
Elliott Wave theory says markets move in repeating waves, five with the trend and three against it. The American accountant Ralph Nelson Elliott developed it in the 1930s. Its five-wave move comes with three hard rules. On a crypto chart it names every swing, but where one wave ends and the next begins is open to interpretation.
Its Fibonacci levels showed no edge in our test on Phemex's own daily closes. Across 29 crypto perpetuals on Phemex from February 2024 to the mid-November close, swings ended in the 38.2%-61.8% zone about as often as on shuffled price data, each coin's own daily moves in random order. Pullbacks that stopped in that zone were followed by a higher close ten days later less often than an average day.
|
How deep the swing went
|
29 Phemex crypto perpetuals
|
Baseline
|
|
Under 38.2% of the move before it
|
92 of 793 swings (11.6%)
|
9.3% on shuffled prices
|
|
38.2%-61.8%, the Fibonacci zone
|
118 swings (14.9%)
|
15.3% on shuffled prices
|
|
61.8%-100%
|
184 swings (23.2%)
|
25.7% on shuffled prices
|
|
Over 100%, the whole move given back
|
399 swings (50.3%)
|
49.7% on shuffled prices
|
|
Zone pullbacks higher ten days later
|
17 of 43 (39.5%)
|
44.7% of all days
|
What Is Elliott Wave Theory?
Elliott Wave theory reads a price chart as crowd psychology moving in patterns. It rests on the idea that the market forms the same types of patterns on a short timeframe as it does on a long one. A five-wave rise on a daily chart can sit inside one bigger wave on a weekly chart, and analysts call each of those levels a degree.
Elliott set out the idea in his 1938 book The Wave Principle, according to Wikipedia's entry on the Elliott wave principle. He later linked the counts to the Fibonacci sequence, and practitioners use the golden ratio of 1.618 and related ratios as support and resistance levels for waves. The 61.8% and 38.2% levels in this test come from that family.
ChartSchool credits Robert Prechter with popularizing the theory in the 1970s, and it says the theory still draws a wide following among traders.
How Many Waves Are in Elliott Wave Theory?
A full Elliott cycle has eight waves, the first five moving with the trend. In a rising market the three odd-numbered waves push higher while waves two and four pull back. A three-wave correction labelled A-B-C then moves against the trend. In a falling market the same shape runs upside down, with five waves down and three up.
ChartSchool lists three rules for the five-wave move, and a count that breaks any of them is wrong.
-
Wave two never retraces all of wave one, so it never goes beyond the point where wave one began.
-
Wave four never retraces all of wave three, so it never goes beyond the start of wave three.
-
Wave three travels beyond the end of wave one and is never the shortest wave.
Everything else is a guideline, a tendency that can fail. Wikipedia's summary gives two Fibonacci guidelines. In wave two "prices usually do not retrace more than 61.8% of the wave one gains," and wave four "typically retraces less than 38.2% of wave three."
Checking the third rule takes a hand count of every wave, so the test below left it alone.
The wave-two rule is easier to use, because it gives you the clearest line on the chart. If a pullback closes below the point where wave one started, the count is wrong by definition, which makes that point the natural home for a stop. Knowing how to use stop-loss orders in crypto trading turns that line into an exit you don't have to watch.
|
Signal
|
What it tells you
|
BTC on Phemex, February 2024 to mid-November 2026
|
|
Five waves with the trend, three against
|
Elliott's basic cycle
|
30 turning points on 10% swings
|
|
Wave 2 gives back less than 100% of wave 1
|
Break it and the count is wrong
|
10 of 14 pullbacks closed below the rise's start
|
|
The 38.2%-61.8% zone
|
Between the two Fibonacci guidelines
|
3 of 28 swings, against about 16% on shuffled data
|
|
Wave 3 is never the shortest
|
The rule for the middle wave
|
Not testable without a hand count
|
|
A pullback that stops in the zone
|
The entry the test used
|
One case, in March 2024: 2.06% lower ten days later
|
|
The latest rise
|
How much of it was given back
|
About 80% of the 30 June to 21 September rise
|
Does Elliott Wave Work in Crypto?
Not on Phemex's crypto charts, at least for the part of the theory a computer can test.
A machine can't count waves the way an analyst does, so we used a zigzag, a line that marks a new swing each time a daily close reverses by a set percentage. We set it at 20% for 29 crypto perpetuals and at 10% for Bitcoin, each market from February 2024 or its later listing.
ChartSchool's ZigZag page warns that the tool "has no predictive power, because it draws lines based on hindsight." The test bought only on the day a swing was confirmed, so it used no later prices.
A swing's depth is how much of the previous move it gave back. About 15% of the 793 swings ended in the Fibonacci zone, and random price paths built by shuffling each coin's own daily moves landed there about as often.
Buying those pullbacks did worse than buying on an average day. Of 43 pullbacks that stopped in the zone, 17 closed higher ten days later, about 40% against about 45% of all days on the same coins.
Pullbacks of other depths fared a little worse again, at about 37%. Before counting, we required the zone to beat both by five points, and it didn't. Most of the 29 coins fell over the test period, so even across all days the median ten-day move was down about 1.5%.
