Direct Answer
A Kagi chart plots one continuous vertical-and-horizontal line with no time axis. The line keeps extending in its current direction as price moves, and only reverses to draw a new leg once price moves against it by a stated reversal amount. The line's thickness changes too: it draws thick (yang) whenever price is above the prior swing high and thin (yin) whenever price is below the prior swing low.
Key Takeaways
- Reversal amount is the only threshold parameter, expressed as a fixed point/dollar amount or a percentage, it sets how far price must move against the line before a new leg is drawn.
- The line's thickness, not its slope alone, carries the traditional signal: thick (yang) above the prior swing high, thin (yin) below the prior swing low.
- A shift from thin to thick, or thick to thin, is the traditional Kagi buy/sell signal, a pattern-based technique with real limitations, not a reliable predictor.
- Like Renko and point and figure, Kagi has no fixed time axis, so it filters out minor price noise the same way both do, just with a different visual encoding.
How do Kagi charts work?
A Kagi chart is built from a single rule set applied continuously to price: keep extending the current line in its current direction as price moves, and only draw a new leg in the opposite direction once price reverses by the stated reversal amount. The reversal amount can be a fixed point/dollar amount or a percentage of price, the same two choices available for box size on a point and figure chart.
The yin/yang line-thickness signal
Every time the line reverses, it establishes a reference point: a shoulder (a swing high, set at the top of a rising leg just before it reverses down) or a waist (a swing low, set at the bottom of a falling leg just before it reverses up). The line's thickness changes when price crosses one of these reference points during a later leg: the line switches to thick (yang) the moment price rises above the most recent shoulder, and switches to thin (yin) the moment price falls below the most recent waist. Between those crossings, the line simply continues at whatever thickness it already had, regardless of which direction it's currently moving.
Common mistake
The common mistake is assuming the line's direction (rising or falling) is the same thing as its thickness (thick or yang, thin or yin). It isn't. A line can be falling while still thick, it only turns thin once it drops below the prior waist, and it can be rising while still thin, until it climbs back above the prior shoulder. Direction changes on every reversal; thickness changes only when a prior high or low is actually broken.
Worked Example: Tracing a Kagi Line
Illustrative numbers, not live market data.
Reversal amount: $3. The table below walks a seven-observation synthetic price series through the construction rules, tracing both the line's direction and its thickness.
| Obs. | Price | Line action | Thickness |
|---|---|---|---|
| 1 | $100 | Line starts at 100 | Thin (no reference point yet) |
| 2 | $106 | Price rises 6; line extends up to 106 (same direction, no reversal test needed) | Thin (no prior shoulder to cross) |
| 3 | $103 | Price falls 3 from the 106 high, meeting the $3 reversal threshold; line reverses down to 103. This sets 106 as the first shoulder | Thin (unchanged) |
| 4 | $109 | Price rises 6 from 103, meeting the reversal threshold; line reverses up to 109, climbing back through the 106 shoulder on the way | Switches to thick at the 106 crossing |
| 5 | $104 | Price falls 5 from 109, meeting the reversal threshold; line reverses down to 104. This sets 103 as the first waist. Price stays above 103, so no waist is broken | Thick (unchanged) |
| 6 | $112 | Price rises 8 from 104, meeting the reversal threshold; line reverses up to 112, a new high | Thick (reinforced) |
| 7 | $101 | Price falls 11 from 112, meeting the reversal threshold; line reverses down to 101, falling through the 103 waist on the way | Switches to thin at the 103 crossing |
Two thickness changes happen in this seven-observation series, and both happen mid-leg rather than at the start of a new reversal. At observation 4, the line is already moving up (it reversed up at that same observation), but the switch to thick specifically happens where the rising leg passes 106, the previous shoulder, not at 103 where the leg began. The same logic applies in reverse at observation 7: the leg reverses down at 112, but the switch to thin happens where the falling leg passes 103, the previous waist, on its way to 101.
