Elliott Wave theory claims markets move in a repeating rhythm: five waves in the direction of the larger trend, three waves against it. It is the most ambitious framework in technical analysis — a complete grammar of price — and also the easiest to fool yourself with. Both facts matter.
The basic structure
A trending move, in Elliott terms, unfolds as five waves: 1, 3 and 5 push with the trend, while 2 and 4 are corrections between them. The whole advance is then corrected by three waves against it, labelled A-B-C. The claimed engine is crowd psychology: wave 1 is early conviction, wave 2 the doubt that almost erases it, wave 3 the broad recognition move — usually the longest and most powerful — wave 4 consolidation as early money exits, and wave 5 the late-crowd push that runs out of buyers, setting up the A-B-C unwind.
The theory is fractal: each wave subdivides into the same pattern at a smaller degree. Wave 3 of a daily count contains its own five waves on the 4-hour chart, and so on down. This nesting is elegant — and it is also the escape hatch that lets a broken count survive by being relabelled at a different degree.
The three rules
Everything else in Elliott is a guideline, but three rules are absolute. If any breaks, the count is wrong:
- Wave 2 never retraces more than 100% of wave 1. A new low below the start of the move kills the bullish count outright.
- Wave 3 is never the shortest of waves 1, 3 and 5.
- Wave 4 never enters the price territory of wave 1 (in the strict form).
These rules are the framework's most practical gift, because they convert a fuzzy narrative into hard invalidation points. "Long, unless price breaks below the wave 1 start" is a testable statement with a stop attached — whatever you think of the theory behind it.
Corrections, briefly
Elliott's taxonomy of corrections is large — zigzags (sharp A-B-C), flats (sideways), triangles (converging) and their combinations. The honest summary for a beginner: corrections are where wave counting goes to die. They overlap, extend and mutate, and even committed Elliott practitioners freely admit that counting a correction in real time is closer to weather forecasting than arithmetic. The common practical takeaways are simpler: corrections tend toward well-known retracement depths (wave 2 deep, wave 4 shallow), and they alternate in character — if wave 2 was sharp, wave 4 tends to be sideways.
The subjectivity problem
Here is the criticism, stated plainly: give the same chart to five wave analysts and you can get five different counts, each internally valid. Because the framework is fractal and its wave definitions are flexible, almost any price path can be labelled after the fact — which makes the theory nearly unfalsifiable as a whole, even though each specific count is falsifiable. That asymmetry is why Elliott commentary always sounds confident and why its practitioners always have an "alternate count" ready. Treat any wave analysis you read — including your own — as a scenario with an invalidation level, never as a forecast.
What it is actually good for
Used modestly, three things survive scrutiny. First, the impulse/correction distinction: learning to see the difference between direct, energetic movement and overlapping, reluctant movement is genuinely valuable, and Elliott trains that eye. Second, invalidation discipline: a count forces you to name the price at which you are wrong before you enter. Third, the wave-3 mindset — the idea that the middle of a recognized trend is where the easy distance is, and that the fifth push in an old trend is where the crowd is largest and the fuel lowest. You can hold all three ideas without believing markets obey a universal five-count.
If you go deeper, go deeper skeptically: paper-count in real time, log your invalidations, and score yourself the way you would score any system. The framework earns exactly as much trust as your own logged results give it — no more.