Dalal Street Blogs provides information for educational purposes only and does not constitute investment advice. Please consult a SEBI-registered investment adviser before making any investment decisions.
CAUTION: Dalal Street Blogs publishes market news and educational content only. We do not promote, recommend or endorse any stock, and nothing here is a solicitation to buy or sell securities. Please consult a SEBI-registered adviser before investing.
Technical Analysis

Concealing Baby Swallow: The Four-Candle “Bottom” That Usually Keeps Falling

Concealing baby swallow: a barn swallow feeding its chick in Chitlang, Nepal

Photo: Barn swallow feeding its chick at Chitlang, Hetauda, Nepal by Supradeepta Neupane (Supradai), CC BY 4.0 via Wikimedia Commons

By Dalal Street Blogs Editorial9 min read

The concealing baby swallow has the best name in the candlestick dictionary and one of the worst track records. It is a four-candle pattern that shows up after a decline, made entirely of black (down) candles, and candle theory says it should mark a bullish reversal. Then you look at the testing. In more than 4.7 million candle lines, Thomas Bulkowski found just four of them, and three of those four kept falling. So this is a pattern worth understanding mostly so you know when to ignore it.

We’ll go through how it’s built, how software like TA-Lib actually detects it, what the tiny sample says, and a worked example with made-up prices so you can see the levels that matter.

In this article
  1. How the concealing baby swallow pattern is built
  2. How TA-Lib detects it, rule by rule
  3. What the testing says, and why the sample is the real story
  4. A worked example with illustrative prices
  5. How we’d actually use it
  6. Common mistakes with this pattern
  7. Key takeaways

How the concealing baby swallow pattern is built

Bulkowski’s concealing baby swallow page lays out the identification guidelines in three lines:

  • Number of candle lines: four.
  • Price trend leading to the pattern: downward.
  • Configuration: four black candles. The first two are long black marubozu candles. The third gaps open lower but trades up into the body of the second day, leaving a tall upper shadow. The fourth candle engulfs the third, shadows included, so it has a higher high and a lower low.

Walk through it as a trader would see it on a daily chart. Day one and day two are heavy, one-way selling days. A marubozu has little or no shadow at either end, which means the stock opened near its high and closed near its low. Two of those back to back tell you sellers controlled the whole of both sessions.

Day three is where things get interesting. The stock opens with a gap below the previous close, then buyers push it up into the previous day’s body. They can’t hold it. The day still closes in the red, but that long upper shadow is the first sign of someone stepping in.

Day four is the “swallow”. It trades above day three’s high, then sells off to a low beneath day three’s low and still closes below its own open. The big black candle swallows the smaller one, the “baby”, whole. In theory, that’s the reversal signal: a final round of selling after buyers have already shown up once.

Theory calls it a bottom. In Bulkowski’s small sample, the selling usually just kept going.

The difference from a plain engulfing candle

Plenty of patterns have one candle engulfing another. What sets this one apart is the setup. You need two clean black marubozu candles first, a gap-down third day that still reaches back into the prior body, and every one of the four candles has to close lower than it opened. If the last candle is white, you’re looking at something else. Bulkowski points out that the bearish three-line strike looks similar but ends in a white candle. We’ve covered that pattern in our three line strike guide.

How TA-Lib detects it, rule by rule

Most retail scanners and many Python screeners lean on the open-source TA-Lib library, and its CDLCONCEALBABYSWALL function is worth reading because it turns the loose description into hard tests. Its “must have” list reads:

  • First candle: black marubozu (very short shadows).
  • Second candle: black marubozu (very short shadows).
  • Third candle: black candle that opens gapping down but has an upper shadow that extends into the prior body.
  • Fourth candle: black candle that completely engulfs the third candle, including the shadows.