Common mistake
The common mistake is treating every direction reversal as a thickness change. In this example, the reversal at observation 5 (109 down to 104) doesn't change the line's thickness at all, it stays thick, because price never falls far enough to break the 103 waist. A Kagi line can reverse direction several times in a row while staying the same thickness the whole time.
What is the traditional Kagi buy/sell signal?
The traditional signal used in Kagi analysis is the thickness change itself, not the direction reversal. A shift from thin to thick, the line climbing back above its most recent shoulder, is read as a traditional buy signal. A shift from thick to thin, the line falling back below its most recent waist, is read as a traditional sell signal. Because thickness only changes when a genuine prior swing high or low is broken, this framing is meant to filter out the more frequent direction-only reversals that don't clear a prior extreme.
As with point and figure's count-based price targets. This is a long-standing, traditional technical-analysis technique with real, documented limitations, not a reliable predictor of future price. A thickness change confirms that a prior high or low has been broken, which is a description of what already happened, not a forecast of what happens next. The signal can also lag noticeably behind the underlying reversal that produced it, since price has to travel back through the prior shoulder or waist before the thickness itself changes.
What is the yin/yang line-thickness signal?
The traditional Kagi signal is the moment the line's thickness itself changes: a shift from thin to thick is read as a traditional buy signal, and a shift from thick to thin is read as a traditional sell signal. This is a pattern-based heuristic from technical-analysis literature, not a statistically validated predictor, and it can occur well after the reversal that produced it.
How is a Kagi chart different from a Renko or point and figure chart?
All three remove the time axis and filter minor price noise using a fixed threshold, but they encode information differently. Renko uses uniform bricks, point and figure uses columns of X's and O's with a separate box size and reversal amount, and Kagi uses a single continuous line whose only variables are the reversal amount and the resulting thick/thin line-thickness state.
Misconceptions Versus Reality
| Misconception | Reality |
|---|---|
| The line's direction (up or down) and its thickness (thick or thin) are the same signal | They're independent: direction changes on every reversal, but thickness only changes when price actually breaks a prior swing high or low |
| A thickness change happens at the same point as the direction reversal that eventually causes it | As the worked example shows, the thickness change happens where the leg crosses the prior shoulder or waist, which is often a different price than where the leg itself started |
| The yin/yang thickness signal is a reliable, statistically validated buy/sell indicator | It's a traditional pattern-based heuristic from technical-analysis literature with real documented limitations, including lag behind the underlying price move |
| Kagi, Renko, and point and figure all encode the same information, just visually different | They share the no-time-axis, threshold-filtering approach, but encode different information: Renko uses uniform bricks, point and figure uses X/O columns with two separate parameters, and Kagi uses one line with a thickness state |
Risks, Limitations, and Exceptions
- Choice of reversal amount materially changes the resulting chart, a smaller amount produces more frequent reversals and thickness changes; a larger amount filters out more movement but reacts more slowly.
- The thickness signal can lag the underlying price reversal that produced it, since price must travel back through a prior shoulder or waist before thickness itself changes.
- Because the chart has no time axis, two Kagi charts covering very different real-world time spans can look visually similar, which can mislead a reader who assumes line length implies elapsed time.
- The traditional yin/yang buy/sell framing is a pattern-based heuristic with documented failure rates, not a statistically validated forecasting method, treat it as one input, never a standalone signal.
- The worked example in this guide uses illustrative, deterministic numbers, not live market data.
Line Length Says Nothing About How Long It Took
The reading habit worth unlearning here is the assumption that horizontal distance means duration. A Kagi chart has no time axis, so a long stretch of line can represent an afternoon or a year, and two Kagi charts covering wildly different periods can look almost identical. Any impression of pace, of a move being fast or grinding, is being supplied by an axis that is not there.
The thickness is where the traditional signal lives, and it is worth being precise about what triggers it. The line turns thick once price is above the prior swing high and thin once it is below the prior swing low, so a thickness change requires price to travel back through a previously established shoulder or waist. That is a structural event, and it lags the price reversal that produced it.
The reversal amount governs everything else. A smaller value produces frequent legs and frequent thickness changes; a larger one filters more movement and responds more slowly. Both charts describe the same price series, and the yin and yang signals they generate can disagree entirely.
