Marubozu candlestick: the shadowless candle, rules and real data

A marubozu is a candle that is all body. It opens at one extreme of the session and closes at the other, with little or nothing sticking out above or below. The Japanese word is usually translated as "close-cropped" or "bald", a reference to the missing shadows. A green, or white, marubozu opens at the low and closes at the high. A red, or black, one opens at the high and closes at the low.

Where the hammer and its relatives are about rejection, a long shadow showing that one side pushed and failed, the marubozu is about the absence of any pushback. One side controlled the session from the first trade to the last. That makes it one of the most intuitive candles to read, and also one that is easy to over-read: a single decisive day says a lot about that day and much less about the next week.

This page covers the shape and its two colors, the precise TA-Lib test, a structural fact about how Finabase labels these candles (every marubozu is also flagged three other ways), two real cases, and what 1,227 marubozus on 60 large US stocks and ETFs did next.

The two marubozus

White marubozu. The first trade of the day was the low, and the last trade was the high. Buyers never let the price dip below the open and never gave back any of the gain into the close. After a decline, it is read as an abrupt shift in control. Inside an advance it reads as continuation.

Black marubozu. The mirror image. The first trade was the high and the last was the low. Sellers had the session the whole way. After a rally it is read as a warning; in a decline, as more of the same.

Nison describes both as expressions of strength in the body's direction, with their meaning depending on context. The textbook ideal has literally no shadows. In practice almost every daily candle has a few cents of shadow, so any usable detector has to allow some tolerance.

The exact rules Finabase uses

Finabase marks marubozus with TA-Lib's CDLMARUBOZU, measured on daily bars against the stock's own recent candles:

The output is the candle's color times 100: +100 if the close is at or above the open, −100 if it is below. That is the whole test.

Some details are worth spelling out:

• "Long" is modest. The body only has to beat the recent average body, not double it. On a stock that has been trading in tiny candles, a fairly ordinary day can qualify.
• "No shadow" is relative to range, not to body. A stock whose candles have averaged a $10 range can show up to $1 of shadow on each side and still pass.
• There is no trend test, no gap test and no volume test. The function looks at one candle and its ten predecessors' sizes.

One candle, four labels

Because the marubozu's conditions are the strictest of a family of body-size tests, any bar that passes it tends to pass the looser ones too. In our 29,760-bar sample:

The first two overlaps are guaranteed by the code: a candle that is short on both shadows is short on either one. In fact the bars that are both a closing marubozu and a belt hold are exactly the marubozus. The long-line overlap is not a logical necessity, but in our data it held every time. On a Finabase chart with all patterns turned on, a marubozu day can therefore carry four markers, and the Signals tab counts four votes for it.

What the absence of shadows says

Shadows are the record of disagreement within a session: prices that traded but were not accepted by the close. A marubozu has almost none, so for one session there was no meaningful disagreement about direction. A white marubozu after a decline says buyers arrived at the open and kept arriving. A black marubozu after a rally says sellers did the same.

That one-sidedness often comes from information rather than from chart psychology: an earnings report, a guidance change, a macro release, an index flow. The candle shows the reaction, not the reason. Whether the reaction continues depends on how much of the news the market has already priced in, which a candle cannot measure.

Does it need confirmation?

Marubozus are often treated as their own confirmation, because the body already shows commitment. Traders who want more evidence typically ask for one of these:

• Follow-through: the next session closes beyond the marubozu's extreme (above the high for white, below the low for black). Our test uses this definition.
• Holding the midpoint: the next sessions do not retrace more than half of the body. A deep retracement suggests the move was a one-day event.
• Volume: a marubozu on heavy volume carries more weight than one on a quiet day.

Follow-through was common. 324 of 672 white marubozus (48%) and 234 of 555 black ones (42%) closed beyond their extreme the next day, much higher rates than for reversal candles with long shadows.

Two real marubozus

UnitedHealth, August 8, 2025. UNH had fallen 10.7% in ten sessions. On August 8 it opened at $244.79, 42 cents above its low, and closed at $250.89, 19 cents below its high. The $6.10 body beat the recent $5.12 average body, and both shadows were under the $0.97 limit. The next session closed at $252.37, above the high, so it confirmed. A week later, after a regulatory filing showed Berkshire Hathaway had taken a stake, the stock gapped sharply higher: +21.2% after five sessions and +22.5% after ten. The marubozu caught the turn; the size of the move came from news no chart could anticipate.

