Doji — the indecision candle you must never read in isolation

On the twenty-second of January 2025, on the daily GBP/USD chart, a candle appeared that Anna recognised instantly — a textbook doji with a zero body, sitting precisely beneath the 1.2480 resistance level that had been tested five times across the preceding three months. The open printed at 1.2462, the close at 1.2461, the lower shadow reached down to 1.2420 and the upper shadow stopped at 1.2483. Anna did not open a position. She waited for the next candle, because she knew that a doji without confirmation of the breakout direction is an informational pattern, not an entry signal. Twenty-four hours later a bearish candle closed at 1.2398 — forty pips below the doji low. Short entry on the close of the confirming candle, stop loss above the doji high, target at the previous support of 1.2200. The position closed in four sessions for a profit of three hundred and eighty pips. This article explains why the doji is one of the most misread candlestick patterns and how to read it in the context that turns a warning into a tradeable setup.

What a doji is and what a zero body actually signals

A doji is a candle in which the open and the close are virtually identical — they differ by one or two pips, or do not differ at all. Visually it looks like a horizontal line cutting through vertical shadows. The name comes from the Japanese word meaning "mistake" or "error", which in classical candlestick tradition referred to the fact that a theoretical candle with open equal to close was a statistical aberration in the era of hand-drawn rice charts from the eighteenth century.

The session mechanics of a doji are as follows: during the time unit, both sides of the market — buyers and sellers — made active attempts to push price, but neither managed to keep the upper hand by the end of the period. The shadows record the range of those attempts; the zero body records the final equilibrium. It is precisely this equilibrium, following an earlier dominance by one side, that turns a doji into a potential harbinger of a change in market dynamics.

Steve Nison, who introduced candlestick patterns to Western markets in 1991 with "Japanese Candlestick Charting Techniques" published by the New York Institute of Finance, described the doji as one of the most important warning candles of the Japanese analytical system. In his follow-up book "Beyond Candlesticks" from 1994, he developed the thesis that a doji gains meaning only in context — the pattern itself, without the right location, is informationally empty.

Four types of doji — anatomy of the differences

The classification of doji in technical analysis rests on the proportions and placement of the shadows relative to the zero body. Four core types are distinguished, each with its own signal strength and interpretation.

The classic doji has shadows of comparable length — upper and lower stretch symmetrically around the horizontal line of the body. The silhouette resembles a plus sign. Session mechanics: the market moved in two directions with roughly equal force, and the final balance reflects the exhaustion of both sides. A classic signal in the role of a warning ahead of a trend reversal, but requiring confirmation of direction by the next candle.

The long-legged doji has both shadows very long — two or three times longer than in a classic doji. Visually it is an elongated cross, in which the vertical line dominates the horizontal. Mechanics: the volatility during the session was extreme, both sides covered significant distances, but neither held its position. The long-legged doji is the strongest signal of uncertainty and equilibrium in the entire doji family. Bulkowski, in "Encyclopedia of Candlestick Charts" (Wiley, 2008), reports that a long-legged doji at the top of an upward move after six consecutive green candles has historically delivered a reversal hit rate of around fifty-eight percent — a small but positive statistical edge.

The dragonfly doji has a long lower shadow, with the open, close and high all sitting at a single point — the very top of the candle. The silhouette forms the letter "T". This is, in essence, a doji with the anatomy of a bullish pin bar: sellers drove price deep down, but buyers reclaimed the entire territory by the close. A dragonfly doji at a meaningful support level is one of the strongest reversal signals in a downtrend.

The gravestone doji is the mirror image of the dragonfly — a long upper shadow, with open and close at the session low. An upside-down "T". Mechanics: buyers tried to lift price, but were pushed back to the starting point. A gravestone doji at the top of an uptrend, in a zone of tested resistance, is a reversal signal of strength comparable to a shooting star.

Contextual location — where the doji matters and where it does not

The rule of contextual location is the same as with the pin bar: the very same graphical pattern in different parts of the chart generates signals of radically different reliability. A doji in the middle of consolidation, with no structural anchor, is essentially market noise and should not generate positions regardless of whether the next candle confirms direction. A doji at the top of a six-week upward move, sitting precisely under a resistance level tested three times in the previous months, is an institutional-grade warning watched by both retail traders and the research desks of investment banks.

