grab page describes.
Which does not mean move the stop somewhere arbitrary. It means knowing that the obvious stop is obvious, and either accepting the risk of being swept or placing it further away and sizing smaller. Both are defensible; not thinking about it is not.
Volume is the only non-price input available. The classical reading is lighter participation on the second low than the first, and expansion on the break of the middle high. It is the element most often skipped, because it is the one that disqualifies formations you want to trade.
Aggregate the bars and the pattern disappears into a single candle with a long lower wick. Nothing about the market changed; the resolution did. Any pattern that survives only at one bar length is partly a property of your chart settings.
A gap above the middle high confirms the pattern at a price nobody could transact at. The formation “worked” and the entry did not exist — a distinction that closing-price backtests quietly lose.
Each attempt pays one round trip of spread and commission — modest against a formation that takes weeks, which is the standing advantage of slow patterns over fast ones.
And the order book holds no memory of the first low. Whatever orders stopped price there are long gone. If the second low holds, it is because new orders arrived at a level people could see, which is the ordinary self-fulfilling mechanism of any visible price.
What a double bottom is not
It is not support. It is two touches. Support is a claim about where orders are, and two touches is the minimum evidence for it, not proof.
It is not complete at the second low. The break of the middle high is the pattern.
It is not an inverse head and shoulders. That requires a deeper middle trough. Here the two lows are level and there is no third.
And it is not defined numerically anywhere. No tolerance, no minimum separation, no minimum depth of the intervening recovery. Every one of those is supplied by whoever is looking.
How this connects
Two lows at a similar level with a rally between them. The double top is the mirror, built the same way.
Support and resistance is the mechanism. The pattern is really one level tested twice, and the shape is a way of noticing that.
The neckline is the trigger - the high between the two lows - which makes this a breakout trade with everything that implies.
A liquidity grab is the common alternative reading. A second low slightly below the first, then a sharp reversal, is the same picture with a different story attached.
And reversals are claims about the future, which two lows and a rally cannot contain on their own.
When it fails
The most common outcome is a third low. At that point the structure was never a double bottom — it was the floor of a range, and the two “equal lows” were the first two touches of a boundary that had several more coming.
The second failure is the tolerance stretch. The second low comes in 3% below the first, which is called “a slightly lower double bottom” and traded anyway. Once the tolerance moves after the fact, the pattern cannot be wrong, and a pattern that cannot be wrong is not informing anything.
A third is entering before confirmation. Buying the second low is buying a level that has held once before, with no break to confirm it. That is a support trade, which is a fine thing to do — it just is not this pattern, and the risk is different.
A fourth is the sweep of the second low. Price dips a few ticks under it, takes out the obvious stops, and then recovers and completes the formation anyway. Everyone positioned correctly is out before the move, which is a recurring cost of trading at the most visible price on the chart.
And a fifth is ignoring the timeframe the pattern lives on. A double bottom on a five-minute chart is two lows separated by an hour. It carries none of the weight of the same shape on a daily chart, and the name does not distinguish them.
The original data
On the S&P 500 between 1950 and 2026, 46.9% of trading days closed lower than the day before, and even ten-day stretches ended lower 41% of the time. Every down close followed by an up close leaves a local low behind it, so a daily chart of the index is crowded with candidates.
That churn is the whole argument for writing your tolerance down. With down closes arriving on nearly half of all days, a chart of any length holds a great many local lows, and pairs of them at similar prices are not rare events — they are the expected state of a series that oscillates. A pattern defined without a numeric tolerance will therefore always be findable, on any chart, in any market, including one generated at random. Choosing the tolerance in advance is the difference between the pattern constraining your decisions and decorating them.
Related
Double top is the mirror formation, with its own liquidity problem at the two equal highs. Chart patterns is the parent page on named shapes and how much of them the observer supplies. And reversals covers the event this pattern claims to mark.
Write down a tolerance before you look: the second low has to be within a stated percentage of the first, or it does not count. It sounds trivial, and it sharply cuts the number of double bottoms you can find on any chart — the ones that survive are the ones worth looking at.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money. Some links on this page earn a commission if you buy through them. It costs you nothing and it does not decide what appears here or in what order — how these pages are made is set out in our methodology.