Impulse vs Corrective Elliott Waves: Rules

An Elliott impulse wave is a specific five-wave motive structure; a corrective wave is a countertrend structure that can take several forms. The useful distinction is not simply “five waves up, three waves down.” You must define the next-larger trend, the wave degree, the pivots that set each endpoint, and the rule that would invalidate the count.

That distinction matters because the same visible move can receive different labels at different degrees. A five-wave decline may be an impulse within a larger downtrend, or it may be wave A of a larger upward-trend correction. A three-wave rally may be a completed correction, one leg of a triangle, or only part of a combination. Elliott Wave terminology describes a proposed hierarchy of price movements; it does not make the future count known in advance.

Key takeaways

A standard impulse must satisfy three structural rules; common Fibonacci relationships are only guidelines.
“Corrective” does not mean every correction is a simple three-leg A-B-C.
Motive and corrective labels are always relative to a stated wave degree and next-larger trend.
Candidate counts should be frozen before later candles are revealed.
Keep a preferred count, an alternate count, and a precise invalidation condition.
Do not turn a wave label into an automatic entry, stop, target, or profitability claim.

Educational note: Elliott Wave analysis is interpretive technical analysis. It can organize chart observations, but it cannot guarantee a forecast or identify the intentions of market participants.

Which Elliott Wave Page Should You Use?

ChartMini separates broad Elliott Wave theory from the narrower classification problem covered here.

Question	Owner
What is Elliott Wave Theory and the broad 5-3 cycle?	Elliott Wave Theory Guide
How do I classify impulse versus corrective waves and record invalidation?	This guide
How do I draw and interpret Fibonacci retracements?	Fibonacci Retracement Guide
How do I define swing highs, swing lows, and trend structure first?	Market Structure Guide
How do I practice pattern recognition without seeing future candles?	Chart Replay Pattern Recognition

This page does not provide a Wave 3 entry system, a Wave 5 exit signal, fixed retracement percentages, or a market forecast.

Elliott Wave Terminology: Motive Is Broader Than Impulse

A common beginner shortcut treats motive wave and impulse wave as identical. They are related, but not identical in standard Elliott terminology.

A motive wave progresses in the direction of the next-larger trend.
An impulse is one type of motive wave with a five-wave internal structure and three hard rules.
A diagonal is another motive structure. It can contain overlap that would invalidate a standard impulse.
A corrective wave moves against the direction of the next-larger trend and can form a zigzag, flat, triangle, or combination.

This means “price moved strongly” is not enough to call a move an impulse. Speed, candle size, volume, and visual steepness may be observations, but the impulse label depends on structure.

Direction is relative to degree

Assume a weekly chart is in a larger upward move. A daily decline can be corrective relative to that weekly rise. Inside that daily decline, an hourly five-wave drop can still be motive relative to the hourly degree.

Therefore, every Elliott Wave record should state:

the analysis timeframe;
the proposed wave degree;
the direction of the next-larger count;
the exact cutoff candle;
the pivots used as wave endpoints.

Without these fields, “impulsive decline” and “corrective rally” can mean different things to different analysts.

The Basic Impulse and Correction Sequence

The familiar educational diagram shows:

waves 1-2-3-4-5 progressing with the next-larger trend;
waves A-B-C correcting part of that progress.

This is a model, not a promise that every market segment will display a clean eight-wave sequence.

In a rising example

Wave 1 advances.
Wave 2 retraces part of wave 1.
Wave 3 advances.
Wave 4 retraces part of wave 3.
Wave 5 advances.
Wave A declines.
Wave B rises against wave A.
Wave C declines.

In a falling example, the directions are reversed.

The labels describe relationships within a count. They do not independently reveal whether a move is caused by optimism, fear, institutions, algorithms, or another participant group.

The Three Hard Rules of a Standard Impulse

A rule determines whether a proposed standard impulse remains structurally possible. A guideline describes a common tendency but can be absent without invalidating the count.

