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

Three Stars in the South Pattern: A Rare Three-Candle Bullish Reversal

Three stars in the south pattern: the Milky Way and Southern Cross region above an observatory dome at night

Photo: Under the Southern Cross by ESO/B. Tafreshi (twanight.org), CC BY 4.0 via Wikimedia Commons

By Dalal Street Blogs Editorial9 min read

Most three-candle bullish reversals that Indian traders look for are the familiar ones: a morning star, maybe a three inside up. The three stars in the south pattern sits in the same family and almost never shows up. When it does, the textbooks call it a bullish reversal. The catch is that the data behind that claim is painfully thin, and the move after the pattern often goes almost nowhere.

Thomas Bulkowski, who catalogued 103 candlestick types in his work on thepatternsite.com, found just nine examples of this setup in the entire sample he studied. That is not a typo. Nine. So before we start hunting for it on Nifty or mid-cap charts, we should be honest about what the pattern actually is, how to identify it, and what those nine cases can and cannot tell us.

In this article
  1. What the Three Stars in the South Pattern Looks Like
  2. What the Performance Numbers Actually Say
  3. A Worked Example on an Illustrative Daily Chart
  4. How Traders Usually Wait for Confirmation
  5. Where This Pattern Tends to Matter, and Where It Does Not

What the Three Stars in the South Pattern Looks Like

The identification rules, as Bulkowski sets them out, are strict. There are three candle lines, and they have to appear after a downward price trend. Day one is a tall black candle with a long lower shadow. In other words, sellers were in charge for most of the session, but buyers managed a meaningful bounce from the low before the close. The body stays black, so the session still finished below the open. The long lower wick is the first hint that selling pressure is starting to meet resistance.

Day two should look similar to day one, only smaller, and it must print a higher low. The second candle’s lower shadow has to bottom above the first candle’s low. That higher low is not a side note. It is the rule that shows selling is not pushing as far as it did the previous day.

Day three is a black marubozu that sits inside the high-low range of day two. A marubozu, as StockCharts’ candlestick dictionary puts it, is a candlestick with no shadow extending from the body at the open, close, or both. The Japanese word is often translated as close-cropped or shaved. In this pattern, that third candle is still black, so it closed below its open, but it did so without poking outside the previous day’s range. The tight range is the point: selling is still there, but it has been squeezed into a narrower box.

Bulkowski’s own sign-off on the identification section is “Good luck finding one.” That is not a joke. The combination of a long lower shadow, a higher low on a smaller black candle, and then a black marubozu fully inside the prior range is rare on daily Indian stock charts, and rarer still if you insist on a clean prior downtrend.

A pattern can stop a fall and still fail to start a rally.

What the Performance Numbers Actually Say

On paper, this is a bullish reversal. In Bulkowski’s tests it also behaved as a bullish reversal 86% of the time. He calls that the best reversal rate among the 103 candle types he studied. Read that sentence twice, because the next one undercuts it. Those percentages rest on nine samples, which he notes is 19,991 fewer than he likes to see. He is unusually blunt: anything he says about this pattern is likely wrong because the sample is too small.

The overall performance rank, which measures the move ten days after the breakout across market conditions and breakout directions, comes in last: 103rd out of 103. The best average move ten days after the breakout in his tables is a drop of 3.64% in a bull market after a downward breakout, which ranks 23rd. He treats a move of 6% or higher in ten days as a good one, so even the “best” case here is roughly half that bar. The best percentage of cases meeting the price target is 50%, and that figure is for a bear market with an upward breakout. Frequency rank is 99, which simply confirms how scarce the pattern is.

There is a useful tension in those numbers. A high reversal rate and a dead-last performance rank can sit together. Price can stop going down and still refuse to go up by much. That is exactly what the data, thin as it is, is telling us.

His worked example on the site is also a warning, not a textbook print. The third day in that example is a four-price doji rather than a clean black marubozu, and the breakout is downward: price closes below the bottom of the pattern and the setup acts as a continuation, not a reversal. With only nine cases to choose from, even the illustration is imperfect.

A Worked Example on an Illustrative Daily Chart

Suppose we are looking at a mid-cap stock that has been slipping for two weeks. We will use round numbers so the arithmetic is easy to check. These prices are illustrative; they are not a live quote.