Bitcoin alone gave the same answer. Only three of its 28 swings at the 10% setting ended in the zone, fewer than on shuffled data. The one pullback among them came in March 2024 and closed about 2% lower ten days later.
|
Daily close
|
BTC on Phemex
|
What happened
|
|
23 February 2024
|
50,753.7
|
The rise began
|
|
13 March 2024
|
73,118.0
|
The top of the rise
|
|
19 March 2024
|
61,969.0
|
The low, about 50% of the rise given back
|
|
25 March 2024
|
69,933.1
|
The first close at least 10% above the low, so the swing was confirmed
|
|
4 April 2024
|
68,492.0
|
Ten days on, 2.06% below the entry
|
|
6 November 2024
|
75,572.9
|
The first close above the March top
|
This is a count of daily closes, not a trading record. It paid no fees or funding and set no stops, its ten-day windows overlap, and 43 signals is a small sample.
Is Elliott Wave Theory Reliable?
As a way to describe a chart, yes. As a forecast, the test found nothing to lean on. A mechanical test can show how deep swings went against chance and if stopping at a Fibonacci depth led anywhere. It can't test an analyst's own count, because a count that gets relabelled whenever it breaks can never fail on prices.
The wave-two rule shows why. Half of the 793 swings on those 29 perpetuals gave back more than the whole move before them. Shuffled data did the same. Under the rules none of those swings could be a wave two or a wave four. An analyst who meets one gives it another label, and the count survives.
ChartSchool is candid about this. It says Elliott Wave "is not a trading technique" and that "there are no specific rules of entry or exit."
Bitcoin's latest big swing shows how much room the labels leave. It rose from about 58,600 at the 30 June 2026 close to about 86,600 on 21 September, then gave back about 80% of that rise by the mid-November close at 64,124.
Labelling that rise a wave one is a judgement the theory leaves to the analyst. The rules only say what would disqualify it, a close below the June start, and Bitcoin hadn't made one.
Before you trade a wave count, backtest a Bitcoin strategy on past closes the way this test did.
Our view is that Elliott Wave earns its place as a vocabulary for where a market sits in a move. The Fibonacci depths added nothing on these charts, and the line at the start of wave one is what we'd trade around.
Elliott Wave vs Wyckoff: What Is the Difference?
Elliott Wave counts the shape of price moves. The Wyckoff method reads supply and demand from price and volume, mostly inside the trading ranges that form between those moves.
ChartSchool's Wyckoff tutorial traces the method to Richard Demille Wyckoff, an early 20th-century pioneer of technical analysis, and counts both him and Elliott among the five "titans" of technical analysis. Wyckoff's method rests on three laws. Supply and demand sets the direction, cause and effect sizes the move out of a range, and effort versus result compares volume with how far price travels.
|
Question
|
Elliott Wave
|
Wyckoff
|
|
What it reads
|
The shape of price swings
|
Price and volume in trading ranges
|
|
Its basic unit
|
Five waves with the trend, three against
|
Accumulation and distribution ranges
|
|
Its hard rules
|
Three rules for the five-wave move
|
Three laws, from supply and demand to effort versus result
|
|
How it sets targets
|
Fibonacci ratios, as guidelines
|
Point and Figure counts
|
The two answer different questions. Wyckoff asks if large players are building or unloading a position in a range. Elliott asks how far along a trend the market has come.
How to Use Elliott Wave on Phemex
Phemex's futures chart has no Elliott Wave indicator, but its drawing tools let you label a count by hand. The Patterns group holds an Elliott Impulse Wave (12345) tool and an Elliott Correction Wave (ABC) tool, and the indicator list carries Zig Zag, a close cousin of the swing tool behind the test.
-
Open the BTC/USDT futures chart and click D to switch to daily bars, the timeframe the test used.
-
If the drawing toolbar is hidden, open it from the small tab on the chart's left edge.
-
Pick Elliott Impulse Wave (12345) from the Patterns group and place its points on the swing lows and highs of the move.
-
Draw Fib Retracement from the Fibonacci group across wave one to see where 38.2% and 61.8% fall, as guides only.
-
Tick TP/SL on the order panel and put the stop beyond the start of wave one before you open the trade.
A stop at the start of wave one can sit a long way from your entry. Using the average true range to set stop-losses lets you compare that distance with Bitcoin's usual daily range.
Futures use leverage, and a leveraged position is closed automatically if losses push its margin below a set minimum. That forced close is a liquidation, and it pays to read what liquidation means in crypto trading before a wave count talks you into size.
Frequently Asked Questions
What are the three Elliott Wave rules?
The second wave can't erase the first, the fourth can't erase the third, and the third must travel past the end of the first without being the shortest. ChartSchool lists those three for the five-wave move. Every Fibonacci level on top of them is a guideline.
Who developed the Elliott Wave principle?
Ralph Nelson Elliott, an American accountant, developed it in the 1930s and published it in his 1938 book The Wave Principle. ChartSchool credits Robert Prechter with making it popular in the 1970s.
Can you use Elliott Wave on its own?
ChartSchool says many traders avoid it "because of the apparently subjective nature of how it may be applied," and that its advocates add other indicators to trade it. On Phemex's crypto closes, its Fibonacci zone alone did worse than an average day.
Bottom Line
Elliott Wave is most useful for naming where a market might be in its cycle, with the start of wave one as the line that tells you the count was wrong. On these crypto charts the Fibonacci zone carried no edge, so a pullback's depth alone gave no reason to buy it.
Disclaimer: This article is for informational purposes only and is not financial advice. Cryptocurrency trading involves substantial risk. Always do your own research before making investment decisions.
