The on neck candlestick pattern is a two-candle bearish continuation pattern: in a downtrend, a long bearish candle is followed by a smaller bullish candle that opens below the first candle's low and closes right at that level. Buyers manage a bounce, but it only reaches the bottom of the prior session, the "neck", and classical candlestick analysis expects the decline to resume.

This guide covers how to identify and trade this candlestick pattern, and what happened after it. Our test bench found 742 on-neck patterns on daily charts since 1983 and compared each one with random days on the same assets.

What is the on neck candlestick pattern?

This candlestick pattern belongs to a family of four setups that share the same first session and the same gap down, and differ only by where the next session ends: on-neck, in-neck, thrusting and piercing line. The on-neck has the weakest bounce of the four, which is why classical analysis reads it as the most clearly bearish one.

Steve Nison, in Japanese Candlestick Charting Techniques (1991), presents the on-neck, in-neck and thrusting lines as weaker relatives of the piercing pattern: the white session does not recover enough of the black one to suggest a turn in the market.

Anatomy of the on neck pattern

Long black candlein a downtrend
Closes at prior low
White candleopens below the prior low

Median shape of all 699 usable daily occurrences from all markets, found by our test bench.

The first candle: a long bearish candle

The first session is a large black (red) candle that extends the downtrend and ends near its bottom. "Long" means a real body more than 1.3 times the average of the 10 previous sessions, 1.6 times on a textbook pattern, stricter than the open source TA-Lib library. On candle shape alone, at TA-Lib's own long-body threshold, our rule finds 93% of the cases that library flags.

The second candle: a small bullish candle

The next session starts with a gap under the first one's lowest price, then rallies. It is a white (green) candle, short in the textbook version, since it only climbs back to where the gap began.

Defining the closing price requirements

The second candle closes at or near the previous candle's low, without entering its real body. That is the whole difference with its twin: in the in-neck candlestick pattern, the close is a little higher, at or slightly above the first session's close. Our bench uses the strict version: after a downtrend (the last close before the pattern at least one average true range under the close 10 bars earlier, and in the lower half of the range of the 40 previous closes), a long black candle, then a white candle that opens under the first candle's low and closes at that low, within a small tolerance (the TA-Lib "equal" threshold, a fraction of the recent average range).

The chart below is not drawn by hand. It is the median shape of the on-neck cases our bench detected, with the faded bars showing the decline before it, and the lines of the trade we test: a short entry under the two sessions, a stop-loss above its high, and a target at twice the risk.

Bearish setup on the median shape of 699 daily occurrences from all markets, with entry, stop loss (SL) and 2:1 target (TP)

How to identify the on neck pattern on a chart

Look for four conditions, in this order:

Trend: price is falling, with lower highs and falling troughs. Without a downtrend, there is nothing to continue. Our bench is looser here: it only checks that the close before the pattern is at least one average true range under the close 10 bars earlier, and in the lower half of the range of the 40 previous closes.
First session: a large black candle that ends near its bottom.
Open of the second session: under the first session's lowest price.
End of the second session: back at or near the previous session's bottom, not above its final level. Check the numbers, not the picture: a few ticks above makes it an in-neck candlestick pattern.

The pattern appears only after a decline. When two similar candles appear in a sideways range, they are not an on-neck pattern.

It is not a frequent candlestick pattern. On daily charts it appears about once every 10,615 eligible bars, more often than the in-neck (once every 11,565) but much less than the thrusting pattern (once every 1,254). It needs a gap at the open, so it is mostly found on stock charts: 673 of the 742 daily cases come from US equities. On crypto, which trades around the clock, we found none in the whole history of daily charts.

Market psychology behind the on neck formation

The first session shows sellers in control: a large red session that ends near the bottom. The next session starts even lower, a sign that selling continued overnight. Then buyers step in and push the market back up, but only to the level where the prior session ended its fall. They fill the gap and stop there. Classical candlestick analysis reads this as a feeble rally: buying pressure was enough to recover the overnight loss, not to reclaim any part of the first session, and the bearish trend should resume.

It is a coherent story about two candles. Whether it predicts anything is a question for data analysis, and we come to it in the statistics section.

Is the on neck pattern bullish or bearish?