These figures come from Bulkowski's empirical studies and from independent analyses run on samples of major currency pairs across 2020 to 2024. The hit rate of a doji is lower than that of a pin bar in comparable conditions — a fact that reflects a fundamental difference: a pin bar has the direction of the bounce written into its anatomy, while a doji needs external confirmation of the direction of the breakout. For that reason, the doji is often treated as a warning signal preceding the actual entry setup, rather than a stand-alone entry trigger.

False doji versus true doji — when a candle only looks like a doji

Day-to-day trading practice reveals a phenomenon that Nison labelled the "false doji" — candles with a technically zero body that are nevertheless informationally empty, because they form under conditions devoid of any genuine equilibrium between market forces. Telling false from true doji is critical, since it directly affects the decision to wait for confirmation and to open a position.

False doji form most often in three situations. First, during hours of thin liquidity — the Asian session on European pairs, national holidays in the major financial centres, weekends on certain instruments. In those windows the candle range is small, volume is minimal, and the equality between open and close may simply reflect the absence of flow rather than a genuine balance of forces. Second, in the middle of a strong trending move, where a brief graphical pause does not alter the dominant market dynamic — the next candle frequently continues the prior move and the doji turns out to be a pause rather than a reversal. Third, immediately before high-volatility macroeconomic releases, when the market deliberately withdraws liquidity in anticipation of headlines, and the resulting doji is a structural artefact rather than a genuine indecision signal.

A true doji requires three conditions met simultaneously: the candle forms during a period of normal liquidity (European or American session for major pairs); it appears after an exhausting earlier trending move or at the vicinity of a meaningful support or resistance level; and it forms on a volume no lower than the twenty-session average. Meeting all three conditions gives a high probability that the zero body really hides a genuine market equilibrium rather than a structural artefact.

Entry rules, stop loss and profit targets

A doji is not a stand-alone entry signal in the classical Nison interpretation — it requires confirmation of direction by the next candle. The entry rule based on a doji works as follows: after the doji closes, we wait for the next candle and open a position in the direction in which the breakout resolves. If a doji forms at the top of an upward move under resistance, and the next candle closes as a bearish candle below the doji low, the short entry is taken at the close of the confirming candle or one pip below its low.

The stop loss is always placed above the doji high (for a short position) or below the doji low (for a long position), with a five to ten pip buffer. The buffer guards against stop-hunting behaviour. In Anna's trade from our opening example, the doji high stood at 1.2483, so the stop loss was placed at 1.2493 — ten pips above the extreme of the pattern.

• First target — the next support or resistance level in the path of the move. In Anna's trade, the first target was the 1.2280 support, where price had bounced three times in the previous quarter. The reward-to-risk ratio came to 1:2.7.
• Second target — the next support, or the low of an earlier consolidation. In this case 1.2200, a floor tested four times during the autumn of 2024. Reward-to-risk ratio 1:3.8.
• Trailing stop once the first target is reached — move the stop loss to break-even and trail the position along the 20-period EMA on the H4 timeframe.
• Position size — classically one percent of capital per trade. With a sixty pip stop loss and a 10,000 euro account, this translates into a position size equivalent to a rounded micro-lot on GBP/USD.

The key distinction from pin bar trading: in a pin bar, the stop loss is placed beyond the extreme of the long shadow, because that shadow defines the strength of the rejection. In a doji, the stop loss is placed beyond the extreme of the entire candle, because both sides of the shadows mattered for the session mechanics. The average doji stop loss is therefore larger than a pin bar stop loss, which automatically lowers the reward-to-risk ratio achievable on the same price scenario.

Confluence — when a doji becomes an A-grade signal

A doji in isolation delivers a hit rate hovering around fifty-five percent — a small statistical edge, but not large enough to build a long-term strategy on. Confluence with other analytical factors lifts the hit rate to around sixty-five percent, and in selected configurations to seventy percent.

The first layer of confluence is structural levels — support or resistance tested several times across the previous weeks or months. A doji in the vicinity of a level tested three times carries more weight than a doji in a random spot. In Anna's trade, the 1.2480 resistance had been tested five times in the three months that preceded the signal — first-order structural confluence.

