Named strategies with backtest + live shadow track records. Pick what to allocate to.
Each named strategy is backtested with walk-forward purged cross-validation, has its information coefficient (IC) tracked over a rolling 90-day window, and runs in a live shadow book whose returns are recorded to the cryptographically attested audit chain. Cards below let you re-run a backtest, inspect the live IC, and allocate to any strategy.
Validation: each strategy passes walk-forward purged cross-validation (no peek-ahead) with rolling 90-day IC tracking. Strategies must clear ICIR > 0.5 (credible threshold) and show stable performance across regimes (risk-on / risk-off / transition).
Live shadow: all strategies run continuously in a paper-traded shadow book before any capital is allocated. Returns are recorded to the attested audit chain (/audit).
Risk gates: strategies that drop below 60% confidence band, exhibit ICIR < 0.2, or show regime-dependent breakdowns (IC drop > 20% in any regime) are auto-deprioritized; allocation suggestions reflect this.
Limitations: backtests assume mid-price fills and do not model slippage on illiquid names; strategies tagged "live" use real returns only after 30 days of stable shadow performance.
See: Scoring · 5-gate Playbook · Walk-forward validation · Calibration