Bullish harami and bearish harami patterns

Harami candlestick pattern: the definition

A harami is read on two consecutive candles. StockCharts' ChartSchool introduction to candlesticks speaks of the 'harami position': a candlestick that forms within the real body of the previous candlestick. The first usually has a large real body, and the second a smaller real body than the first.

The same page gives the meaning of the word: according to it, harami means 'pregnant' in Japanese, and the image is that of the second candle nestled inside the first. StockCharts' candlestick pattern dictionary sums the pattern up in one line: a two-day pattern in which the second day has a small body completely contained within the range of the previous body, and is the opposite colour.

What the pattern describes is a pause. A tall candle moved price, and the next period opens and closes without leaving that body. StockCharts reads the small candle as consolidation. That is all the two candles contain: they say neither who stopped pushing, nor which way price will leave.

Its name changes with where it appears. StockCharts puts it this way: all harami look the same, and it is the preceding trend that makes them bullish or bearish. After a decline, the pattern is read as a potential bullish reversal; after an advance, as a potential bearish reversal. The word that matters is 'potential', and the measurements quoted below show why.

Identifying a harami: the criteria, and where sources disagree

Three sources were read for this lesson: StockCharts' ChartSchool, Thomas Bulkowski's pages on thepatternsite.com and TradingView's help pages on its candlestick pattern indicators. They agree on the essentials: two candles, a first one with a large body, a second one with a small body sitting inside the first. They disagree on three points, and each one changes the number of harami you will count on the same chart. Pages consulted in October 2026.

The body alone, or the whole candle? For StockCharts, the shadows of the second candle do not have to be contained within the first, though it is preferable. Bulkowski writes to ignore the shadows: only the open and the close of the second candle must fall within the body of the first. He adds that one edge of the small body may be at the same price as an edge of the large one, but not both. TradingView does not use the same word from one page to the next: in its help page on the bullish harami, the small candle is entirely encompassed within the range of the large one; in its help page on the bearish harami, within its body.

The colour of the second candle. StockCharts' dictionary asks for the opposite colour to the first. Its articles on reversal patterns nevertheless accept all four combinations: white then white, white then black, black then white, black then black. Bulkowski and TradingView fix the colours: for the bullish harami, a tall bearish candle, black or red, then a small bullish one, white or green; for the bearish harami, the reverse.

The trend before the pattern. Bulkowski makes it an identification criterion: a downward price trend before the bullish harami, an upward one before the bearish harami. StockCharts also asks for a trend to reverse, and gives examples of guidelines to establish it: a price below its 20-day exponential moving average, lower and lower peaks and troughs, a price below its trend line. TradingView writes that the pattern is usually preceded by a trend, and leaves the choice in the settings of its indicators: compare price with its 50-period simple moving average, additionally require that average to sit the same way relative to the 200-period average, or take no account of the trend at all.

None of these definitions is the right one. But a harami counted sometimes with the shadows, sometimes without, sometimes with a trend filter, sometimes without, fits into no record. Pick one definition, write it down, and hold it: that is the rule the lesson on candlestick charts sets for every pattern whose authors do not give the same definition.

The bullish harami pattern

The bullish harami appears after a decline. On Thomas Bulkowski's page on the bullish harami, the criteria fit in three lines: two candles, a downward price trend before, a tall black candle followed the next day by a white candle nestled within its body. TradingView describes it as a small green candle encompassed within a red candle, usually preceded by a downward trend.

In theory, it is a bullish reversal. Bulkowski measured what actually happens. He calls a breakout the first close above the top of the pattern or below its bottom. For the bullish harami, the breakout is upward, so against the decline, 53% of the time, and downward 47% of the time. He concludes that the pattern functions almost randomly, and that you probably cannot guess the breakout direction.

His other numbers point the same way. Out of 103 candlestick patterns ranked, the bullish harami has a frequency rank of 25, which makes it a common pattern, and an overall performance rank of 38, which Bulkowski calls decent but not outstanding. According to his glossary, that rank orders the price changes recorded 1, 3, 5 and 10 trading days after the breakout. Its best average move ten days after the breakout is a rise of 4.05%, obtained in a bear market after an upward breakout; Bulkowski only considers a move good from 6% up.

These numbers have a population, and it is poorly known. The page states that they are based on hundreds of trades, and refers to Bulkowski's book, Encyclopedia of Candlestick Charts, for the detail. It gives neither the market nor the period measured, and its criteria are written in days. Nothing says that 53% carries over to your market and your timeframe.

Bulkowski gives two sorting leads, referring to that book: bullish harami that appear, in his words, 'within a third of the yearly low' are the ones that perform best, and tall candles are to be preferred. These are hypotheses to check against your own records, not buying rules.

The bearish harami pattern

The bearish harami is the same pattern after a rise. On Thomas Bulkowski's page on the bearish harami: two candles, an upward price trend before, a tall white candle followed by a small black one whose open and close fall within the body of the white candle. That page is where he defines the breakout: a close above the top of the pattern or below its bottom. TradingView describes the bearish harami as a small red candle encompassed within the body of a green candle, usually preceded by an upward trend.

