Abandoned Baby Candlestick Pattern: Rules, Trading and Tested Results

The abandoned baby candlestick pattern is a rare three-candle reversal pattern: a long candle in the direction of the trend, a doji separated from it by a price gap, shadows included, then a candle in the opposite direction, separated by a second gap. The middle candle is left alone on the chart, like an island, which gives the formation its name. In simple terms, the pattern represents a sudden change of direction: a strong move, a pause, then a strong move the other way.

This guide will help you learn how to identify both versions, why the setup is read as a trend reversal, how traders enter, place a stop-loss and take profit, and what happened after it in actual markets. Our test bench scanned full price history on stocks, ETFs, indices, commodities, crypto and forex. It found only 20 bullish and 21 bearish occurrences on daily charts, and compared each one with random dates on the same assets.

What is the abandoned baby candlestick pattern?

This candlestick pattern is a three-bar reversal signal. Steve Nison, who introduced Japanese candlestick charting to Western traders in Japanese Candlestick Charting Techniques (1991), presents it as a rare form of the doji star: the middle candle is a doji whose shadows do not overlap the shadows of the candles on either side. In Western technical analysis, the closest equivalent is the island reversal, where a group of bars is isolated by a gap on each side.

On a chart, the pattern looks like a lone doji floating between a long candle and a candle in the opposite direction. There are two versions. The bullish abandoned baby appears after a downtrend and signals a potential upward reversal. The bearish abandoned baby appears after an uptrend and signals a potential downward move. Both say the same thing in technical analysis terms: the market moved hard in one direction, stalled, and then moved hard in the other direction, leaving the middle session stranded. This does not mean the trend must turn; it means momentum has stalled and a potential trend reversal is possible.

Our bench follows the rule of the open-source TA-Lib library (CDLABANDONEDBABY), without its minimum size for the third candle: a long first candle, a doji separated from it by a full price gap, and a third candle separated from the doji by a full price gap in the other direction, closing at least 30% of the way into the body of the first candle.

How to identify the abandoned baby pattern

Check four things in order: the trend before the pattern, the first candle, the middle candle and its two windows (the Japanese term for a gap), and the close of the third candle.

Structure of a bullish abandoned baby

• Trend: prices have been falling. Our rule requires a downtrend before the first candle.
• Candle 1: a long black candle, in the direction of the decline.
• Candle 2: a doji, with an opening and closing price nearly equal, whose high is below the low of candle 1. The doji indicates indecision after the decline.
• Candle 3: a white candle, of any size, whose low is above the high of the doji, closing at least 30% into the body of candle 1.

The figure below is not drawn by hand. It is the median shape of the occurrences our bench detected, with the faded candles showing the decline before the pattern and the levels we test: entry above the pattern high, stop-loss under the doji low, target at twice the risk.

Structure of a bearish abandoned baby

• Trend: prices have been rising before the first candle.
• Candle 1: a long white candle that extends the uptrend.
• Candle 2: a doji whose low is above the high of candle 1, with an opening price above the previous close.
• Candle 3: a black candle, of any size, whose high is below the low of the doji, closing at least 30% into the body of candle 1.

The bearish abandoned baby pattern is the mirror image. Here the test sells below the pattern low, with the stop-loss above the doji high.

Size of the pattern

On the median daily occurrence, the three candles span 5.51% of price for the bullish version and 5.30% for the bearish one, about two times the average true range (ATR) of the asset (2.17 and 2.60 ATR). The formation is large, so the stop is far from the entry.

Why it is so rare, and absent from crypto and forex

Two full gaps in a row are unusual for any candlestick pattern. On daily charts, our bench found one bullish pattern every 382,474 eligible bars and one bearish every 367,195. For comparison, the morning star, which only needs the star to clear the body, appears once every 2,592 daily bars. Most occurrences come from the stock market: 15 of the 20 bullish and 14 of the 21 bearish daily cases are US stocks, where each session opens after a night of news.

Crypto and forex are continuous markets. Bitcoin trades around the clock, currency pairs trade 24 hours on weekdays, and a new candle almost always opens where the previous one closed. Without a jump between sessions, the middle candle cannot be isolated. With its strict definition, our bench found no occurrence at all in forex market data, and none on crypto charts either, whether daily, 4-hour or hourly, on hundreds of coins. If a forex or crypto chart shows one, check whether the gap is real or an artifact of a low-volume session or a data feed.

Key differences between abandoned baby and morning/evening star patterns

This candlestick pattern belongs to the same family as the morning star and the evening star. The difference is how far the middle candle is separated.

• Morning and evening star: the middle candle (the star) has a small body that gaps beyond the body of the first candle. Shadows can overlap.
• Doji star: the star is a doji, but only two candles are needed, and the shadows can still overlap.
• Abandoned baby: the star is a doji and nothing overlaps, neither bodies nor shadows, on both sides.

Every occurrence is therefore also a morning or evening doji star, but very few doji stars are abandoned babies. The island reversal follows the same logic with one or more bars of any shape isolated by two windows. On our data, the more demanding definition leaves too few cases to measure any advantage: see the reliability section below.

