Navigation
DashboardProbesRealness Loop › Iter 28 · #14 slice 1

iteration 28 · 2026-07-24 · incrementality axis · bounded slice 1 · laptop-drives-bigblack

📐 #14 Spanning intercept CHECKPOINT

The increment/mechanism phase opens. The Huberman-Kandel spanning test turns the feature into a strategy return and asks whether its alpha survives being spanned by the shipped factor set. Slice 1 grounds the regression + coverage + null-FPR + substitution; the power gate sits at the HAC-t boundary because the strategy returns are long-memory — exactly the case §7 routes to a block-bootstrap t (slice 2).

0.955
CI coverage (target ≥.90)
0.0
null-FPR (spanned)
0.22 / 0.83
substitution B / C
0.725
power@5bps (HAC-t boundary)
Preflight (resource-only): load1 4.08 · 37 GiB available · si/so ~0 · ClickHouse active, readonly=2. 5c/5G/no-swap capped; single-thread BLAS. Reused the #12 wide slice for the basis returns.

Slice-1 gate status

Gate (§7 row 14, slice 1)Result (seed 20260723)Target
Coverage & null-tcoverage 0.955 · null-t mean −0.05 std 0.99 · FP 0.04≥.90 & FP≤.06PASS
Null-FPR (spanned feature)0.0≤ .05PASS
Substitutionspanned B \|t\| 0.22 (collapses) · incremental C survives 0.833B<1 & C≥.8PASS
Power @ α≈5bps0.725 (achieved α 4.19bps, HAC-t median 4.99)≥ .8boundary

The diagnosis — the strategy returns are long-memory (HAC-t is the wrong SE)

The power miss is not a magnitude problem and not fixable by more data:

This is exactly what §7 row-14 anticipates: "null FPR≤.05 across Hurst/t(3)/block — else block-bootstrap t." The HAC-t is the wrong standard error for the long-memory strategy return.

Slice 2 (terminal): the §7-specified block-bootstrap t (moving-block bootstrap of the intercept) as the α standard error → expected to restore power≥.8 and substitution-C; plus null-FPR across Hurst / t(3) / block regimes, stability across a frozen NW-lag set (kills lag p-hacking), and the omitted-premium Harden (the basis is only as complete as pre-registered → basis-PREREG + sensitivity-as-basis-grows, the #13-incomplete-Z analogue on the return object). Then #15 mechanism-intensity → #16/#17/#18.