On Monday the stock opens at ₹250, which is also the day’s high, slides to ₹236, and closes at ₹242. That is a tall black candle (close below open) with an ₹8 body (250 minus 242) and a ₹6 lower shadow below the close. Tuesday opens at ₹244, which is its high, dips to ₹239, and closes at ₹241. The body is smaller at ₹3, the candle is still black, and the low of 239 is above Monday’s low of 236, so the higher-low rule is met. Wednesday opens at ₹241.50 and closes at ₹240, with the session’s high and low both inside Tuesday’s 244–239 range, and with little or no wick. That is our black marubozu squeezed inside the prior day.

The pattern’s high is ₹250 and its low is ₹236, a 14-rupee range. A common measured-move idea, used only as a yardstick, is to add that range to the top of the pattern: 250 + 14 = ₹264. In Bulkowski’s tables, the best reading on meeting a price target is 50%, and that is in a specific market and breakout combination, so we should treat ₹264 as a reference, not a forecast.

Now the practical part. If Thursday opens and then closes at ₹237, below the ₹236 low of the pattern, the breakout is downward, just as in Bulkowski’s example. The three-candle structure has not failed the identification rules; the follow-through has simply gone the other way. If, instead, Thursday closes at ₹252, above the pattern high of ₹250, we have an upward breakout and a candidate long only after that close, not during Wednesday’s tight black candle.

A high reversal rate on nine samples is a clue, not a trading system.

How Traders Usually Wait for Confirmation

Nothing in the identification rules says you have to buy the close of day three. Given how often the post-breakout move is modest, and given that even the documented example broke downward, waiting for a close above the high of the three-day range is the conservative read. On our worked example, that means a close above ₹250, not a hope that Wednesday’s black marubozu marks the exact bottom.

Volume is not part of Bulkowski’s formal identification list for this pattern. Some traders still like to see selling volume shrink on days two and three, because that would fit the story of selling pressure being squeezed. That is an overlay, not a rule in the source we are using, so it should be treated as optional context.

A stop is often parked just below the pattern low — ₹236 in the example, or a tick or two under it to allow for a wick. Risk per share is then roughly 250 minus 236, or ₹14, if you enter only after a close above the high. Position size follows from that distance, not from the 86% reversal headline. If a ₹14 risk is too wide for the account, the trade is simply too large, however rare and “bullish” the candles look.

It also helps to remember what this pattern is not. It is not a morning star, which StockCharts describes as a long-bodied black candle, a short middle candle that gaps down, and a long-bodied white candle that gaps up and closes above the midpoint of the first day’s body. It is not a hammer, which StockCharts describes as a single candle that falls well below the open and then rallies to close well above the intraday low, during a decline. And a black marubozu on its own is just a session with little or no wick; the three stars in the south pattern only appears when that shaved black candle is the third day, sitting inside day two, after a downtrend. If you want the broader toolkit around these candles, our guide to technical analysis for Indian stocks is a better starting point than hunting for a nine-sample rarity.

Where This Pattern Tends to Matter, and Where It Does Not

The further-reading notes on Bulkowski’s page point to his wider studies: candles near the yearly low perform best, and tall candles appear at turning points. Those are general observations, not statistics for this pattern. Still, if you ever do find a clean print, it is more interesting after a several-week decline, close to an obvious support zone, than halfway through a messy range.

On liquid names — Nifty 50 constituents, heavyweight Bank Nifty stocks — a perfect three-day print is uncommon, which is consistent with a frequency rank of 99. On thinner mid-caps and small-caps, odd three-day sequences appear more often, but so do gaps, circuit limits and wide spreads, which can invalidate a “marubozu inside the prior range” read by the time the auction is over. Intraday charts will throw up lookalikes; most of them fail the higher-low rule or print a wick on day three. We would not force the label.

The honest use of the three stars in the south pattern is as a rare warning that a decline may be losing depth, not as a standalone buy signal. The 86% reversal rate is the best in Bulkowski’s set, and he says so. He also ranks the follow-through last, warns that nine samples are not enough, and shows an example that continued down. That combination should slow us down, not speed us up.

If you do get a clean identification after a downtrend, mark the pattern high and low, wait for a close beyond one of those levels, and size the trade off the opposite end of the range. Everything else — the name, the rarity, the headline reversal rate — is commentary.

Sources: Thomas Bulkowski, “Three Stars in the South” (thepatternsite.com); StockCharts ChartSchool, Candlestick Pattern Dictionary (Marubozu definition). The worked example prices are illustrative assumptions, not market data.

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. Candlestick patterns are subjective and do not guarantee future results.

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