Why this pattern signals a trend reversal

To understand the signal, follow the price action over three sessions of the bullish version. Sellers are in control and push prices downward on a long black candle, a sign that the decline continues. The next session opens lower again, but the selling stalls: the doji shows buyers and sellers in balance for a moment, far below the prior day. Then the third session opens above the baby and rallies. In the classic reading, traders who sold on the gap down are now losing, and their buying to cover adds to the upward move. The shift in market sentiment is visible on the chart itself, indicating that the other side has taken control.

The bearish version tells the opposite story, with a downward shift in sentiment: buyers push a strong white candle, the next session opens higher but goes nowhere, then the market opens lower and sellers take over, leaving late buyers trapped above.

That reading is plausible, and it is why the rule asks the third candle to close well inside the first body. But it describes three sessions; whether they change the trend is a question for data. The reliability section shows what happened next.

How to trade the abandoned baby pattern

The classic approach for the bullish setup is to buy when prices trade above the high of the pattern, with a stop-loss under the low of the doji and a target set as a multiple of the risk. For the bearish setup, sell below the pattern low, with the stop above the doji high. This is what our bench simulates on every occurrence, with a target at twice the risk (2:1).

Conservative entry strategy

Two entry options exist. The conservative trader places a buy stop just above the highest point of the three candles and waits for prices to trade through it. This is the version we measured. On daily charts, the bullish entry was triggered on 85.0% of occurrences; in 15.0% of cases, the doji low broke first and the setup was canceled before any trade; a setup that touches neither level within the 20 bars the order stays active, or before the data end, expires, and none did here. For the bearish version, 76.2% triggered and 23.8% were canceled.

Aggressive entry strategy

The aggressive trader chooses to enter at the close of the third candle, as soon as possible once the pattern is complete. The entry is closer to the stop and the position is larger for the same risk, but there is no confirmation: you are in the trade even in the cases where the conservative order would never have filled. We have not tested this version separately.

Profit targets

A target at twice the risk is the version we test. Because the pattern is about two ATR wide, that target sits far away. On daily charts, only 7 bullish trades reached it and 10 were stopped, too few to give a useful median time. Some traders take profit at the last high or low before the pattern instead, based on the previous structure of the chart.

Using technical indicators for confirmation

Can a technical indicator confirm the signal? Here is what we tested and what we did not.

Combining with volume analysis

One rule wants high trading volume on the third candle, as a sign that the new side is committed, with the next candle expected to confirm the move. Our data cannot test it: with only 20 bullish and 21 bearish daily occurrences, the bench does not split them by volume, since each group would hold a handful of cases.

Using relative strength index (RSI) for verification

Another filter wants the relative strength index (RSI) below 30 (oversold) before a bullish setup, or above 70 (overbought) before a bearish one, and some add a MACD crossover that confirms the signal. We have not tested an RSI or MACD filter on this pattern. With so few occurrences, any filter shrinks the sample further: test it on your own market before relying on it.

Moving average trend filter

We split daily occurrences by the position of prices relative to the 200-day moving average. With so few occurrences, the bench does not report this split for the abandoned baby: each group would be far below the 30 cases needed to read a rate.

Risk management and stop-loss placement

• Size the position on the distance from entry to stop. The pattern is about two ATR wide, so a standard position size would risk too much.
• Expect gaps against you. A market that jumped twice in three sessions can jump again, and a stop order fills at the next available price, not at your level: a loss can then exceed 1 R.
• Plan the management. A trade can move well in your favor and still come back to the stop, as the natural gas ETF (UNG) case in the failed setups below shows.
• Keep the risk per trade small: one pattern is not a reason to concentrate a trading account, and a written trading plan helps you stick to it.

Reliability and success rates

The abandoned baby candlestick pattern is too rare to prove anything on its own. With only 20 bullish and 21 bearish daily occurrences, every row, US stocks included, is below the 30 cases our bench needs before reading a result: the numbers below are shown, not interpreted. Here are the numbers from our test bench, before fees and other trading costs.

Read the table in pairs of columns to understand what the numbers indicate. "Moved as expected" means followed by a higher price 10 sessions later for the bullish version and lower for the bearish one. 60.0% of bullish occurrences were higher 10 days later, against 55.1% of random days on the same assets: random days rise more often than they fall, so the comparison matters more than the raw figure. The average excess return over the rest of the market after 10 bars was −1.24%, with a 95% confidence interval from −5.05% to +1.70%: a range this wide, on so few cases, says nothing either way.

The 2:1 setup is just as thin. The bullish setup reached its target in 41.2% of 17 trades, against 30.4% for the same entry, stop and target on random dates, for an expectancy of +0.18 R per trade before costs. The bearish setup reached its target in 50.0% of 16 trades, against 26.0% on random dates (+0.57 R against −0.11 R). On fewer than 30 trades, a few lucky or unlucky cases are enough to move these figures a long way: they are shown, not interpreted.

