Elliott Wave Theory: Impulse and Corrective Waves and Why Counts Get Revised

Elliott wave theory is one of those ideas that sounds neat on paper and gets messy the moment you open a live chart. The claim is simple: prices move in a five-wave push with the trend, then a three-wave correction against it, and the same shape repeats on every timeframe. The hard part is the counting. Two experienced analysts can look at the same Nifty chart and label it differently, and both may revise their count a week later. This guide walks through impulse and corrective waves, the three rules that can’t be broken, the softer guidelines, a worked example with made-up numbers, and why revisions are part of the method rather than a sign it has failed.

Where Elliott wave theory comes from

The theory is named after Ralph Nelson Elliott (1871–1948), an American accountant who studied market price movements and built his set of tools in the 1930s. He set it out in The Wave Principle in 1938, summarised it in a series of articles in Financial World magazine in 1939, and covered it most fully in Nature’s Laws: The Secret of the Universe in 1946, according to the Wikipedia entry on the Elliott wave principle, which cites the standard Elliott texts.

His core idea was about crowd psychology. Traders as a group swing between optimism and pessimism, and those swings leave repeating footprints in price. That’s why Elliott practitioners talk about each wave having a “personality”: the news, volume and sentiment you’d typically see during that stretch of the move.

Impulse waves: the five-wave move with the trend

An impulse (also called a motive wave) moves in the direction of the larger trend. In a rising market it has five legs:

• Wave 1: the first leg up. It’s rarely obvious when it starts, because the news is usually still bad and most people assume the old downtrend is intact.
• Wave 2: a pullback that corrects wave 1. It often retests the earlier low, and bearish talk comes back quickly.
• Wave 3: usually the biggest and strongest leg. Corrections inside it are short and shallow, and by its middle the crowd tends to join in.
• Wave 4: a sideways or choppy correction. It’s often frustrating to trade because the trend seems to stall.
• Wave 5: the last push. The news is upbeat, latecomers pile in, and momentum indicators frequently start to diverge from price.

Waves 1, 3 and 5 move with the trend; waves 2 and 4 move against it. In a falling market the picture simply flips: five waves down, three up. As the CFI explainer on Elliott waves lays out, each of the trend legs breaks down into five smaller waves of its own, and each of the two pullbacks breaks into three.

Corrective waves: A, B and C

After the five-wave impulse comes a correction, labelled with letters instead of numbers. The classic version is a three-wave A-B-C. Waves A and C move against the larger trend and each subdivides into five smaller waves; wave B is a counter move that subdivides into three.

Corrections are harder to read than impulses, and this is where a lot of counts go wrong. Wave A often gets dismissed as a routine dip in an ongoing bull market. Wave B then rallies and convinces many people the uptrend is back. Wave C is the one that hurts: it falls in five waves and is typically at least as large as wave A.

Not every correction is a tidy zigzag. Elliott corrective patterns come in a few families, mainly zigzags, flats and triangles, and these can also combine into more complex corrections. Triangles usually turn up in wave 4 and only very rarely in wave 2.

Impulses are usually the easy part to spot. Corrections are where most Elliott counts fall apart, because a wave B rally can look exactly like a new trend.

The three rules of Elliott wave theory you can’t bend

Almost everything in Elliott analysis is a tendency. Three things are not. A valid impulse count has to respect all three:

1. Wave 2 never retraces more than 100% of wave 1. If price falls below the start of wave 1, that wasn’t wave 2.
2. Wave 3 is never the shortest of waves 1, 3 and 5. It doesn’t have to be the longest, but it can’t be the smallest.
3. Wave 4 never enters the price territory of wave 1. In an uptrend, the low of wave 4 has to stay above the top of wave 1.

These rules are useful precisely because they give you a hard line. Break one, and the count is wrong. You don’t argue with it; you relabel.

Guidelines: what usually happens, not what must happen

On top of the rules sit guidelines. They help you guess where a wave might end, but they’re probabilities, and treating them as rules is a common beginner mistake.

• Wave 2 depth: prices usually don’t retrace more than 61.8% of wave 1.
• Wave 3 length: wave 3 often extends wave 1 by a ratio of 1.618 to 1.
• Wave 4 depth: wave 4 typically retraces less than 38.2% of wave 3.
• Alternation: waves 2 and 4 tend to look different. A sharp, simple wave 2 often points to a slower, sideways and more complex wave 4, and the other way round.
• Volume: volume tends to be lighter in wave 2 than in wave 1, and lighter in wave 5 than in wave 3.

Those 61.8%, 38.2% and 1.618 figures come from the Fibonacci sequence. Elliott himself linked his wave counts to Fibonacci numbers, and practitioners use the ratios to set likely support and resistance. If you want the mechanics of drawing those levels, our piece on Fibonacci retracement on Nifty covers it step by step.

Wave degrees: why the same chart has several counts at once

Elliott waves are fractal. Every wave is part of a bigger wave and is made up of smaller ones. Practitioners give each level a name, called its degree. The usual order runs from Grand Supercycle (multi-century) through Supercycle, Cycle, Primary, Intermediate, Minor, Minute and Minuette, down to Subminuette, which plays out over minutes. Degrees are defined by shape rather than by a fixed size or duration, so two waves of the same degree can look very different in length.

For a trader, the practical point is that you’re always counting at more than one level. The wave 3 you see on a daily chart may be wave 5 of something on the hourly chart, and only wave 1 on a weekly view.

