Hammer Candle: Count Them Before You Trust Them
A hammer is a candle with a small body at the top of its range and a lower wick at least twice the body's length, appearing after a decline. It is read as selling that was pushed back, and the shape is a description of the price path rather than evidence about who was trading.
Covered on this page: TradingView.
How it works
A hammer is a candle whose body sits at the top of its range with a long tail hanging beneath it. Price fell during the session, then came back and closed near where it opened.
The usual definition is a lower wick at least twice the body’s length and little or no upper wick. Color is not part of the definition, though a green hammer is generally treated as slightly stronger than a red one.
The context requirement is the part that gets dropped. A hammer is defined as appearing after a decline. The identical shape in the middle of a range is not a hammer in any meaningful sense — it is a bar with a wick, and there is no rejection of anything because there was no prior move to reject.
The story, and what the story actually rests on
The narrative is that sellers pushed price down, buyers stepped in, and the recovery to the close shows the buyers won. It is a satisfying story and it explains the shape neatly.
What the shape actually records is where price went, in what order. Open, low, close. It does not record how many contracts traded at the low, whether the recovery was one large buyer or the absence of further sellers, or whether the whole thing happened in the last two minutes of the session.
This distinction matters because the story is the reason people size positions on it. “Buyers defended this level” is a claim about participants; the candle is a claim about four prices. Those are different kinds of statement, and only one of them is on the chart.
Volume is the closest thing to evidence available. A hammer on heavy participation means a lot of contracts changed hands during that rejection; a hammer on nothing means the price path happened in a vacuum. That is a real distinction and it is the only one the chart can support.
In practice: the counting exercise
Apply the standard definition mechanically to a long daily history and hammers turn up regularly — common enough to matter and rare enough to notice. The count is the first number worth having.
Suppose 40 of them appear, and 22 are followed by a higher close: 55%.
Now set that beside every day in a real record. From 1950 to 2026 the S&P 500 closed above the prior day’s close on 53.1% of trading days, pattern or no pattern. So the hypothetical hammer’s 55% sits about two points above a baseline, on a sample of 40.
Both halves of that comparison are essential and the second one is almost never published. A pattern statistic without a base rate is unreadable: 55% sounds like an edge until you know that doing nothing in particular scores 53.1% on the broad index.
And the sample size settles it. Forty occurrences cannot distinguish a two-point difference from chance — a single hammer falling the other way would take the rate to 52.5% and erase it entirely. This is a demonstration of method, not a finding about hammers, and the same arithmetic applies to every pattern statistic quoted anywhere without an occurrence count next to it.
The shape is also timeframe-dependent in a way that undermines it. A daily hammer is a summary of a session that, on a five-minute chart, contains its own trend, range and reversal. Aggregate the bars differently and the hammer disappears or appears. Nothing about the market changed.
A gap down that recovers through the session draws a textbook hammer mechanically. The “rejection” is the market reopening at a price the prior session’s participants never saw, which is a completely different event from buyers defending a level.
Each one traded pays a full round trip in costs. Forty attempts is 40 payments of the spread and commission, which have to come out of whatever the two-point edge is worth before anything is left.
What a hammer is not
It is not a reversal. It is a candle that sometimes appears at reversals, along with appearing in a great many other places.
It is not the same as a doji. A doji has almost no body at all; a hammer has a small but real body at the top of the range.
It is not distinguishable from a “hanging man” by shape. The identical candle after an advance is given the opposite name and the opposite meaning. If the shape carried the information, the context would not have to.
And it is not a level. The low of the hammer is often used as a stop, which is reasonable, but the candle itself does not mark support — the wick shows where price went, not where orders are waiting.
How this connects
A small body with a long lower wick, which makes it a near neighbor of the dragonfly doji - the difference is only how much body is allowed.
The wick records rejection. Price traded well below and did not stay, which is a genuine fact about the period.
It is part of the candlestick patterns vocabulary, so everything true of candlesticks applies, including the arbitrariness of the session boundary.
Reversals are the attached claim - and one bar cannot contain a statement about the next one.
And support is what turns it into a setup. The same shape at a level tested several times is a different event from the same shape in the middle of a range.
When it fails
In a range the lower boundary manufactures hammers. Price approaches the floor, turns, and closes near the top of its bar — which is the definition. They occur constantly and carry no more information than “price bounced in a range again,” which you already knew.
The second failure is ignoring the context clause. A hammer without a preceding decline is a bar with a wick. Screeners that flag the shape alone return hundreds of these.
A third is treating the wick low as a hard floor. Placing a stop a tick below it puts the stop at the most obvious price on the chart, which is exactly where the liquidity sits.
A fourth is the definition drift. Two-to-one wick-to-body, three-to-one, upper wick allowed or not — every source uses slightly different thresholds, so “hammers work” and “hammers do not” can both be true of different definitions on the same chart.
And a fifth is the single-bar frame itself. One candle summarizes a whole session into four numbers, discarding the sequence inside it. A hammer built by a slow grind down and a sharp late recovery and a hammer built by an instant flush and an all-day drift back look identical and are not the same event.
The original data
The figure to hold every hammer statistic against is the base rate: 53.1% of S&P 500 trading days from 1950 to 2026 closed above the day before. A hit rate quoted for any candle has to clear the equivalent number for its own instrument and period, with the occurrence count printed next to it.
The point of those three figures together is the shape of the argument, not a result. A hit rate, an occurrence count and a base rate belong in one sentence, and a rate a few points above the base on a few dozen occurrences is what chance produces routinely. That is how easily a pattern statistic appears out of nothing, and it is the reason to demand an occurrence count and a base rate before believing any figure about any candle, including the hypothetical ones on this page. Run the same three numbers on your own instrument; the code to do it is a dozen lines.
Related
Candlestick patterns covers the wider vocabulary and where it came from. Candlesticks explains what the four prices in a bar actually are. And reversals is the event a hammer is supposed to mark, treated on its own terms.
I count a pattern before I trust it, and the hammer is the one I would start with. Get a number for how often the shape appears and a number for what happens next, then ask whether the gap between them supports the confidence you are trading it with.
— 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.