Candlesticks
Beyond Candlesticks: New Japanese Charting Techniques Revealed
The reason to own this is Part Two: the exact construction rules for kagi, renko, and three-line break charts, three Japanese trend charts nobody in the West had documented in 1994. Part One is a candle refresher you already bought if you read Nison's first book, so decide on Part Two alone.
The one idea
A candlestick chart draws a new mark every session whether the market moved or not. That is useful for reading the fight between buyers and sellers day by day, and useless for answering the simpler question of which side is actually winning. Nison's answer in this book is a family of charts that refuse to draw anything until price does something. No new high, no new low, no new mark. Time comes off the horizontal axis entirely. A quiet week produces nothing at all.
Three charts do this in three different ways. The three-line break lets the market decide when a reversal has happened: after three same-color blocks in a row, price has to break through all three before the color flips. Renko uses a fixed brick size, so a five-point renko only ever draws five-point bricks and simply discards the leftovers. Kagi uses a fixed reversal amount and, uniquely, changes the thickness of its single continuous line: the line thickens when it clears the previous high (a shoulder) and thins when it breaks the previous low (a waist). Thick means buyers own it, thin means sellers do.
The payoff is that these charts only require the closing price. That means you can chart a mutual fund, a bond yield, or anything else that publishes one number a day, which a candle chart cannot do because it needs an open, high, low, and close. Nison's own recommended workflow, which he attributes to a member of the Nippon Technical Analysts Association, is to read the trend off the kagi or three-line break and then wait for a candle signal before actually placing the order. Big chart for direction, small chart for timing.
What it actually teaches
The book teaches four separate tools; here is how they fit together as one workflow, with the book's own numbers.
PICK THE SENSITIVITY KNOB FIRST
Every chart in Part Two has exactly one input and it changes everything. Three-line break: how many lines must be broken (two-line is jumpy, three-line is standard in Japan, five- and ten-line are for position traders). Renko: the brick size. Kagi: the reversal amount, either fixed (a $1 kagi, a 24/32 bond kagi) or, better, a percentage. The percentage kagi is the one genuinely elegant idea here: at 3 percent, a $50 stock needs $1.50 to reverse, and if it runs to $70 the reversal amount scales itself to $2.10. Nison gives one concrete recommendation in the whole book, from an NTAA kagi expert: 3 percent for stocks, with 5 percent popular among longer-term traders. Everything else is 'find it through trial and error.'
THREE-LINE BREAK: LET THE MARKET SET THE REVERSAL
Start at a base price. Close higher than the top of the current range, draw a white block in the next column from the old high to the new close. Close lower than the bottom, draw a black block. Close inside the range, draw nothing. Price must exceed the prior high or low, not merely touch it. Once three same-color blocks stack up, the rule tightens: new highs keep extending white blocks by even one tick, but to flip to black, price must close below the low of the last three white blocks. In Nison's worked example the base is 135, three white blocks are complete at session 12 (close 158), and the flip level becomes 132. A close at 156 in session 20 breaks below 158 (the low of the last three whites by then) and draws the black turnaround line. Eight to ten consecutive same-color lines is his overextension warning.
RENKO: FIXED BRICKS, LEFTOVERS DISCARDED
Choose a brick size. A five-point renko draws only five-point bricks, however far price moved. From a base of 135, a close at 132 is three points and draws nothing. A close at 128 is seven points, which draws one black brick from 135 down to 130, and the last two points simply never appear on the chart. A move that covers two brick heights draws two bricks in two columns. Note the trap: renko says touch or exceed, three-line break says exceed. One trading rule exists and Nison says so plainly: buy the first white brick, sell the first black brick. That is the entire chapter's trading content.
KAGI: READ LINE THICKNESS, NOT DIRECTION
Set a reversal amount (four points in the book's example, base 135). Price moving with the line extends it by any amount, no matter how small. Price moving against it by the reversal amount or more draws a short horizontal inflection line into the next column, then a vertical line the other way. Ignore anything smaller. The signal is not the direction, it is the thickness. When a thin line pushes above the previous high, it turns thick at exactly that price. When a thick line breaks below the previous low, it turns thin at exactly that price. Basic rule: buy the flip to thick, sell the flip to thin. Layered on top are level breaks (wait for two or three prior highs to be taken out before trusting it), the midpoint rule (a pullback that stops above the middle of a long prior line is healthy), and nine record shoulders or nine record waists as a stretched-trend warning.
