Japanese candlesticks, patterns and levels: a trader's cheat sheet

Japanese candlesticks in trading: how to read a single candle

A candle draws the price over one time segment: a minute, an hour, a day. The length of the segment is called the timeframe, the trader chooses it, and a candle is read the same way on any timeframe. The rectangle in the middle is called the candle body. Its edges are the opening and closing prices. The thin lines above and below are the upper and lower shadows, also called wicks. The end of the upper shadow is the high of the period, the end of the lower one is the low.

The color shows where the price went. A close above the open - the candle is green or hollow, the open at the bottom of the body, the close at the top. A close below the open - the candle is red or filled, and everything is the other way around. Colors change in the terminal settings, so on someone else's screenshot first find which color marks growth there.

Example. An hourly candle opened at 100, rose to 106, fell to 99 and closed at 104. On a candlestick chart this looks like: a green body from 100 to 104, an upper shadow to 106, a lower shadow to 99. The candle's range (high minus low) is 7 points.

If the open and close are the same, there is no body at all; instead of a rectangle, a horizontal line is drawn. Such a candle is called a doji. If the price closed exactly at the high, the upper shadow disappears. At the low, the lower shadow disappears.

What the candle doesn't show

A candle doesn't show the order in which the price moved. From the candle in the example, you can't tell what came first, the high of 106 or the low of 99, or how many times the price reversed within the hour. That is only visible on a lower timeframe, for example on 5-minute candles for the same hour. A classic candle also has no volume: trading volume is drawn as a separate histogram below the chart. News and reports are even less present in a candle, although the price moves because of them too.

Candlestick patterns in trading: hammer, engulfing, harami

A pattern is a recognizable combination of one, two or three candles. Candlestick analysis looks for them where a trend may change. A reversal pattern needs a trend to reverse: a hammer in the middle of a sideways range means nothing.

Hammer. In Japanese, takuri, "probing the bottom". The lower shadow is long, roughly twice the body or more. The body sits at the top of the range, and there is no upper shadow or it is very short. After a decline, a hammer is read as a signal of a possible upward trend reversal. The same shape at the top of a rally is called a hanging man and is read as a signal downward.

Engulfing. Two candles in a pronounced trend, at least a short-term one. The second candle's body completely covers the first candle's body, and its color is the opposite. A bullish engulfing after a decline looks like this: a small red candle, followed by a large green one.

Harami. The opposite of engulfing: an unusually long body is followed by a short one. The word means "pregnant" in Japanese; the small body seems to sit inside the large one.

In broker textbooks, three more are added to these: the gravestone doji, the dragonfly doji, the tweezers and the morning star. Interpretations in the sources sometimes differ. For example, Russian Wikipedia calls a long upper shadow a bullish signal, while the Gazprombank Investments breakdown describes a bullish hammer with a long lower shadow. So check the same candle against several sources and against where it sits on the chart.

The chart shows a conditional example. The price falls from 110 to 100.6. The hammer drops to 98 but closes at 101. Next comes a small red candle closing at 100.3, followed by a bullish engulfing closing at 102.4: its body from 100.1 to 102.4 completely covers the red candle's body. Ten candles later the price is at 107.3. In Bulkowski's test, after a hammer the price failed to go up in 40% of cases.

Japanese candlestick patterns on history: what the tests showed

Thomas Bulkowski, author of "Encyclopedia of Candlestick Charts", tested 103 candlestick patterns on historical data. In his test, the hammer produced an upward reversal in 60% of cases. Bulkowski himself writes that 60% is not that far from a random 50%. By reversal, the hammer ranks 26th out of 103 for him; by overall performance, 65th.

Bullish engulfing reversed the price upward in 63% of cases, which is 22nd out of 103. It occurs often, 12th by frequency, and by overall performance it ranks 84th.

Thomas Bulkowski's data (thepatternsite.com), checked 09.11.2026. Reversal here means the direction of the price after the pattern.

