PRELON SWEEP CONTINUATION
Pre-London (00:00–02:00 ET) sweeps an Asia extreme → London continues past it. 81% UP / 75.6% DN when Asia range is tight. Fires ~62% of days.
updated 2026-04-21The setup
Asia (20:00–00:00 ET) builds a range. Pre-London (00:00–02:00 ET) pokes through one side. London (02:00–05:00 ET) continues past that extreme — it doesn’t reverse.
This is the continuation-side of the same dynamic that the AMD framework plays for reversal. We measured both. Both are real; they show up on different subsets of days (AMD-classic fires when Pre-London doesn’t sweep).
Lines ship on our chart HUD as PRELON SWEEP UP / PRELON SWEEP DN.
Entry rules
- Window: 02:00–05:00 ET London session
- Trigger (UP): Pre-London high > Asia high (Pre-London swept Asia H)
- Trigger (DN): Pre-London low < Asia low
- Direction: same as the sweep (continuation, not fade)
- Target: pre-extreme ± 25 points (p50 continuation depth)
- TTL: 180 minutes
Probability by Asia range bucket:
| Asia range | UP prob | DN prob |
|---|---|---|
| tight (<40pt) | 81% | 76% |
| normal (40–80) | 79% | 68% |
| wide (80–130) | 72% | skip |
| xwide (130+) | 76% | skip |
DN side skipped below 65% (wide/xwide) — sample turns unreliable.
Backtest
backtest_prelondon_sweep_continuation.py, n = 1,367 sessions, 5-year MNQ. Replicates Herman Trading’s published numbers within 2%:
| Metric | Ours | Herman | Δ |
|---|---|---|---|
| Pre-London sweeps Asia H | 34.8% | 34.4% | +0.4 |
| Pre-London sweeps Asia L | 27.4% | 27.3% | +0.1 |
| Pre-H → London ext UP | 78.4% | 77.2% | +1.2 |
| Pre-L → London ext DN | 68.5% | 69.6% | −1.1 |
Continuation depth past pre-extreme: UP p50=25pt / p75=48pt. DN p50=31pt / p75=58pt.
Firing cadence: 851 / 1,367 sessions = ~62% of days.
Both-swept edge case
When Pre-London breaches both Asia H and L (n=13, 1.0% of days): 76.9% up / 30.8% dn. Asymmetric up-bias but sample too thin to ship. Model does not fire on both-swept.
Why Tier 1 direction setter
- Replicates an independently-published edge (Herman) within noise — not a data-mined artifact.
- Fires ~62% of days with Tier 1 probability — one of the highest-cadence flagship signals.
- Continuation thesis is the inverse of the AMD-classic reversal; tested empirically, not assumed.
- Exempt from Phase A/B regime-adjust and OB confluence skip — already directionally validated.
History
- 2026-04-21 — Shipped. Per-bucket probability table, 180-min TTL, Tier 1 direction setter.