What is Elliott Wave theory?

By AnalytikCapital · Updated October 3, 2026

Elliott Wave theory is a method of market analysis, developed by Ralph Nelson Elliott in the 1930s, which holds that prices advance in five-wave motive patterns and correct in three-wave patterns, and that these patterns repeat at every scale of time.

Where the theory comes from

Ralph Nelson Elliott (1871-1948) was an American accountant who, in the 1930s, studied decades of stock index data and concluded that price swings were not random but followed a small set of recurring shapes. He described them in The Wave Principle (1938) and later in Nature’s Law (1946). His central claim was that these shapes reflect crowd psychology: optimism and pessimism alternate in a recognizable sequence, and that sequence leaves the same footprint on a chart whether the move lasts minutes or decades.

The theory stayed relatively obscure until A. J. Frost and Robert Prechter published Elliott Wave Principle: Key to Market Behavior in the late 1970s. Prechter’s firm, Elliott Wave International, remains the main reference for the canonical rules today.

The basic 5-3 structure

Elliott divided every trend into two kinds of waves:

• Motive waves move in the direction of the trend of one larger degree and subdivide into five waves, labeled 1, 2, 3, 4, 5. Waves 1, 3 and 5 push the trend; waves 2 and 4 interrupt it.
• Corrective waves move against the larger trend and usually subdivide into three waves, labeled A, B, C.

Together, five waves up and three waves down form a complete eight-wave cycle. In a falling market the same structure is simply inverted: five waves down, three up. The most common motive wave is the impulse; the rarer diagonal is a wedge-shaped motive wave covered below.

The three rules of an impulse

Rules are what make a count falsifiable. If any one is broken, the impulse count is wrong and must be replaced:

1. Wave 2 never retraces more than 100% of wave 1. If price falls below the start of wave 1, there was no wave 1.
2. Wave 3 is never the shortest of waves 1, 3 and 5. It is frequently the longest, but the rule only forbids it from being the shortest.
3. Wave 4 does not overlap wave 1. In a standard impulse, wave 4 does not enter the price range of wave 1.

Sources phrase these rules slightly differently. StockCharts’ ChartSchool, for example, states the third rule as “wave 4 retraces less than 100% of wave 3”, which is a weaker condition than the no-overlap rule used by Elliott Wave International and Wikipedia. Elliott Wave International also notes that brief overlaps can occur in highly leveraged futures markets, usually only on daily or intraday charts.

The no-overlap rule has one important structural exception: diagonals. In an ending diagonal, which typically appears in the fifth-wave or C-wave position, wave 4 almost always overlaps wave 1 and the waves subdivide into threes. Leading diagonals appear in the first-wave or A-wave position and may subdivide either in threes or as 5-3-5-3-5. Overlap is therefore the defining trait of a diagonal, not a violation of the impulse rules, and a count has to commit to one pattern or the other before the overlap test means anything.

Guidelines: what is typical, not required

Guidelines describe what usually happens. Breaking one makes a count less likely, not invalid.

• Alternation. Waves 2 and 4 tend to differ in form: if one is sharp and deep, the other tends to be sideways and shallow.
• Extension. Usually one of the three motive waves is stretched, with clearly visible subdivisions. Which one extends varies by market and degree.
• Channeling. An impulse often travels inside a parallel channel drawn across the ends of waves 2 and 4, which helps estimate where wave 5 may end.
• Fibonacci relationships. Waves often relate to each other by ratios such as 0.618 and 1.618. These are covered on the Fibonacci ratios page.

Corrective patterns

Corrections are harder to read than impulses because they meander. Elliott Wave International groups them into four families:

The difference between a zigzag and a flat depends on whether wave A subdivides into five or into three. That one decision often changes the whole forecast.

Degrees and fractality

Each wave is itself made of smaller waves of the same shapes. Elliott named the levels of this hierarchy degrees, from the multi-century Grand Supercycle down to the Subminuette, which spans minutes. Most references list nine named degrees, and in principle the hierarchy continues beyond them in both directions. A five-wave impulse at one degree becomes wave 1 (or A, or C) of the next degree up.

This nesting is why Elliott analysis is usually multi-timeframe: a count on a daily chart has to agree with the structure on the weekly and hourly charts.

Why counts are subjective, and what an algorithm must decide

The rules exclude counts, but they rarely leave only one. Elliott Wave International states that two or more valid interpretations usually exist at any time; the analyst ranks them by how many guidelines each satisfies. Investopedia’s summary puts the problem more bluntly: several practitioners can produce several counts for the same chart.

Any software that counts waves has to turn those judgment calls into explicit choices:

• Pivot detection. Which highs and lows are real swings and which are noise. A small change in the swing threshold changes the count.
• Degree assignment. Which swings belong to the same degree, so that a wave 3 is compared with the wave 1 of the same level.
• Pattern choice. Impulse or diagonal, zigzag or flat, and which corrective family fits.
• Alternatives. How to rank the valid counts that remain, and when to say no count qualifies.

The criticism

Elliott Wave theory is one of the most debated ideas in technical analysis, and the criticism is serious:

• Subjectivity. Analysts often disagree on where waves begin and end. Benoit Mandelbrot and David Aronson, among others, have argued that the loose definitions leave too much to the analyst’s judgment.
• Hindsight. Counts are much easier to draw on a finished chart than in real time. When a forecast fails, the count can be revised, so the theory itself is rarely put at risk.
• Falsifiability. Because guidelines are flexible and alternates are always available, critics argue that the theory can fit almost any price path. Academic work on automated wave recognition notes that the rules lack precise criteria for systematic testing and that counts can be adjusted to match known outcomes.
• No time dimension. The theory says little about how long a wave should take.
• Evidence on ratios. A 2005 study by Batchelor and Ramyar, cited on Wikipedia, did not find Fibonacci ratios between waves in Dow Jones data more often than chance would predict.

Supporters answer that the hard rules do give objective invalidation levels, and that the method is a way to organize scenarios and risk rather than a forecast of certainty. Both positions are reasonable to hold; this page does not settle the debate.

How AnalytikCapital applies it

AnalytikCapital does not ask an analyst to choose the count. Its engine fixes pivot detection, degree assignment and pattern scoring in code, so the same chart always produces the same count. It enforces the three impulse rules as hard filters, measures the wave 4 overlap on closing prices, and reads the third rule literally: a count is rejected only when wave 3 is shorter than both wave 1 and wave 5, so impulses with an extended first or fifth wave remain possible. Guidelines and Fibonacci ratios only affect how candidates are ranked. The methodology page describes each step and where the engine departs from textbook practice.

Educational content, not financial advice. Trading involves substantial risk of loss.