STRATEGY E039 · jybj.org

SextantFix

Composite-Momentum Reading + Random-Walk Confirmation · H1

Overview

The chart is a noisy sky full of flickers. A sextant does not react to every cloud — it takes one steady reading and fixes your true position. SextantFix works the same way: it reads a composite momentum line for the true heading, then asks one blunt question before it commits — is this an actual directional move, or just random wandering?

The Two Instruments

Timing + Direction

Know Sure Thing (KST)

Instead of a single momentum reading, KST takes the rate-of-change over four lookbacks (4·7·10·15), smooths each, and blends them with weights 1·2·3·4 into one steady line, plus a signal line (SMA 5). A cross up over the signal line opens an up window; a cross down opens a down window.

Trend-Strength Filter

Random Walk Index (RWI)

For several short horizons (up to 5) it compares how far price actually traveled against how far pure randomness would drift (ATR·√n), and takes the strongest reading. A value near or above 0.8 in the traded direction means the move is stronger than random — a genuine trend, not noise.

Signal Reading

signal
How the reading forms: composite momentum crossover, confirmed by random-walk strength.

Execution Flow

flow
Symmetric up / down participation, sharing one risk frame.

Entry / Exit Rules

WindowTriggerConfirmationAction
UpKST crosses up over its signal lineRWI up-strength ≥ 0.8Enter long at Ask
DownKST crosses down under its signal lineRWI down-strength ≥ 0.8Enter short at Bid

Stop = 1.5 × ATR from entry · Target = 1.6 × stop distance (R:R 1.6) · risk 1% of account per window · one window at a time · max 10 windows per day.

Parameters (defaults)

Frequency Self-Check

On simulated H1 data: raw KST crosses ≈ 1.7/calendar-day (≈2.4/trading-day) → after the RWI strength filter trims ~40% + one window at a time → about 0.8 trade per calendar-day (≈1.1–1.2 per active trading day), inside the healthy 1–10 range.
Study-and-observation tool for sharing ideas, not a promise of any outcome. Markets carry real ups and downs. Backtest 1–2 years of H1 in the tester and paper-run before anything else.