Using weekly ATR levels instead of daily, this study tracks how often the Golden Gate (38.2% to 61.8%) completes across multiple trading days. The higher-timeframe filter is the daily Phase Oscillator — specifically the previous day’s closing value, to avoid any look-ahead bias.
The weekly ATR levels define a wider range than daily ATR. Rather than expecting completion within a single session, the Golden Gate is tracked across 1 to 5 trading days after the 38.2% level is hit.
The “Bilbo” filter uses the daily Phase Oscillator (previous day’s close). This eliminates look-ahead bias entirely — the PO state is known before the trading day begins. The full ATR target is the 100% weekly ATR level.
No look-ahead bias: The daily PO value used is always from the previous day’s close. You know the setup conditions before the session opens.
The bearish Bilbo signal (daily PO Low + Falling) is the strongest edge in this dataset. When the daily PO was already deeply bearish entering the session, the weekly Golden Gate completed 96.3% of the time — on day 1 alone. Most of those reached the full weekly ATR.
Most multi-day Golden Gates that complete do so on day 1 — the baseline is 65–72% completion on the first trading day. The bearish Bilbo at 96.3% is effectively a certainty, suggesting deeply bearish daily momentum creates a powerful weekly downside continuation.
Bearish Bilbo is the standout: 96.3% day-1 completion vs 72% baseline is a +24-point edge. Bullish Bilbo is modest at +9 points (73.9% vs 65%). The asymmetry is striking — bearish momentum carries weekly ATR far more reliably.
Bullish weekly Golden Gate completion rates by daily PO state. n=901 total entries. The Bilbo condition (High + Rising) provides a modest edge, but the biggest differentiator is whether the PO is rising or falling.
| Daily PO State | Day 1 | Day 2 | Day 3 | Day 4 | Day 5 | n |
|---|
| Daily PO State | Day 1 | Day 2 | Day 3 | Day 4 | Day 5 | n |
|---|
Bearish weekly Golden Gate completion rates by daily PO state. n=706 total entries. The Bilbo condition (Low + Falling) dominates everything else in this dataset — 96.3% completion flat across all five days.
| Daily PO State | Day 1 | Day 2 | Day 3 | Day 4 | Day 5 | n |
|---|
| Daily PO State | Day 1 | Day 2 | Day 3 | Day 4 | Day 5 | n |
|---|
Beyond the 61.8% Golden Gate completion, how often does price reach the full 100% weekly ATR? The Bilbo filter dramatically improves full ATR rates, especially on the bearish side.
88.9% full weekly ATR by day 4 for bearish Bilbo. When the daily PO is already deeply bearish, the weekly range isn’t just reached — it’s blown through. This is the most actionable multi-day signal in the dataset.
1. Bearish Bilbo multi-day is near-certainty. When the daily PO closed Low + Falling, the weekly GG completed 96.3% of the time on day 1 alone. 81.5% reached the full weekly ATR on that same day. This is the strongest signal across all Golden Gate studies.
2. Most completions happen on day 1. The baseline shows 65–72% of multi-day GGs that ever complete do so on the first trading day. The multi-day window mostly adds marginal completions — the big move happens early.
3. Bullish Bilbo is modest but real. High + Rising daily PO gives 73.9% day-1 completion vs 65% baseline (+9 points). The full ATR edge is more meaningful: 31.3% vs 23.5% baseline on day 1, rising to 52.2% vs 47.3% by day 5.
4. Falling PO without extreme levels underperforms. Mid + Falling shows the weakest completion rates on both sides. The PO direction matters, but it matters most at extremes (the Bilbo condition).
5. No look-ahead bias. The daily PO value is the previous day’s close. You can evaluate the Bilbo condition before the market opens, making this actionable for real trading.