Bearish. The next session is white, which can look like a turning point, but where it ends is what counts: a rally that only reaches the first session's bottom is read as a failed attempt. A real turn needs much more, a close above the midpoint of the first session for a piercing line, or above its open for a bullish engulfing.

How to trade the on neck pattern

Trading this pattern means selling short, or exiting buy trades, in a decline that is already under way. Here is the trading plan our bench tests on every occurrence.

Entry: Sell stop at the pattern low
Stop loss (SL): At the pattern high
Target (TP): Twice the risk (2:1)

The trade our bench tests, drawn on the typical shape. Median shape of all 699 usable daily occurrences from all markets.

A clean case from our bench: ETFsChosen on its shape alone, never on the outcome. ETF prices are adjusted for splits and dividends, so they can differ from the prices quoted at the time.

TextbookKRE March 15, 2023Daily chart0.1% against the expected direction after 10 bars2:1 trade: stopped out, −1.00 R, 85 bars after entry, past this chart

Unless stated otherwise, the figures in this section and the next ones are for daily charts, all markets combined.

Setting entry points

The classic entry, and the one we test, is a sell stop just under the bottom of the two-session formation, so that the market has to confirm the continuation before you go short. In our tests, that entry was triggered in 59.0% of cases; in 40.0%, the market rose above the pattern high first and the setup was canceled before any trade was opened. The remaining setups expired: neither level was reached within 20 bars, or before the end of the data. A more conservative trader waits for a bearish session that ends under the formation.

Determining stop loss placement

The natural stop-loss goes just above the high of the pattern, the top of the first session. If the rally gets there, the bears have lost control. The median on-neck setup spans 6.20% of price, or 1.87 times the average true range (ATR), so the stop is wide: size the trade on the distance from entry to stop and risk only a small, fixed share of the trading account.

Defining take profit targets

The target we test is twice the risk (a 2:1 reward-to-risk ratio); the next support level is a classic alternative that we have not tested. Trades that reached the 2:1 target took a median of 12 days; those that hit the stop took a median of 8 days. Of the trades that hit their stop, 26.2% had first been ahead by 1 R, an argument for a clear exit plan and sound risk management.

On neck vs. in neck vs. piercing patterns

Key differences in closing prices

On-neck: the next session ends at the first session's bottom. Continuation.
In-neck: it ends at or slightly above the previous session's close. Continuation.
Thrusting: it ends well inside the first session, under its midpoint. Continuation.
Piercing line: it ends above the midpoint of the first session. Turn higher.

When the first candle ends exactly at its bottom, the on-neck and in-neck definitions overlap, and the same two sessions can carry both names.

Comparing reliability and trend implications

The three continuation candlestick patterns are tested the same way, with a short trade under the setup. On daily charts, every asset class combined:

Pattern (daily)	Occurrences	Down 10 days later	Random days	Closed trades	2:1 target hit	Same trade, random dates	Expectancy (R)
On-neck	742	49.1%	45.8%	436	27.1%	24.2%	−0.14
In-neck	681	47.1%	45.6%	374	32.4%	27.1%	+0.01
Thrusting	6,762	45.2%	45.6%	3,429	25.2%	26.4%	−0.22

The weakest bounce of the family did not give the best short. None of the three beat random days, and the piercing line, tested as a buy trade, did not either (36.6% of targets hit on 2,320 trades against 35.6% on random dates; its average result per trade, +0.15 R against +0.12 R, is not statistically significant).

Statistics and reliability of the on neck pattern

Our test bench applies the same rule to full price history and compares the outcome of each occurrence with random bars on the same assets. The pattern is bearish, so the expected direction is down and the trade tested is a short.

49.1%vs 45.8% at random

Moved the expected way 10 bars later, compared with random bars.

Measured on 742 daily on neck occurrences across 542 assets. Random bars on the same assets are the honest baseline.

No clear edge

On neck against random bars on the same assets, by market, on daily candles. Bars of almost the same length mean no edge.

Moved the expected way 10 bars laterOn neckRandom bars

All markets
All markets, On neck: 49.1%
All markets, Random bars: 45.8%
US stocks
US stocks, On neck: 49.6%
US stocks, Random bars: 46.0%
ETFs
ETFs, On neck: 43.8%
ETFs, Random bars: 44.4%

2:1 trade hit its targetOn neck tradeSame trade, random dates

All markets
All markets, On neck trade: 27.1%
All markets, Same trade, random dates: 24.2%
US stocks
US stocks, On neck trade: 28.5%
US stocks, Same trade, random dates: 24.1%

All statistics

Full results for the on neck candlestick pattern, market by market

Every occurrence our rule finds on full price history, compared with what the same assets did on random bars. A pattern only earns an edge if it beats random entries after correcting for the thousands of tests we run.

