📋 Verdict
⚠ Static Baseline Outperforms Walk-Forward in Aggregate
The fixed parameter set (Q3 | 09:00 | 5p | Fri 19:00) produced PF 1.63 across the full H1 history versus the walk-forward aggregate PF of 1.12. This is primarily explained by the dominant 2022–2026 bull regime (PF 2.38) which these fixed params happen to match well. However the regime breakdown reveals two clearly losing periods (2013–2015, 2019–2021) where fixed params produced PF < 1.0 — this represents real historical risk of sustained drawdown in unfavourable regimes. The walk-forward's lower aggregate PF partly reflects that it traded through ALL regimes using different adapted parameters rather than the current-cycle-optimal set. The WF-Deployable gate (PF 1.06) shows the performance achievable when only deploying capital in folds the plateau classified as DEPLOYABLE.
🔢 Summary Statistics
| Metric | Static Q3/09:00/5p/Fri19 |
WF — All All forward trades |
WF — Deployable Plateau-gated only |
Static vs WF-All |
|---|---|---|---|---|
| Trades (N) | 92 | 131 | 101 | — |
| Win Rate | 30.4% | 23.7% | 23.8% | ++6.70% |
| Profit Factor | 1.63 | 1.12 | 1.06 | ++0.51 |
| Net Pips | +1246.4p | +422.1p | +173.0p | ++824.30p |
| Expectancy / trade | +13.5p | +3.2p | +1.7p | ++10.30p |
| Max Drawdown | 273.1p | 556.7p | 839.6p | -283.60p (lower=better) |
Static: fixed Q3/09:00/5p/Fri19 applied to full H1 history (2010–2026). WF-All: all forward-window trades across 224 folds regardless of plateau verdict. WF-Deployable: only forward trades from folds where plateau classification = DEPLOYABLE.
📈 Overlaid Equity Curves
Calendar-date x-axis. Regime bands shaded in background. Gold = static fixed params. Dark blue = all WF forward trades. Dashed green = WF trades from DEPLOYABLE folds only.
📅 Annual Performance — Net Pips by Year
Year-by-year breakdown shows regime sensitivity: static parameters produce losses in 2013–2015 and 2019–2021 while WF adapted parameters may have found different setups (or correctly identified UNSTABLE periods).
🎯 Regime Profit Factor Comparison
🔄 Rolling Profit Factor (20-Trade Window)
Dashed gold line = target PF 1.2. Rolling PF below 1.0 indicates a losing 20-trade run. Shows how consistent each approach's edge was over time rather than just the final aggregate.
🧠 Key Analytical Insight
Why is static PF higher than WF in aggregate?
The static fixed parameters were selected from the most recent research window
(the 200-week regime ending 2026). That regime happened to be strongly bullish USDJPY
(2022–2026, PF 2.38). Applying those same parameters to the full 15-year history
produces a high aggregate PF because the recent bull run has the most trades and
highest PF — it dominates the average. But in the losing regimes
(2013–2015 and 2019–2021) the fixed parameters silently degraded to PF < 1.0.
A real-money trader using static parameters would have suffered sustained drawdowns
in those periods with no signal to stop.
What does the walk-forward process actually test?
The WF process is not trying to maximise historical PF — it is testing whether the
process of monthly adaptive re-selection would have remained deployable
over time. The lower aggregate PF (1.12) means the process found valid candidates
in many regimes that produced modest-to-flat forward results — which is expected.
The WF-Deployable gate PF is the most decision-relevant metric:
it shows what actually happened when the plateau said "deploy this."
Current USDJPY conclusion: Both approaches produced net positive equity over the full history. Static showed higher absolute PF but with binary regime sensitivity. The adaptive process showed lower PF but operated through all regimes. Current verdict: RESEARCH ONLY — not yet sufficient aggregate forward evidence for live deployment.