Rule 1: Wave 2 cannot move beyond the start of wave 1

For a rising impulse candidate:

wave 1 begins at a defined low and ends at a defined high;
wave 2 may retrace deeply;
if wave 2 moves below the start of wave 1, that specific impulse count is invalid.

For a falling impulse candidate, reverse the direction.

The endpoint convention must be consistent. Decide whether wave endpoints use the extreme high/low, a closing price, or another rule. Standard Elliott diagrams normally use price extremes. Changing from wicks to closes only after an overlap appears introduces hindsight.

Rule 2: Wave 3 cannot be the shortest of waves 1, 3, and 5

This rule is often misstated as “wave 3 must be the longest.” It does not have to be the longest. It cannot be the shortest among the three actionary waves.

Measure all three waves using the same price convention:

Wave length = absolute value of ending price - starting price

The rule can only be evaluated completely after candidate wave 5 has an endpoint. While wave 3 is still developing, the analyst can record constraints, but cannot know the final comparison with wave 5.

Rule 3: Wave 4 cannot overlap wave 1 price territory

In a standard rising impulse, wave 4 should not move into the price territory of wave 1. In a falling impulse, reverse the direction.

An overlap does not prove that the chart has no Elliott interpretation. It means the proposed standard impulse is invalid under that endpoint rule. Possible alternatives include:

a diagonal rather than an impulse;
a larger correction;
different wave endpoints;
a different wave degree;
a count that started at another pivot.

Do not silently convert the count to a diagonal after the overlap unless diagonal criteria were defined and the revised count is recorded as a new version.

Rules Versus Guidelines

Separating rules from guidelines prevents a preferred ratio or visual expectation from being treated as proof.

Observation	Type	What it means
Wave 2 does not pass the start of wave 1	Hard impulse rule	Violation invalidates that standard impulse count
Wave 3 is not the shortest of 1, 3, and 5	Hard impulse rule	Requires all three lengths to assess fully
Wave 4 does not overlap wave 1 territory	Hard impulse rule	Overlap invalidates a standard impulse, but may suggest another structure
Wave 3 is often visually strong	Guideline	Can help describe a count but is not required in every case
Waves 2 and 4 may alternate in form or depth	Guideline	A tendency, not an invalidation test
Fibonacci ratios may appear between waves	Guideline	A measurement, not confirmation
Volume may differ between waves	Guideline or contextual observation	Depends on market, data source, and timeframe
Sentiment may change through a sequence	Narrative interpretation	Not directly proven by OHLC data

A reproducible count should first pass hard structural rules. Guidelines may then be logged as supporting, neutral, or conflicting evidence.

How to Define Wave Endpoints Before Counting

Wave counting becomes impossible to reproduce when endpoints move each time the count encounters a problem.

Choose a pivot rule

Possible pivot definitions include:

a fixed number of completed bars on both sides of an extreme;
a percentage reversal threshold;
an ATR or range-based threshold;
a manually selected swing that meets a written prominence rule;
a higher-timeframe swing mapped onto a lower timeframe.

No pivot method is universally correct. The important control is to use the same method throughout the sample.

Record endpoint fields

For every labeled wave, record:

Field	Example purpose
Start timestamp	Identifies the exact source candle
End timestamp	Prevents later movement of the endpoint
Start and end price	Allows length and retracement calculations
Wick or close convention	Makes overlap rules reproducible
Pivot confirmation time	Shows when the endpoint became knowable
Wave degree	Prevents mixing nested counts
Status	Candidate, confirmed under rule, invalidated, or alternate

A pivot can occur before it becomes confirmable. This delay must be included in any replay study.

Impulse Wave Classification Checklist

Use this checklist for a proposed standard impulse.

1. State the next-larger trend

Is the proposed five-wave sequence moving with the next-larger count? If the larger direction is unknown, label the sequence as a candidate five-wave movement rather than asserting its role.

2. Freeze five candidate endpoints

Record candidate waves 1 through 5 without changing the pivot method mid-count.

