The Three Elliott Wave Laws, Stress-Tested
The laws of this framework do one job: they kill wrong counts, and what a broken law proves is always the same thing, that the pattern on your chart is not an impulse at that degree. The earlier lesson stated the three rules while building the five-wave shape; this one treats them as working tools, stress-tested against live trading. A law is not a warning, it is a verdict, and knowing what each verdict proves is what turns counting from label-sticking into a procedure.
Laws and Guidelines Are Different Machines
The framework carries two kinds of statements, and confusing them is the most common beginner error. Laws are structural: wave 2 never fully retraces wave 1, wave 3 is never the shortest of the three impulse waves, wave 4 never trades into wave 1 territory. Break one and the count does not weaken, it dies, because each law defines a boundary that an impulse at that degree cannot cross. Guidelines are tendencies: alternation, equality, channeling. They steer expectations and break all the time without invalidating anything.
The distinction has a practical consequence. A guideline failure asks for smaller size. A law failure asks for a new label. Traders who treat laws as strong guidelines hold broken counts and defend them with hope; traders who treat guidelines as laws exit good counts on ordinary noise. The first step in using this framework professionally is knowing which sentences in it are fences and which are advice.
Law One Protects the Origin
Wave 2 never retraces the full extent of wave 1. The rule exists because the start of wave 1 is the foundation of the whole count: it is the price that separates the new trend from the old one. A wave 2 that returns all the way to the origin has not corrected the advance, it has erased it, and a structure that erases its own first wave was never an advance at that degree in the first place.
What a violation proves, specifically: the move you labeled wave 1 was itself a correction, most likely wave A or wave B of a larger counter-trend structure, and the decline you labeled wave 2 is a fresh impulse at one degree higher. The relabeling is not a defeat. It is the framework telling you which structure you are actually inside, and the new label comes with its own, usually tradable, implications. A deep wave 2 that stops one tick above the origin is the same law working in your favor: the origin held, the count stands, and the depth itself is fuel for wave 3.
Law Two Protects the Middle
Wave 3 is never the shortest of the three impulse waves. This law is less about geometry than about honesty in labeling: in a genuine trend, the middle wave carries the crowd's arrival, and crowds produce the longest legs. A wave 3 that finishes shorter than both wave 1 and wave 5 describes a market whose strongest participation arrived first, which is the signature not of a trend but of a terminating pattern or a corrective rally.
What the violation proves: the advance you are counting is probably wave B of a correction, the kind of rally that looks strong early and fades because it was never a trend, or a first five of smaller degree about to be retraced entirely. Either way the correct response is the same, stand down from trend-following expectations and re-read the structure as corrective. The law acts as a recount trigger, and the recount usually finds the real trend living somewhere else on the chart.
Law Three Protects the Boundary
Wave 4 never trades into the price territory of wave 1. This is the boundary law: the territory above wave 1's extreme belongs to wave 3's advance, and a wave 4 that dips back inside it has destroyed the separation between the trend's beginning and its digestion. Overlap is the single most diagnostic event in the whole framework, because it usually appears at the exact moment the crowd is most committed to the trend story.
What the violation proves: the pattern is corrective, not impulsive. The practical reading that follows is nearly always a flat or a zigzag where what you called waves 1 through 5 were really the a, b and c internals of a larger B wave, and the market's next significant move is in the opposite direction to the count you were defending. Two refinements keep the law honest. First, overlap is judged on a structural close, not a two-tick wick; an intraday poke that closes back outside the boundary is a warning, not a verdict. Second, the law applies within one degree at a time; wave 4 at one degree overlapping wave 1 of a smaller degree inside it is not a violation, it is nesting.
| Law | What it protects | What a violation proves |
|---|---|---|
| Wave 2 never fully retraces wave 1 | The origin of the trend | The "impulse" is itself a correction at a higher degree |
| Wave 3 is never the shortest | The trend's middle conviction | The rally is a B wave or a smaller-degree first five |
| Wave 4 never enters wave 1 territory | The boundary of the advance | The pattern is corrective: relabel and reverse the bias |
The kill-switch habit here is the one the Dow lessons built: the level decides, in writing, before the trade.
