The framework
The US Equity Allocation Model is a regime-conditional allocation framework, not a single-factor product. The framework is anchored on the Fama-French factor literature: an equity sleeve holds permanent beta exposure plus a rotating factor overlay whose state is governed by independent macro-stress inputs. The model output is a monthly answer to the question "what allocation should this sleeve hold this month" — not which factor to buy long-term.
Two strategies are published in parallel. Core + Momentum Overlayanchors on cap-weighted S&P 500 with a momentum / minimum-vol overlay rotation, designed for subscribers seeking factor-overlay alpha against an EW4 factor blend. Anti-Concentrationanchors on equal-weight S&P 500 with a dynamic top-7 megacap overlay, designed for mandates with elevated cap-weighted megacap concentration risk. Both are core equityproducts by institutional taxonomy (70% in the broad-market core); “Core + Momentum Overlay” replaces the prior “Aggressive” variant label to avoid mandate-mismatch implications.
Architecture — Core + Momentum Overlay
Core (70%, permanent):SPDR S&P 500 ETF (SPY). Held continuously; never trades.
Overlay (30%, rotating): in the default state, iShares MSCI USA Momentum (MTUM). In the defensive state, iShares MSCI USA Min Vol (USMV).
Defensive trigger (1-of-3): VIX above 30, or rolling 60-month z-score of high-yield OAS above +1.50, or trailing 3-month SPY return below −15%. Any one condition fires the rotation to USMV.
State machine: 30-day cooling quarantine after a defensive-to-default exit. Defensive-to-default transitions require two consecutive calm months before reverting (exit confirmation).
Signal construction
The defensive flag is constructed from independent, observable macro inputs — none are derived from the strategy's own returns. VIX and SPY 3-month return are sourced from yfinance; high-yield OAS is sourced from FRED (ICE BofA US High Yield OAS, series BAMLH0A0HYM2).
The state machine is intentionally hysteretic. Once the defensive state fires, exit requires two consecutive months of calm inputs — a one-month dip below threshold is insufficient. After an exit, a 30-day cooling period prevents immediate re-entry on a second stress wave. This pattern was validated against historical false-positive analogs (the V4_cool30 analysis in the project repo).
Academic anchors
Eugene Fama and Kenneth French — the canonical reference for factor decomposition (size, value, momentum). Carhart 1997 extends to the four-factor model used in the strategy's alpha attribution. Asness, Frazzini and Pedersen on factor regime dynamics. Cohen, Polk and Vuolteenaho on factor predictability. The Anti-Concentration framework is informed by literature on equal-weight index construction and concentration risk in cap-weighted indices.
Full prose and citation list: release notes in the project repo.