Our verdict

Not a reliable signal on its own here. Use it as context, not as a trigger.

10-bar move No clear edge2:1 trade No clear edge

On neck: 742 occurrences tested

Bearish pattern · All markets · Daily · 542 assets · 1983-12 to 2026-10

Moved the expected way after 10 bars

49.1%

vs 45.8% on random bars

Excess return vs its market, 10 bars

−0.06%

95% interval −0.56% to +0.50%

Excess return is signed so that positive favors the pattern. On a bearish setup, it means the asset lagged its market.

2:1 trade hit the target

27.1%

vs 24.2% for the same trade on random dates

Average result per trade

−0.14 R

vs −0.25 R on random dates

What this means. No measurable difference from the rest of its market over the next 10 bars.

Timeline

Entry triggered, after the pattern: 1 day · middle half 1 to 2 · on 438 triggered entries; Stop order beyond the pattern's extreme. 2 of these entries are not among the 436 trades: still open at the end of our data, or left out by the trade rules.
Time to the 2:1 target, when hit: 12 days · middle half 5 to 25 · on 118 trades that hit it
Time to the stop, when hit: 8 days · middle half 3 to 17 · on 302 trades that hit it
Time in the trade, all exits: 9 days · middle half 4 to 21 · on 436 trades

Size and trade path

Pattern range, high to low: 6.2% · middle half 4.2% to 9.2% · on 742 occurrences
Range vs average bar (ATR): 1.87× · middle half 1.55× to 2.22× · on 742 occurrences
Best open profit during the trade: 0.91 R · middle half 0.28 R to 2.01 R · on 436 trades
Stopped trades that were first up 1 R: 26.2% · on 302 stopped trades; Trades a stop moved to breakeven at +1 R would have closed flat instead of at a loss
Entry order triggered: 59.0% · on 742 occurrences

Medians, each on the cases named on its line; "middle half" spans the 25th to 75th percentile; no value under 30 cases.

How we detect it. In a downtrend: a long black candle, then a white candle of any size opening below the prior low and closing at that low. A long body is at least 1.3 times the average body of the 10 candles before the pattern and at least 0.6 times the largest of them. Candlestick counts are not checked against planted examples, so they may not be exhaustive.

All measurements (12 rows)

10-bar move: did price go the expected way?

Figures in gray: fewer than 30 occurrences, not interpreted. Excess return is signed so that positive favors the pattern. The test uses the bootstrap’s standard error and the interval its percentiles: near the 0.05 line, or on a few extreme moves, an interval can leave out zero while the test does not pass, and the test decides. Stock data misses most delisted companies, a survivorship bias we have not sized, which can flatter stock rows.
Market	Occurrences	Pattern	Random bars	Excess	10-bar move verdict
Crypto · 1-hour	26	42.3%	51.6%	−0.08%	Not enough data
Crypto · 4-hour	1	100.0%	50.0%	−1.45%	Not enough data
US stocks · Daily	673	49.6%	46.0%	+0.06%	No clear edge
ETFs · Daily	48	43.8%	44.4%	−1.39%	Weak, below chance
Indices · Daily	14	42.9%	42.9%	−2.01%	Not enough data
Forex · Daily	2	0.0%	51.2%	−1.40%	Not enough data
Commodities · Daily	5	60.0%	50.6%	+2.93%	Not enough data
All markets · Daily	742	49.1%	45.8%	−0.06%	No clear edge
US stocks · Weekly	34	38.2%	42.7%	+1.92%	No clear edge
ETFs · Weekly	4	25.0%	39.9%	−0.63%	Not enough data
Indices · Weekly	2	0.0%	33.2%	+0.11%	Not enough data
All markets · Weekly	40	35.0%	42.0%	+1.57%	No clear edge

