USDJPY — Static Parameter Test 01 — Open-Trap Long
CONFIDENTIAL — KALAYA CAPITAL PROPRIETARY RESEARCH |
Generated: 2026-03-14 20:58 | Symbol: USDJPY |
Period: 2010-02-14 to 2026-03-01 |
Fixed params: Q3 | 09:00 | 5p | Fri 19:00
What is this? This test applies a fixed parameter set — discovered from the recent 200-week regime —
across the entire H1 history back to 2007. This is a diagnostic test. A parameter set optimised on a recent
regime is expected to degrade in older regimes. The purpose is to demonstrate how much it degrades,
and to contrast with the adaptive walk-forward process where parameters are re-selected each month.
Overall Results — Fixed Parameter Set
92
Total Trades
30.4%
Win Rate
+1246.4p
Net Pips
1.63
Overall PF
+13.5p
Expectancy
273.1p
Max Drawdown
STATIC OUTPERFORMANCE Fixed parameters have positive long-run PF ≥ 1.2 across full history.
Results segmented by 3-year market periods. A strategy adapted to recent regimes
tends to perform best in the period it was calibrated on and worse in prior regimes.
Regime
N
Win Rate
Net Pips
PF
Expectancy
Max DD
2007–2009
0
0.0%
+0.0p
0.00
+0.0p
0.0p
2010–2012
18
27.8%
+97.5p
1.26
+5.4p
233.6p
2013–2015
11
27.3%
-7.0p
0.97
-0.6p
97.5p
2016–2018
22
31.8%
+365.8p
1.68
+16.6p
177.5p
2019–2021
14
21.4%
-22.2p
0.91
-1.6p
69.5p
2022–2026
27
37.0%
+812.3p
2.38
+30.1p
112.0p
Rolling Profit Factor (20-trade window)
Shows how the fixed parameter PF evolves through time.
Periods below the red 1.0 line = parameter is losing money in that segment.
Methodology
Fixed parameters: Q3 | 09:00 | 5p | Fri 19:00
• Bucket: Q3 (weekly open in upper-middle quartile of prior week range)
• Entry: Monday 09:00, Long downside_failed feature (low < open, price > open at entry)
• Stop: prior week low − 5p buffer
• Exit: Friday 19:00 (or stop hit if triggered first)
• Universe: USDJPY Q3 weeks where the Long feature fires, full H1 history
Why this test matters:
The walk-forward validation shows the adaptive process produces a positive aggregate PF across regimes.
This static test shows what happens without adaptation — the same parameters, frozen, applied universally.
The contrast between the two tests is the empirical evidence for the value of the monthly re-selection process.
Interpretation guide:
• If static PF ≈ walk-forward PF: the strategy is regime-stable; parameter selection adds little value
• If static PF < walk-forward PF: adaptive selection is genuinely extracting regime value
• If static PF < 1.0: the recent parameters actively lose money in other regimes