DISPARITY INDEX: SIZE THE STRETCH
The forgotten chapter. Disparity index is just the close expressed as a percent distance from a chosen moving average. A 13-week reading of minus 25 percent means price is 25 percent below the 13-week average. Japanese defaults are 5, 9, and 25 days short term and 13 or 26 weeks, 75 or 200 days long term. Nison's Delta weekly example treats plus or minus 10 percent on the 13-week as the overbought and oversold band; his S&P example uses plus or minus 15 percent on a 13-period. Between the extremes, a rising disparity confirms an uptrend and a falling one confirms a downtrend. The 'divergence index' in the same chapter is the identical calculation on a different scale (102 percent divergence equals plus 2 percent disparity), which he admits.
TREND FROM THE NEW CHART, ENTRY FROM THE CANDLE
This is Nison's actual recommendation and the reason both halves of the book exist. Take direction from the kagi or three-line break, which are deliberately slow and will not flip on noise. Then use candle patterns as the trigger in that direction only. His GM example: once a black turnaround line appears, act on bearish candle signals and ignore bullish ones, including a doji that would otherwise look like a bottom. The new charts also solve something candles do not, which is when to get out, because a candle pattern almost never gives you a price target but a turnaround line gives you an exit.
What it looks like Monday morning
Add a kagi or three-line break chart next to whatever you already look at, and use it for one job only: deciding whether you are allowed to be long or short today. On stocks, start with the book's one concrete number, a 3 percent kagi, and stop taking signals against the thickness of the line. Every modern platform draws all three of these charts natively, so the construction chapters exist so you can verify the plot is doing what you think, not so you can draw them by hand.
The second change is smaller and probably worth more. Nison's Chapter 4 says a candle pattern is not a trade until you can name the stop and the target, and that if the risk equals the reward you skip it even when the pattern is textbook. He walks two perfectly good hammers on the S&P where a 15 to 20 point stop was chasing a 20 to 25 point target, both of which worked, and says you still should not have taken them. If you are the person shouting 'doji' across the office, that chapter is the fix.
| Session | Close | Three-line break | Renko (5 pt) | Kagi (4 pt) |
|---|---|---|---|---|
| 1 | 135 | Base price | Base price | Base price |
| 2 | 132 | Black line, 135 to 132 | Nothing (3 pts, under brick size) | Nothing (3 pts, under reversal) |
| 3 | 128 | Black line, 132 to 128 | One black brick, 135 to 130 (last 2 pts discarded) | Thin yin line, 135 to 128 |
| 4 | 133 | Nothing (inside 128 to 135) | Nothing (needs 140 or 125) | Inflection line, then up 128 to 133 |
| 11 | 145 | White line, 139 to 145 | Two white bricks, 135 to 145 | Line extended up |
| 12 | 158 | White line, 145 to 158 (third white) | Two white bricks, 145 to 155 | Line extended to 158 |
| 20 | 156 | Black turnaround line to 156 | Nothing (needs 155 or lower) | Turnaround line down to 156 |
Same data, same day, three different answers: on session 4 the kagi acts, the three-line break waits, and the renko sees nothing at all.
My other book, Japanese Candlestick Charting Techniques, took you to the gate. This book takes you to the house.Steve Nison, Beyond Candlesticks
Where it fails
- You are paying full price for about a third of a book. Part One is roughly 145 pages re-teaching hammers, hanging man, shooting star, dark cloud cover, piercing, engulfing, harami, windows, and stars. Nison says outright the book is 'self-contained' and re-covers this ground for new readers. If you already own Japanese Candlestick Charting Techniques, the genuinely new material is Chapters 5 through 8, roughly 90 pages, plus Chapter 4 on trade management. That is the honest purchase. If you own neither book, buy the first one first, because this one assumes you can already read a candle.
- Not one performance number in the entire book. Every chart is a historical example annotated after the fact with buy and sell arrows. There is no win rate, no average trade, no drawdown, no comparison against buy and hold. The single piece of statistics anywhere in the book is a borrowed 1980s Nippon Technical Analysts Association study of the Nikkei's divergence bands (200-day divergence sits between 102 and 110 percent in a rising market, 95 percent of the time), and it is a study of an indicator's distribution, not of a trading result. Nison states his position directly: 'Consequently, I will not attempt to find the optimum reversal amount.' The construction rules are exact and testable. The claim that trading them makes money is untested inside these covers.