Reversal and profit are not the same thing. Bulkowski measures where the price went, and does not account for the spread and commission. For costs he has a separate metric: the share of patterns after which the price did not move at least 5% in the desired direction. Such a move does not pay for the trade.

Scientists tested more strictly. Duvinage, Mazza and Petitjean in 2013 took 83 candlestick rules and ran them on 5-minute charts of 30 Dow Jones stocks. About a third of the rules beat simply buying and holding the stocks. After accounting for costs, a few rules remained. After adjusting for overfitting to history, none remained, and trading systems built from the best rules did not win either.

Park and Irwin's 2007 review collected 95 modern studies on technical analysis. 56 found profit, 20 found loss, 19 had mixed results. But for most of them the review's authors found weaknesses: the rules were selected with hindsight, and risk and costs were poorly assessed. Lo, Mamaysky and Wang in 2000 taught a program to find patterns in US stocks over 1962-1996. Some patterns, such as head and shoulders, provided additional information and, according to the authors, may have practical value. There is also a caveat that patterns on a chart often exist only in the eye of the beholder.

Another difficulty: candlestick analysis is subjective. Where one trader sees a hammer, another sees just a long shadow. Programs that find patterns on their own work by predefined rules, and the result depends on those rules. More on which methods stood up to scrutiny is written in the article on technical analysis for beginners.

Chart patterns in trading: reversal and trend continuation

A pattern is built from dozens of candles and lines that a trader draws along the highs and lows. Chart patterns in trading are divided by what is expected of them. The triangle, flag, pennant and wedge are considered trend continuation patterns: the move should continue after a pause. Head and shoulders, double top, double bottom and diamond are trend reversal patterns; a change of direction is expected from them. By the method of construction, trading patterns are also divided into chart, candlestick and harmonic.

Head and shoulders - three peaks in a row, the middle one higher than the other two. The shoulders sit at a similar price. The neckline connects the "armpits," that is, the two lows between the peaks. Volume is usually largest at the left shoulder and at the head. The pattern is considered confirmed when a candle closes below the neckline.

According to Bulkowski, head and shoulders at a top in a bull market ranks 9th out of 36 patterns. Price fell by an average of 16%, but only reached its calculated target in 51% of cases. In 68% of cases, after breaking the neckline, price returned to it. The basis is more than 2,800 trades under ideal rules.

The double top and double bottom are simpler: two highs or two lows at roughly the same price. Confirmation is the same, a close beyond the price between them.

Triangle in trading

A triangle in trading is considered a continuation pattern: after a temporary correction, price should move in the previous direction. There is the classic triangle, in which both boundaries converge. Ascending and descending triangles are built from a horizontal line and a slanted line along the trend or against it. As with all patterns, a signal appears only after a candle closes beyond the boundary.

Support and resistance levels: how to find them

Support is a price at which a decline is believed to tend to stop and reverse. It is a demand zone: buyers are located here. Resistance is the same kind of price above, a supply zone. A level is marked by several touches without a breakout, and the more touches there are, the stronger it is considered.

In trading, levels are sought at the highs and lows where the price has already reversed, at the open and close prices of the day, and at round numbers, for example multiples of 50. Large limit orders are often placed at round numbers. A higher timeframe matters more than a lower one. A level from the daily chart is usually respected by the price on the hourly chart too, while a level from the hourly chart is broken through on the daily chart with ease. Therefore, mark only key levels, otherwise the entire chart will be covered with lines.

A broken level changes its role. Broken resistance becomes support, broken support becomes resistance. Such a level is called a mirror level. The price returns to it from above, touches it and moves on. This return is called a retest.

On the chart, the price reaches 110 three times and pulls back. On the fourth time, the candle closes at 110.2, the price rises to 113.5, and then returns and drops to 110.3. From the former resistance, it moves up to 116. The example is hypothetical and hand-picked: on a real chart, the price may pass straight through the former level.

