Operator Monitor
Live state of every validated portfolio + the v2 Locality Core Book. Operator-only — not linked from public nav.
Repricing Core book v1
5-position concentrated regime-conditional book (GLD / TIP / LMT / CCJ / short TLT). BETA ↔ REPRICING switcher gated by K stress count. Cycle overlay halves gross during RECESSION.
Repricing Core book v2 — Locality
Cross-pillar top-3 selection on multi-order locality signal. 49-candidate universe across 6 pillars. 15% vol-target, RECESSION × 0.5 cycle overlay. Live picks: CME, ICE, CCJ.
Single-Name Sleeve
28-name thematic basket with 15% vol-target. Late-cycle expression of the Repricing thesis where it dominates SPY-VT with half the drawdown.
Cycle-Beta Sleeve
6-ETF regime-rotation sleeve. EARLY/EXPANSION → 100% SPY; LATE → 60/40; RECESSION → 25/65/10.
All Weather Wrap
Equal-weight Core + Single-Name + Cycle-Beta. The strongest single-product Sharpe in the family — +0.29 Sharpe lift purely from diversification across regimes.
Regime Rotation Sleeve
12-ETF rotation gated by the Inflation Regime classifier. Six locked allocations, one per regime label. Monthly rebalance, 6/6 walk-forward gates pass.
Recently validated
- v2 Core Book — Locality (2026-05-01) — cross-pillar Druckenmiller-style top-3 with multi-order CausalPath signal. Walk-forward Period A (2018-2022, pre-regime): Sharpe 0.54 ~ baselines. Period B (2023-2026, post-Russia regime FIRED): Sharpe 1.89 vs SPY 1.02 (+598bps annualized). MaxDD -8.8% vs SPY -18.1%. MSTR correctly rotated in only when Money pillar fired post-2022. Live: CME, ICE, CCJ.
- Locality signal substrate —
bci_compute/locality_signal.py(multi-order signal),locality_signal_loader.py(RegimeContext loader from full data substrate),food_pillar.py(Food pillar percentile time series).