Harami Cross Candlestick Pattern: Bullish and Bearish Rules, Trading and Tested Results

The harami cross candlestick pattern is a variant of the harami candlestick pattern, made of two candles: a long candle in the direction of the trend, followed by a doji whose tiny body sits completely within the real body of the previous candle. After a downtrend it is a bullish harami cross; after an uptrend, a bearish harami cross. In technical analysis, it is read as a sign of hesitation signaling a possible reversal.

This guide covers the definition, how to identify the pattern on a chart, the psychology behind it, how it differs from the standard harami pattern, how to build a trading plan around it with confirmation, indicators and risk management, and what actually happened after it. We applied one rule to full price history and found 11,016 bullish and 14,362 bearish cases on daily charts alone, across the US stock market, ETFs, indices, commodities, crypto and forex, and compared the outcome with random bars on the same assets.

What is the harami cross candlestick pattern?

The harami cross is a two-candle reversal pattern from Japanese candlestick charting, one of the classic candlestick patterns of technical analysis. "Harami" is the Japanese word for pregnant: the long first candlestick carries the second candle within it. "Cross" refers to the shape of that second candlestick, a doji, whose opening and closing prices are almost equal, so that it looks like a plus sign. Steve Nison, who introduced Japanese candlestick patterns to Western traders in Japanese Candlestick Charting Techniques (1991), treats it as a variant of the harami that deserves more attention, because the second candlestick shows complete indecision and uncertainty.

The pattern has four features: a clear trend before it, a long first candle in the direction of that trend, a doji as the second candle, and that small body completely within the first body. Only the bodies count: the shadows of the second candlestick may extend beyond the first body.

Like other harami patterns, it appears after a trend and indicates that the move is losing strength: a smaller candle within a large one. The pattern is related to the inside bar, but it is not the same: an inside bar keeps the whole second candle within the range of the first, while a harami cross only needs the second body within the first body.

The psychology is a story of lost momentum. In a bullish harami cross candlestick pattern, the first session belongs to the sellers: the long red candlestick ends near its bottom, and selling pressure looks strong. In the next session, price starts above that close, within the prior range, and neither side wins: the market ends almost where it began. This candle represents uncertainty: the sellers have lost their strength, but buyers have not taken over yet. In the bearish version, the same story follows a rally. Nison reads this pause as a warning that the trend may change. Whether it leads to a reversal more often than a random session is a question for data, answered below. Read as a bullish reversal, it suggests the bearish trend may be ending; read as a bearish reversal, it warns that a large bullish candle may have been the last push higher.

How to identify the harami cross pattern

Our test bench applies the definition without judgment: after a trend (the last close before the pattern is at least one average true range (ATR) below, or above, the close 10 bars earlier, in the matching half of the 40 previous closes), a long candlestick, then a second candle with almost no body contained in the first. "Long" and "doji" are measured against the 10 bars before, as in the open-source TA-Lib library. To spot it on a chart, look for a large candle followed by a much smaller one with almost no body, after a clear move. The candlestick charts below are not drawn by hand: each is the median shape of the clean cases our bench detected, with the lines of the trade we test.

Structure of the bullish harami cross

• The market is in a downtrend: a series of falling candlesticks.
• The first candlestick is a long bearish candle (red or black), often closing near its low.
• The second candle is a doji that opens and closes within the previous body.
• Its color does not matter.

On the median daily case in our bench, the two candlesticks span 4.33% of price, or 1.41 times the average true range (ATR). The bullish harami cross candlestick pattern is a potential bullish reversal signal on any chart: the sellers who dominated the first session could not move the close away from the open in the second.

Structure of the bearish harami cross

• The market is in an uptrend.
• The first candlestick is a long bullish candle (green or white), often closing near its high.
• The second candle is a doji within the previous body.

The bearish harami cross is the mirror image, a potential bearish reversal pattern at the end of a rise: the buyers stop, the market hesitates, and the end of the rise is possible, but likely only if the next session falls. It can serve as a warning to protect gains in a bullish trend.

Differences between a standard harami and a harami cross

The harami and the harami cross share the same structure. The only difference is the second candle: in a standard harami pattern, it is a small real body of either color; in a harami cross pattern, it is a doji, with almost no body at all. The bullish harami pattern and the bullish harami cross pattern both need a falling market and a long red candle first.

Nison's reading makes this variant the stronger of the two harami patterns, because the doji shows more hesitation; taken as a claim of reliability, it can be tested. Our data does not support that claim. On daily charts, the bullish harami cross reached its 2:1 target 36.0% of the time against 36.2% for the same trade on random dates, with an expectancy of +0.10 R; the standard bullish harami reached it 36.8% of the time against 36.3%, with +0.11 R. The two harami candlestick patterns are almost identical: on daily charts, both bullish 2:1 trades did about as well as the same trade on random dates in average result per trade (+0.11 R against +0.10 R for the harami, +0.10 R against +0.10 R for the harami cross), a gap that stays within the range of chance in our tests (the bench rates both trades No clear edge). The harami cross is also canceled more often: the standard bullish harami triggered its entry in 51.5% of daily cases, against 44.4% for the bullish harami cross. The other side tells the same story: the standard bearish harami pattern had an expectancy of −0.14 R per trade before costs on daily charts, against −0.15 R for this variant.