There is also scientific verification. Carol Osler of the Federal Reserve Bank of New York in 2000 took the support and resistance levels published by six currency dealers. She found strong evidence that these levels help predict where an intraday trend will break. The strength varied across currencies and across dealers. In 2003, Osler studied actual currency orders. Take-profits cluster exactly at round numbers, hence the bounces off levels. Stop-losses sit just beyond round numbers, which is why the price flies further so quickly after a breakout.

False breakout: why the price returns beyond the level

A breakout is the moment when the price passes through a level and stays beyond it. A false breakout looks the same, except the price then returns to the previous range: there was not enough momentum to hold. It can fit within a single candle, when the price passed the level and returned by the close. There can be a false breakout of a level, a false breakout of a range, and a single candle with a long wick beyond the level.

There are several reasons. Large participants collect the stop orders of novices placed just beyond the level. Unexpected news comes out. Or the breakout lacks volume. Osler's data on orders explains the mechanics: stops lie in a cluster just beyond a round number, the price knocks them out, and the move ends. That is why a false breakout is called a trap.

Example. The price moves between 103 and 110 for two weeks. A trader buys at 110.5 when the price passes the level and places a stop at 108. The candle reaches 111.5 but closes at 109.2. The next one drops to 107.4 and takes out the stop. A few candles later, the price is already at 103.5.

It is impossible to distinguish a false breakout from a true one at the moment of entry; it is only visible in hindsight. Bulkowski calls such cases failed patterns: the price broke out of a pattern, moved no more than 10%, and reversed. Failure can be stronger than success. In a bull market, after 286 triple bottoms the price rose on average by 37%, while after 145 failed triple tops it rose by 42%.

That is why the entry point, stop-loss and take-profit are determined before the trade. A short stop reduces the loss on the occasion when the breakout turns out to be false. How to calculate the stop and position size is covered in the article on stop-loss and take-profit.

Fibonacci in trading: where 38.2% and 61.8% come from

In the Fibonacci sequence, each number equals the sum of the two previous ones: 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89. The ratio of adjacent numbers tends to 1.618, the golden ratio, and its reciprocal equals 0.618. A trader needs the ratios between the numbers. If you divide 55 by 89, you get 0.618. A number one place away gives 0.382 (34 by 89), two places away gives 0.236 (21 by 89). The 78.6% level is the square root of 0.618. The 50% and 100% levels are not derived from the sequence; they are added because many traders consider them psychologically important.

Fibonacci levels are drawn on a clear trend using two points. In a rise, these are the minimum where the move begins and the maximum where it ends; after that the terminal calculates the retracement levels itself. In a sideways market there is little use in the tool. Example: the price rose from 100 to 120. A 38.2% retracement is 120 minus 7.64, that is 112.36. A 61.8% retracement is 107.64.

On the chart the retracement ended at 108.5, between the 50% and 61.8% levels. Which one of them worked can only be decided in hindsight. Critics write that the significance of such levels cannot be confirmed by data. Arthur Merrill in the book Filtered Waves found no standard correction depth. Large tests on thousands of stocks showed that a retracement to 38%, 50% or 62% is no more likely than any other. The broker CMS Prime states outright: there is no law of nature by which prices observe these ratios, and the weight of the levels rests on the fact that many people watch them. Charles Schwab adds that no technical tool predicts price reliably.

Japanese Candlesticks: Steve Nison's book and the history of the method

Rice exchanges in Japan had operated since 1730, these are "Dojima" in Osaka and "Kuramae" in Tokyo, opened under the reform of Shogun Tokugawa Yoshimune. The inventor of the candlestick chart is considered to be the rice trader Homma Munehisa, who began trading in 1750. His 1755 book "The Golden Source of the Three Monkeys" is written about money management and psychology; there are no candlestick patterns in it. Steve Nison himself in the book "Beyond Candlesticks" writes that Homma hardly used candlestick charts. According to his version, candlesticks appeared later, at the beginning of the Meiji period, at the end of the 19th century.

