SPX / ES / SPY Donchian Breakout — Structural Validation

Research Question

Before asking whether a Donchian breakout can be traded through 0DTE options, does the signal have a genuine structural edge at the underlying index point level — and does that edge hold across SPX, ES, and SPY?

Methodology

The research engine runs in seven phases (0–7):

• Point-level signal discovery on long-history index data.
• Risk parameterization and regime filtering.
• Walk-forward analysis and baseline testing.
• Option-aware proxy validation using Black-Scholes, a delta proxy, and VIX-scaled implied volatility.
• Real SPXW quote-chain validation with 1-second execution replay.
• Phase-by-phase acceptance gates, with Rust acceleration via PyO3 for the heavy replay work.

Findings

By measuring at the index point level first, the framework cleanly separates “is there a signal” from “can the option layer monetize it.” The same methodology applied to ES and SPY provides cross-asset corroboration of where the structure is and is not robust, which directly informed the negative-result conclusion in the dedicated SPX 0DTE study.

Follow-up Work

• Reuse the point-level harness as the gate before any new option-layer simulation.
• Extend regime filters and re-test on the most liquid linear proxies.
• Keep the option-proxy and real-quote layers explicitly separated in all reporting.