Generated 2026-10-05 | frozen eqmom baseline; book beta from own returns; borrow 0.5%/yr + 5bps on |dh|; C7 = stress-only SPY hedge.
| variant | ret | sharpe | maxDD | downvol | CVaR5 | calmar |
|---|
| BASE | +124.4% | 0.51 | -48.0% | 18.0% | -2.11% | 2.59 |
| C7 | +125.2% | 0.57 | -36.4% | 14.5% | -1.87% | 3.43 |
| C7+state | +105.3% | 0.53 | -36.5% | 14.0% | -1.77% | 2.89 |
| state_hedge | +99.8% | 0.50 | -41.7% | 14.9% | -1.93% | 2.39 |
| static_0.3 | +82.6% | 0.46 | -42.9% | 13.3% | -1.69% | 1.92 |
| variant | ret | sharpe | maxDD | downvol | CVaR5 | calmar |
|---|
| BASE | +279.2% | 0.98 | -34.1% | 17.8% | -2.18% | 8.18 |
| C7 | +233.8% | 1.05 | -25.7% | 14.1% | -1.92% | 9.12 |
| C7+state | +217.2% | 1.07 | -20.4% | 13.0% | -1.75% | 10.66 |
| state_hedge | +252.7% | 1.13 | -20.4% | 13.1% | -1.79% | 12.40 |
| static_0.3 | +206.9% | 1.02 | -25.4% | 13.7% | -1.76% | 8.14 |
| variant | ret | sharpe | maxDD | downvol | CVaR5 | calmar |
|---|
| BASE | +236.6% | 1.49 | -26.0% | 18.3% | -2.48% | 9.08 |
| C7 | +215.2% | 1.51 | -19.0% | 16.7% | -2.40% | 11.33 |
| C7+state | +171.7% | 1.39 | -19.1% | 15.6% | -2.33% | 8.99 |
| state_hedge | +176.0% | 1.40 | -19.9% | 15.7% | -2.37% | 8.86 |
| static_0.3 | +171.7% | 1.41 | -21.4% | 15.9% | -2.20% | 8.04 |
| segment | dSharpe (C7+state - C7) | block-bootstrap p |
|---|
| OOS | +0.01 | 0.465 |
| holdout | -0.12 | 0.020 |
- OOS: C7 Sharpe 1.05 vs BASE 0.98; C7+state 1.07 (p=0.47).
- Holdout: C7 1.51 vs BASE 1.49; C7+state 1.39 (p=0.02).
Promotable only if the state-augmented hedge beats the incumbent C7 on Sharpe OR materially cuts CVaR5/maxDD with SR within -0.10, on the holdout, beyond the bootstrap null. Otherwise the hedge channel closes the same way as the exposure channel.