Broadcom, December 19, 2024. AVGO had run up 28.1% in ten sessions. On December 19 it opened at $227.91, 14 cents under the high, and closed at $218.32, 74 cents above the low, with a $9.59 body against a $6.45 average. That is a black marubozu at the top of a big rally, the textbook setup for a reversal warning. The next session closed at $220.79, above the low, so it never confirmed. The stock recovered: +10.7% after five sessions and +8.3% after ten. One forceful down day inside a strong trend was not enough to end it.

What 1,227 marubozus did next

We ran CDLMARUBOZU over daily bars from October 30, 2024 to October 29, 2026 for 60 large US stocks and ETFs, then looked at five-session outcomes in the direction of the candle's color.

Confirmed outcomes are measured from the confirmation close. Returns are plain price changes, so a positive figure after a black marubozu means prices rose on average.

• White marubozus leaned in their direction. 55.5% higher against a 53.1% baseline, and 60.5% for confirmed ones. That is a modest tilt, consistent with short-term momentum in a rising market, not a strong signal.
• Black marubozus did not. 47.9% lower against 46.7%, and prices rose on average after them. Strong down days in this period were more often bought than extended.
• Context was mostly continuation. Only 29.4% of white marubozus followed a 10-session decline, and 32.7% of black ones a 10-session rise. Most appeared in the direction of the existing move.

Marubozu and its body-size cousins

The marubozu is the rarest and most demanding of the group, which is part of why it carries a little more information than the others.

An educational outline

How the candle is commonly used, described for learning only.

1. Decide whether it is continuation or reversal. In the direction of the trend it confirms strength; against a long trend it is a potential turn.
2. Use the candle's far end as the reference. For a white marubozu, the open (its low) is where the day's buyers started. A close below it erases the session.
3. Many traders avoid chasing the close. A marubozu often closes at an extreme; waiting for a pullback toward the body's midpoint gives a closer reference point.
4. Illustrative numbers: a white marubozu from $40.00 to $43.00. An entry at the $41.50 midpoint with invalidation at $39.90 risks $1.60 a share; a $320 budget is 200 shares.

How Finabase shows marubozus

• Chart markers. Add the Marubozu study (the UNH link above opens it). Markers appear on every flagged bar, white and black alike. Expect them to coincide with Closing Marubozu, Belt Hold and Long Line Candle markers if you turn those on too.
• Scanner. The S&P 500 candlestick scan splits detections by the sign of TA-Lib's output: white marubozus appear under Bullish (t=bullish), black ones under Bearish (t=bearish).
• Signals tab. In the Candlestick gauge, the marubozu votes Buy or Sell according to the color of its most recent detection in the last eight bars. In our sample it was voting Buy on 15.8% of days, Sell on 12.4% and neutral on the remaining 71.8%. Because the same bar also triggers three related patterns, one strong day can move the gauge by up to four votes.

Frequently asked questions

Is a marubozu bullish or bearish?

It takes the color of its body. A white (green) marubozu is bullish for that session, a black (red) one bearish. Whether it signals continuation or reversal depends on the trend before it.

Does a marubozu need zero shadows?

The textbook ideal has none. TA-Lib allows each shadow to be up to 10% of the average daily range of the previous ten candles, which is a few cents to a dollar or so on most large stocks.

What is the difference between a marubozu and a closing marubozu?

A marubozu has very short shadows at both ends. A closing marubozu only needs a very short shadow at the close end, so it may have a long shadow where it opened. Every marubozu is also a closing marubozu, but most closing marubozus are not marubozus.

Are marubozus more reliable than other candles?

In our data the white marubozu showed a small edge and the black one none. Neither is reliable enough to act on without the broader context.

Summary

The marubozu is a session without dissent: a long body and almost no shadows. TA-Lib defines it with three size tests and scores it by color, with no view of trend. It fired on about 4% of daily bars in our sample and was always accompanied by the closing marubozu, belt hold and long line labels. White marubozus leaned slightly upward over the next week, black ones showed no edge. Read it as a strong statement about one day, and look to the trend for what it means.