Piercing Pattern: Exact Rules, Real Examples, and Why It Barely Exists on Crypto
A piercing pattern is a failed capitulation. A downtrend prints a solid red bar, the next session opens below that bar's entire range — under the wick, a fresh panic low — and then buyers reverse the whole ambush, closing more than halfway back up the red body. The Japanese name is kirikomi-sen (切り込み線), the "cutting-in line," for the way the recovery slices into the prior candle.
Textbooks present it beside the hammer as a standard bottom signal. The rule Vike's engine actually runs is stricter than most books admit, and on a market with no opening bell it is nearly extinct — I measured one daily detection per 20,000 bars. Both facts change how you should use it, so this page leads with them.
What the piercing pattern means
The sequence matters more than the shape. Sellers ended the first session in full control. The second session opened with the worst print yet — a gap through the previous low, the moment stop-losses and panic sells actually execute. And from that low, demand absorbed everything and marched price back through the majority of the prior day's losses.
That round trip is a trap springing shut on late sellers. Everyone who sold the open is underwater by the close, and everyone who held short through the gap watched their best price evaporate. When it appears after a sustained decline, it is one of the few single-glance candles that records both capitulation and absorption.
That recovery has a ceiling, though. Close at or beyond the red bar's open and the pattern stops being a piercing — that finish belongs to the bullish engulfing, and the two definitions never overlap.
How to identify a piercing pattern: the exact rules
batch_piercing checks. It reads open, low and close, and never loads the high column at all.Four conditions, in detector order:
- The first bar is red — a genuine down session.
- The second bar is green.
- It opens strictly below the first bar's low — beneath the whole wick, not just the body.
- It closes above the midpoint of the red body, yet below its open.
How Vike computes it.batch_piercinginvike-indicatorsreads three columns — open, low, close — and never touches a high. It stamps+100whenopen < prior low,close > (prior open + prior close) / 2, andclose < prior openall hold across a black-then-white pair. Nothing measures body size, shadows, volume, or the trend behind the pattern.
Every threshold is strict. An open exactly on the prior low fails; so do closes landing precisely on the midpoint or on the prior open — ties belong to other patterns or to nothing.
One detection per 20,000 daily bars
Now the measurement that reframes the pattern. On crypto, a bar opens where the previous one closed — exactly, 37% of the time across my 480,125-bar daily scan — and the previous close always sits inside the previous range. Opening below the prior low therefore demands a real gap through the wick, which a liquid 24/7 market almost never supplies.
Daily numbers are stark: 122,296 red-then-green pairs, 100 opened below the prior low, 58 recovered past the midpoint, and 24 cleared the final fence. All 24 were thin, small-cap pairs; not one major ever printed the pattern in nine years of tape. The 34 pairs that recovered too far became engulfings and left the definition — the fences work exactly as the source says.
So the live examples on this page come from the 4-hour scan. Across 1,495,766 bars it found 443 detections, with a median qualifying gap of just 0.070% — shorter bars give a between-bar jump less time to be filled in, and that is the only reason the pattern exists at scale on this venue. The rule itself is untouched — I refuse to soften a clause to manufacture examples, because then the page would be describing a different pattern than the engine trades.
Piercing vs dark cloud cover
Dark cloud cover is the same trap upside down: a green bar, a gap above its high, and a red close driven back through the body's midpoint.
In the crate they are two separate functions, mirrored clause for clause, and the mirror swaps an entire price column: batch_piercing loads open, low, close — highs are invisible to it — while batch_dark_cloud_cover loads open, high, close and cannot see a low. The bullish member literally cannot read an upper wick. That column swap is the same construction as the morning and evening stars — and a different kind of symmetry than engulfing, where one function serves both directions with a sign.
How to trade a piercing pattern
Location first, always. A piercing into a prior swing low or a level that already produced a bounce is a setup; the same two bars mid-range are a statistic. The COW detection above — an 82.2% recovery of the red body — fired at the bottom of a week-long slide and ran 11.2% in five bars.
Demand the confirmation close. The bar after the pattern should close above the two-bar high. TUT's detection technically resolved upward, but only 2.1% — a reminder that a valid pattern can still pay almost nothing without momentum behind it.
Put the stop under the gap low. The pattern's second bar printed the lowest price of the whole structure; below it, the capitulation you bought never actually ended.
Know the honest base rate. Of 394 graded 4-hour detections, 181 rose at least 2% within five bars and 213 fell instead. On raw tape this pattern loses more often than it wins — 46% — and that number is the strongest argument on this page for treating it as a location tool rather than a signal.
Where the piercing pattern fails
Look at the failed chart above — it is typical of the breed. PLUME gapped 0.099% under its prior low on 2026-07-12, recovered 69.8% of the red body — a textbook print — and then fell another 5.4% within five bars. The downtrend simply did not care.
Two structural weaknesses drive the failure rate. First, the only venues that can produce the qualifying gap are thin ones, and thin books mean the "recovery" half of the pattern can be a handful of orders rather than real absorption. Second, a strict midpoint recovery still leaves the close below the previous session's open — by definition the buyers have not yet retaken even one full bar of ground, which is a weaker claim than an engulfing makes.
The mistake I made when I first traded these was reading the gap-and-recover as a guarantee of exhaustion. It is testimony about one session's auction. My scan says the market overrules that testimony slightly more often than it honours it, so every piercing trade I take now leans on the level it printed at, not the print itself.
Frequently asked questions
Is the piercing pattern bullish or bearish?
Bullish only — the engine emits +100 and nothing else. Its bearish mirror is dark cloud cover, a separate function reading a different price column.
Does the piercing pattern require a gap down? Under Vike's rule, yes, and a demanding one: the green bar must open strictly below the prior bar's low. Books that only require an open below the prior close describe a much looser pattern than the engine flags.
Where must the recovery finish? Strictly past the halfway point of the red body, and strictly short of its open. Reach the open and the pair becomes a bullish engulfing instead — the two patterns are mutually exclusive by construction.
How is a piercing pattern different from a hammer? Bar count and mechanism. A hammer is one bar whose own low gets rejected; a piercing is two bars where the second's open is the rejected extreme. Both claim a failed flush — the piercing just documents it across a session boundary.
This is educational material, not financial advice. Candlestick patterns describe probabilities, not certainties, and every real trade carries real risk — size positions so a wrong read costs little.