Browse the library

Candlesticks · Two-candle patterns

In-neck candlestick pattern: a bearish continuation line

How the in-neck line forms, the narrow TA-Lib band that separates it from on-neck and thrusting lines, and what 31 real in-necks on US stocks did next.

On this page
  1. Anatomy of an in-neck line
  2. The exact rules Finabase uses
  3. Why traders expect the decline to continue
  4. How to confirm an in-neck line
  5. Two real in-neck lines
  6. What the data says
  7. In-neck versus similar patterns
  8. An educational outline for an in-neck setup
  9. How Finabase shows in-neck lines
  10. Frequently asked questions
  11. Summary

The in-neck line is a two-candle pattern that shows up inside a decline. A long black (red) candle is followed by a white (green) candle that opens below the black candle's low, rallies through the session, and then stalls right at the black candle's close. The rebound gets back to the "neck", the closing price of the previous day, and no further.

It belongs to a small family of bearish continuation patterns, together with the On-neck candlestick pattern: when the old low becomes a ceiling and Thrusting candlestick pattern: rules, examples and real results lines. All three have the same first two steps: a long down day, then a gap lower at the next open and a recovery. They differ only in how far that recovery gets. In the in-neck version buyers manage to close the session level with, or a hair above, the prior close, which looks like progress but stops exactly where yesterday's sellers finished.

This page explains the pattern, the precise and very narrow band TA-Lib uses to detect it on Finabase's charts, why traders expect the decline to resume, two real cases from Intel and Mastercard, and the results from our two-year study, which are striking but rest on only 31 cases.

Anatomy of an in-neck line

In-neck line in a decline (illustration, prices not real) Closes at the black close Opens below the prior low
An idealized in-neck line in a decline. Prices are illustrative.

Read the two candles from left to right.

  1. The long black candle. The decline continues with a session that loses substantially more than a typical day.
  2. The gap down. The next session opens below the black candle's low, so overnight selling has extended the move.
  3. The recovery. Buyers take over during the session and push price back up, forming a white body.
  4. The stall at the neck. The white candle closes at the black candle's close, or just above it, without getting meaningfully into the black body.

The Japanese name is usually given as iri kubi, "entering the neck". The image is of a rally that reaches the neck of the previous candle and is held there.

The exact rules Finabase uses

Finabase detects in-neck lines with TA-Lib's CDLINNECK. The function is a pure shape test on two candles:

Condition in TA-Lib Default setting
First candle Black, with a body longer than the average body of the 10 candles before it
Second candle White (close at or above open)
Gap at the open Second open strictly below the first candle's low
Close: floor Second close at or above the first candle's close
Close: ceiling Second close no more than 5% of the average high–low range of the prior 5 candles above it
Output −100 on the second candle; the function never returns a positive value

The neck family side by side

The in-neck, on-neck and thrusting lines share the first three rows of that table. What separates them is where the white candle closes:

Pattern Where TA-Lib requires the white candle to close
On-neck Within ± 5% of the average range of the black candle's low
In-neck From the black candle's close up to 5% of the average range above it
Thrusting More than 5% of the average range above the black close, up to the midpoint of the black body
(Piercing) Above the midpoint of the black body; a separate bullish pattern

To put numbers on the in-neck band: a stock with an average daily range of $4 lets the white candle close anywhere from the black close to 20 cents above it. A close one cent below the black close is not an in-neck at all.

Where TA-Lib departs from the textbook

There is no downtrend test. The source comment says the pattern is meaningful in a downtrend and then says the function does not consider it. Our data includes in-necks formed after ten-session rallies.

Black candles that close at their low create overlaps. When the first candle closes on its low, the black close and the black low are nearly the same price, and a single white candle can sit inside both the in-neck band and the on-neck band. Six of our 31 in-necks were also flagged as on-necks.

Why traders expect the decline to continue

The story is one of a counterattack that runs out of breath at an obvious level. After a big down day and a lower open, buyers step in, perhaps short sellers covering or bargain hunters, and the session turns green. But the rally stops at the prior close, the price at which yesterday's sellers were content to leave the market. Buyers did not manage to recover any of the previous day's loss.

The classic literature, Nison included, groups the in-neck with the on-neck as a signal that lower prices should follow. The in-neck's close reaches a little higher than the on-neck's, so some writers treat it as the slightly less bearish of the two. Either way, the underlying idea is that a weak, failing rally inside a downtrend confirms the sellers' control rather than challenging it.

How to confirm an in-neck line

Since the pattern predicts continuation, confirmation means sellers resuming:

  • A close below the white candle's low in the next session or two. That is the confirmation rule in our test.
  • A gap down at the next open.
  • A failed retest of the neck: price drifts back up toward the black close and is turned away again.
  • Momentum still pointing down, for example an Relative Strength Index (RSI): how to read and use it that remains under 50 and is falling rather than diverging upward.

The pattern is invalidated if the following sessions close back above the black candle's open or, more conservatively, above its midpoint.

Two real in-neck lines

Intel (INTC) daily, February to April 2025: an in-neck line before the tariff sell-off $20 $25 In-neck 2025-03-31 −10.6% after 10 sessions Mar Apr
Intel (INTC): an in-neck line on 2025-03-31 after an 11.6% ten-session decline. Ten sessions later the stock was 10.6% lower.

Intel, March 31, 2025. INTC had fallen 11.6% in ten sessions. On March 28 it closed at $22.71 after opening at $23.35. On March 31 it opened at $22.38, below the previous low of $22.56, rallied, and closed at $22.71, exactly the prior close. That is an in-neck in its purest form. Two days later the US announced sweeping new tariffs, and within five sessions INTC was 13.8% below the in-neck close; ten sessions later it was still down 10.6%. The pattern was right, though the size of the move came from news no candle could foresee.

