Candlestick Patterns Guide: The 14 Patterns That Actually Work
The 14 candlestick patterns a prop-firm trader should actually memorize – with anatomy, statistical evidence, context rules, and the traps that fool beginners.

01Anatomy of a candlestick
Each candle encodes exactly four pieces of information for a given time window: open, high, low and close. The body (the thick rectangle) spans open to close. The wicks or shadows (the thin lines above and below) extend to the period's high and low.
- Bullish candle – close higher than open; typically drawn green or hollow.
- Bearish candle – close lower than open; drawn red or filled.
- Long body – strong directional conviction within the period.
- Long wick – price tested an extreme but was rejected by the close.
- No body (doji) – open equals close; indecision or balance.
The tape of candles is the story of buyers vs. sellers. Reading the story is easier than trying to memorise names.
02Single-candle patterns (4 worth learning)
These appear as a single bar and always require confirmation from the next bar and from location. They are intent signals, not execution signals.
- Hammer – small body at the top of the range, long lower wick (≥ 2× body), little or no upper wick. Appears at the bottom of a downtrend. Says: sellers drove price down during the period, buyers absorbed and pushed it back up.
- Shooting Star – mirror of the hammer. Small body near the low, long upper wick, little lower wick. Appears at the top of an uptrend.
- Doji – open and close within a few ticks of each other; wicks can be small or large. Signals indecision. A doji after a long trending move is meaningful; a doji in the middle of noise is not.
- Marubozu – full-body candle with no wicks (or tiny ones). Signals extreme conviction. A bullish marubozu at a breakout level is one of the cleanest signals in the toolkit.
03Two-candle patterns (6 with edge)
Two-candle patterns show a dynamic – a reversal or a continuation – that a single candle cannot. They still need location and confirmation.
- Bullish Engulfing – a small red candle followed by a green candle whose body fully engulfs the prior body. Ideal at the end of a downtrend. Bulkowski ranks it among the top-performing reversal patterns in his 2008 testing.
- Bearish Engulfing – mirror of the above. Red body engulfs prior green body at the top of an uptrend.
- Piercing Line – red candle followed by a green candle that opens below the prior low and closes above the prior candle's midpoint. Bullish reversal; weaker than engulfing because it does not fully swallow the prior body.
- Dark Cloud Cover – mirror of the piercing line; bearish reversal.
- Tweezer Top / Bottom – two candles with near-identical highs (top) or lows (bottom). Clean signal of a failed probe of that level. Best when the second candle is a hammer or shooting star.
- Harami (Inside Bar) – a large candle followed by a much smaller candle contained entirely within the prior body. Signals a pause – often the start of a reversal, but equally often a continuation. Always use with a clear invalidation level.
04Three-candle patterns (4 worth knowing)
Three-candle patterns require the most setup time but deliver the strongest context. All four below are statistically significant in Bulkowski's study.
- Morning Star – a bearish candle, then a small-body indecision candle (often a doji) that gaps down, then a bullish candle closing deep into the first candle's body. The classic bottom reversal.
- Evening Star – mirror of the morning star at a top.
- Three White Soldiers – three consecutive long-body bullish candles, each closing near its high, each opening within the prior body. Strong continuation or reversal-into-continuation signal.
- Three Black Crows – mirror of the above; three long bearish candles in succession.
Approximate "success rate" (reaching a measured target before stop) from Bulkowski's 2008 US-equities study. Not FX-calibrated.
Source: Bulkowski (2008), Encyclopedia of Candlestick Charts
05Location is more important than the pattern
If you study only one section of this article, make it this one. A pattern at a meaningless level is noise; the exact same pattern at a significant level is tradeable. Significant levels include:
- Prior swing high / low – where the trend last reversed.
- Horizontal support / resistance zone – a level tested at least twice before.
- Key moving averages – 50-EMA, 200-SMA, VWAP for intraday.
- Fibonacci retracement – 0.382, 0.5, 0.618 of a prior impulse.
- Previous session's high/low or the daily open – intraday mean-reversion levels.
A hammer at the 50-EMA after a pullback in a confirmed uptrend is a high-probability setup. A hammer in the middle of a sideways range at no particular level is not.
06Confirmation: the candle after the candle
Beginners act on the pattern the moment it forms. That is premature; the pattern is a hypothesis, not a trade. Professional traders wait for confirmation, which takes one of three forms:
- Price confirmation – the next candle closes in the direction of the pattern. For a bullish engulfing at support, you want the following candle to close green.
- Structure confirmation – a lower-timeframe break of a micro-level in the expected direction.
- Volume confirmation – the pattern candle prints on noticeably higher volume than recent average (more useful on futures and equities than on FX).
