vrp_z = z90(log(BVOL/RV30)) | held window [t,t+1] close-close | IS<2025-01-01/OOS>= | 15bps/flip | 30d block placebo x2000.
| leg | exposure | IS net SR | OOS net SR | OOS B&H | net DD | flips |
|---|
| vrp_z>0 (PRIMARY) | 53% | 2.11 | 0.14 | -0.21 | -40% | 78 |
| rv_z<0 (low-vol) | 47% | 2.12 | 0.19 | -0.21 | -41% | 32 |
| bvol_z>0 | 44% | 1.38 | -0.69 | -0.21 | -46% | 72 |
| vrp_z<0 (reverse) | 43% | -0.01 | -0.81 | -0.21 | -54% | 77 |
| MOM21 z>0 (control) | 47% | 1.17 | 0.57 | -0.21 | -38% | 86 |
- vrp_z IC: 1d +0.036, 5d +0.052, 21d +0.041
- partial IC(vrp_z, fwd1 | mom21) = +0.0371; | rv_z, mom21 = +0.0499
- block placebo vrp_z>0: OOS obs 0.14 vs null -0.31/p95 0.51 -> p 0.178
- DSR(K=4) OOS net vs B&H -0.21: 0.30 | haircut +0.00
| leg | exposure | IS net SR | OOS net SR | OOS B&H | net DD | flips |
|---|
| vrp_z>0 (PRIMARY) | 52% | 1.41 | 0.03 | -0.21 | -64% | 66 |
| rv_z<0 (low-vol) | 46% | 1.77 | 0.41 | -0.21 | -61% | 24 |
| bvol_z>0 | 48% | 0.96 | -0.31 | -0.21 | -57% | 122 |
| vrp_z<0 (reverse) | 44% | -0.59 | -0.53 | -0.21 | -77% | 65 |
| MOM21 z>0 (control) | 48% | 0.81 | 0.24 | -0.21 | -55% | 90 |
- vrp_z IC: 1d +0.041, 5d +0.094, 21d +0.136
- partial IC(vrp_z, fwd1 | mom21) = +0.0370; | rv_z, mom21 = +0.0333
- BTC PRIMARY vrp_z>0: OOS net-vs-BH +0.36, placebo p 0.178, IS edge +2.32, partial IC(|mom21) +0.037
- REJECT -> VRP/low-vol timing does not clear the placebo / momentum-control bars as a daily overlay; feed stays research infrastructure. PASS would graduate to shadow only.