In theory, it is a bearish reversal. As measured by Bulkowski, it is a bullish continuation 53% of the time: slightly more than one time in two, price first closes above the top of the pattern, and the rise carries on. He calls that result near random, and writes not to depend on the pattern to reverse the primary trend.

The rest is weaker than for the bullish harami. Frequency rank 26, so a pattern just as common; overall performance rank 72 out of 103. Its best average move over ten days is a drop of 4.01%, in a bear market after a downward breakout, again short of the 6% Bulkowski considers a good move.

He states where the pattern is, in his view, most likely to act as a reversal: at the top of an upward retrace within a downward primary trend. When price then breaks out downward, it rejoins the primary trend. Put differently: in that case the primary trend carries the reading, and the pattern only dates the end of the retrace.

The harami cross: when the second candle is a doji

The harami cross is a harami whose second candle is a doji. StockCharts' dictionary defines it that way, and defines the doji as a candle whose open and close are virtually equal: it looks like a cross or a plus sign. The pattern exists in a bullish version, after a decline, and in a bearish version, after a rise.

On paper, it is the strong version of the pattern. StockCharts writes that the smaller the body of the second candle, the more likely the reversal, and that the chances increase if it is a doji. TradingView's help page on the bearish harami cross says the doji shows indecision, and that the pattern hints that the trend might reverse.

The measurements say something else. On Bulkowski's page on the bullish harami cross, the pattern is a bullish reversal in theory, and in practice a bearish continuation 55% of the time: price keeps falling. Its overall performance rank is 50 out of 103, its frequency rank 47. On his page on the bearish harami cross, the pattern meant to reverse a rise is a bullish continuation 57% of the time, with an overall performance rank of 80 out of 103 and a frequency rank of 45. Bulkowski sums it up this way: the bearish harami cross works better than the bearish harami, but not in the direction that traders expect.

With Bulkowski, its definition is also stricter than that of the plain harami: the doji must fit, shadows included, between the high and the low of the first candle, and he notes that a reader of chart patterns will recognise it as an inside day. TradingView's help page on the bullish harami cross, for its part, asks for a doji entirely encompassed within the body of the first candle. Two authors or two platforms therefore do not count the same harami crosses.

Harami vs inside bar vs engulfing candle

Harami and inside bar. The inside bar compares extremes: its high and its low sit within those of the previous candle. The harami compares bodies, and StockCharts specifies that the shadows of the second candle do not have to be contained. The two tests therefore do not overlap. A small candle whose body sits within the large body, but with a shadow that goes above the previous high, is a harami without being an inside bar. Conversely, after a candle with a small body and long shadows, a candle can sit between its extremes without its body being inside the other's: that is an inside bar, not a harami.

Harami and engulfing candle. They are inverse patterns. In the engulfing candle, the body of the second candle fully covers that of the first; in the harami, the body of the first is the one that contains that of the second. StockCharts' article on bullish reversal patterns describes both patterns with those two definitions.

Harami, spinning top and doji. The second candle of a harami is often a candle you know under another name. StockCharts notes that doji and spinning tops have small bodies, and can therefore form in the harami position. The lesson on the spinning top describes that candle when it stands alone.

Bearish harami and shooting star. Both are read after a rise as a potential bearish reversal. The bearish harami takes two candles; the shooting star fits in a single one, with a small body and a long upper shadow. It has its own lesson, with its criteria and its numbers.

Confirmation, invalidation and limits of the harami

Confirmation. For StockCharts, most reversal patterns require confirmation, and without it the pattern is neutral: at best it indicates a potential support or resistance level. For a bullish harami, confirmation is further upside follow-through: a gap up, a long white candlestick or a high-volume advance, within one to three days after the pattern. For a bearish harami, it is the same thing downward, within the same window, according to its article on bearish reversal patterns.

Invalidation. None of the sources read gives an invalidation level specific to the harami. The clearest marker is the breakout as Bulkowski defines it: after a bullish harami, a close below the bottom of the pattern is a downward breakout, and the decline has continued instead of reversing. After a bearish harami, a close above the top of the pattern says the same thing for the rise.

Limits. The first lies in the numbers: 53%, 55% and 57% are close to one chance in two, and for three of the four variants it is continuation that wins, not the announced reversal. The second lies in the definition: shadows counted or not, colour imposed or not, trend required or not. An indicator that prints a harami on your chart applies one of those definitions, not all of them. The third lies in duration: StockCharts points out that candlestick patterns are short-term patterns, usually effective for one to two weeks, and that they are read alongside other aspects of analysis.

In your journal. None of these readings is an order to buy or sell: they are conventions, to be tested on your market and your timeframe with the method in the article on how to backtest a trading strategy. Tradoshi does not detect candlestick patterns: you spot the harami on your chart, with the definition you chose. The journal serves the next step: you tag each trade taken on that reading, for example 'bullish harami, bodies only, downward trend', and the dashboard filters by tag to show what it has given you, personally.

Frequently asked questions