The market moves in waves of crowd psychology

Ralph Elliott noticed that markets do not move in straight lines. They move in repeating waves driven by swings between greed and fear — and the same shapes appear at every zoom level.
Green Call here! 🌊 Price is crowd emotion drawn on a screen, and crowds do not change their mind smoothly. They surge, doubt, surge harder, doubt again. Elliott's contribution was noticing that the doubting and surging happen in a countable rhythm — and that the rhythm looks the same on a 5-minute chart and a monthly one.

A full cycle is eight waves. First a 5-wave impulse moves with the main trend: waves 1, 3 and 5 push forward while waves 2 and 4 pull back. Then a 3-wave correction, labelled A-B-C, moves against it.
Impulse, then correction, over and over — and each of those waves is itself made of smaller waves following the same pattern. That fractal quality is what makes the theory elegant and what makes it slippery.

Most Elliott mistakes come from ignoring the three rules that make an impulse valid. These are not guidelines — break one and the count is simply wrong.
Wave 2 never retraces more than 100% of wave 1. Wave 3 is never the shortest of waves 1, 3 and 5. Wave 4 never overlaps the price territory of wave 1.
If your labelling breaks any of them, do not argue with the chart — recount. This is the single most useful discipline in the whole theory, because it gives an otherwise subjective method a hard failure condition.
Wave 3 is typically the longest and most powerful of the impulse — it is where the crowd finally agrees on the direction. If your count has a puny wave 3, you have almost certainly mislabelled the start.
Waves give you the shape; Fibonacci gives you the measurements. The two are used together almost universally.
Wave 2 often retraces 50–61.8% of wave 1. Wave 3 frequently extends to 161.8% of wave 1. Those ratios turn a qualitative count into projected levels — places to look for an entry, and places to expect a wave to end.

Elliott Wave is a framework for organising what you see, not a machine for predicting what comes next. Two skilled analysts can label the same chart differently and both be reasonable.
That subjectivity is not a reason to dismiss it — but it is a reason to trade it the same way you would trade anything else: with a defined invalidation level and a stop. The rules above conveniently give you one. If wave 4 overlaps wave 1, your read was wrong and you should already be out.
Historical charts are always beautifully countable, because you already know how the move ended. Counting in real time, on the right-hand edge, with three plausible labels available, is a completely different exercise. Judge the tool by the second one.
Five up, three back. Learn the rules, measure with Fibonacci, and always keep the stop that the count itself defines. 🏄
