Trend

Fibonacci Retracement indicator: settings and signals

Fibonacci Retracement

Fibonacci levels are horizontal lines at fixed fractions of a chosen price move, used as reference points for the depth of a pullback.

What the indicator looks like on a real chart

Closing price

The levels are horizontal lines rather than a calculated series: they are drawn once from the high and low of a chosen section and do not change afterwards. Hence the tool's defining feature: the result depends entirely on which extremes the trader selected.

Data: EUR/USD, daily bars, 2025-06-16 — 2026-07-20. The indicator is computed over 400 bars; the chart shows the last 180. Quote source: Binance Spot REST API (api.binance.com/api/v3/klines).

Effectiveness by market regime
Regime	Score
Trending market	48
Range	30
High volatility	33
Low volatility	41

What the indicator actually measures

Unlike all the other tools in the set, nothing here is computed as bars arrive. The trader selects two extremes, and the levels are drawn once as fractions of the distance between them: 23.6%, 38.2%, 50%, 61.8%, 78.6%.

The numbers come from the ratios of the Fibonacci sequence: 61.8% is the limit of the ratio of adjacent terms, 38.2% is its square. The 50% level has no relation to the sequence and was added from Dow practice, where half a move was considered a typical correction depth.

The main feature of the method is subjectivity. The result is entirely determined by which extremes are selected, and on any chart they can be selected in a dozen ways. Two traders on the same instrument will get different grids of levels, and both will be able to point after the fact to a level that "worked".

Formula

Level = High − (High − Low) × Fibonacci ratio

A level equals the high minus a fraction of the range of the move. There is no market quantity in the calculation — only the geometry of the selected segment, so the levels are not updated as new bars arrive.

Standard settings

38.2% / 61.8%

Key retracement levels

50%

Psychological level

161.8%

Extension level

Standard parameters
Parameter	Value
Key retracement levels	38.2% / 61.8%
Psychological level	50%
Extension level	161.8%

The standard set of fractions took hold by tradition. The 38.2% and 61.8% levels are considered key, 50% auxiliary. Extensions beyond 100% (127.2%, 161.8%) are used as target references rather than as support.

Implementation code (Python)

fibonacci.py

def fib_levels(high, low):
    diff = high - low
    return {
        '38.2%': high - diff * 0.382,
        '50%': high - diff * 0.5,
        '61.8%': high - diff * 0.618,
    }

When the indicator stops working

Limitations

It depends entirely on the subjective choice of construction points — the reproducibility of the method is low.
The levels are static and account for nothing that happens after construction.
It does not lend itself to rigorous testing: with enough lines on a chart, price will inevitably react near one of them.

Common mistakes in use

Drawing the grid after the fact on a move that has already happened and concluding that the levels work. This is the classic survivorship bias.
Placing several grids from different extremes on the chart at once — with dense line coverage any move will end up "at a level".
Ascribing a mystical meaning to the numbers. The ratios are taken from a mathematical sequence, but no market mechanism enforces them.
Using the levels as a standalone signal instead of as a reference that requires confirmation from price behavior.

Who needs this and why

For traders who use the levels as auxiliary references within their own markup rather than as a source of signals. It requires honesty with oneself: the method is easily bent to the desired conclusion.

Frequently asked questions

Why exactly 61.8% and 38.2%?

These are the limiting ratios of adjacent terms of the Fibonacci sequence and their derivatives. The 61.8% level is the limit of the ratio of a term to the next one, 38.2% is its square. The 50% level has no relation to the sequence and entered the set from Dow theory practice.

Do Fibonacci levels actually work?

There is no rigorous proof, and the method resists testing because of the subjectivity of construction. The most plausible explanation for the observed reactions is a self-fulfilling effect: enough participants watch the levels for orders to accumulate near them.

From which points should I draw the grid?

From the extremes of the move whose correction you are assessing. The problem is that the choice of that move is subjective, and this is where the main weakness of the method lies. The minimal discipline is to define a rule for selecting extremes in advance and not to change it to fit the desired result.

Other «Trend» indicators

MACD

EMA

SMA