🧭 Meta-Regime 15-MODULE COMPOSITE

Methodology · Bloomberg-Gap Meta-Regime

The 15 Bloomberg-Gap modules each emit a regime label for their slice of the market (vol, gamma, credit, funding, retail flow, news, central bank, global cross-section, commodities, insider behavior, etc.). This page fuses them into a single meta-state.

Step 1 — Module polarity. Each module's regime label maps to a polarity in {−1, 0, +1}. Polarities are direction-aware: e.g., OVERSHORT funding is bullish-for-spot (squeeze risk), CAPITULATION retail is contrarian-positive, MELTUP_PRONE credit is short-term positive but tail-risk-elevated.

Step 2 — 7 Dimensions. Each module belongs to one dimension; dimension score = mean of its constituents.

• VOL — dealer-gex, vix-curve · stress vs calm
• RISK_ON — credit, crypto-funding, retail · risk appetite
• LIQUIDITY — dix, options-flow · market-making capacity + flow density
• POLICY — cb-stance · Fed restrictiveness (negative = hawkish)
• REFLATION — commodity-curves, global-markets · inflation/cyclical bias
• SMART_MONEY — insider-clusters, 13F, finra-short · institutional positioning
• FUNDAMENTALS — earnings-nlp, news-velocity · earnings tone + media flow

Step 3 — Composite score. Weighted mean × 100 (range −100..+100). VOL and RISK_ON get higher weights (1.5×) because they're the most predictive of near-term returns.

Step 4 — Meta-regime classifier (rule-based).

• MELT_UP — vol low + risk-on high + fundamentals strong + liquidity supportive
• GOLDILOCKS — vol calm + reflation contained + policy not too hawkish
• LATE_CYCLE — risk-on continues BUT policy hawkish + reflation elevated (classic pre-correction)
• NORMAL — balanced, no extremes
• DEFENSIVE — 2-3 dimensions turning negative
• RISK_OFF — vol up + risk appetite weakening
• CRISIS — 5+ dimensions negative or vol explosion

Why this is useful: regime changes in any one module can be noise. Concurrent shifts across multiple dimensions are signal. Bloomberg's market-state summary across the top of every terminal does exactly this; here we get the same lens for our 15 proprietary modules, with Telegram alerts on regime change.

Refresh: Lambda runs every hour at :15 (after most constituent modules refresh). Snapshots stored 168 hours back.