A worked Elliott wave example (hypothetical numbers)

Let’s run through a count on an imaginary NSE stock. These prices are made up purely to show the arithmetic; they don’t describe any real company or index.

Step 1: mark the first two waves. Say the stock bottoms at ₹200 and rallies to ₹260. That’s a candidate wave 1 of ₹60. It then pulls back to ₹225. The pullback is ₹35, which is 35 ÷ 60 = about 58% of wave 1. It stays above ₹200, so rule 1 holds, and it’s inside the usual 61.8% guideline.

Step 2: project wave 3. Using the 1.618 guideline, a typical wave 3 would be 1.618 × ₹60 = about ₹97 long. Added to the wave 2 low of ₹225, that gives a rough target near ₹322. Suppose the stock actually runs to ₹330. Wave 3 is then ₹105, or 1.75 times wave 1. A little longer than the guideline, which is perfectly normal.

Step 3: check wave 4. A 38.2% retracement of a ₹105 wave 3 is about ₹40, which would put a typical wave 4 low near ₹290. Say the stock dips to ₹295. That’s a ₹35 pullback, roughly a third of wave 3. More importantly, ₹295 is above the wave 1 top of ₹260, so rule 3 holds.

Step 4: wave 5 and the length check. The stock then climbs from ₹295 to ₹350, a ₹55 move. Our three trend legs are ₹60, ₹105 and ₹55. Wave 3 is the longest, so rule 2 is comfortably satisfied. If momentum indicators make a lower high while price makes this new high, that’s the kind of divergence Elliott traders expect near the end of wave 5. (Our guide on how to read RSI divergence explains how to spot it.)

Step 5: what invalidates the count. Now picture a different path. After the ₹330 high, the stock slides to ₹255 instead of ₹295. That’s below the ₹260 wave 1 top, so this can’t be wave 4 of the impulse we labelled. Either the move to ₹330 wasn’t wave 3, or the whole structure from ₹200 is something else, perhaps a correction rather than an impulse. The count has to be redone, and any trade built on “wave 5 is coming” loses its reason to exist.

That last step is the useful bit. The invalidation level, ₹260 in this example, is a natural place to think about a stop-loss, because it’s the price at which your whole reading of the chart is proven wrong.

Why Elliott wave counts get revised

If you follow Elliott analysts for a while, you’ll notice counts change often. There are a few honest reasons for that.

You only know a wave is finished after it’s finished. A wave 3 can extend much further than expected, and what looked like wave 5 turns out to be a smaller wave inside a still-running wave 3.

Corrections are ambiguous. A flat, a zigzag and a triangle can look alike in their early stages, and complex corrections can chain several patterns together.

Degree is a judgement call. Because degrees are defined by form rather than size, deciding whether a move is a Minor or an Intermediate wave involves interpretation, and a change at one degree ripples through the rest of the count.

Rules break the current count. When price crosses an invalidation line like the ₹260 level above, there’s no choice but to relabel.

A revised count isn’t automatically a failure. The real problem is a count that never commits to an invalidation level, because then it can never be wrong.

The criticism worth taking seriously

Elliott wave theory has serious critics, and you should know their arguments before you put money behind a count. The mathematician Benoit Mandelbrot, who modelled prices using fractal geometry, cautioned that wave prediction depends more on the chartist’s subjective judgement than on anything objective or replicable. The technical analyst David Aronson argued that the method’s loose rules and its many nested waves let it fit almost any stretch of market history after the fact. And a study by Roy Batchelor and Richard Ramyar found no significant difference between how often Fibonacci price and time ratios showed up in Dow Jones Industrial Average cycles and how often you’d expect them by chance. All three are summarised, with sources, in the Wikipedia entry linked earlier.

We think that’s a fair warning rather than a reason to ignore the theory altogether. Elliott works best as a way to organise what you’re seeing and to define where you’re wrong, not as a crystal ball. Treat a count as one scenario with a stated invalidation price, keep an alternate count in mind, and confirm with other tools such as trend, momentum and volume.

Using Elliott wave theory on Indian charts

A few practical habits help when you apply this to NSE stocks or the Nifty:

• Start on a higher timeframe. Count the weekly or daily chart first, then zoom in. Intraday noise makes small-degree counts very unreliable.
• Write down the invalidation level. For every count, note the price that would break one of the three rules.
• Keep a second count. Many Elliott traders carry a preferred count and an alternate. When the preferred one fails, the alternate is often already waiting.
• Don’t force five waves. If you can’t see a clean impulse, the move may be corrective. Calling every rally a wave 3 is how people get caught at tops.
• Combine it with other evidence. Support and resistance, moving averages and divergence give the wave count something outside itself to lean on.

For the wider toolkit these fit into, see our pillar guide to technical analysis for Indian stocks.

Elliott wave theory in brief

Markets, in Elliott’s view, move in five waves with the trend and three against it, and the pattern repeats at every degree. Three rules are fixed: wave 2 can’t go below the start of wave 1, wave 3 can’t be the shortest trend wave, and wave 4 can’t overlap wave 1. Everything else, including the Fibonacci ratios and alternation, is a tendency. Counts get revised because waves are only confirmed in hindsight and corrections are ambiguous, so the most valuable thing a count gives you is a clear price at which it’s wrong.

Disclaimer: This article is for education only and is not investment advice. The prices in the worked example are hypothetical and do not refer to any real stock or index. Please do your own research or consult a SEBI-registered adviser before making investment decisions.