Pre-reg: PRIMARY = 1.25x on OK longs iff VIX[D-1] > 252d median. IS < 2025-01-01, OOS >= 2025-01-01. Non-OK days BLOCKED. Costs per unit-notional flip. Control = press the calm tail (must lose).
Window: 2022-12-02 .. 2026-09-13 (1382 days). Status mix: OK 1365, BLOCKED 17; PRIMARY boost days (full): 97; OOS: 23. Old F4-veto would skip 23 OOS long-days.
| Segment | Variant | LongDays | Ret | Sharpe | Net10 Sharpe | Turnover |
|---|---|---|---|---|---|---|
| IS | baseline 1.0x | 256 | +86.70% | +1.03 | +1.00 | 19 flips |
| IS | VIX-press 1.25x (PRIMARY) | 256 | +81.28% | +0.93 | +0.90 | 25 flips |
| IS | CREDIT-press 1.25x | 256 | +86.70% | +1.03 | +1.00 | 19 flips |
| IS | CALM-press control | 256 | +116.05% | +1.10 | +1.07 | 26 flips |
| OOS | baseline 1.0x | 104 | +68.08% | +1.13 | +1.10 | 16 flips |
| OOS | VIX-press 1.25x (PRIMARY) | 104 | +69.87% | +1.11 | +1.08 | 18 flips |
| OOS | CREDIT-press 1.25x | 104 | +73.83% | +1.17 | +1.14 | 18 flips |
| OOS | CALM-press control | 104 | +84.60% | +1.15 | +1.12 | 20 flips |
| FULL | baseline 1.0x | 360 | +213.81% | +1.07 | +1.04 | 34 flips |
| FULL | VIX-press 1.25x (PRIMARY) | 360 | +207.95% | +1.00 | +0.98 | 42 flips |
| FULL | CREDIT-press 1.25x | 360 | +224.54% | +1.09 | +1.06 | 36 flips |
| FULL | CALM-press control | 360 | +298.83% | +1.12 | +1.09 | 46 flips |
OOS verdict (PRIMARY, gross)
- lift -0.02, bootstrap p 0.333
- beta vs BH: +0.01 | corr: +0.20 | resid Sharpe +0.32 (ret +1.78%, maxDD -3.83%)
- OOS blocked-day fraction 2.7%
- DSR(K=3) 0.92 | haircut +0.59
Verdict: REJECT (lift -0.02 < +0.10; p 0.333 NOT < 0.05; resid +0.32; maxDD -18.93% vs base -18.42% (within +5pp); blocked 2.7%; DSR(K=3) 0.92; calm-press lift +0.01 >= PRIMARY)
Reading
- PASS means proceed to walk-forward + shadow sizing (F2b path), not live. The calm-press control separates stress-timing from leverage.
- On REJECT: ledger entry, stress-sizing parked with the veto family.
Time: 3s. Boost 1.25x mirrors carry MULT_BOOST.