Nison brought the method to the West. His book Japanese Candlestick Charting Techniques came out in 1991. The second edition was published by Prentice Hall Press on November 1, 2001, it has 320 pages, and the hardcover from the publisher Penguin Random House costs $115 as of 09.11.2026. In Russian, this book, Japanese Candlesticks by Steve Nison, is known under the title "Японские свечи: графический анализ финансовых рынков".

Nison's second book, "Beyond Candlesticks", came out in Russian from the publisher Alpina PRO, it has 272 pages. The description says that candlesticks show not only the direction but also the strength of the move and combine well with other types of technical analysis. In the bibliography of the Russian Wikipedia, next to Nison stand "The Secret Code of Japanese Candlesticks" by F. Tudela, G. Morris's book "Japanese Candlestick Charting Techniques. The Method of Analyzing Stocks and Futures, Tested by Time" and "Encyclopedia of Candlestick Charts" by Bulkowski.

Books about Japanese candlesticks in trading describe patterns on other people's charts. You will still have to test the rule on your own trades. The word "Japanese" in the title only says where the chart was invented. It shares only its name with Japanese spark plugs for cars, even iridium ones. And Japanese phonk in a playlist has nothing to do with charts at all.

How to test a pattern before risking real money

For beginners, Japanese candlesticks in trading are a convenient chart language. The entry rule still has to be assembled from them and tested. Write it down so that another person could repeat it: which pattern, on which timeframe, at which level, where the entry, where the stop, where the exit. Then run 30 trades on it on a demo account and record each one in a journal: date, pattern, level, result. After 30 trades it will become clear how much the rule loses in a bad streak. How a demo differs from a cent account and a prop challenge is covered in the article where to start: demo, cent account or prop.

The next step is risk per trade. Suppose it equals 1% of the account. On a $10,000 account, a stop-loss then costs $100. Three stops in a row - $300, or 3%. With your own money that is an unpleasant streak. On a prop account with a 3% daily limit, such a streak eats up the entire day's limit.

Editorial calculation, a scenario, not a forecast.

A prop firm lets someone who has passed a paid evaluation, a challenge, trade with its account. The trader trades on a simulated account, and the prop firm pays the payout from its own funds under the terms of the contract. Every challenge has a daily loss limit: if the account goes into the red for the day by more than the limit, the challenge is failed. In the PropMarketCap catalog as of 09.11.2026, 66 of 136 challenges from seven prop firms have a daily limit of 3% at the first stage.

Share of PropMarketCap catalog challenges by first-stage daily loss limit, snapshot 09.11.2026, shares rounded to whole numbers.

In two-stage challenges on $10,000 the limits differ. At Hash Hedge and Era Trade such a challenge costs $99, the daily limit is 5%, the overall limit is 10%. At The Funded Way the cheapest program is Starter way at $92, but the daily limit in it is 3%, and the overall limit is 6%. If your daily pattern produces streaks of four stops, a program with a 3% limit will close the account before the rule has a chance to play out. It is convenient to check your risk against the program limits when you select a challenge in the calculator.

With your own money the risk is higher. According to data from the European regulator ESMA for 2018, 74-89% of retail accounts lost money on CFDs. The average loss per client ranged from 1,600 to 29,000 euros. On a prop account the trader risks only the challenge fee, but it can be lost just as quickly. How prop trading works as a whole is described in the article what is prop trading.

Summary

Japanese candlesticks, figures, levels and Fibonacci describe the chart and suggest where to place the entry and stop. They do not provide a guarantee: a hammer reverses the price in 60% of cases versus 50% for a coin flip, and a false breakout is only visible in hindsight. Test the rule based on the pattern on 30 trades in a journal and check the risk against the daily limit. How to read the challenge rules is covered in free PMC Academy lessons: 4 of 28 lessons are open without payment. Once the rule passes the test, choose a challenge in the calculator.