TimelessMarket Theory
Educational only — not financial advice. Fibonacci levels are reference zones, not predictions. The evidence that they beat arbitrary levels is thin — use them with care.
Concept · Definitive Guide

Fibonacci Retracements

Famous ratios, drawn on charts — and an honest look at whether they actually work.

Overview

A Fibonacci retracement takes a single price swing — from a swing low to a swing high — and divides it with a set of horizontal lines at 23.6%, 38.2%, 50%, 61.8% and 78.6%. The idea is that after a strong move, price often pulls back to one of these levels before continuing, so they act as candidate areas of support or resistance.

Fibonacci is one of the most popular tools in technical analysis — and one of the most debated. This guide explains exactly where the ratios come from, how to use the levels, and gives an unusually honest account of how weak the evidence really is.

Origins & history

How it works

Fibonacci retracement — levels drawn from swing low to swing high 0% (swing high) 23.6% 38.2% 50% 61.8% 78.6% 100% (swing low) bounce at 61.8% impulse pullback Levels are reference zones, not guarantees — note that 50% is not actually a Fibonacci ratio.
Levels are drawn from the swing low to the swing high; price pulls back and, in this case, finds support near the 61.8% level before resuming. Note that 50% is included by convention even though it is not a Fibonacci ratio. (Illustrative.)

The trading ratios come straight from the sequence:

61.8% = 0.618 (inverse of the golden ratio φ) 38.2% = 1 − 0.618 (also any number ÷ the number two places ahead) 23.6% = a number ÷ the number three places ahead 78.6% = √0.618 50% = a convention only — NOT a Fibonacci ratio

You anchor the tool on a clear swing: low-to-high in an uptrend (to find pullback support), or high-to-low in a downtrend (to find bounce resistance). Beyond 100%, extensions at 127.2% and 161.8% are used to project targets. The single most important honesty point sits right in the toolbox: the widely watched 50% level is not a Fibonacci number at all — it survives purely because traders find halfway retracements useful.3

Why they sometimes "work"

There is no known mechanism by which the golden ratio governs markets. The most credible explanation for any real effect is the self-fulfilling prophecy: because so many traders draw the same levels from the same obvious swings, orders cluster there, briefly creating the very support or resistance the tool predicts.3 That makes Fibonacci most useful not on its own, but as one input that gains weight when it lines up — confluence — with independent evidence like a prior support level, a moving average, or a round number.

Honest assessment

Strengths

Used well, Fibonacci gives a consistent, objective way to mark candidate pullback zones and to set structured profit targets via extensions. As a confluence tool — "is a Fib level also a real support level?" — it can help you find higher-quality areas to act, and it standardises where many traders are watching.

Evidence rating: weak. Rigorous studies have not shown that Fibonacci levels predict turns better than arbitrary levels; where they appear to work, the effect is most plausibly self-fulfilling. It is widely accepted, even among advocates, that Fibonacci should not be used as a standalone strategy.3

Weaknesses & failure modes

Professional uses vs. retail misuses

How careful traders use it

  • As a confluence filter — a Fib level that coincides with real support/resistance, never alone.
  • To set structured targets with extensions (127.2%, 161.8%).
  • Drawn only from obvious, agreed-upon swings the whole market can see.

Common retail misuses

  • Treating a Fib level as a guaranteed turn and trading it blindly.
  • Re-drawing the tool until a level "fits" recent price (curve-fitting).
  • Stacking Fib fans, arcs, and time zones until the chart is noise.

Going deeper

Extensions project targets beyond the move (127.2%, 161.8%, 261.8%). Confluence is the real skill: combine Fib with support/resistance, market structure, and moving averages. Relatives: the ratios also appear in Elliott Wave and in harmonic patterns (Gartley, butterfly), which formalise Fib relationships into shapes — powerful-looking, but inheriting all the same evidence problems. Weaker cousins: Fibonacci fans, arcs, and time zones have even less support than horizontal retracements.

Practice

Quiz 1 — Which commonly used Fibonacci level is not actually a Fibonacci ratio?

The 50% level. It's included by convention because halfway retracements are common, but it does not come from the Fibonacci sequence. (61.8%, 38.2%, 23.6% and 78.6% do.)

Quiz 2 — What's the most credible reason Fibonacci levels sometimes hold?

The self-fulfilling prophecy: so many traders draw the same levels that orders cluster there, briefly creating support or resistance. There's no evidence of a natural mechanism.

Quiz 3 — What's the safest way to use Fibonacci?

As a confluence tool — only act where a Fib level coincides with independent evidence (real support/resistance, a moving average, structure). Never as a standalone signal.

This concept in the knowledge graph

PrerequisitesSupport & resistance, trends
UnlocksConfluence-based entries & extension targets
RelatedElliott Wave, harmonic patterns, moving averages
Opposing viewThat the levels have no edge beyond being self-fulfilling — a view much of the evidence supports

Resources

References (primary where possible)

  1. The Fibonacci sequence & golden ratio (Leonardo of Pisa, Liber Abaci, 1202) — overview.
  2. Fibonacci retracement levels, extensions & strategy — Britannica Money.
  3. The mixed evidence and self-fulfilling-prophecy critique (and the 50% convention) — Investopedia; analysis.