In the code, those translate into specific comparisons:

  • All four candles must close below their open.
  • On candles one and two, both the upper and lower shadows must be “very short”.
  • Candle three’s open and close must both sit below candle two’s open and close. That is how the code checks the gap down.
  • Candle three’s upper shadow must be longer than the “very short” threshold, and its high must be above candle two’s close, so it reaches into the prior body.
  • Candle four’s high must be above candle three’s high, and its low must be below candle three’s low.

What counts as “very short”? TA-Lib’s default candle settings define a shadow as very short when it is shorter than 10% of the average high-low range of the previous 10 candles. So if a stock’s last 10 sessions averaged a ₹20 range, a shadow under ₹2 passes.

Two things the code does not do. It does not look at the trend before the pattern, so a hit in a sideways stock still gets flagged. And when it finds the pattern, it returns +100, the positive value TA-Lib uses for bullish signals. That number reflects candle theory, not test results. If your screener colours it green, don’t read that as a buy signal.

What the testing says, and why the sample is the real story

Here is Bulkowski’s summary from the same page. All ranks are out of 103 candlestick patterns, with 1 the best:

  • Theoretical performance: bullish reversal.
  • Tested performance: bearish continuation 75% of the time.
  • Frequency rank: 103, dead last.
  • Overall performance rank: 101.
  • Best percentage meeting price target: 67%, in a bull market with a downward breakout.
  • Best average move in 10 days: a fall of 7.10%, also bull market and downward breakout.
  • Best 10-day performance rank: 3.

Now put those numbers next to the sample size. Four patterns. A 75% continuation rate on four samples means three went down and one went up. That’s not a statistic you can build a strategy on, and Bulkowski says as much. He left the pattern’s statistics out of his Encyclopedia of Candlestick Charts because four wasn’t enough, and his advice is that this is “probably one you can ignore”, since you’ll see only a handful in your lifetime.

The ranks also need context. The 101 overall rank looks awful, but he explains why: his overall measure averages four categories (bull and bear markets, up and down breakouts), and he only found concealing baby swallows in two of them, so zeros were filled in for the rest. Meanwhile the 7.10% average 10-day drop is genuinely strong. He considers anything above 6% good, and that move ranks 3rd out of 103. One bad number and one great number, both coming from a sample too small to trust.

Four examples can’t tell you whether a pattern works. They can only tell you it’s rare.

What we take from this: the concealing baby swallow is not a reliable bottom signal. If anything, the little evidence there is points the other way. Its sister patterns built from heavy black candles, like three black crows, at least have enough samples to discuss.

A worked example with illustrative prices

The prices below are made up for teaching. They don’t describe any real NSE or BSE stock.

Say a mid-cap stock has been sliding for a few weeks. Its average daily high-low range over the last 10 sessions is ₹20, so TA-Lib’s very-short-shadow limit is ₹2 (10% of ₹20). Then come four sessions:

DayOpen (₹)High (₹)Low (₹)Close (₹)What it shows
1500501479480Black marubozu, shadows of ₹1 each
2479480459.50460Black marubozu, shadows of ₹1 and ₹0.50
3455466448450Gap-down open, ₹11 upper shadow reaching above ₹460
4464468440442Black candle, high above ₹466 and low below ₹448

Check each rule. Days 1 and 2 have shadows under ₹2, so both are marubozu by TA-Lib’s default. Day 3 opens at ₹455 and closes at ₹450, both below day 2’s open of ₹479 and close of ₹460, which is the gap. Its high of ₹466 is above day 2’s close of ₹460, so it reaches into the prior body, and its upper shadow of ₹11 (₹466 minus the ₹455 open) is well over the ₹2 limit. Day 4 trades from ₹468 down to ₹440, taking out both ends of day 3’s ₹448 to ₹466 range. All four closes are below their opens. It qualifies.

Where the breakout levels sit

Bulkowski counts a breakout when price closes beyond the pattern’s range. On his three-line strike page he spells out that an upward breakout is a close above the top of the candle pattern, and his price target is the pattern’s height projected up from the top or down from the bottom.

