Direct Answer
A swing chart filters price action down to a directional line connecting only the significant swing highs and swing lows in a series of bars or candles, ignoring the smaller moves in between. It's typically overlaid on standard candlestick data to highlight market structure, higher highs and higher lows in an uptrend, lower highs and lower lows in a downtrend, rather than functioning as a fully standalone chart type.
Key Takeaways
- Swing points are usually identified with a simple N-bar fractal rule, or a minimum-size filter based on percentage move or Average True Range.
- Swing identification is inherently parameter-dependent: different N values or size thresholds produce different swing points from the same underlying price data.
- A swing chart is most often built on top of an existing candlestick series rather than replacing it outright, so the underlying bar type, including Heikin-Ashi, still matters.
- Reading swing structure (higher highs/higher lows vs. lower highs/lower lows) is a description of what already happened, not a forecast of what happens next.
What Is a Swing Chart?
A swing chart answers a narrower question than a full price chart: of all the back-and-forth movement in a security's price, which points actually mattered? It does this by discarding every bar except the ones marking a genuine change in direction, the swing highs (local peaks) and swing lows (local troughs), and drawing a line connecting only those points in sequence.
This makes swing charts conceptually related to how Point and Figure, Renko, and Kagi charts each filter out noise in their own way, but there's an important distinction: those three are standalone chart-construction methods that replace the underlying bar structure entirely. A swing chart is typically overlaid on or derived from standard candlestick data specifically to highlight market structure, rather than replacing the candles themselves. You can still see the individual candles underneath; the swing line is an added layer marking which candles were structurally significant.
Why Traders Use Swing Points
Swing highs and swing lows are the raw material for describing trend structure in plain, falsifiable terms. Instead of relying on a moving average's slope or a subjective "this looks like an uptrend" impression, a trader can point to a specific sequence of swing highs and swing lows and say precisely why the structure looks bullish, bearish, or directionless, see the market-structure section below.
How Are Swing Highs and Lows Identified?
There are two common approaches, and they can be combined.
The N-Bar Fractal Rule
The most widely used method is a simple bar-counting rule:
Swing high (N-bar rule) = a bar whose high is higher than the highs of the N bars immediately before it and the N bars immediately after it.
Swing low (N-bar rule) = a bar whose low is lower than the lows of the N bars immediately before it and the N bars immediately after it.
A common choice is N = 2, sometimes called a 5-bar fractal (the pivot bar plus 2 on each side). A larger N requires a more pronounced turning point to qualify, which produces fewer, larger swings; a smaller N flags more, smaller swings.
The Minimum-Size Filter
An alternative or complementary method ignores the bar-counting pattern entirely and instead requires a candidate swing to move by at least a minimum size before it counts, commonly expressed as a percentage of price (e.g., at least a 3% move from the prior swing point) or as a multiple of Average True Range (ATR), which scales the threshold to the security's own recent volatility instead of using a fixed percentage. A move smaller than the threshold is treated as noise and folded into the existing swing rather than starting a new one.
Why Swing Identification Is Inherently Subjective
Both methods require a parameter chosen in advance, N for the fractal rule, or the size threshold for the minimum-size filter, and that choice is not dictated by the price data itself. This is a real limitation, not a minor technical detail: the exact same underlying price series will produce a materially different set of swing highs and swing lows depending on which N or which threshold is selected. Two traders looking at the same chart with different swing settings can reasonably disagree about where the last significant swing point actually was. Swing points are a useful simplification of price action, not an objective fact about it.
Common Mistake
The common mistake is treating a chart's default swing setting as if it were the "correct" one. There is no universally correct N or threshold, the right setting depends on the timeframe and the trader's own tolerance for noise versus lag, and changing it after the fact to make a prior pattern look cleaner is a form of hindsight bias.
Worked Example: Identifying Swing Points With a 2-Bar Rule
Illustrative, hand-verified numbers, not live market data.
The table below applies the N = 2 fractal rule to a nine-day synthetic price series. A day qualifies as a swing high only if it has at least 2 bars on each side to compare against and its high exceeds the highs of both bars on each side; a swing low follows the same logic using lows.
