A dollar-neutral, daily cross-sectional REVERSAL basket across ~10 established, confirmed-full-history Binance USD-M majors (BTC, ETH, SOL, BNB, XRP, DOGE, ADA, LTC, LINK, BCH) — the natural diversifier to the portfolio's momentum-heavy book. Each rebalance it ranks majors by their recent (≈5-day) VOLATILITY-NORMALIZED return, goes LONG the biggest relative LOSERS and SHORT the biggest relative WINNERS, in equal dollar notional, betting on the short-term overreaction reverting. This is deliberately the STRUCTURALLY-SAFE form of mean-reversion: every time-series MR this factory tried (15m z-fade, 4H volume-climax) blew up catastrophically (-99.96%, -97.7%) on fee drag + single-asset gap-through tail losses, but a CROSS-SECTIONAL, dollar-neutral, RANK-based reversal is fundamentally different — it is self-hedging (long-loser/short-winner nets out market moves), has NO single-asset stop to gap through, and rebalances WEEKLY (not intraday) to control fees. There is positive evidence this class works here: CrossSectional7DayReversalV3 is in the portfolio and is NOT in the failure list (unlike all the time-series MR). It stays data-safe (established majors only — broad-alt/HL baskets die on data), market-neutral (improving the 87%-long skew), and is engineered to dodge BOTH recurring sizing failures: NO vol-targeting leverage scalar and NO uncapped inverse-vol weights — equal-ish capped sizing with gross hard-capped at 1.0× (the leverage-explosion fix), and a single frozen lookback (the overfit fix). It complements rather than duplicates the existing 7-day reversal by using a shorter ≈5-day vol-normalized formation and weekly rebalance. It is genuinely orthogonal to the factory's validated momentum factor (opposite sign), the highest-value diversifying sleeve available on the validatable surface.