Two more warnings. On so few cases, the result also moves over time: the bullish excess return after 10 bars was −2.25% in the first half of the history and −0.23% in the second (a descriptive split, not in the tables). And every market, US stocks included, produced fewer than 30 occurrences: the statistics section lists them all as "Not enough data", a sign of how rare the pattern is. The full table, with every market and timeframe, is in that section, and the method is in our methodology.

These results agree with academic work on candlesticks: Marshall, Young and Rose, in "Candlestick technical trading strategies: Can they create value for investors?" (Journal of Banking & Finance, 2006), found that candlestick strategies did not add value on large US stocks.

Real examples on daily charts

The galleries in this section show real daily cases, picked by our bench (best and worst results, textbook shapes, most recent, random draws), with their actual outcome.

A bullish setup that reached its target

On July 13, 2022, M&T Bank (MTB) printed a black candle. The next session was a doji whose high (131.90) stayed below the prior low (133.19). On July 15, the third candle opened above the doji, at 133.54, and closed at 136.38. On July 18, the share price opened above the pattern high, so the buy stop filled at 138.63, with a stop-loss at 129.91 under the doji and a target at 156.07. The target was reached on August 5, 2022, 14 sessions after entry, and MTB was 13.6% higher 10 sessions after the pattern.

A bearish setup that reached its target

On December 14, 2023, Ready Capital (RC) printed a doji whose low (8.66) sat above the high of the prior white candle (8.56). The next day the stock gapped down and closed at 8.26. The sell stop was triggered on December 18, with a stop at 8.85 and a target at 6.69, reached on February 28, 2024, 48 sessions after entry.

Failed setups

On December 8, 2025, the US natural gas ETF (UNG) completed a bearish pattern: a white candle that ended at 15.66, a doji at 16.37 whose low (16.30) sat above it, then a black candle that closed at 15.07. On December 9 the fund opened at 14.60, below the pattern low, so the sell stop filled there, with a stop at 17.00 above the doji and a target at 9.80. By January 15, 2026, the price had fallen as low as 9.95, about 1.9 R in favor and just short of the target, and it was 21.4% lower 10 sessions after the pattern, a very large move for this fund; it then rallied and hit the stop on January 30: a full 1 R loss.

In April 1985, the FTSE 100 printed a bearish pattern: a white session to 1,304.0, a doji whose low (1,305.4) stayed above it, then a black session that closed at 1,299.7. The index opened at 1,287.2 on April 23, below the 1,290.9 pattern low, so the short filled at that open, with a stop at 1,313.0. The index dipped only to 1,282.7, was 0.9% higher 10 sessions after the pattern, and hit the stop on May 7: a full 1 R loss. And on July 19, 2021, a bearish setup on Otter Tail (OTTR) was canceled before any trade: the share price never broke the pattern low and traded above the doji high on July 29.

Limitations and common pitfalls to avoid

• Calling a morning star an abandoned baby. If a shadow of the doji overlaps a shadow of its neighbors, it is a doji star.
• Seeing the pattern on 24-hour markets. On crypto and forex charts, an apparent window can come from a thin session or a data problem.
• Trusting the rarity. Among candlestick patterns, a rare signal feels special, but rare also means too few cases to measure: our bench cannot rate either version.
• Ignoring the size. The setup spans about two ATR: a position sized like any other risks twice as much.
• Forgetting gap risk. A stop does not protect against a market that opens beyond it.

Pros and cons

The pros: a strict, objective definition of the candlestick pattern, easy to code and to check; a clear invalidation point beyond the doji; and the strictest definition in the doji star family. The cons: you will rarely see one, the stop is wide, and on our data both versions are too rare for the bench to measure their result per trade.

Tools to find the pattern

Stock screeners and charting platforms can flag candlestick patterns, and the open-source TA-Lib library provides the CDLABANDONEDBABY function our rule is based on, with a default penetration of 30% for the third candle. Check which definition your tool uses: some only require a gap between bodies, which counts some morning stars as abandoned babies and changes the results. A scanner across a large list of stocks is the only practical way to find this pattern: on US stocks, our bench found one bullish occurrence every 487,954 daily bars, more bars than the full daily history of any single stock in our data. Intraday charts did not help on our bench either: hourly crypto charts gave no occurrence at all, and we have no intraday stock data to test.

Frequently asked questions

Summary of key points

• This candlestick pattern is a three-bar reversal signal: a long trend candle, an isolated doji, then a candle in the other direction, with nothing overlapping on either side of the doji.
• It is very rare: one occurrence every 382,474 daily bars for the bullish version, against one every 2,592 for the morning star.
• It is absent from our crypto and forex data: these markets trade continuously and rarely gap.
• On 17 closed trades on bullish daily occurrences, the 2:1 trade reached its target 41.2% of the time against 30.4% on random dates, but fewer than 30 trades are too few to read a result.
• The bearish version is just as rare: 21 daily occurrences and 16 closed trades, too few to measure.

Final thoughts on risk management

Treat this candlestick pattern as a good place to define risk, not as proof that the trend has turned, and compare it with other reversal patterns you trade. Put the stop beyond the doji, size the position on that distance, allow for overnight jumps, and keep the risk per trade small. Trading involves risk, and nothing here is investment advice.

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 fixed, 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.