3. Check internal direction

In a rising impulse candidate:

waves 1, 3, and 5 progress upward;
waves 2 and 4 move against them.

Reverse the directions for a falling count.

4. Apply all three hard rules

Record pass or fail separately for each rule. Do not replace a failed rule with “close enough.”

5. Inspect subdivisions cautiously

Elliott analysis expects smaller structures inside larger waves. But lower-timeframe data can produce many possible subdivisions. Record the lower timeframe and the pivot rule rather than selecting only the subdivision that supports the preferred count.

6. Keep at least one alternate count

An alternate might classify the same sequence as:

three waves rather than five;
a diagonal;
part of a correction;
a different degree;
an unfinished wave.

7. Define invalidation before revealing more data

The invalidation condition may be a price boundary, a rule violation, or a structural development incompatible with the preferred count.

Diagonals: The Main Overlap Exception

Diagonals are motive structures but are not standard impulses. They matter because beginners often preserve an invalid impulse by saying, after the fact, that the overlap “must be a diagonal.”

A controlled process should define diagonal criteria before using the label. Record:

whether the diagonal is proposed as leading or ending;
the location in the larger count;
whether trendlines converge or diverge;
whether wave 4 overlaps wave 1;
the internal subdivision convention being used;
the point that invalidates the diagonal interpretation.

The presence of overlap alone does not prove a diagonal.

Corrective Waves Are More Varied Than A-B-C

Corrective structures generally move against the next-larger trend and are often more difficult to classify in real time. A complete correction is not always a single three-leg pattern.

Zigzag

A zigzag is commonly described as 5-3-5:

wave A subdivides as a five-wave motive move;
wave B subdivides correctively;
wave C subdivides as a five-wave motive move.

This explains why a five-wave movement against a larger trend does not necessarily mean the larger trend has reversed. It may be wave A of a broader zigzag.

Flat

A flat is commonly described as 3-3-5:

wave A is corrective;
wave B is corrective;
wave C is motive.

Different flat variants depend on how far B and C travel relative to earlier endpoints. Those relationships should be measured and labeled rather than forced to match a preferred percentage.

Triangle

A triangle is commonly labeled A-B-C-D-E, with each leg corrective. It can contract or expand depending on the endpoint geometry.

A triangle should not be declared complete merely because two trendlines can be drawn. Record each leg, the boundary rule, and the condition that ends or invalidates the triangle count.

Combination

Combinations connect corrective structures with one or more X waves. Common labels include:

W-X-Y;
W-X-Y-X-Z.

Combinations are one reason corrections can continue longer and become more complex than a simple A-B-C expectation.

Why “Corrections Are Never Fives” Needs Context

Elliott Wave educational material often states that a correction is not a simple five-wave structure. That does not mean no five-wave segment can appear inside a correction.

Examples:

wave A of a zigzag can subdivide into five waves;
wave C of a zigzag or flat can subdivide into five waves;
the entire correction includes additional legs and is not complete after the first five-wave segment.

Therefore, classify the degree and role of the five-wave move before concluding that the larger trend has changed.

Impulsive Versus Corrective Price Action

Traders often use “impulsive” informally to mean fast, directional, or composed of large candles. That informal use is not the same as an Elliott impulse.

Informal observation	Elliott classification question
Large directional candles	Does the structure subdivide and satisfy impulse rules?
Rising volume	Is the volume source comparable and is the structure valid?
Little overlap	Are five qualifying pivots actually present?
Choppy movement	Is it a correction, diagonal, lower-degree sequence, or simply unclassified?
Deep retracement	Does it violate the relevant count or remain within the rule?
Strong breakout	Which wave degree and endpoint rule make the label reproducible?

A move can look forceful yet fail the structural rules. A slower move can still be part of a valid motive sequence.

The Impulse-Correction-Impulse Sequence

Bing data for this page includes the query impulse correction impulse. A practical way to study this sequence is to separate observation from prediction.