The cost of waiting for confirmation is giving up a few pips of entry; the benefit is filtering out failed patterns, which Bulkowski's data shows is roughly 25–45 % of all occurrences.
07The five most common traps
Most losing candlestick trades lose for the same handful of reasons:
- Taking patterns inside a range. Reversal patterns work at trend extremes, not inside consolidation. The hammer in the middle of a range is noise.
- Ignoring higher-timeframe context. A bearish engulfing on the 5-minute inside a strong daily uptrend is usually a bear trap.
- Entering on the signal candle. The pattern is the hypothesis. Without confirmation, you are statistically taking a 55 % edge, not a 78 %.
- Placing stops inside the pattern. Stop below the hammer's low, not halfway up the wick. The wick is the level the market tested – a close below it is the invalidation.
- Treating every doji as significant. A doji after a 20-bar grind means nothing. A doji after a 10-candle trend at resistance means something.
08A practical workflow for prop-firm traders
Integrating candle patterns into a prop-firm evaluation strategy in four steps:
- Bias from the higher timeframe. Define trend direction on the daily or 4-hour. Only take reversal patterns in the direction of the higher-timeframe trend; treat counter-trend patterns as continuation fades.
- Mark significant levels. Horizontal S/R, key moving averages, the prior day high/low, the session opening range.
- Wait for price to reach a level. Only scan for candle patterns when price is at one of your pre-drawn zones. This eliminates 90 % of bad signals.
- Enter on confirmation. One of the three confirmation rules fires; enter with a stop beyond the pattern; target a measured move or a prior swing.
Combine this with the position sizing framework and the stop-loss guide to translate the visual pattern into a trade ticket.
Sources & further reading
Citations are checked against primary regulators and academic sources. External links open in a new tab; we're not responsible for third-party content.
- Japanese Candlestick Charting Techniques (2nd ed., 2001) – Steve Nison, Prentice Hall · accessed Apr 18, 2026
- Encyclopedia of Candlestick Charts (2008) – Thomas N. Bulkowski, Wiley · accessed Apr 18, 2026
- Candlestick Charting Explained (3rd ed., 2006) – Gregory L. Morris, McGraw-Hill · accessed Apr 18, 2026
- The Origins of Candlestick Charting (historical overview) – CME Group Education · accessed Apr 18, 2026
Frequently asked questions
Do candlestick patterns actually work?
Some do, most modestly. Bulkowski's large-sample study found that a minority of patterns – bullish engulfing, morning star, three white soldiers and their bearish mirrors – showed statistically significant edges in US equities 1995–2006. Many other patterns taught in books have near-coin-flip performance. The edge comes from pairing the valid patterns with location and confirmation, not from rote pattern recognition.
How many patterns should a trader actually memorise?
Between 10 and 20. Trying to memorise all 103 patterns in the candlestick literature is a distraction. The 14 covered in this guide – four single, six two-candle, four three-candle – will cover 95 % of setups you will ever take. Master these; ignore the rest unless a specific one becomes part of your personal strategy.
Which timeframe is best for candlestick patterns?
Daily and 4-hour are the most reliable on FX and equities. 15-minute and 5-minute patterns are tradable on futures during active sessions but carry more noise. Sub-1-minute candle patterns are effectively random – the signal-to-noise is too low. Match timeframe to holding period: a day trader uses 15-minute patterns with 1-hour context; a swing trader uses daily patterns with weekly context.
Does a candle pattern on FX behave the same as on stocks?
Not exactly. The Bulkowski statistics are from US equities where each day has a defined open, close, and volume footprint. FX is 24-hour, with different sessions having structurally different behaviour – an engulfing candle on the H1 during Asia tells you less than the same pattern during London open. Always match the pattern to the session's normal volatility.
Should I use candlestick patterns with or without indicators?
With context, always. The most reliable combinations are: pattern + horizontal S/R, pattern + moving average, pattern + Fibonacci retracement. Oscillators like RSI or Stochastic can add a filter (e.g., "only take bullish reversal patterns when RSI(14) < 30 on the same timeframe") but are not required.
How do I avoid getting stopped out on a valid pattern?
Two practical rules. First, the stop goes beyond the invalidation level of the pattern – below the hammer wick, not halfway up. Second, size the position so the stop distance equals your intended dollar risk. If the resulting size feels too small, the setup is too wide for your account – pass the trade, don't tighten the stop.
Pass the Evaluation Up to $150K –Start in Minutes
Pass one evaluation. Trade a simulated funded account. Earn up to a 90% Reward Coefficient on your simulated performance. Receive your Performance Reward in ~8 hours on average.
- One-time fee
- Refundable on second Performance Reward
- No subscription