| Day | High | Low | Classification (N = 2) |
|---|---|---|---|
| Day 1 | 50.00 | 49.20 | Not enough neighbors (needs 2 bars on each side) |
| Day 2 | 51.50 | 50.30 | Not enough neighbors |
| Day 3 | 52.80 | 51.60 | Swing high, 52.80 exceeds Days 1-2 and Days 4-5 highs |
| Day 4 | 52.10 | 50.90 | No pivot, Day 6's high (53.60) exceeds it |
| Day 5 | 51.40 | 50.10 | Swing low, 50.10 is below Days 3-4 and Days 6-7 lows |
| Day 6 | 53.60 | 51.20 | No pivot, Day 7's high (55.90) exceeds it |
| Day 7 | 55.90 | 53.40 | Swing high, 55.90 exceeds Days 5-6 and Days 8-9 highs |
| Day 8 | 54.70 | 52.80 | Not enough neighbors |
| Day 9 | 53.90 | 52.10 | Not enough neighbors |
Day 3 qualifies as a swing high because its high of 52.80 is greater than both Day 1 (50.00) and Day 2 (51.50) before it, and both Day 4 (52.10) and Day 5 (51.40) after it. Day 5 qualifies as a swing low because its low of 50.10 is less than both Day 3 (51.60) and Day 4 (50.90) before it, and both Day 6 (51.20) and Day 7 (53.40) after it. Day 7 qualifies as a swing high because its high of 55.90 exceeds Day 5 (51.40), Day 6 (53.60), Day 8 (54.70), and Day 9 (53.90). Days 4 and 6 look like plausible turning points at a glance, but each is disqualified by a bar just past it that goes further in the same direction, Day 4's high is exceeded by Day 6, and Day 6's low doesn't undercut Day 4's low, which is exactly the kind of case where the N-bar rule earns its keep over an eyeballed reading of the chart.
The resulting swing sequence for this series is: swing high at Day 3 (52.80) → swing low at Day 5 (50.10) → swing high at Day 7 (55.90). Because Day 7's swing high (55.90) is higher than Day 3's swing high (52.80), this short sequence describes a higher high, one building block of an uptrend structure, covered next.
Common Mistake
The common mistake is applying the rule only to the "obvious" turning points on the chart and skipping the mechanical check on borderline bars like Day 4 and Day 6 above. The N-bar rule is only as reliable as its consistent, mechanical application, selectively applying it produces a different, less reproducible set of swing points than applying it uniformly to every bar.
How Do Swing Points Reveal Market Structure?
Once a series of swing highs and swing lows has been identified, comparing each new pivot of a given type to the prior pivot of the same type describes trend structure in plain, checkable terms:
- Higher highs and higher lows, each new swing high exceeds the last swing high, and each new swing low exceeds the last swing low. This sequence describes an uptrend structure.
- Lower highs and lower lows, each new swing high falls short of the last swing high, and each new swing low falls short of the last swing low. This sequence describes a downtrend structure.
- Mixed sequence, for example, a lower high following a higher low, or vice versa. This describes a range or a potential change in structure, rather than a clean trend in either direction.
In the worked example above, the two-swing-high sequence (52.80 then 55.90) is a higher high on its own; a full read on trend structure would also need the swing-low sequence to confirm, and a longer run of data than nine bars to be meaningful in practice.
Swing Charts and the Underlying Bar Type
Because a swing chart is a layer identifying turning points rather than a replacement for the underlying bars, it can be built on top of any bar type, including Heikin-Ashi candles. Heikin-Ashi already smooths price into a trend-following series with its own averaging formula; running swing-point identification on top of Heikin-Ashi bars instead of standard candles can produce a cleaner-looking (but not necessarily more accurate) swing sequence, since some of the noise Heikin-Ashi already smooths away never reaches the swing-detection step. Traders who use swing structure as part of a broader plan for entering and exiting positions around those pivots may find related context in Swoopr Investment's trading strategies guide.
Misconceptions Versus Reality
| Misconception | Reality |
|---|---|
| A swing chart is a standalone alternative chart type like Renko or Point and Figure | A swing chart is typically an overlay identifying turning points on top of existing candlestick (or other bar-type) data, not a replacement for the underlying bars |
| Swing highs and lows are an objective fact about the price series | Swing points depend on a chosen parameter, the N in an N-bar rule, or a size threshold, and different reasonable choices produce different swing points from the identical price data |
| A wider N always produces "better" swing points | A wider N filters more noise but also reacts later to genuine turning points; the right setting is a tradeoff, not a universal default |
| Identifying a higher-high, higher-low structure predicts the trend will continue | Swing structure describes what already happened in the price series; it does not forecast whether the next swing will continue or break the pattern |
Risks, Limitations, and Exceptions
- Swing point identification is parameter-dependent, changing N (fractal rule) or the size threshold (minimum-size filter) changes which bars qualify as swing points, sometimes substantially.
- The N-bar fractal rule requires N confirming bars on each side, meaning a genuine swing point cannot be confirmed until N bars after it have printed, there is an inherent lag between when a swing actually occurred and when it can be confirmed.
- Neither method is standardized across charting platforms; two platforms' "swing high" indicators can disagree on the same data if their default N or threshold differs.
- Swing structure is a descriptive summary of price history, not a trading signal or a prediction of future price direction by itself.
- The worked example in this guide uses illustrative, deterministic numbers, not live market data, and is limited to nine bars, real swing analysis typically uses a longer history.