A Worked Example: The Recount Procedure
A hypothetical instrument, invented numbers throughout. A decline has carried it from 260 to 200, and a rally follows that a trader labels as a new five-wave advance: wave 1 from 200 to 236, wave 2 down to 212, a deep pullback that holds well above the origin, wave 3 up to 252, the longest leg so far, and wave 4 pulling back to 234. The count feels strong. It is dead: wave 4's low at 234 sits two points inside wave 1's territory above 236, and the boundary law has fired.
The recount follows a fixed order. First, name the violation precisely: overlap at the wave 4 position, law three. Second, ask what structure the failure proves: not an impulse, so the advance from 200 is corrective. Third, relabel: the prior decline from 260 to 200 was wave A of an expanded flat, and the entire rally from 200 to the top at 244 was wave B, its internal tangle of apparent impulses and overlaps being exactly what B waves look like from inside. Fourth, project the consequence: wave C of the flat travels from the 244 top, and with wave A having covered 60 points, a C equal to A lands near 184. The trader who held the broken count through this sequence rode from long at 240 to a forecast of 184 without ever knowing why the market felt wrong. The trader who obeyed the law flipped bias at 234, twenty points earlier, with a defined stop above the B top.
Blunt version: the law saved nobody any analysis. It saved them from arguing with a price.
Practical Discipline
The procedure only works if the levels exist before the market reaches them. Write the three law prices the moment a count is adopted: the wave 1 origin, the wave 3 extreme, and the wave 1 territory boundary for wave 4 to respect. An entry without written law prices is not a count, it is a mood. The prices also settle the wick-versus-close question in advance: decide, before the test, whether your instrument and timeframe honor closes only, or whether a structural print through the level counts, and then apply the answer without renegotiating.
Degree consistency is the final habit. The laws operate within one degree at a time, and most false violation calls come from reading a smaller degree's overlap as a larger degree's failure. Count the degree you are trading, check its laws, and check the degree above only for context. A count that survives its own degree's laws is tradable even when the degree above looks messy, because the higher-degree mess is, quite literally, a different chart problem. The levels written in advance do the deciding; the trader only obeys.
Wave Law Questions
Do the laws apply on every timeframe?
Yes, at every degree, from multi-year advances down to intraday runs. The laws define what an impulse is, and an impulse at any degree has the same boundaries. What changes with timeframe is noise, not the law: smaller degrees produce more wick-tests of the boundaries, which is exactly why the close-versus-print decision belongs in the plan before the test, not during it.
What if wave 4 only touches wave 1 territory by a tick?
A structural print through the boundary kills the count, and a wick that closes back outside leaves it standing but wounded. The honest handling is the one decided in advance. Traders who honor closes treat the tick as a warning, tighten expectations, and watch the next closes; traders whose method reads prints relabel immediately. The unforgivable version is deciding after the fact which standard to use.
Can a guideline ever become a law?
No. The framework's guidelines, alternation, equality, channeling, stay tendencies at every degree, however reliable they feel in a given market. The moment a tendency is treated as a boundary, ordinary guideline failures start killing valid counts, and the trader's error rate rises for reasons the market never caused.
What do I do first when a law breaks?
Name which law fired, then relabel before forming any new opinion. The violation itself is the information: it proves the pattern was corrective at your degree, which points to the structures that fit, flats, zigzags and combinations, and away from every trend-following read. The recount comes before the new forecast, always, because the new forecast is only as good as the label it sits on.
The laws reject what cannot be, and the guidelines rank what remains. The next lesson takes the framework's working tendencies, alternation, channeling and equality, shows how each one steers a live count, and where each one quietly fails.