2:1 trade: target hit and average result

R = initial risk of the trade. Entry by an order at the trigger level, filled at the open when a bar opens beyond it. Exit at the stop or the target, or at the open when a bar opens beyond either one; otherwise at the close after 100 bars. When one bar touches both the stop and the target, the loss is counted. No fees or slippage. n/a = not measured. Per trade: average result, against the same trade on random dates. Figures in gray: fewer than 30 closed trades, not interpreted.
Market	Trades	Target hit	Same trade, random dates	Per trade	2:1 trade verdict
Crypto · 1-hour	21	38.1%	46.0%	+0.14 R vs +0.38 R	Not enough data
Crypto · 4-hour	1	100.0%	100.0%	+2.00 R vs +2.00 R	Not enough data
US stocks · Daily	397	28.5%	24.1%	−0.10 R vs −0.26 R	No clear edge
ETFs · Daily	28	7.1%	19.3%	−0.81 R vs −0.34 R	Not enough data
Indices · Daily	7	28.6%	38.9%	−0.10 R vs +0.17 R	Not enough data
Forex · Daily	0	n/a	n/a	n/a vs n/a	Not enough data
Commodities · Daily	4	25.0%	37.5%	−0.26 R vs +0.13 R	Not enough data
All markets · Daily	436	27.1%	24.2%	−0.14 R vs −0.25 R	No clear edge
US stocks · Weekly	17	17.6%	23.2%	−0.47 R vs −0.31 R	Not enough data
ETFs · Weekly	2	50.0%	0.0%	+0.50 R vs −1.01 R	Not enough data
Indices · Weekly	1	0.0%	0.0%	−1.00 R vs −1.00 R	Not enough data
All markets · Weekly	20	20.0%	20.3%	−0.40 R vs −0.39 R	Not enough data

Data updated November 5, 2026 · code version c2a1683f · Read the methodology. We measure every pattern on crypto and forex (1-hour, 4-hour, daily and weekly charts) and US stocks, ETFs, indices and commodities (daily and weekly charts). On 1-hour and 4-hour forex charts the pattern only forms at the weekly open, across the weekend gap, so we do not publish those forex rows. Any other market or timeframe from that list missing from the tables above had no occurrence.

What you often read

“The on neck candlestick pattern is a reliable bearish continuation signal.”

What our data shows

On 742 daily occurrences, it moved in the expected direction 49.1% of the time, against 45.8% for random bars: no clear edge, since the average excess return is not statistically significant.

Table quoted in the text (verdicts and every market: see All measurements)

Market and timeframe	Patterns tested	Down 10 bars later	Random bars	Closed trades	2:1 target hit	Same trade, random dates	Expectancy (R)
All markets, daily	742	49.1%	45.8%	436	27.1%	24.2%	−0.14
US stocks, daily	673	49.6%	46.0%	397	28.5%	24.1%	−0.10
ETFs, daily	48	43.8%	44.4%	28	7.1%	19.3%	−0.81
Crypto, hourly	26	42.3%	51.6%	21	38.1%	46.0%	+0.14
All markets, weekly	40	35.0%	42.0%	20	20.0%	20.3%	−0.40

Read the table in pairs of columns. In the first row, 49.1% of on-neck occurrences saw a drop over the next 10 days, against 45.8% of random days on the same assets: the bearish candlestick pattern preceded a decline a little more often than chance. The excess return in the expected direction is −0.06%, with a 95% confidence interval from −0.56% to +0.50%. Zero is inside: no clear edge.

The short reached its 2:1 target in 27.1% of 436 trades, against 24.2% for the same entry, stop and target on random dates, for an expectancy of −0.14 R per trade before costs, against −0.25 R at random. Shorts on random dates lost money on these assets, mostly US stocks that rose over our data (a sample that misses most delisted companies, a survivorship bias we have not sized), and this setup did not change that; it lost a little less, a gap that is not statistically significant. The 10-bar excess return changed sign between the two halves of the history (+0.24% and −0.35% of excess return). On weekly charts and hourly crypto, too few trades closed to read the short. On daily ETFs, only 28 trades closed, too few to read, but the 10-bar move did worse than chance: an excess return of −1.39% in the expected direction on 48 patterns, rated weak, below chance, a result that passes a simple test but not our correction for multiple testing. On daily indices, the short hit its target 28.6% of the time against 38.9% on random dates, for −0.10 R against +0.17 R per trade, but on only 7 trades, too few to judge: rated Not enough data; on 4-hour crypto charts the bench finds a single on-neck, too few to show; daily commodities were rated Not enough data too, and daily forex had too few cases to judge. The method is published in our methodology.

