Fibonacci Retracement

The seven canonical retracement levels (0/23.6/38.2/50/61.8/78.6/100%) of the most recent confirmed swing leg, recomputed each time a new pivot confirms.

Quick reference

Item	Value
Family	Fibonacci
Input type	Candle (uses high, low)
Output type	FibRetracementOutput (7 prices)
Output range	bracketed by the leg's start and end prices
Default parameters	none (swing threshold 5%, baked)
Warmup period	2 (two confirmed pivots)
Interpretation	Pullback support/resistance inside the last swing

Formula

last two confirmed pivots define a leg start -> end:
  level(r) = end + r * (start - end)
for r in {0, 0.236, 0.382, 0.5, 0.618, 0.786, 1.0}

level_0 sits on end (the most recent confirmed extreme) and level_1000 on start (the prior pivot); the interior ratios are the classic Fibonacci pullbacks. Pivots are confirmed by the shared non-repainting 5% swing tracker. See crates/wickra-core/src/indicators/fib_retracement.rs.

Parameters

None. The swing threshold 0.05 is a baked-in family constant (pattern_swing.rs); the seven ratios are fixed. FibRetracement::new is infallible.

Inputs / Outputs

const _: fn(&mut wickra::FibRetracement, wickra::Candle) -> Option<wickra::FibRetracementOutput> =
    <wickra::FibRetracement as wickra::Indicator>::update;

• Python. update((o,h,l,c,v,ts)) → (level_0, level_236, level_382, level_500, level_618, level_786, level_1000) or None during warmup; batch(high, low) → (n, 7) ndarray (rows are NaN until ready).
• Node. update(high, low) → { level0, level236, level382, level500, level618, level786, level1000 } or null; batch(high, low) → flat number[] of length n*7 (NaN warmup).
• WASM. update(high, low) → object (same camelCase keys) or null.

Warmup

warmup_period() == 2. Two confirmed pivots are required to define a leg; before that update returns None. Pinned by tests accessors_and_metadata and no_output_before_two_pivots.

Edge cases

• Levels of a confirmed down leg are exact (test retracement_levels_of_a_down_leg).
• Levels refresh once a newer leg confirms — only the latest two pivots are kept (test levels_refresh_on_a_new_leg).
• reset clears all state (test reset_clears_state).
• Streaming equals batch (test batch_equals_streaming).

Examples

Rust

use wickra::{Candle, FibRetracement, Indicator};

fn main() -> Result<(), Box<dyn std::error::Error>> {
    // A swing high at 200 confirmed, then a swing low at 100: a 100-point leg.
    let bars = [
        (199.8, 200.0, 199.8, 199.8),
        (100.0, 198.0, 100.0, 100.0), // confirms high @200 (one pivot)
        (101.0, 110.0, 101.0, 101.0), // confirms low @100 (leg complete)
    ];
    let mut fib = FibRetracement::new();
    let mut last = None;
    for (ts, (o, h, l, c)) in bars.iter().enumerate() {
        last = fib.update(Candle::new(*o, *h, *l, *c, 1.0, ts as i64)?);
    }
    let v = last.unwrap();
    println!("{} {} {}", v.level_0, v.level_618, v.level_1000); // 100 161.8 200
    Ok(())
}

Python

import wickra as ta

bars = [
    (199.8, 200.0, 199.8, 199.8, 1.0, 0),
    (100.0, 198.0, 100.0, 100.0, 1.0, 1),
    (101.0, 110.0, 101.0, 101.0, 1.0, 2),
]
fib = ta.FibRetracement()
print([fib.update(b) for b in bars][-1])
# (100.0, 123.6, 138.2, 150.0, 161.8, 178.6, 200.0)

Node

const wickra = require('wickra');
const fib = new wickra.FibRetracement();
let last = null;
for (const [h, l] of [[200.0, 199.8], [198.0, 100.0], [110.0, 101.0]]) {
  last = fib.update(h, l);
}
console.log(last.level0, last.level618, last.level1000); // 100 161.8 200

Streaming

fib = ta.FibRetracement()
for o, h, l, c, v, ts in candle_feed:
    levels = fib.update((o, h, l, c, v, ts))
    if levels is not None:
        golden = levels[4]  # 61.8% — watch for a bounce here

Interpretation

1. Pullback map. After a strong leg, traders expect price to retrace to one of these levels before continuing; 38.2%, 50% and 61.8% are the most watched.
2. Non-repainting anchor. The leg only shifts when a new pivot confirms at the 5% threshold, so the levels are stable intra-swing — they do not redraw on every bar like a discretionary fib drawn by eye.
3. Pairs with the Golden Pocket. GoldenPocket narrows the 61.8-65% sub-zone of the same leg.

Common pitfalls

• Leg selection is mechanical. The tool always uses the latest leg; for the dominant recent swing use AutoFib instead.
• Threshold sensitivity. Tiny wiggles below 5% never form pivots, so on very quiet series the leg can lag the visual swing.

References

• Fischer, R. Fibonacci Applications and Strategies for Traders (1993).

See also

• FibExtension, GoldenPocket, AutoFib, FibConfluence.
• Indicators-Overview — full taxonomy.