In the early 1930s an American accountant named Ralph Nelson Elliott, bedridden after an illness contracted in Central America, spent his recovery charting decades of Dow data by hand. In 1938 he published The Wave Principle, arguing that crowd behaviour pushes prices forward in five waves and corrects them in three. Nearly a century later his labels still cover charts in every trading room, usually drawn three different ways by three different people.
The five-three structure
The core claim is simple. In the direction of the larger trend, price advances in a five-wave impulse: waves 1, 3 and 5 push with the trend, waves 2 and 4 pull against it. Against the trend, price corrects in three waves, labelled A, B and C. Each wave subdivides into smaller versions of the same pattern, so a wave 3 on the daily chart contains its own five waves on the 4-hour chart, and so on down through the degrees.
That fractal quality is why the theory feels so complete. Any chart, any timeframe, any market can be labelled. It is also the first warning sign: a framework that can describe everything after the fact needs strict rules to say anything useful in advance.
The three rules that cannot bend
Elliott analysis separates rules from guidelines, and the rules are few. Wave 2 may not retrace beyond the start of wave 1. Wave 3 may not be the shortest among waves 1, 3 and 5, and in practice it is usually the longest and most violent. Wave 4 may not overlap the price territory of wave 1 in a standard impulse.
Everything else, alternation between sharp and flat corrections, channel projections, the personality of each wave, is guideline territory. The rules matter because they define invalidation. If your count says a wave 2 is finishing and price then trades through the start of wave 1, the count is dead and the stop is obvious. Used this way, Elliott gives you something most discretionary methods lack: a price at which you are provably wrong. That discipline pairs naturally with plain market structure reading, which asks the same question with fewer labels.
Fibonacci and wave targets
Elliott's followers, most influentially through the work popularized in the 1970s and 1980s, tied the wave structure to Fibonacci ratios. Wave 2 commonly retraces 50 to 61.8 percent of wave 1. Wave 3 often extends to 161.8 percent of wave 1. Wave 4 tends to hold the 38.2 percent retracement of wave 3. None of these are laws; they are observed tendencies that give the analyst zones to watch rather than exact prices.
In day-to-day trading this is the part of the theory that survives contact with reality best. You do not need a full count to act on the idea that a strong impulse is usually followed by a partial retracement into a measurable zone. A standard Fibonacci retracement drawn on the last clear impulse captures most of the practical value with a fraction of the ceremony.
Where the theory breaks down
The honest criticism is not that Elliott is wrong; it is that it is under-determined. At almost any moment, several counts satisfy all three rules simultaneously. Is this a wave 4 of a larger impulse or wave B of a developing correction? The chart cannot tell you until it has already moved, and by then the losing counts have quietly been redrawn. Analysts call this recounting; sceptics call it curve fitting in real time.
Corrections are the worst offenders. Elliott catalogued zigzags, flats, triangles and combinations, and a sideways market can be labelled as half of them at once. Currency pairs, which spend most of their life in overlapping ranges, are particularly cruel to wave counters. Trending equity indices are kinder, which is where Elliott built the theory in the first place.
A wave count is a scenario with an invalidation price, never a prediction. The moment you catch yourself defending a count instead of trading the level, the theory has stopped working for you.
Using waves without becoming a wave trader
There is a productive middle ground between full nine-degree labelling and dismissing the whole thing. Take three ideas and leave the rest. First, strong moves subdivide: after an initial impulse and a shallow pullback, continuation is the higher-probability path, which is the same insight that drives trend following. Second, third waves are where the money is; the middle of a trend is more forgiving than picking its start or end. Third, overlapping, choppy structure is corrective until proven otherwise, so trade it smaller or not at all.
Traders who want the crowd-behaviour framing without the labelling burden often get further with the Wyckoff method, which describes accumulation and distribution in terms of ranges and breakouts rather than numbered waves. Both schools are trying to read the same thing: who is in control and how far they have already pushed. Elliott just insists on more paperwork.
Whatever the method, the risk rules do not change. A count, however elegant, sizes no position and rescues no account. Position sizing and hard stops do, and they work the same whether you believe in waves or not.
"I have never met two wave analysts holding the same count on the same chart. That tells you exactly what a count is: an opinion with labels, and the only part I trade is the invalidation price."
— Alex Onta, Executive Director, SINGUARD
Key Takeaways
- Elliott Wave describes markets as five-wave impulses with the trend and three-wave corrections against it, repeating fractally across timeframes.
- Only three rules are absolute: wave 2 never exceeds the start of wave 1, wave 3 is never the shortest, and wave 4 does not overlap wave 1.
- Multiple valid counts almost always coexist, so treat any count as a scenario with a hard invalidation price rather than a forecast.
- The durable, tradeable core is small: impulses retrace into Fibonacci zones, third waves carry furthest, and overlapping chop is corrective.
Frequently Asked Questions
Does Elliott Wave theory actually work?
It works as a framework for organizing price structure, and it fails as a prediction machine. The three hard rules give you clear invalidation points, which is genuinely useful for risk control. The problem is that several valid counts usually exist at the same time, so two competent analysts can hold opposite views from the same chart. Treat a wave count as a scenario with a defined invalidation level, never as a forecast.
What are the three rules of Elliott Wave?
Rule one: wave 2 never retraces more than 100 percent of wave 1. Rule two: wave 3 is never the shortest of waves 1, 3 and 5. Rule three: wave 4 does not overlap the price territory of wave 1 in a standard impulse. Everything else in the theory is a guideline that can bend. If a rule breaks, the count is wrong and must be redrawn.
Is Elliott Wave useful for forex trading?
Partially. Currency pairs spend long stretches in ranges and overlapping corrections, which are the hardest structures to count. Most practical FX traders use Elliott ideas selectively: the expectation that a strong impulse is followed by a partial retracement, and Fibonacci zones for where that retracement might end. Full nine-degree wave labelling on a 5-minute EURUSD chart is more hobby than edge.