Morning star candlestick pattern: rules, examples and test results

The morning star is a three-candle bottom pattern. A long red candle extends a decline, a small-bodied candle opens below it and goes nowhere, and then a strong green candle drives back up into the first candle's territory. Read left to right, the three sessions show sellers in control, sellers running out of conviction, and buyers taking over.

The name comes from the Japanese charting tradition that Steve Nison introduced to Western traders in 1991: the morning star is the planet Venus, which appears just before sunrise. The small middle candle is the "star", and the green candle after it is the dawn. Its mirror image at market tops is the Evening star candlestick: how to spot a three-day top.

This page explains how to recognize the pattern, the precise test TA-Lib applies, an important quirk in how Finabase runs that test, two real examples, and the results of a two-year test on 60 large US stocks and ETFs. The short version of the test: the shape alone did not help.

Anatomy of a morning star

Each of the three candles has a job:

1. A long red candle. The decline is still in force, and this session confirms it with a close well below the open. "Long" means long for this stock: a body bigger than its recent typical session.
2. The star. The next session's real body (the box between open and close) is small and sits entirely below the red body. Its color does not matter. A small body after a big one says the selling pressure lost its push: the market opened lower and could not travel far in either direction.
3. A long green candle. Buyers take over and close the session well inside the first candle's body. The deeper that close, the more of the first day's damage has been repaired. Nison's guideline is a close at least halfway into the first body.

A gap between the star and the third candle strengthens the picture for most authors, because it isolates the star at the low. When both gaps are extreme, so that even the shadows don't touch, the formation becomes an Abandoned baby candlestick: the rare double-gap reversal. When the star is a doji, with open and close almost equal, it is a Morning doji star: the doji version of the bullish star pattern.

The exact rules Finabase uses

Finabase detects candlestick patterns with TA-Lib, the open-source technical analysis library. Its CDLMORNINGSTAR function measures each candle against the stock's own recent candles rather than fixed percentages. In plain English, it returns +100 on the third candle when all of these hold:

Where TA-Lib differs from the textbook:

• No trend test. The function's own source comment says the user should judge whether the pattern appears in a downtrend; the code never checks. In our data only 37.1% of morning stars came after a 10-session decline.
• A loose definition of "small". The star only has to be no bigger than an average body. Some flagged stars are ordinary-looking candles.
• No second gap. TA-Lib requires the star's body to gap below the first body, but not the third candle to gap above the star.
• A 30% close, not 50%. At the default, the third candle has to recover 30% of the first body, a lower bar than Nison's halfway mark.

Why Finabase's markers never show a morning star

The last rule is where Finabase's configuration matters. TA-Lib's optInPenetration parameter is a fraction of the first candle's body. The value 0.3 means "close at least 30% of the way back up into the first body".

Finabase passes optInPenetration: 10 to this function and to six others. For five of them (the evening star, both doji stars, the abandoned baby and dark cloud cover) the parameter is a minimum depth, so 10 makes them just as unreachable as this one. Mat hold is the exception: there the parameter is a maximum, so 10 only loosens it. The value is inherited from the legacy Stock Master app and kept verbatim so the new platform computes exactly what the old one did. Read as a fraction, 10 means 1,000%: the third candle must close ten first-body lengths above the first candle's close.

Put numbers on it. Say candle 1 opens at $105 and closes at $100, a $5 body. At the default, candle 3 must close above $101.50. At Finabase's setting it must close above $150, a 50% gain in one session from a stock trading near $100. That essentially never happens: across the 29,760 daily bars in our study, the setting produced zero morning stars.

So, to be explicit: Finabase's chart markers, scanner and Signals-tab gauge will not show this pattern. Everything on this page about real occurrences comes from re-running the same detector at TA-Lib's default of 0.3. The "Try it in Finabase" chart link above does exactly that, because the chart accepts a penetration value in its study link.

The psychology behind the shape

Think of the three sessions as a short negotiation.

On day one, sellers dominate. Price opens and falls steadily into the close, and anyone who bought that day is underwater. The mood is that the decline has further to run.

On day two, the market opens lower again, below the whole of day one's body, which looks like more of the same. But the session stalls. Neither side can move price far from the open, and the day ends with a small body. A falling market that stops falling, even for a day, is new information.

On day three, buyers show up in size. Price rises all session and closes well into the first day's range. Traders who sold short on day one are now losing money and may buy to cover, adding fuel. The pattern's logic is that the balance of power changed within three sessions, and the third candle is the evidence.

The weak link is the second day. A pause can be a turning point, or it can be the market catching its breath before the next leg down. The third candle tells you which story buyers are betting on, not which one will win.

Confirmation

Because the third candle already does a lot of work, many traders treat a morning star as more self-confirming than a single-candle pattern like the Hammer candlestick pattern: meaning, rules and reliability. Others still want one more piece of evidence:

• A close above the third candle's high in the next session or two. This is the rule our test uses.
• Volume on the third candle well above recent average, a sign the buying was broad.
• Location at a level that already mattered: prior support, a long-term moving average, the bottom of a trading range.
• Momentum turning: an oscillator such as the Relative Strength Index (RSI): how to read and use it rising out of oversold territory or showing a bullish divergence.