Mastercard (MA) daily, November 2024 to February 2025: an in-neck line that buyers overran $500 $525 $550 $575 In-neck 2025-01-13 +8.4% after 10 sessions Dec Jan 2025 Feb
Mastercard (MA): an in-neck line on 2025-01-13 after a 5.8% decline, followed by an 8.4% rise within ten sessions.

Mastercard, January 13, 2025. MA had slipped 5.8% over ten sessions. A long black candle on January 10 (open $512.12, close $504.67) was followed on January 13 by an open at $501.00, under the prior low, and a close at $504.91, 24 cents above the black close, inside a band that was about 41 cents wide for MA at the time. Textbook shape, modest downtrend. It failed completely: MA rose on each of the next four sessions and was 8.4% higher ten sessions later. A trader waiting for a close below the white candle's $499.92 low never got one.

What the data says

We ran CDLINNECK across two years of daily bars (October 30, 2024 to October 29, 2026) for 60 large US stocks and ETFs. Over 29,760 symbol-days it fired only 31 times, on 20 of the 60 symbols. For a bearish pattern the useful question is how often price was lower afterwards, so the baseline is the 46.7% of all days that were lower five sessions later.

Group Cases Lower after 5 sessions Avg. 5-session return Avg. 10-session return
Any trading day (baseline) 29,760 46.7% +0.35% +0.66%
Every in-neck line 31 64.5% −1.63% −0.82%
After a 10-session decline (textbook setting) 21 71.4% −2.26% —
After a 10-session rise 10 50.0% −0.32% —
Confirmed, measured from the confirming close 8 50.0% −0.26% —

On the face of it this is the best-behaved pattern of the neck family. Nearly two out of three in-necks were followed by lower prices a week later, and inside an existing decline the share rose above 70%, with an average five-session loss of more than 2%. That lines up with the textbook.

In-neck versus similar patterns

Pattern Second candle opens Second candle closes Usual reading
On-neck Below the first candle's low Back at the first candle's low Bearish continuation
In-neck Below the first candle's low At or just above the first candle's close Bearish continuation
Thrusting Below the first candle's low Inside the black body, below its midpoint Weak bearish, disputed
Piercing line Below the first candle's low Above the midpoint of the black body Bullish reversal
Bullish engulfing At or below the first close Above the first candle's open Bullish reversal

The sequence from on-neck to piercing is a ladder of how much ground buyers recover. The further the white close gets into the black body, the more the reading shifts from continuation toward reversal.

An educational outline for an in-neck setup

A sketch of how the pattern might frame a bearish trade or, more commonly, a reason not to buy a bounce. It is not a recommendation.

  1. Require a real decline first, since the textbook pattern is a continuation.
  2. Mark the black candle's close (the neck) and the white candle's low. These are the two levels that matter.
  3. Look for continuation: a close below the white candle's low.
  4. Place the invalidation above the neck zone, for instance above the black candle's midpoint, and size the position from that distance.

How Finabase shows in-neck lines

  • Chart markers. The In Neck study, switched on by the INTC link at the top of this page, marks each white candle where TA-Lib returned −100.
  • Scanner. Detections are sorted by the sign of TA-Lib's output. Because CDLINNECK only ever returns −100, in-neck lines always appear on the scanner's Bearish list, whatever the trend before them.
  • Signals tab. In the candlestick gauge on a symbol's Signals tab, each of 61 patterns votes once: Buy if its latest non-zero reading in the past 8 bars was positive, Sell if negative, Neutral if none. An in-neck within that window is a Sell vote. The gauge score is (buy − sell) ÷ (buy + sell + the smaller of the neutral count and the larger side); below −0.2 reads Sell, below −0.6 Strong Sell.

Frequently asked questions

Is the in-neck pattern bullish or bearish?

Bearish. It is a continuation pattern: a rally inside a downtrend that stops at the previous close, after which the decline is expected to resume.

What is the difference between in-neck and on-neck?

The closing level of the white candle. On-neck closes back at the black candle's low. In-neck closes slightly higher, at or just above the black candle's close.

What if the white candle closes further into the black body?

Then it is no longer an in-neck. Below the midpoint TA-Lib calls it a thrusting line; above the midpoint it becomes a piercing line, which is read as bullish.

Why is the in-neck so rare?

The band is narrow. The white candle must gap below the prior low and then close within a few cents of the prior close, which happens in only a tiny share of sessions: 31 times in 29,760 in our study.

Summary

The in-neck line is a gap-down open followed by a rally that dies at the prior close, a compact picture of buyers failing to win back any of a big down day. TA-Lib draws its band tightly, which keeps the pattern rare. In our small sample it behaved as the books say, with 64.5% of 31 cases lower a week later and stronger results inside existing declines, but 31 cases are not enough to trust. Use it as a reason to doubt a bounce, and let the white candle's low and the neck above it define where the idea is right or wrong.

Sources and further reading

  1. Steve Nison, Japanese Candlestick Charting Techniques (New York Institute of Finance, 1991)
  2. Thomas N. Bulkowski, Encyclopedia of Candlestick Charts (Wiley, 2008)
  3. TA-Lib source code — CDLINNECK and the default candle settings
  4. Finabase chart data — daily bars for 60 US stocks and ETFs, 2024-10-30 to 2026-10-29; patterns detected with TA-Lib CDLINNECK