VWAP REGIME
Where price sits relative to session VWAP at the 10:30 and 12:00 ET snapshots predicts RTH close direction. 75/70% at 10:30, 80/79% at 12:00 — strongest pre-SB regime filter.
updated 2026-04-21The setup
Session VWAP is the volume-weighted average price from RTH open. At a fixed snapshot time (10:30 or 12:00 ET), where price sits relative to VWAP predicts the close with remarkable stability. Above VWAP at 10:30 → RTH closes BULL 74.9%. Below → closes BEAR 69.8%. Push to 12:00 and the rates climb to 80.1% / 79.1%.
This is our strongest pre-Silver-Bullet directional filter. Lines ship on our chart HUD as VWAP REGIME BULL / BEAR.
Entry rules
- Window: fires after 10:30 ET
- Snapshot 1 (10:30): close − VWAP difference
- Snapshot 2 (12:00): same; used for persistence/flip
- Regime: BULL if diff ≥ +2pt, BEAR if ≤ −2pt, NEUTRAL otherwise
- Target: ny_high + 30 (bull) or ny_low − 30 (bear)
- TTL: through RTH close
Tiered probability:
| Condition | Bull | Bear |
|---|---|---|
| |d10:30| ≥ 20pt | 76% | 74% |
| |d10:30| ≥ 10pt | 75% | 70% |
| |d10:30| ≥ 2pt | 73% | 68% |
| 10:30 & 12:00 same side, both ±10+ | 80% | 79% |
| 10:30 & 12:00 same side ±5pt+ | 84% | 83% |
| 12:00 flips vs 10:30 | −10% penalty | — |
Backtest
audit_vwap_1030.py, n = 1,366 RTH sessions, 5-year MNQ:
| Snapshot | Threshold | ABOVE → BULL | BELOW → BEAR |
|---|---|---|---|
| 10:30 | ±2pt | 73.1% | 67.9% |
| 10:30 | ±10pt | 74.9% | 69.8% |
| 10:30 | ±20pt | 76.2% | 73.7% |
| 11:00 | ±10pt | 75.7% | 71.5% |
| 12:00 | ±10pt | 80.1% | 79.1% |
| 12:00 | ±20pt | 82.0% | 81.8% |
| 14:00 | ±20pt | 89.6% | 86.5% |
| 10:30 & 12:00 same side ±5pt | 84.1% | 82.9% |
Avg close-points: ABOVE +80 to +112pt. BELOW −84 to −129pt. This is a directional trajectory signal, not just a close-above/below bias — price genuinely drifts in the regime direction.
Why two snapshots
10:30 gives early read; 12:00 confirms or flips. Persistence across both is worth +5 to +7 probability points. A 12:00 flip is a real warning — we apply a −10% penalty rather than ignore it.
14:00 rates are even higher (89/86%) but too late to be actionable — most of the move is already in by then.
Why Tier 1 direction setter
- Largest, cleanest directional filter measured across all audits.
- Two independent snapshots add robustness.
- Exempt from regime-adjust skip list (this model is the regime).
- Exempt from OB confluence boost (already priced in).
History
- 2026-04-21 — Shipped. Added to Phase A/B regime skip and OB confluence skip to avoid double-counting.