Single-instrument long-short strategy on AVAXUSD.HYPERLIQUID 1-hour bars implementing a DELAYED-ENTRY volatility expansion mechanism. This is the direct empirical follow-up to the documented LINK 4H ATR-Expansion failure (af7ef6b9), whose abandon-postmortem contained the explicit recommendation: 'If pursuing volatility-expansion mechanisms further, redesign as a delayed-entry variant — Wait for ATR expansion → then wait for a PULLBACK before entering (vs entering at peak vol).' The mechanism: when a 1H bar's true range exceeds 2x its trailing 20-bar average TR (vol regime change signal), the strategy DOES NOT enter immediately (which is the empirically-failed 'chase the expansion' mode that buys the local extreme). Instead, it tags the expansion bar's direction (bullish/bearish), waits up to 4 hours for price to pull back toward the expansion bar's midpoint, and only THEN enters in the expansion direction at a much better price. This decouples regime detection (high signal: vol cluster confirmed) from entry timing (favorable price after fade). Single-instrument keeps verification clean (avoids the COIN-M dynamic-instrument resolution failure mode). 1H timeframe gives ~60-120 trades/year (vs the 12-25/year of daily strategies), letting walk-forward windows have ~200-400 IS trades — addressing the over-parameterization issue flagged in the breadth-gated daily failure. AVAX on Hyperliquid hits three under-represented quotas simultaneously: HYPERLIQUID venue (4.2% vs >=20%), long-short direction (12.9% vs target), and mid-30m-2h horizon (11.5% vs target). Avoids every documented session failure mode: not classical TA pattern, not first-moment momentum already done, not ratio reversion, not vol-expansion-late-entry (the WHOLE POINT is to fix that failure mode), zero supp-data dependency, no breadth-correlation regime risk, no multi-instrument P&L complexity.