First candidate impulse: Label only after five endpoints and the hard rules can be evaluated.
Candidate correction: Classify the countertrend structure without assuming it must end after three visible swings.
Second candidate impulse: Require a new structural count rather than assuming trend resumption because the correction appears complete.

Record three separate questions:

Did the first impulse remain valid?
Which correction type best fit the frozen data?
When did the next motive sequence become classifiable?

This avoids using the label “correction” as a promise that the prior direction will resume.

Waves Within Waves: Managing Degree

Elliott Wave proposes nested structures. The challenge is that a chart can support several plausible degrees.

Use a degree ladder

A simple record can use functional labels instead of trying to memorize every traditional degree name:

Degree A: next-larger context;
Degree B: primary count being evaluated;
Degree C: subdivisions used to inspect Degree B.

For each degree, record timeframe, pivot rule, and cutoff candle.

Do not mix degrees to rescue a count

Common degree errors include:

using a daily pivot for wave 1 and an hourly pivot for wave 4;
counting five lower-timeframe fluctuations inside only one higher-timeframe candle;
changing degree after a rule violation without preserving the old count;
comparing wave lengths measured from different adjustment methods.

If the degree changes, create a new count version.

Fibonacci Ratios: Measurement, Not Confirmation

Fibonacci ratios are widely associated with Elliott Wave analysis. They can be recorded as relationships between selected endpoints, but they do not prove that the endpoints are correct.

Possible measurements include:

retracement of wave 2 relative to wave 1;
length of wave 3 relative to wave 1;
retracement of wave 4 relative to wave 3;
length of wave 5 relative to wave 1;
length of wave C relative to wave A.

A controlled ratio workflow

Select wave endpoints under the pivot rule.
Freeze the count.
Calculate ratios from those endpoints.
Record the actual values.
Mark whether the values were supporting, neutral, or conflicting under the study's predefined ranges.
Do not move endpoints until a preferred Fibonacci percentage appears.

The Fibonacci Retracement Guide owns the broader drawing and interpretation workflow. When the comparison depends on which wick, close, body, or confirmed pivot supplied the endpoints, use the Fibonacci anchor and calculation rules and freeze that anchor method before evaluating the wave count.

Volume, Momentum, and Sentiment

Volume and momentum are often described as wave “personalities.” They can provide context, but they are not universal impulse rules.

Before recording volume:

state whether the data is exchange volume, tick volume, broker-specific volume, or another measure;
state the session and timeframe;
compare like-for-like bars;
avoid treating one volume pattern as mandatory across stocks, futures, spot FX, CFDs, and crypto venues.

A volume bar does not identify institutions, retail traders, market makers, or “smart money.” Momentum divergence also does not prove that wave 5 has ended. These observations should remain separate from the hard structural count.

Preferred Count, Alternate Count, and Invalidation

The strongest protection against hindsight is to record uncertainty explicitly.

Preferred count

The interpretation that best fits the current frozen data under the selected rules.

Alternate count

A plausible competing interpretation. It should use the same chart data and explain which endpoints or degree differ.

Invalidation

The exact event that makes a count structurally impossible under its rules.

Count field	Example format
Preferred count	Rising standard impulse, degree B
Alternate count	A-B-C correction inside degree A
Cutoff candle	Timestamp
Pivot method	Two completed bars on each side
Price convention	Wick extremes
Preferred invalidation	Wave 2 crosses wave 1 start
Alternate invalidation	Candidate C fails its structural definition
Next review point	After the next confirmed pivot

Do not delete invalidated counts. Keeping them allows you to measure how often a method changes labels.

Candidate, Confirmed, Invalidated, or Unresolved

Avoid using only “right” and “wrong.” A wave count develops over time.

Candidate

The pattern is plausible, but later endpoints or subdivisions are not yet available.

Structurally confirmed under the selected rules

The completed sequence satisfies the rules used by the study. This does not guarantee a future outcome.

Invalidated

A hard rule or predefined structural condition was violated.

Unresolved

More than one count remains plausible, or the available data does not support a reliable classification.