It is also not the same as the doji star, where the doji opens with a gap away from the previous candlestick instead of within it. That detail matters in our data: the bullish doji star reached its 2:1 target 37.7% of the time against 36.7% on random dates, for an average result of +0.18 R against +0.16 R per trade: a wider gap, but one our bench still rates no clear edge. Other related candlestick patterns include the engulfing pattern, where the second body covers the first, and the three inside up and down, a harami confirmed by a third candle.

How to trade the harami cross pattern

The classic trading plan buys a break above the high of the two candlesticks for a bullish harami cross (or sells a break below their low for the bearish one), with a stop-loss on the other side and a target set as a multiple of the risk. This is the trading plan our bench simulates on every occurrence. Before trading it, consider the potential reward against the distance to the stop: the formation alone indicates a pause, not a new direction, and a stricter plan waits for a higher open the following session (a lower open for the bearish version).

Setting up your trade entry

Managing risk with stop-loss and take-profit

The natural stop-loss goes below the bottom of the two candlesticks (above their top for the bearish version). If that point breaks, the prior move has resumed. Because the first candlestick is long, the stop is wide, a potential problem for a modest trading account: size the position on the distance from entry to stop so that a full loss costs a small, fixed share of your trading account. One target is twice the risk (2:1), another the next resistance (the next support for a short). In our daily data, bullish trades that reached the 2:1 target took a median of 10 bars, and those that hit the stop a median of 5 bars. Of the bullish trades that hit their stop, 22.8% had first been up 1 R, an argument for a plan to protect open profit.

Combining with technical indicators

Because the harami cross candlestick pattern alone has little edge, and that candle alone indicates only a pause, it is worth checking other indicators and trading tools before making a decision.

• Using MACD and RSI for confirmation. One rule takes a bullish harami cross only when the RSI is below 30 (oversold) or forms a bullish divergence, and the bearish version when the RSI is above 70; another waits for a MACD crossover in the direction of the trade. We have not tested these indicators: test them on your own market before relying on them.
• Analyzing volume during the pattern. One volume rule wants low trading volume on the cross-shaped candle and more volume on the confirmation candle. We measured one version: bullish cases on volume above 1.5 times the 20-day average reached the 2:1 target 36.1% of the time with +0.15 R, against 35.8% with +0.09 R on normal volume. The gap is minor, and this breakdown is descriptive (not corrected for multiple testing).
• Trend filter. Above or below the 200-day moving average, the result was the same: bullish cases reached the 2:1 target 36.4% of the time above it and 35.6% below it, with +0.12 R and +0.08 R. The long-term trend did not matter for the bullish version.
• Support and resistance. The pattern at a support level marked in advance, or at resistance for the bearish version, gets more weight in classic technical analysis. We have not measured this filter.

Harami cross: our statistics by market

Is the pattern an edge on its own? Not for the 2:1 trade, and only on one line for the move 10 bars later (the bearish version on daily US stocks). On daily charts, the bullish harami cross pattern was followed by a rise slightly more often than random bars on the same assets, but the difference does not survive our correction for multiple testing. Here are the numbers from our test bench: the blocks just below follow the Bullish and Bearish switch, and the table quoted in the text, further down, is for the bullish side.

Read the columns in pairs: this is how we compare candlestick patterns with chance. "Up 10 bars later" is the share of cases where price had risen 10 bars after the pattern; "random bars" is the same measure on random bars of the same assets. In the first row, 56.0% were followed by a rise, against 53.9% for random days: the typical stock rises more often than it falls, so a figure above 50% proves nothing alone. The excess return over the rest of the market after 10 bars is +0.18%, with a 95% confidence interval from +0.01% to +0.37%: zero lies just outside it, but with so many patterns and markets tested, a result this close to zero does not survive our correction for multiple testing, so the bench rates it no clear edge. The excess return also weakened between the first half of the history (+0.28%) and the second (+0.09%). The "2:1 target hit" columns use the plan above; expectancy is the average result in R (multiples of the amount at risk), before fees. The full method is in our methodology.

On indices, the 2:1 target was hit 49.3% of the time against 38.6%, but on only 67 trades, and the gap in average result per trade does not survive our statistical tests. On weekly charts, 61.2% of cases were up 10 weeks later against 56.4%: rated No clear edge: the excess return over the rest of the market (+1.07%) does not survive our correction for multiple testing, and it changed sign between the first half of the history and the second. The 2:1 trade on those weekly cases did not do measurably better than on random dates (39.6% against 38.2%).

