Harami is an old Japanese word for “pregnant”, and once you see the pattern the name makes sense: a big candle followed by a small one tucked completely inside its body. The big move ran out of energy, and the next day the stock went nowhere.
It’s one of the most common two-candle patterns, which makes it easy to test. Our scanner has recorded 17,603 bullish and 15,000 bearish harami patterns across 559 US stocks and ETFs since October 2023.
What it looks like
- Day one: a long candle.
- Day two: a small body that sits entirely inside day one’s body. The wicks don’t matter.
- Bullish harami: day one is a big down day. It hints that the selling is fading.
- Bearish harami: day one is a big up day. It hints that the buying is fading.
- Harami cross: the same thing, but day two is a doji.
How often it worked
| After a… | Cases | Higher 1 day later | Higher 5 days later | Higher 20 days later |
|---|---|---|---|---|
| Bullish harami | 17,603 | 52.6% | 53.4% | 55.4% |
| Bullish harami cross | 5,496 | 53.2% | 53.5% | 55.5% |
| Bearish harami | 15,000 | 52.8% | 53.3% | 54.8% |
| Bearish harami cross | 4,901 | 51.7% | 53.6% | 55.5% |
| Any day | 413,576 | 51.9% | 53.8% | 55.6% |
All four versions land within about one point of a random day. The bearish harami is very slightly weaker over 20 days (54.8% against 55.6%), which is the right direction, but only just.
Notice also that the bullish and bearish versions produced almost the same numbers. If a pattern and its opposite lead to the same result, the pattern isn’t telling you much about direction.
By trend
| Before the pattern | Bullish harami, higher 5 days later | Bearish harami, higher 5 days later | Any day |
|---|---|---|---|
| Fell more than 5% in 10 days | 53.1% (4,270) | 55.8% (1,299) | 54.0% |
| Moved less than 5% | 53.2% (10,929) | 52.5% (8,577) | 53.1% |
| Rose more than 5% | 54.2% (2,148) | 53.3% (4,805) | 53.7% |
The textbook setups are the bullish harami after a drop and the bearish harami after a rally. Neither stands out.
With confirmation
A bullish harami followed by a higher close was up five days later 53.7% of the time, about a point better than an ordinary up day (52.7%). That’s right at the edge of the margin of error, so call it a maybe.
One stock can look very different
On individual tickers, small samples can make the harami look much better or worse than it is. AMD, for example, was higher five days after 18 of its 30 bullish haramis, 60%. With only 30 cases, the margin of error is about 18 points either way, so that’s consistent with luck. The AMD page on the scanner shows the case count next to the percentage for exactly this reason.
How I’d use it
As a sign that a strong move has paused, nothing more. A bullish harami after a selloff in a stock you like is a reasonable time to start watching for an entry, perhaps with a cash-secured put below day one’s low. A bearish harami after a run-up is a reasonable time to check your position size. Neither is a reason to trade by itself.
Recent examples
In October 2026 the scanner flagged 333 bullish and 464 bearish haramis. Bullish ones on October 23 included AMD, AMAT and INTC; bearish ones included ACN and ADSK. Today’s list is on the candlestick scanner, and every pattern is compared in Do Candlestick Patterns Work?
Data: daily prices for 559 US stocks and ETFs, October 2023 to October 2026, patterns detected by TA-Lib. Educational only, not financial advice.