The Most Recent Swing Is Always Provisional
An N-bar fractal rule requires N confirming bars on each side, which has an implication worth sitting with: the newest swing point on your chart is not confirmed yet, and it may not survive. Whatever structure you are reading at the right-hand edge, the part that matters most for a current decision, is the least settled part of the chart. That gap between when a swing occurred and when it can be confirmed is built into the method rather than being a limitation of a particular setting.
The parameter dependence compounds it. A different N, or a different minimum-size threshold based on percentage or Average True Range, produces a different set of swing points from identical data, sometimes substantially different. Two platforms with different defaults will disagree about the same chart, and neither is malfunctioning.
Because of that, fix your definition and apply it consistently rather than adjusting until the structure looks clean. Structure read through a setting chosen to produce a tidy sequence of higher highs and higher lows is structure you selected.
And remember the overlay is built on an underlying series. If those bars are Heikin-Ashi or another derived format, the swing points inherit that smoothing and no longer mark real traded extremes. Swing structure summarises what price has already done, which is useful for orientation and is not a forecast.
Frequently Asked Questions
What Is a Swing Chart?
A swing chart filters price action down to a directional line connecting only the significant swing highs and swing lows in a series of bars or candles, ignoring the smaller back-and-forth moves in between. It is typically overlaid on or derived from standard candlestick data to highlight market structure, such as higher highs and higher lows in an uptrend, rather than functioning as a fully standalone chart type the way Renko or Point and Figure do.
How Are Swing Highs and Lows Identified?
The most common method is a simple N-bar fractal or pivot rule: a swing high is a bar whose high is higher than the highs of N bars on either side, and a swing low is a bar whose low is lower than the lows of N bars on either side. An alternative method filters swings by a minimum size, such as a percentage move or a multiple of Average True Range, discarding any swing smaller than that threshold regardless of the bar-count pattern.
How Do Swing Points Reveal Market Structure?
Once swing highs and lows are identified, comparing each new pivot to the prior one of the same type describes trend structure in plain terms: a sequence of rising swing highs and rising swing lows describes an uptrend, a sequence of falling swing highs and falling swing lows describes a downtrend, and a mix of the two, such as a lower high following a higher low, describes a range or a potential change in structure.
How many bars should the swing lookback be?
There is no value derived from theory, only a tradeoff. A short lookback such as two bars marks many swing points, most of which are minor and some of which vanish from relevance quickly. A longer one marks fewer and more durable points but confirms each one later. The setting should follow the structure you are trying to describe rather than the other way round, and a chart is not comparable to another with a different setting.
What is the difference between a bar-count swing rule and a percentage swing filter?
A bar-count rule asks whether a high is the highest of the surrounding N bars, so it can mark a turn that moved very little in price. A percentage filter asks whether price has retraced a set amount, so it ignores small turns regardless of how many bars they took. The two disagree most in slow grinding trends, where the bar rule finds swings the percentage filter rejects, and in fast volatile ranges, where the opposite happens.
Are swing highs and lows the same as pivot points?
No, although the words overlap in casual use. Pivot points in technical analysis are levels computed by formula from a prior period high, low and close, and they exist before price reaches them. Swing highs and lows are located from realised price structure and only exist once the surrounding bars have printed. One is a projected level, the other is an observation about what already happened.
Can one bar be both a swing high and a swing low?
Under a plain bar-count rule an outside bar can satisfy both conditions at once, since its high exceeds the neighbouring highs and its low undercuts the neighbouring lows. Implementations resolve this differently: some assign the bar to whichever direction the sequence was already in, others exclude outside bars from consideration. It is worth knowing which rule your chart applies, because the two produce different structure counts.
Do swing charts behave differently on intraday timeframes?
The rule itself is timeframe agnostic, but the input is not. Intraday data contains far more small reversals, so a lookback that produces a readable structure on a daily chart can produce a dense cluster of swing points on a five-minute chart. Practitioners usually compensate by lengthening the lookback or adding a minimum-size filter rather than by reading the raw output.
What happens to swing points after a stock split or dividend adjustment?
A split or dividend adjustment rescales the historical price series, which leaves a percentage-based swing filter unchanged because the proportions are preserved. A fixed-currency minimum-size filter is affected, since the same absolute threshold now represents a different fraction of price across the adjusted history. Swing structure that looked consistent on an unadjusted chart can change once the adjustment is applied.
References
Swing-point identification follows long-standing, publicly documented technical-analysis conventions (the N-bar fractal/pivot rule and volatility- or percentage-based swing filters). Key reference sources include:
- CME Group, Technical Analysis Basics: cmegroup.com/education/courses/technical-analysis.html: exchange-published education covering chart-pattern and trend-structure fundamentals this guide builds on.
- CFA Institute, Technical Analysis Reading Materials: cfainstitute.org: Technical Analysis: professional-standard coverage of trend and reversal-pattern identification methodology.
The nine-day 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.