Fibonacci Extensions

A three-point projection tool that extends Fibonacci ratios beyond a completed swing to mark possible price objectives past the prior extreme.

What it is

Fibonacci extensions are a projection tool that answers a different question from Fibonacci retracements. A retracement asks 'how far might a pullback travel within a completed swing'; an extension asks 'how far might the next leg travel beyond it'. The tool takes three user-chosen points — the start of an impulse move (A), its end (B), and the end of the subsequent pullback (C) — and projects levels above or below C at fixed multiples of the A-to-B distance, most commonly 61.8%, 100%, 127.2%, 161.8%, and 261.8%. The math is elementary: each level is C plus (B minus A) times a ratio. There is nothing mystical in it, and it is worth being honest about the ratios themselves: no verified market mechanism makes price obey Fibonacci numbers. To the extent extension levels 'work', the most defensible explanation is that they are self-fulfilling focal points — widely watched round-ish targets where many participants happen to place orders — not market physics. Extensions are a way of drawing consistent reference levels, nothing more. This page is educational only, not financial advice, and all trading involves risk of loss.

How it works

The construction is a three-point measurement. First identify an impulse swing: point A at its origin and point B at its extreme. Then identify point C, where the corrective pullback ends. The tool measures the vertical distance of the impulse, R = B − A (negative for a down-move), and plots horizontal levels at C + R × ratio for each chosen ratio. In an uptrend, the 100% extension sits one full impulse-length above the pullback low — the symmetry level where the second leg equals the first (also called a measured move or AB=CD). The 61.8% level lies short of that, while 127.2%, 161.8%, and 261.8% lie progressively beyond it. The ratios derive from the Fibonacci sequence: 0.618 is the limit of adjacent-term ratios, 1.272 is the square root of 1.618, and so on — internally consistent mathematics whose relevance to markets is asserted by convention rather than demonstrated by evidence. Critically, the output depends entirely on which three points you pick. Two competent chartists will often anchor A, B, and C differently and produce different levels from the same chart, and it is easy in hindsight to select the anchors that make past levels look prophetic. The tool is deterministic; the inputs are not.

How traders read it

• 100% extension: the second leg has matched the first in size — the 'measured move' symmetry point, and arguably the least arbitrary level of the set.
• 61.8% extension: a projection short of full symmetry, sometimes watched when the second leg is expected to be weaker than the first.
• 127.2% and 161.8%: levels beyond symmetry that some traders use as areas to reassess or scale out — reference points, not magnets.
• 261.8% and beyond: extreme projections that are only occasionally reached; their rarity is a property of how far away they are, not of any special power.
• Confluence caution: an extension level coinciding with a prior high or a round number is often cited as 'confirmation', but stacking subjectively anchored levels can manufacture confluence anywhere.
• Whatever the level, price is under no obligation to react there — treat extensions as a consistent measuring convention, not a forecast.

Common settings

The standard ratio set is 0.618, 1.0, 1.272, 1.618, and 2.618, with some platforms adding 1.414 (√2) and 2.0. There are no lookback parameters — the 'settings' are the three anchor points themselves, which is exactly where the subjectivity lives. Common conventions: anchor A and B on the wicks of the impulse swing, C on the wick of the pullback extreme, and keep the same anchoring convention (wicks vs closes) across all your charts so levels are at least internally consistent.

Strengths

• Provides forward reference levels beyond the prior extreme, where tools like retracements and prior support/resistance offer nothing.
• Transparent arithmetic: every level is reproducible from three prices and a ratio — nothing hidden, nothing fitted.
• The 100% level encodes swing symmetry (measured moves), a concept older than Fibonacci charting and simple to reason about.
• Widely watched, which is the honest source of whatever relevance the levels have as shared focal points.
• Forces explicit definition of the swing you are analyzing, which is a useful discipline in itself.

Pitfalls to watch

• Anchor subjectivity: different but reasonable choices of A, B, and C yield different levels; the tool is deterministic only after the subjective inputs are fixed.
• Hindsight bias: it is trivial to re-anchor after the fact so that a level 'nailed' a turn — historical examples in trading content are usually selected this way.
• No causal mechanism: markets have no obligation to Fibonacci ratios; treating the levels as physics rather than convention overstates them.
• Level clutter: plotting many ratios guarantees price will 'respect' some of them by chance, which then gets read as validation.
• Wick-vs-close inconsistency: mixing anchoring conventions across charts silently moves every level and makes comparisons meaningless.
• No predictive power: an extension level is a measurement of past swings projected forward; reaching or reversing at a level is never assured, and trading against these levels carries full risk of loss.

Pine v6 example

//@version=6
indicator("Fibonacci Extension Example", overlay = true)

// Three-point projection: impulse A -> B, pullback ends at C.
// Anchor selection is subjective — these inputs make that explicit.
priceA = input.float(100.0, "Point A (impulse start)")
priceB = input.float(110.0, "Point B (impulse end)")
priceC = input.float(105.0, "Point C (pullback end)")

impulse = priceB - priceA  // negative for a downward impulse

ext0618 = priceC + impulse * 0.618
ext1000 = priceC + impulse * 1.000
ext1272 = priceC + impulse * 1.272
ext1618 = priceC + impulse * 1.618

plot(ext0618, "61.8% extension",  color = color.gray)
plot(ext1000, "100% extension (measured move)", color = color.blue)
plot(ext1272, "127.2% extension", color = color.orange)
plot(ext1618, "161.8% extension", color = color.teal)

Pro tip: Audit your own anchoring before trusting any extension level: pick a past swing, have the levels drawn independently (or re-draw them yourself a week later without looking), and compare. If your A-B-C choices are not reproducible, your levels are not either — and any apparent accuracy was hindsight. The 100% symmetry level is the most defensible single projection because it assumes the least. Educational context only: extension levels are shared reference points, not predictions, and no drawing tool removes the risk of loss.

Educational only — not financial advice, not a recommendation to trade. No indicator is predictive; trading involves substantial risk of loss.