The bearish harami cross pattern gave a mixed picture. On daily charts (14,362 cases), 46.8% were lower 10 bars later against 45.2% for random bars, a gap rated promising: it passes a simple test, but not our correction, and it is much weaker in the second half of the history (+0.26% in the first, +0.05% in the second); on US stocks alone (13,485 cases), the excess return of +0.17% in the expected direction is rated proven edge. The 2:1 short trade, opened on 5,699 of them, reached its target 28.5% of the time against 27.2%, with an expectancy of −0.15 R against −0.19 R. On hourly crypto (5,960 cases, 2,404 closed trades), the expectancy was −0.20 R against −0.07 R. In a bullish stock market, selling on this signal lost money, about as much as a short on any random day.

Harami cross case studies

The gallery in the Real examples section shows real cases picked by our bench (best and worst results, textbook shapes, most recent cases and typical draws, where a market has them). The cases written up below are on daily charts: the XLU utilities ETF, the S&P 500 index and the UNG natural gas ETF. None of those cases reached its 2:1 target, which is not surprising: on daily charts across all markets, only 28.5% of closed bearish trades and 36.0% of bullish ones reach it.

Bullish harami cross example, canceled before a rebound (XLU). On February 15 and 16, 2022, after a decline, the XLU ETF printed a red candlestick from 29.00 to 28.68, then a cross-shaped candle from 28.72 to 28.75. The buy stop sat above the high (29.11), with a stop at 28.47 and a 2:1 target at 30.39. Price never reached the entry: on February 17 it fell to 28.42, below the bottom, and the pattern was canceled. XLU was 4.6% above the pattern close 10 bars later: here the cancellation rule cost the trader a rise.

Bullish harami cross example, entry then stop (S&P 500). On July 8 and 9, 2015, the S&P 500 completed the pattern: a red candle from 2,077.66 to 2,046.68, then a cross from 2,049.73 to 2,051.31, with a top at 2,077.66 and a bottom at 2,044.66. The buy stop was triggered on July 10, with a 2:1 target at 2,143.66. Ten sessions later the index was 2.5% higher and had traded up to 2,132.82, just short of the target; it then faded and fell through the stop on August 20: a full 1 R loss.

Bearish example, entry then stop (UNG). On September 12 and 13, 2012, after a sharp rise, the UNG ETF printed a green candlestick from 319.36 to 328.96, then a near-doji that opened at 326.24 and closed at 325.44. The sell stop below the bottom (317.6) was triggered on September 14, with a stop at 330.4. Price fell to 302.24 on September 19, far from the 292 target, then turned up, and the stop was hit on September 26. Ten sessions after the pattern, UNG was 3.7% higher.

Pros and cons of trading the harami cross

• Pro: simple to identify. It is easy to look for on any chart: two candlesticks and objective rules based on opening and closing levels.
• Pro: an early warning when a turn does come. The pattern forms on the reversal bars themselves, earlier than a moving average crossover or a trendline break would signal the turn. On its own, it did not predict turns reliably: the 2:1 trade never did measurably better than on random dates, and only one 10-bar result survives our correction for multiple testing, the bearish version on daily US stocks (see the verdict).
• Pro: a natural stop that appears with the pattern. The other end of the pattern, the extreme of its two candles, defines where the idea is wrong.
• Con: few signals per asset. It occurs about once every 770 daily bars for the bullish version, every 590 for the bearish one.
• Con: most setups never trigger. 55.3% of daily bullish cases were canceled before entry.
• Con: no clear edge and many false signals. On daily charts, all markets together, neither version showed an edge that survives our correction for multiple testing, so it should not be traded alone.
• Common mistake: trading the pattern itself. Buying at its close means entering before any confirmation.
• Common mistake: ignoring the context. The bullish version needs a downtrend and the bearish one an uptrend. A harami pattern in a sideways range is not a reversal pattern.
• Common mistake: confusing it with a doji star. A candle that gaps away from the previous body is a different pattern, with different results.

Frequently asked questions about the harami cross

Summary of key points

• A harami cross is a long candle followed by a smaller doji within its body, after a trend: bullish after a downtrend, bearish after an uptrend.
• It differs from the standard bullish harami pattern and its mirror only by the shape of its body, and did not perform better in our data.
• On 11,016 daily bullish cases, 56.0% were up 10 bars later, against 53.9% for random bars, but the average excess return over the rest of the market does not survive our correction for multiple testing.
• On daily charts, the bullish 2:1 trade reached its target 36.0% of the time against 36.2% for the same trade on random dates.
• Most setups are canceled before entry: 55.3% for the bullish version, 59.9% for the bearish one.

Risk warning. This page is educational and is not investment advice. No harami candlestick pattern, and no other candlestick pattern, guarantees a result, and any trade can lose the full amount at risk.

About the author and the data. Written by Amaury, founder of PatternsWizard, who is passionate about finance, quantitative trading and data analysis. He built the PatternsWizard test bench, which detects each of these patterns with fixed, published rules on full market history: crypto, forex, US stocks, ETFs, indices and commodities. Trades and win rates are compared with random bars on the same assets, excess returns with the rest of the market, with error margins and a correction for multiple testing, and the whole pipeline is checked on random-walk prices, where nothing can be predicted. Statistics on this page were generated on November 5, 2026. The full methodology explains every step.