- These are trend-following charts and the book shows them getting chopped up without ever costing it. Nison labels the whipsaws himself. The Intel renko exhibit has three losing round turns marked B1-S1, B3-S3, and B4-S4. The Merrill Lynch 3 percent kagi has losers at B2-S2 and B3-S3. He acknowledges that in a sideways market 'the buy and sell signals can induce losses.' What is missing is any filter, any stop rule, any position-sizing rule, and any sense of how often this happens. The offered remedy is to adjust the sensitivity by trial and error, which is another way of saying the book hands you the losses and no method for reducing them.
- Signals arrive late by design, and the workaround reintroduces discretion. Nison concedes it on the Wal-Mart tweezers top: the candle chart flagged the top through late March while the kagi did not confirm until early April. He writes that kagi, renko, and three-line break charts 'are not for those who are trying to pick exact tops or bottoms.' Worse is the close-confirmation gap. Because the charts run on closes, a $2 reversal amount can be $4 through by the time the bell rings, so you surrender $2 of the move. His fix is to buy a light position intraday when the level trades, add at the close if confirmed, and dump it if not, which is a discretionary scaling plan bolted onto a mechanical chart.
- The one input that determines every signal is left entirely to you. Change a three-line break to a two-line break and you get many more trades on identical data; go to five lines and you get far fewer. Nison prints the same Ford weekly data as a two-line, three-line, and five-line chart to demonstrate exactly this, then declines to say which is right. The whole book's concrete guidance on parameter choice is one NTAA member's habit of using 3 percent on stocks and the observation that 5 percent 'appears popular for longer term traders.' That is a personal preference relayed secondhand, not a finding, and it is the number a reader will end up trading.
- The 1994 price conventions do not survive translation. Bonds are quoted in 32nds (the kagi examples use a 24/32 reversal), stocks trade in eighths (45.125, 47.625), and every fixed-dollar setting in the book, the $1 kagi, the $2 renko, the $3 gold renko, is calibrated to 1994 price levels and 1994 volatility. Decimalization, tighter spreads, and index levels an order of magnitude higher have made all of those specific numbers dead on arrival. Only the percentage kagi survives the move intact, which is a good argument for using it and ignoring the fixed-amount examples.
- Renko gets a chapter and one rule. Nison writes that 'the renko charts are more limited. The only trend reversal signals with renko charts are with the emergence of a bullish white brick or bearish black brick.' Nine pages, two annotated exhibits, one signal. There is also a rounding problem he never confronts: in a five-point renko, a run from 100 to 107 prints one brick and the last two points vanish from the chart forever. On the 40-to-70 dollar stocks in his examples that is noise. On a high-priced stock or an index, the discarded remainder is real money you cannot see.
- Day traders get one of the three tools. The book states that three-line break and renko charts 'are not normally used on an intra-day basis,' speculating that they have simply been less successful there. Only kagi is presented intraday, with a five-minute S&P example, and Nison notes it requires a trader who has 'the time and the data' for tick-by-tick construction. If you trade inside the day, two-thirds of Part Two is background reading. The book also assumes screen time, a data feed, and the patience to hold through a chart that may not print a mark for a week.
- Some of the named patterns are folklore, not method. The 'black shoe, white suit, and a neck' sequence gets several pages and a Japanese saying ('Buy when the neck emerges from the white suit with black shoes'). Strip the costume metaphor and it reduces to 'wait for a second confirming line after the reversal,' which Nison already stated plainly two pages earlier as the extra-confirmation rule. The 'two-paired chimney' is a double top. The 'three-Buddha' is a head and shoulders, which he says himself. Naming familiar things in Japanese does not add information.
Who it is for
Buy it if
You already know candles, trade off daily or weekly closes, and keep getting shaken out because you have no objective definition of the prevailing trend. Part Two gives you three mechanical ways to define it and the exact rules to verify your platform is drawing them correctly. It is also the only book that explains why a kagi line changes thickness, which is the single most useful idea in it.
Skip it if
You have never read a candlestick book. Start with Japanese Candlestick Charting Techniques instead, because this one assumes the vocabulary and then repeats a compressed version of it anyway. Skip it too if you want evidence: there are no statistics, no backtests, and no equity curves here, only annotated 1994 charts chosen after the outcome was known.
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.