Golden Pocket

The 0.618-0.65 "optimal trade entry" band of the most recent confirmed swing leg — the narrow retracement zone many swing traders watch for continuation.

Quick reference

Item	Value
Family	Fibonacci
Input type	Candle (uses high, low)
Output type	GoldenPocketOutput (low, mid, high)
Output range	inside the last swing leg
Default parameters	none (swing threshold 5%, baked)
Warmup period	2 (two confirmed pivots)
Interpretation	High-probability continuation entry zone

Formula

last two confirmed pivots define a leg start -> end:
  edge_a = end + 0.618 * (start - end)
  edge_b = end + 0.650 * (start - end)
  low  = min(edge_a, edge_b)
  high = max(edge_a, edge_b)
  mid  = (low + high) / 2

Sorting the two edges makes low <= high regardless of swing direction. See crates/wickra-core/src/indicators/golden_pocket.rs.

Parameters

None. The swing threshold 0.05 and the 0.618/0.65 band are baked-in constants. GoldenPocket::new is infallible.

Inputs / Outputs

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

• Python. update((o,h,l,c,v,ts)) → (low, mid, high) or None; batch(high, low) → (n, 3) Matrix (NaN warmup).
• Node. update(high, low) → { low, mid, high } or null; batch(high, low) → flat number[] length n*3.
• WASM. update(high, low) → object (low/mid/high) or null.

Warmup

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

Edge cases

• Band of a confirmed down leg is exact (test zone_of_a_down_leg).
• Band stays sorted (low <= high) for an up leg where the edges flip (test band_is_sorted_for_an_up_leg).
• reset clears all state (test reset_clears_state).
• Streaming equals batch (test batch_equals_streaming).

Examples

Rust

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

fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Leg 200 (high) -> 100 (low): the 61.8-65% band sits at 161.8-165.
    let bars = [
        (199.8, 200.0, 199.8, 199.8),
        (100.0, 198.0, 100.0, 100.0),
        (101.0, 110.0, 101.0, 101.0),
    ];
    let mut gp = GoldenPocket::new();
    let mut last = None;
    for (ts, (o, h, l, c)) in bars.iter().enumerate() {
        last = gp.update(Candle::new(*o, *h, *l, *c, 1.0, ts as i64)?);
    }
    let v = last.unwrap();
    println!("{} {} {}", v.low, v.mid, v.high); // 161.8 163.4 165
    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),
]
gp = ta.GoldenPocket()
print([gp.update(b) for b in bars][-1])  # (161.8, 163.4, 165.0)

Node

const wickra = require('wickra');
const gp = new wickra.GoldenPocket();
let last = null;
for (const [h, l] of [[200.0, 199.8], [198.0, 100.0], [110.0, 101.0]]) {
  last = gp.update(h, l);
}
console.log(last.low, last.mid, last.high); // 161.8 163.4 165

Streaming

gp = ta.GoldenPocket()
for o, h, l, c, v, ts in candle_feed:
    zone = gp.update((o, h, l, c, v, ts))
    if zone is not None and zone[0] <= c <= zone[2]:
        pass  # price inside the golden pocket — continuation watch

Interpretation

1. Tight entry band. Narrower than a full retracement grid; the 0.618-0.65 zone is where continuation entries cluster after a pullback.
2. Subset of retracement. It is the level_618-to-0.65 slice of the same leg reported by FibRetracement.

Common pitfalls

• Not a signal by itself. Marks a zone, not a trade; combine with momentum or candlestick confirmation.
• Latest leg only. Uses the immediate last leg; for the dominant swing pair it with AutoFib.

References

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

See also

• FibRetracement, AutoFib, FibConfluence.
• Indicators-Overview — full taxonomy.