Fed Balance Sheet ROC Buy Overlay — WSHOSHO ROC5 >4.8 (qm-4bx.19)
Rejected qm-4bx.19 · 2026-08-12
Decision: NO-SHIP frozen — investigate-only; do not admit to catalog/Meta. Commit: 6575fa16 · Date: 2026-08-12 · Payload: research/findings/fed-bal-roc-qm-4bx.19.json · Repro: python -m research.scripts.auditfedbalroc --write
TL;DR Fed balance-sheet liquidity impulse (WSHOSHO 5-day ROC >4.8%) is a buy-only auxiliary overlay that forces max long SPY for a holding window. After maximal FRED ingest, weekly→daily forward-fill with publication lag L=7–10d, and daily nextclose+L causality via runbacktest, the signal is sparse (57 triggers, 0.96% of days), crisis-concentrated (GFC/COVID), and brittle to falsification
Provenance
Equity vs S&P (aligned, rebased to 1.0)
Aligned window, both rebased to 1.0 at start — Aligned window, both rebased to 1.0 at start for fair comparison · S&P dashed, experiment solid. X = Date, Y = Equity (rebased to 1.0).
Signal driver & thresholds (x-aligned)
Parameter grid: grid_lookback_threshold_hw20
Threshold / holding window grid — see payload for full grid.
Signal timeline (sparse triggers)
Triggers when overlay takes over base strategy
Stats
| CAGR | 4.68% | Sharpe | 63.36% |
|---|---|---|---|
| Sortino | 92.53% | maxDD | -18.41% |
Disposition
# Fed Balance Sheet ROC Buy Overlay — WSHOSHO ROC5 >4.8 (qm-4bx.19) **Decision:** `NO-SHIP frozen` — investigate-only; do not admit to catalog/Meta. **Commit:** `6575fa16` · **Date:** 2026-08-12 · **Payload:** `research/findings/fed-bal-roc-qm-4bx.19.json` · **Repro:** `python -m research.scripts.audit_fed_bal_roc --write` ## TL;DR Fed balance-sheet liquidity impulse (WSHOSHO 5-day ROC >4.8%) is a buy-only auxiliary overlay that forces max long SPY for a holding window. After maximal FRED ingest, weekly→daily forward-fill with publication lag L=7–10d, and daily `next_close+L` causality via
Follow-on work
- Extend vintage via ALFRED point-in-time to test revision bias (currently `latest`).
- Try M2-ROC placebo only as explicit placebo, not as synthetic balance sheet.
- If revived, pre-register a three-sleeve defensive base (e.g., published Risk-Off 200DMA) and evaluate OOS 60/36 with Deflated Sharpe.
- Weekly native 1W/2W directly vs daily 5/10 with L=7 re-audited on trading calendar (currently calendar-day shift).
- Do not invent pre-2002 balance sheet without cited archive.
Errors & data gaps
- Extend vintage via ALFRED point-in-time to test revision bias (currently `latest`). - Try M2-ROC placebo only as explicit placebo, not as synthetic balance sheet. - If revived, pre-register a three-sleeve defensive base (e.g., published Risk-Off 200DMA) and evaluate OOS 60/36 with Deflated Sharpe. - Weekly native 1W/2W directly vs daily 5/10 with L=7 re-audited on trading calendar (currently calendar-day shift). - Do not invent pre-2002 balance sheet without cited archive.
Follow-up
gh issue create --title "Follow-up Fed Balance Sheet ROC Buy Overlay WSHOSHO ROC5 48 qm-4bx19" --body "Follow-up to qm-4bx.19: Decision: NO-SHIP frozen — investigate-only; do not admit to catalog/Meta. Commi" --label "priority:p2" --label "pipeline:research" --label "agent:ready"