Common mistakes when trading the on neck pattern

Ignoring the prevailing trend

We split on-neck setups by the 200-day moving average. Above it (a short-term decline inside a long-term uptrend), the market was down 10 days later in 42.2% of cases, for an expectancy of −0.39 R; under it, 51.4% and −0.06 R. Shorting against a rising long-term trend looked worse. This breakdown is descriptive, not corrected for multiple testing.

Trading without volume confirmation

A classic rule asks for heavy trading activity on the pattern. On our data, setups printed on volume above 1.5 times the 20-day average were down 10 days later in 52.4% of cases, against 47.6% on normal activity, and the 2:1 short reached its target more often (30.5% against 25.6%). Heavy activity looked better on both counts, with an expectancy above that of normal activity (+0.05 R against −0.22 R). This breakdown is descriptive too, not corrected for multiple testing.

Failing to use additional technical indicators

Combining with technical analysis

Classical technical analysis reads two candles in their context: momentum, support levels, the long-term direction. Here is how the on neck pattern fits into that kind of trading analysis, and which signals we could test.

Using moving averages for context

Integrating RSI or MACD for confirmation

One variant only shorts when the RSI is not yet oversold, so the decline has room to run, or when the MACD stays under its signal line. We have not tested these technical indicators with this pattern: test them on your own market before relying on them. The open-source TA-Lib library offers a CDLONNECK function to scan for it; check which rule your trading software uses for the second close.

Limitations and risks of the on neck pattern

False signals in ranging markets

In a sideways market there is no decline to continue, so the pattern's definition does not apply. The bench only counts setups after a decline, and even there no edge appeared.

The need for pattern confirmation

A sell stop under the formation is a built-in filter, and it filtered out 40.0% of setups. Even so, the trades that did enter did not beat random shorts.

A real example of the on neck pattern

Here is one case from a daily index chart, shown as an illustration, not as evidence. On March 3, 2020, the Nikkei 225 printed a big black session that ended at its bottom, 21,082.7. On March 4 it opened at 20,897.2 and closed at 21,100.1, back at that level. The sell stop at 20,862.1 was triggered on March 6, with a stop-loss at 21,719.8 and a target at 19,146.6, reached on March 10, two sessions after entry. The index was 20.7% down 10 sessions after the pattern, in the March 2020 crash: an extreme move, not a typical one, and the same two sessions also meet the in-neck definition.

The Real examples section shows more on-neck patterns picked by our bench, with their actual outcome.

Frequently asked questions

Is the on neck pattern a reliable reversal signal?

It is not a reversal pattern at all: it is a bearish continuation candlestick pattern. As a continuation setup, it preceded a decline a little more often than a random day on our data (49.1% against 45.8% on daily charts), but with an average excess return of −0.06%, within the margin of error.

How does the on neck pattern differ from the thrusting pattern?

The thrusting pattern's last session ends well inside the first session, under its midpoint. The on-neck pattern's last session ends at the first session's bottom, below its close.

Can the on neck pattern be used in forex and crypto trading?

Rarely. Crypto and forex trade continuously and hardly ever gap: we found 2 on-neck cases on daily forex and none on daily crypto. Hourly crypto charts had 26 (21 closed trades), too few to read.

What timeframes are best for identifying this pattern?

None stood out. Daily stock charts give the most occurrences. Hourly crypto charts give only 26 and weekly charts 40, too few closed trades to read the short, and the weekly 10-bar move showed no clear edge (35.0% down after 10 weeks against 42.0%, an average excess return of +1.57% that does not pass our statistical test).

Summary of key points

The on-neck pattern is a big black candle followed by a white one that opens under its bottom and closes back at that level, in a downtrend.
On 742 daily occurrences, 49.1% were down 10 days later, against 45.8% for random days.
The 2:1 short reached its target in 27.1% of 436 trades against 24.2% on random dates (−0.14 R against −0.25 R before costs): rated No clear edge, a gap that does not pass our statistical test.

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 setup with 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. This page is educational and is not financial advice: every position carries a risk of loss, and past results on our bench do not guarantee future ones.