Treat the traditional buy and sell framing as what it is: a pattern-based heuristic with documented failure rates rather than a validated forecasting method. It reads well on a chart of the past, which is exactly the property that makes it easy to overtrust.
Frequently Asked Questions
How do Kagi charts work?
A Kagi chart plots one continuous vertical-and-horizontal line with no time axis. The line keeps extending in its current direction as price moves, and only reverses to draw a new leg once price moves against it by a stated reversal amount. The line's thickness changes too: it draws thick (yang) whenever price is above the prior swing high and thin (yin) whenever price is below the prior swing low.
What is the yin/yang line-thickness signal?
The traditional Kagi signal is the moment the line's thickness itself changes: a shift from thin to thick is read as a traditional buy signal, and a shift from thick to thin is read as a traditional sell signal. This is a pattern-based heuristic from technical-analysis literature, not a statistically validated predictor, and it can occur well after the reversal that produced it.
How is a Kagi chart different from a Renko or point and figure chart?
All three remove the time axis and filter minor price noise using a fixed threshold, but they encode information differently. Renko uses uniform bricks, point and figure uses columns of X's and O's with a separate box size and reversal amount, and Kagi uses a single continuous line whose only variables are the reversal amount and the resulting thick/thin line-thickness state.
How is the Kagi reversal amount chosen?
It is set either as a fixed price increment, as a percentage of price, or from a volatility measure such as ATR. The choice governs how many lines the chart contains: a small reversal produces a busy chart that records minor swings, a large one produces a sparse chart that only records substantial ones. There is no correct value, and changing it is not a cosmetic setting because it rewrites the entire chart from the same data.
What are shoulders and waists on a Kagi chart?
A shoulder is a turning point at the top of a Kagi move, where the line reversed downward; a waist is the corresponding turning point at a bottom. They matter because the thickness rule is defined against them: the line thickens when price exceeds the previous shoulder and thins when it falls below the previous waist. Without identifying the shoulders and waists, the thickness changes cannot be interpreted.
Does a Kagi chart have a time axis?
Not a uniform one. The line advances horizontally only when a reversal occurs, so a week of quiet trading can occupy less width than a single volatile session. Distances measured left to right describe how many reversals happened, not how much time passed. Anything that depends on duration has to be read from the underlying price data rather than from the Kagi chart itself.
Should a Kagi chart be built from closing prices or from highs and lows?
The traditional construction uses closing prices, and that remains the common default. Building from intraday highs and lows registers reversals that a close-only chart never sees, producing more lines and different shoulder and waist locations from identical underlying data. Neither is wrong, but the two are not interchangeable, and a chart that does not state which it uses cannot be compared with one that does.
Why does the same Kagi chart look different on two platforms?
Three settings account for most of it: the default reversal amount, whether the price field is the close or the high and low, and the date the series starts. Because Kagi construction is path dependent, a different starting bar can put the chart on a different sequence of shoulders and waists that persists for a long stretch. Matching charts across platforms means matching all three, not just the reversal amount.
Are Kagi thickness signals lagging?
By construction, yes. A thickness change requires price to have already exceeded a prior shoulder or fallen below a prior waist, both of which are levels set earlier in the chart. The signal therefore reports a move that has happened rather than one that is starting. That is a property of the definition rather than a flaw, but it rules out treating a thickness flip as an early warning.
References
Kagi charting follows a long-standing, publicly documented construction convention used across technical-analysis literature and charting-platform documentation. Key reference sources include:
- Chicago Mercantile Exchange (CME Group), Technical Analysis Education: cmegroup.com/education: background on chart-based technical analysis conventions used across exchange-listed markets.
- Nasdaq, Market Activity: nasdaq.com/market-activity: historical price data conventions this illustrative example's construction rules are built on.
The seven-observation worked example in this guide uses a clearly labeled, deterministic illustrative dataset, not live market data. This content was reviewed by the Swoopr Editorial Team in August 2026.