As the data below shows, in our sample confirmation lifted the hit rate from 42.4% to 45.9%. That is still under the 53.1% of an average day.

Two real morning stars

AT&T, January 14, 2025. AT&T had slipped 4.6% over ten sessions. On January 10 it closed lower at $21.69. On January 13 it opened below that close and finished at $21.56, a body of seven cents. On January 14 it closed at $21.80, above the first candle's close by more than the required 30% of its body. The next session closed at $21.91, above the pattern's high of $21.82, which confirmed it. Five sessions later the stock was 2.4% higher, and after ten it was up 11.1%.

Note how small these candles are. AT&T's daily bodies were often under 1% of its price, so a 19-cent red body counted as "long". TA-Lib's relative definitions make the pattern work on quiet stocks and volatile ones alike, which also means a morning star on a low-volatility name can be a very small event in dollar terms.

Intel, April 3, 2025. Intel had fallen 6.4% in ten sessions. A red candle on April 1, a small star on April 2 and a green candle on April 3 that closed at $22.43 satisfied every rule. The very next session the stock fell 11.5% to $19.85, below the entire pattern, during the market-wide sell-off that followed the US tariff announcement of April 2, 2025. Ten sessions after the pattern, Intel was 15.6% lower.

The lesson is not that the pattern was "wrong". Three candles describe the balance of buyers and sellers in one stock over three days. They cannot anticipate a shock that hits the whole market. A trader with a stop below the star's low would have been out at a defined loss on the first day.

What morning stars did next: the test

We ran TA-Lib's morning star detector, at the default penetration of 0.3, over daily bars from October 30, 2024 to October 29, 2026 for 60 large US stocks and ETFs, 29,760 day-samples in all. For each morning star we measured the return over the next five and ten sessions, and compared it with every day in the sample.

The result is unflattering. In a period when the typical stock in the sample rose more often than it fell, morning stars were followed by a lower close five sessions later more often than not, and the average return was slightly negative. Waiting for confirmation, then measuring from the price you could actually have paid, improved the hit rate a little but not to the level of a random day.

Two details help explain it. First, 63% of the detections did not follow a 10-session decline, because TA-Lib doesn't check for one, so many were "bottoms" in the middle of sideways or rising stretches. Second, the pattern often fires after a sharp three-day round trip, which is exactly the kind of volatile stretch where the next move is hard to call.

Morning star versus similar patterns

The morning star and the hammer tell a similar story at different speeds. A hammer packs the sell-off and the recovery into one session. A morning star spreads them over three, which gives a slower, more visible handover.

An educational outline for trading it

This is how practitioners typically structure the pattern. It is not a recommendation to trade, and the numbers are illustrative.

1. Check the setting. Is there a real decline into the pattern, and is the star at a level that mattered before?
2. Grade the third candle. How deep into the first body did it close? A close past the midpoint is the traditional standard.
3. Decide on confirmation. Either act at the third candle's close or wait for a close above its high.
4. Place the invalidation point. The star's low is the natural line. If price closes below it, the selling that the star seemed to exhaust has resumed.
5. Size from the risk. The distance between entry and stop, times the share count, is what you are prepared to lose.

How Finabase shows the morning star

• Chart markers. The Morning Star study exists in the chart's candlestick list, but at Finabase's penetration of 10 it does not mark anything. The study link at the top of this page sets the penetration to 0.3, so the chart marks morning stars the way TA-Lib's default does.
• Scanner. The candlestick scanner sorts each detection by the sign of TA-Lib's output. A morning star always scores +100, so it would sit on the Bullish list, but with the penetration of 10 that list stays empty for this pattern.
• Signals tab. The Candlestick gauge on a symbol's Signals tab (for example AAPL's) lets each of the 61 patterns vote Buy, Sell or Neutral from its most recent detection in the last eight bars. The morning star always votes Neutral.

Frequently asked questions

How reliable is the morning star?

On our data, not very. At TA-Lib's default setting, 42.4% of 151 measurable morning stars were higher five sessions later, against 53.1% for all days. Treat it as a prompt to look at the chart, not as a signal on its own.

Does the middle candle have to be red or green?

No. The star's color is not part of the definition, in TA-Lib or in most books. What matters is that its body is small and sits below the first candle's body.

What is the difference between a morning star and a morning doji star?

The star. In a Morning doji star: the doji version of the bullish star pattern the middle candle is a doji, with an open and close almost equal. Many traders consider the doji version a stronger sign of indecision, but in our sample it fared worse.

Why doesn't Finabase mark morning stars on my chart?

Because Finabase runs TA-Lib with a penetration of 10 instead of 0.3, a value carried over from the legacy app. That requires a third-day close ten first-body lengths above the first close, which did not happen once in our 29,760-bar sample. A chart link with the penetration set to 0.3 shows the pattern.

Summary

The morning star is a clear story in three sessions: a strong down day, a stalled day below it, and a strong up day that reclaims much of the loss. TA-Lib turns that into precise rules, but without a trend test, and Finabase's inherited penetration of 10 means its own markers never fire. Run at TA-Lib's default, 155 morning stars on large US stocks did worse than an ordinary day over the next week, confirmed or not. Use the pattern as a reason to study the chart, insist on a real decline and a meaningful level, and decide in advance where the idea is wrong.