21 Candlesticks Every Trader Should Know
Dr. Melvin Pasternak · 2006 · 120 pages
Pasternak picks 21 candlestick patterns out of roughly 100 in circulation and gives each a name, a shape, and one real chart example. It is a naming guide, not a tested method, and it never once shows a signal that failed.
The one idea
The book's whole premise is that most candlestick material is encyclopedic and unusable, so Pasternak trims it to the 21 patterns he says occur most often and matter most for short-term reversals: four dojis, hammer and hangman, bullish and bearish engulfing, dark cloud cover, piercing, morning and evening star, shooting star, inverted hammer, harami, marubozu, high wave, spinning top, three black crows, three white soldiers, and tweezers. Each gets a shape diagram, a plain-English story about who won the day's fight between buyers and sellers, and one annotated stock or index chart.
His organizing idea is that candles are an 'anticipatory' signal, meaning they show up before momentum indicators like stochastics or CCI confirm the move, which in turn show up before a trendline or moving-average break confirms it days later. The pitch is that reading the candle first buys you a head start on the trade.
Underneath the pattern catalog he layers in the tools he actually uses alongside candles: a 4-day and 9-day moving average crossover for short-term trend direction, a 50-day and 150-day moving average for intermediate and long-term trend, Bollinger bands (John Bollinger's bands are built to hold 19 of every 20 closes, so a close outside them flags an extreme), and stochastics and CCI for overbought/oversold confirmation. The candle patterns are the trigger; these are the context he checks before acting on one.
What it actually teaches
The 21 candles aren't a sequence, they're grouped by how many trading days each pattern takes to complete, with a confirmation rule and a gap taxonomy layered on top.
ONE-DAY SIGNALS
The four dojis (common, long-legged, gravestone, dragonfly), hammer and hangman, the full marubozu, and spinning top/high wave. A doji is any candle where the open and close land at nearly the same price, so no real body shows. To count as an 'official' hammer or hangman, the book's stated rule is that the lower shadow must be at least twice the height of the real body; hammer appears after a downtrend, hangman after an uptrend, and the two look identical, told apart only by where they sit in the trend.
TWO-DAY SIGNALS
Bullish/bearish engulfing (the second candle's real body must fully consume the first candle's real body), dark cloud cover and piercing (the second candle must close at least halfway back into the first candle's real body to count; less than that gets renamed 'on-neck,' 'in-neck,' or 'thrusting'), harami (a small candle tucked inside the prior day's large real body), and tweezers (two candles whose highs, or whose lows, land at essentially the same price).
THREE-DAY SIGNALS
Morning star and evening star (a large trend candle, then a small-bodied gap 'star,' then a third candle that closes well back into day one's real body, at which point Pasternak says no further confirmation is needed), three black crows (three straight lower closes after a down day), and three white soldiers (three straight higher closes after an up day, each opening in the upper half of the prior day's range).
THE RULE OF TWO
His stated confirmation rule for the single- and two-candle signals: never act on one candle alone, wait for at least part of the next trading day to confirm it. He allows that a gap in the signal's direction on the open can substitute for waiting out the full day.
GAPS AS A FOURTH LAYER
He splits gaps into four types: area/common (inside a consolidation, usually filled fast), breakaway (ends a consolidation on heavy volume, may stay unfilled for months), continuation/measuring (roughly the midpoint of a straight-line move, used to project a price target by adding the prior move's size to the gap), and exhaustion (the third gap in a row in one direction, expected to fill within about a week). He then layers in Steve Nison's rule that Japanese theory treats a gap, or 'window,' as a continuation signal to be traded in the direction it broke.
What it looks like Monday morning
A reader who has never looked at a candlestick chart walks away able to name and draw the 21 shapes, which is genuinely the book's stated goal, right down to a 'draw them from memory' exercise in the closing pages. On Monday morning that means being able to label what just printed on a chart instead of staring at an unfamiliar shape.
It does not hand over a system. There's no position size, no stop placement beyond 'watch the next day,' and no rule for when two candle signals disagree. A reader who wants an actual trading plan still has to build the entry, exit, and risk rules themselves; this book only supplies the vocabulary and the general instruction to wait for confirmation before acting.
Candles are your personal sentry providing you with consistent early warnings of impending trend change.Dr. Melvin Pasternak, 21 Candlesticks Every Trader Should Know
Where it fails
- Every worked example is a winner. All roughly twenty case studies (AVID, TXU, CAL, MRK, AMR, $DDX, $SOX, EGOV, PLCM, MACR, $BTK, BNI, and more) are hand-picked instances where the pattern worked. There is no losing trade shown anywhere in the book, no hit-rate claim, and no count of how often a doji or engulfing candle failed to reverse the trend. A reader has no way to tell from this book whether these patterns beat a coin flip.
- Only two of the 21 candles get a hard numeric rule. The hammer/hangman shadow-to-body ratio (at least 2:1) and the engulfing/piercing 'must cover the prior real body' rule are the only quantified thresholds in the whole set. Everything else, doji, star, spinning top, high wave, marubozu, is defined by words like 'small real body' or 'large upper shadow' with no stated size, leaving pattern recognition entirely to the reader's eye.
- It's an unreleased manuscript, not the finished book. The PDF's own cover page is stamped UNRELEASED MANUSCRIPT with a stated page count of approximately 120, but the file runs 83 pages and has no copyright page, ISBN, or printing statement. There's no way to confirm which numbers or wording survived into the printed edition, which is the version anyone buying today actually gets.
- The supporting indicators are used, never taught. Stochastics, CCI, and Bollinger bands appear on nearly every chart with specific thresholds cited (stochastics below 20 is oversold, above 80 overbought; CCI above +100 overbought and above +200 'highly' overbought), but the book never explains how these are calculated or why those particular cutoffs were chosen. A reader who doesn't already use these tools is expected to take the levels on faith.
- Every case study sits inside one narrow window. The named examples (Dow, Nasdaq, $SOX, $BTK) cluster around a single roughly six-month stretch of 2005, a relatively range-bound, low-volatility period. The book never shows how these same candles performed in a sharp trending market or a high-volatility regime, so the reader has no evidence the patterns generalize beyond the market conditions the author happened to be charting when he wrote it.
Who it is for
Buy it if
You have never learned to read a candlestick chart and want a compact, named glossary of the patterns you'll actually see most often, illustrated with real (if cherry-picked) stock and index charts. It also works as a quick refresher if you already trade but keep mixing up hammer versus hangman or the three star patterns.
Skip it if
You already know standard candlestick terminology, or you're looking for a tested, quantified trading system with defined entries, stops, and a track record across different market regimes. Nothing here is backtested, and most of the 21 patterns are defined by eye rather than by rule.
As an Amazon Associate I earn from qualifying purchases. It costs you nothing extra, and it does not change what gets recommended or what the verdict says.