“Unresolved” is a legitimate result. Forcing every chart into a completed Elliott count increases hindsight and selection bias.

A Repeatable Elliott Wave Classification Checklist

Record instrument, venue, provider, price type, session, timezone, and adjustment method.
Choose the analysis timeframe and next-larger context.
Define the pivot confirmation rule.
Define wick-versus-close endpoint treatment.
Freeze the last visible candle.
Record the proposed wave degree.
Mark candidate endpoints with timestamps and prices.
Classify the candidate as impulse, diagonal, zigzag, flat, triangle, combination, or unresolved.
Apply hard rules before Fibonacci, volume, or momentum guidelines.
Record internal subdivisions only from a stated lower timeframe.
Write the preferred count.
Write at least one plausible alternate count.
Define invalidation for each count.
Save the chart state before revealing later candles.
Reveal one completed candle or one confirmed pivot at a time.
Record count revisions as new versions.
Separate structural validity from any hypothetical trade outcome.
Include invalidated and unresolved cases in the final sample.

Elliott Wave Replay Worksheet

Field	What to record
Instrument and provider	Exact market and source
Timeframe and session	Candle construction
Adjustment method	Split, dividend, or futures-continuation handling
Cutoff candle	Last information visible at decision time
Next-larger direction	Up, down, range, or unresolved
Wave degree	Functional or traditional degree label
Pivot rule	Exact confirmation logic
Endpoint convention	Wick, close, or another method
Candidate structure	Impulse, diagonal, zigzag, flat, triangle, combination, unresolved
Wave endpoints	Timestamps and prices
Hard-rule checks	Pass, fail, or not yet measurable
Preferred count	Frozen interpretation
Alternate count	Plausible competing interpretation
Invalidation	Exact price or structural condition
Fibonacci observations	Actual ratios, not fitted endpoints
Volume/momentum observations	Data source and timeframe
Next revealed data	Candle or confirmed pivot
Revision	No change, relabel, invalidated, unresolved
Version	Preserve every prior count

Example Without a Trade Recommendation

Suppose a rising sequence has candidate endpoints:

start: 100;
wave 1 end: 112;
wave 2 end: 105;
wave 3 end: 131;
wave 4 end: 116;
wave 5 end: 139.

Under a wick-extreme convention:

wave 2 remains above the start of wave 1;
wave lengths are 12, 26, and 23 for waves 1, 3, and 5, so wave 3 is not the shortest;
wave 4 remains above the wave 1 endpoint of 112, so no overlap occurs.

The sequence can be recorded as a structurally valid candidate standard impulse at the selected degree. That classification does not establish that price must next complete a particular correction, reach a Fibonacci target, or produce a profitable trade.

An alternate count might interpret the same movement at another degree. Both counts should remain in the record until one is invalidated.

Common Elliott Wave Counting Errors

Counting before pivots are confirmable

A turning point is easiest to see after later candles have formed. Record the confirmation delay instead of acting as though the endpoint was known at the extreme.

Repeating the same impulse rule twice

The three rules are distinct. The old version of this article accidentally repeated the wave 3 rule before correcting itself. A checklist prevents this error.

Saying wave 3 must be the longest

The hard rule is that wave 3 cannot be the shortest among waves 1, 3, and 5. “Usually strongest” or “often extended” is a guideline.

Treating all A waves as five-wave structures

Wave A in a zigzag can be motive, while wave A in a flat or triangle is corrective. Classification depends on the larger correction type.

Treating every correction as A-B-C

Triangles and combinations require additional legs and labels.

Using Fibonacci to choose the endpoints

Ratios should be calculated after endpoints are selected. Moving the endpoint to produce 61.8% or 1.618 creates circular confirmation.

Converting every overlap into a diagonal

Overlap invalidates a standard impulse, but does not automatically validate a diagonal.

Hiding alternate counts

Publishing only the count that later worked conceals real-time uncertainty.

Claiming a universal success rate

A success rate is meaningless without a fixed instrument set, timeframe, pivot rule, count rule, cutoff procedure, outcome definition, costs, and complete sample. Do not transfer a result from one study to all Elliott Wave analysis.

