Three-candle patterns are reversal (and occasionally continuation) signals read across a sequence of three sessions rather than a single bar. The two canonical reversals are the Morning Star (bullish, marks the bottom of a downtrend) and the Evening Star (bearish, marks the top of an uptrend); the Three White Soldiers and Three Black Crows are the momentum-exhaustion cousins. Each encodes a small narrative: an established trend (candle 1), a stall or indecision (the small-bodied "star," candle 2), then a decisive thrust the other way (candle 3). Because they require three specific bars to line up, they are rarer than one- or two-candle signals — which makes their published statistics noisier and easier to over-trust.

How they're formed

Morning Star (bullish reversal, after a downtrend) — per Bulkowski's identification rules: 1. A tall black/down candle continuing the downtrend. 2. A small-bodied candle of any color whose body gaps below the first body — the "star." Indecision; sellers no longer dominate. 3. A tall white/up candle that gaps above the second body and closes at least midway into the body of the first candle.

Evening Star (bearish reversal, after an uptrend) — the mirror: 1. A tall white/up candle continuing the uptrend. 2. A small-bodied candle that gaps above the prior body (the star). 3. A tall black/down candle that opens below the star and closes at least midway down the body of the first candle.

The third candle's penetration is the load-bearing condition: the deeper it closes into candle 1's body, the stronger the signal is conventionally held to be; the halfway-close minimum is the standard threshold (Bulkowski, StockCharts). When the middle candle is a true doji (open ≈ close), the variants are the Morning Doji Star / Evening Doji Star — read as stronger indecision but, per Bulkowski, materially rarer.

Three White Soldiers — three consecutive tall up-candles, each opening within the prior body and closing near its high; read as a bullish reversal after a decline. Three Black Crows — the bearish mirror (three tall down-candles after an advance).

> Note on gaps: textbook stars require gaps around the star's body. Modern 24-hour and high-liquidity markets gap intraday far less, so many screeners relax the gap to a body-overlap test. This is a real definitional fork — base-rate stats assume the strict definition.

How they're used in practice

The style-agnostic core uses are: (1) reversal confirmation — the completed third candle is the trigger, not the star itself; (2) location filtering — the patterns are treated as meaningful only at a prior swing extreme or support/resistance, not mid-range; (3) context stacking — practitioners weight a star far more when it coincides with a trendline test, a moving average, oversold/overbought oscillator readings, or a volume spike on candle 3. The doji variant and deeper third-candle penetration are used as a crude confidence dial. Specific entry/stop/target and hold-period mechanics belong to the Swing Trading branch — defer there; this node defines the signal, not the trade.

Standing & evidence

Treat the two evidence streams separately and honestly.

Bulkowski's measured base rates (from Encyclopedia of Candlestick Charts, hundreds-of-thousands-of-instances sample, strict definitions):

• Morning Star — acts as a bullish reversal 78% of the time (reversal rank 6 of 103); overall performance rank 12; frequency rank 66 (uncommon).
• Evening Star — bearish reversal 72% of the time (reversal rank 10); overall performance rank 4 ("top notch"); frequency rank 71 (rarer than the morning star).
• Three White Soldiers — reversal ~82% (rank ~3 among top reversal candles); Three Black Crows — reversal ~78% (rank ~7).
• The doji variants are reliable-ish but materially rarer (e.g. Bulkowski ranks the Evening Doji Star ~81 of 103 on frequency), so their stats rest on smaller samples.

Two cautions on those numbers. First, a high "reversal rate" is partly definitional — the pattern is defined as a thrust that closes back into candle 1, so it almost tautologically "reverses" the immediate prior move; the meaningful question is forward edge, not the labeling rate. Second, three-candle patterns are rarer, so even Bulkowski's large database yields smaller per-pattern samples than for common single candles — the ranks are directionally informative but the precision should not be over-read.

The academic finding is less flattering. Marshall, Young & Rose (2006), testing 28 candlestick patterns on Dow Jones 30 stocks (1992–2002) with a bootstrap that compares signals against randomly generated OHLC series, found no statistically significant value — candlestick strategies did not beat random for large U.S. stocks over their window. This is the most-cited of several studies reaching broadly similar conclusions: candlesticks carry little reliable tradeable edge after transaction costs. Bulkowski's percentages describe what happened next in a database; they are not a claim of risk-adjusted, cost-adjusted profitability.

Strengths & limitations

• Strengths: the three-bar structure encodes a clear, mechanical exhaustion-then-thrust narrative; the third-candle penetration rule makes the signal objective and confirms after the turn rather than predicting it; location-filtered, it's a usable confirmation trigger.
• Limitations / failure modes: rare (so few opportunities, noisy stats); regime-dependent (a star against a strong trend is routinely overrun); timeframe-sensitive (intraday stars are mostly noise); the relaxed/no-gap variants used by most screeners are not the strict patterns the base rates were measured on. The #1 misuse is trading the star in isolation, mid-range, with no confirmation of candle 3's close — i.e. acting on indecision before the third candle validates it. Net: a confirmation overlay, not a standalone edge.

System relevance

Cross-links: the single-candle building blocks (doji, long-bodied candles) and the two-candle reversals (engulfing) are sibling nodes under Candlestick Patterns; operational swing entry/stop/target mechanics live in the Swing Trading branch. For Delvantic's Augustus trade-setup agent, a completed star is at most a secondary confirmation input layered on regime, level, and volume context — never a primary trigger. Given the academic null result and small-sample fragility, Augustus should down-weight an isolated star and defer the profitability judgment to live data plus Cairn's measured track record rather than to the textbook reversal percentage.

Sources

• Bulkowski, thepatternsite.com — Morning Star (https://thepatternsite.com/MorningStar.html), Evening Star (https://thepatternsite.com/EveningStar.html), Top 10 Reversal Candlesticks (https://thepatternsite.com/CandleReverse.html), Pattern Rank (https://www.thepatternsite.com/rank.html). Source of the 78%/72% reversal rates, ranks, and identification rules.
• StockCharts ChartSchool — Morning/Evening Star identification and third-candle penetration convention.
• Marshall, Young & Rose (2006), "Candlestick technical trading strategies: Can they create value for investors?", Journal of Banking & Finance — academic null result on Dow 30 (https://www.sciencedirect.com/science/article/abs/pii/S0378426605002116).
• Investopedia — Morning/Evening Star and Three White Soldiers / Three Black Crows definitions (cross-check).