The second layer of confluence is the higher timeframe. A daily doji that lines up with a gravestone doji on the weekly chart, or with a significant resistance on the monthly chart, carries institutional weight. In practice this means that, before opening a position based on a daily doji, the trader checks whether the same price zone produces analogous signals on D1, W1 and M1. The alignment of three timeframes is a setup that appears only a handful of times each year — and it is precisely this setup that drives the statistics showing a seventy percent hit rate.

The third layer of confluence is technical indicators — most commonly the RSI oscillator in overbought or oversold territory, MACD divergences, extreme stochastic readings. A doji at the top of an uptrend, with RSI above seventy and a bearish divergence on MACD, is a setup in which every analytical element points to weakening upward momentum. That still does not guarantee a reversal, but it shifts the probability another few percent in the trader's favour.

"The doji is a candle of equilibrium in a sea of dominance. Its strength lies not in the pattern itself but in the contrast with the prior price history. A doji at the top of three weeks of bullish dominance is a warning that cannot be ignored. A doji in the middle of consolidation is visual noise that most often simply means the market does not know where it is going." — Steve Nison, Japanese Candlestick Charting Techniques, New York Institute of Finance, 1991, pp. 152–158.

Five most common mistakes in trading the doji

The doji looks like a simple pattern to master — learn to recognise the zero body and the strategy is essentially ready. In reality, all the hit-rate numbers cited above assume the trader avoids five classical traps that beginners walk into almost without exception.

• Opening a position without confirmation of direction. The classic doji and the long-legged doji require the next candle to confirm the direction of the breakout. Entering on the close of the doji itself amounts to guessing direction, and the hit rate of that guess does not deviate meaningfully from fifty percent. Without confirmation, the trader is flipping a coin.
• Trading a doji in the middle of consolidation. A doji without confluence with a support or resistance level, or with an exhausting trending move, is an informationally empty candle. The hit rate of such signals falls below the break-even threshold once spreads and commissions are taken into account.
• Doji on low timeframes. M1, M5 and M15 generate dozens of doji every session, because the small range of those candles makes accidental equality between open and close easy to occur. A doji carries informational value from the one-hour timeframe upwards, and works best on H4, Daily and Weekly.
• Stop loss inside the body or the shadows. A classic mistake consisting of placing the stop loss "more safely" than the rule prescribes — five pips above the body rather than above the extreme of the upper shadow. That placement guarantees the stop will be taken on the first retest of the zone by market makers monitoring the clusters of protective orders.
• Ignoring volume and the hour of formation. A doji during the thin liquidity hours of the Asian session on GBP/USD is often a structural artefact rather than an indecision signal. The trader should always check whether the candle formed under normal volume conditions, and ideally during the European or American session.

Conclusions

A doji is a candle with a zero body, in which the open and close are virtually identical. It signals an equilibrium between buying and selling pressure following an earlier period of dominance by one side of the market. The four core types — classic doji, long-legged doji, dragonfly doji and gravestone doji — differ in the anatomy of their shadows and in signal strength, but they all belong to the same family of indecision patterns.

The pattern itself is a conditional signal whose hit rate hovers around fifty-five percent in isolation and rises to sixty-five or seventy percent when three conditions are met: the right contextual location, confirmation of direction by the next candle, and confluence with support or resistance levels or with signals from higher timeframes. A doji in a random spot on the chart, unconfirmed, on a low timeframe is essentially market noise. A doji at a meaningful resistance after an exhausting upward move, confirmed by a bearish candle closing below the pattern low — that is the kind of setup Nison and Bulkowski both describe as one of the strongest reversal signals of Japanese candlestick analysis.

In practical trading, three rules apply: wait for the next candle to confirm direction (with the exception of dragonfly and gravestone doji, which have direction baked into their anatomy), place the stop loss several pips beyond the extreme of the entire candle rather than beyond the body, and trade doji only on the H4, Daily and Weekly timeframes. The five most common mistakes are: entering without confirmation, trading a doji in the middle of consolidation, dropping down to noisy lower timeframes, placing the stop inside the candle, and ignoring the context of volume and time of formation. Eliminating these traps is the bulk of the work for any trader who wants to use the doji as part of a regular strategy.

Related reading: Japanese candlesticks — the basics for the broader family of candlestick patterns; pin bar reversal trading for a comparison with the strongest rejection pattern; most important candlestick patterns — the full survey that complements the doji family.