Treating a wave label as an order instruction

Wave labels do not define position size, liquidity, spread, slippage, fees, borrow availability, margin, or execution risk.

How to Practice Elliott Wave Classification With ChartMini

ChartMini can hide future historical candles so you can record a count before the later chart is visible.

A controlled exercise:

Open ChartMini.
Choose one instrument, timeframe, and session.
Freeze the visible lookback and write the next-larger context.
Apply one pivot rule.
Mark the preferred and alternate counts.
Record all endpoints and invalidation conditions.
Save the count version.
Advance one completed candle or wait for one newly confirmed pivot.
Mark the count unchanged, revised, invalidated, or unresolved.
Keep every version and every failed count.
Compare the same rules across a larger sample.

ChartMini supports lightweight historical candle replay. It does not automatically count Elliott Waves, select a wave degree, validate Fibonacci relationships, simulate a full live order book, model slippage and commissions, or prove profitability. Historical replay is not evidence that future prices will follow the same structure.

Official and Educational Sources

StockCharts ChartSchool: Introduction to Elliott Wave Theory — explains the broader motive-plus-corrective cycle and the distinction between impulse and diagonal motive waves.
StockCharts ChartSchool: Identifying Elliott Wave Patterns — covers impulse structures and sharp versus sideways corrections.
Elliott Wave International Waveopedia: Impulse — summarizes the structural definition of an impulse.
Elliott Wave International Waveopedia: Corrective Waves — describes the variety and complexity of corrective structures.
TradingView: Elliott Wave Theory and Tools — documents its impulse and correction drawing tools and notes that an automated pattern can be rebuilt as chart points change.
CFTC: Commodity Trading Systems Sold on the Internet — explains limitations of hypothetical and simulated results and warns against guaranteed-profit claims.

These sources explain the conventional Elliott Wave framework. They do not establish that one wave count, Fibonacci ratio, timeframe, or trading rule predicts future price reliably.

Frequently Asked Questions

What is the difference between an impulse wave and a corrective wave?

In Elliott Wave terminology, an impulse is a five-wave motive structure that progresses in the direction of the next-larger trend and follows three structural rules. A correction moves against the next-larger trend and can take several forms, including zigzags, flats, triangles, and combinations. A complete correction is not always a simple A-B-C shape.

What are the three rules of an Elliott impulse wave?

For a standard impulse, wave 2 cannot move beyond the start of wave 1, wave 3 cannot be the shortest of waves 1, 3, and 5, and wave 4 cannot overlap wave 1 price territory. A diagonal is a different motive structure and can allow overlap, so it should not be labeled as a standard impulse.

Are all corrective waves three waves?

No. A correction may be labeled A-B-C, but its internal structure depends on the type. A zigzag is commonly described as 5-3-5, a flat as 3-3-5, a triangle as five corrective legs, and a combination may use W-X-Y or W-X-Y-X-Z. A five-wave move against a larger trend may be only one leg of a broader correction.

Do Fibonacci ratios confirm an Elliott Wave count?

No. Fibonacci relationships are guidelines or measurements, not hard proof that a count is correct. A count should first satisfy its structural rules and use clearly defined pivots. Ratios can be recorded afterward as supporting or conflicting observations without moving the wave endpoints to fit a preferred percentage.

Why do Elliott Wave analysts produce different counts?

Different counts can result from different timeframes, pivot rules, wave degrees, chart adjustments, cutoff candles, diagonal assumptions, and correction classifications. A reproducible analysis records a preferred count, at least one plausible alternate count, and the price or structural condition that invalidates each one.

Can ChartMini automatically identify Elliott Waves?

No. ChartMini provides historical candle replay for manual chart-reading practice. It does not automatically label Elliott Waves, select the correct wave degree, validate a count, reproduce live execution, or prove that a wave method is profitable.

Elliott Wave Impulse vs Corrective Waves: Rules, Counts, and Invalidation