  • Pattern high: ₹501 (day 1). Pattern low: ₹440 (day 4).
  • Height: ₹501 − ₹440 = ₹61.
  • Upward breakout: a close above ₹501, which is about 13.3% above the ₹442 close. Upside target: ₹501 + ₹61 = ₹562.
  • Downward breakout: a close below ₹440. Downside target: ₹440 − ₹61 = ₹379, a drop of about 13.9% from ₹440.

Notice the asymmetry. The stock closed at ₹442, just ₹2 above the bottom of the pattern and ₹59 below the top. A downward breakout is one bad session away. An upward one needs a rally of more than 13% just to confirm. That geometry alone explains part of why so many of these resolve lower.

Also look at the target. A ₹61 projection is big for a stock that moves ₹20 a day. For comparison, Bulkowski’s best average 10-day move for this pattern is a 7.10% fall, which from ₹440 would take the stock to roughly ₹409, nowhere near ₹379. Bulkowski notes that tall patterns have a harder time reaching their targets than short ones, so a tall four-candle pattern like this one should come with a modest target if you trade it at all.

How we’d actually use it

Given the evidence, treat a concealing baby swallow as context, not a trigger.

If you hold the stock: don’t take the pattern as a reason to average down. Two marubozu days and a failed rally attempt on day three show the selling hasn’t finished. A close below the day four low is a reasonable point to review your stop.

If you’re waiting to buy: wait for an upward breakout, a close above the top of the pattern, before treating the decline as over. In our example that’s a close above ₹501. Until then, the “bottom” is only a theory.

If you trade short in F&O: a close below the pattern low lines up with the direction three of Bulkowski’s four samples took, but four samples is four samples. Put the stop above the pattern high or at least above day four’s high, size the position for that distance, and use a target smaller than the full height.

And if you scan for it, expect silence. A daily CDLCONCEALBABYSWALL scan can sit empty for very long stretches. That’s not a bug in your code. It’s what a frequency rank of 103 out of 103 looks like.

Common mistakes with this pattern

Trusting the bullish label. Both the theory and TA-Lib’s +100 output call it bullish. The test results don’t back that up.

Ignoring the trend. TA-Lib won’t check it, so you have to. Four black candles in a range-bound stock aren’t a concealing baby swallow in Bulkowski’s sense.

Loosening the marubozu rule. If you let candles with real shadows through, you’ll start “finding” the pattern more often, and you’ll be trading something with even less evidence behind it.

Reading too much into four samples. The 67% target hit rate and the 7.10% average fall sound precise. They come from a handful of trades. Keep your own log if you trade it, and judge it on your results.

Key takeaways

  • A concealing baby swallow is four black candles after a decline: two black marubozu, a gap-down candle with a tall upper shadow reaching into the prior body, then a black candle that engulfs it, shadows included.
  • TA-Lib’s CDLCONCEALBABYSWALL treats a shadow as very short when it’s under 10% of the 10-candle average range, returns +100 when it finds the pattern, and doesn’t check the prior trend.
  • Bulkowski found only four in more than 4.7 million candle lines. Three acted as bearish continuations, and he calls it a pattern you can probably ignore.
  • If you do act on it, wait for a close beyond the pattern’s high or low, and keep the target modest.

For the bigger picture on reading candles, start with our candlestick patterns for beginners guide.

Sources: Thomas Bulkowski, Concealing Baby Swallow and Bearish Three-Line Strike (ThePatternSite.com); TA-Lib CDLCONCEALBABYSWALL source and default candle settings. All example prices are illustrative assumptions.

Disclaimer: This article is for information and education only and is not investment advice. Please do your own research or consult a SEBI-registered investment adviser before making any investment decision.

Leave a Reply

Join the conversation

Commenting is open to registered readers. Sign up with just your email — no password needed, and it takes under a minute.

Sign up to comment Log in