Institutional Quantitative Models
All models run on our shared model engine once per bar and publish canonical signals. Subscribe to models for paper simulation, API webhook delivery, or connected execution.
Model Specialization & Multi-Regime Architecture
Engineered across the complete 5.8-year macroeconomic cycle (2021–2026). Utilizes adaptive holding duration with dynamic volatility trailing stops (letting winning trades run up to 6–8 days during macro expansions) while automatically enforcing defensive de-risking in bear and chop markets.
Specialized for modern institutional post-ETF market dynamics (2025–2026). Engineered with a tactical short-horizon capture window (typically ~48h) and strict conviction hurdles to prevent profit giveback in range-bound chop, yielding up to 80% win rate in sideways markets.
Full-cycle macro regime adaptive trend engine with dynamic ATR trailing stops.
Full-cycle macroeconomic quantitative engine engineered across the complete 2021–2026 market cycle. Abandons rigid 48h exits in favor of adaptive ATR trailing stops, allowing winning trends to run for up to 6–8 days during macro expansion while enforcing swift tactical de-risking in chop regimes.
Calibrated cross-sectional opportunity ranking across top-10 liquid digital assets.
Allocates capital dynamically to the top-3 highest expected-edge opportunities among 10 liquid assets. Utilizes ensemble dispersion uncertainty sizing and throttles gross exposure during macro derivatives stress.
Dynamic opportunity sizing for top-tier conviction breakouts.
Concentrates allocations strictly into high-conviction signals with isotonic probability calibration, cutting cash drag while maintaining capital preservation.
Uncertainty shrinkage and tactical short downside hedging.
Applies conformal prediction bounds to compress position exposure during volatility and initiates tactical short hedges during market stress regimes.
Conservative baseline hurdle for SOL/USDT with sentiment decay.
Baseline reference hurdle combining technical breakout thresholds with sentiment half-life decay on SOL/USDT, establishing the benchmark for semantic overlay models.
Fast-decay sentiment overlay minimizing momentum profit giveback.
Fast half-life sentiment decay overlay (tau = 18h) that locks in gains swiftly as news catalyst momentum fades, before mean-reverting pullbacks occur.
High conviction hurdle (z >= 0.55) minimizing trade frequency for maximum accuracy.
Enters positions only under extreme statistical consensus across technical momentum and sentiment factors to achieve highest win rate.
Intermediate conviction hurdle balancing frequency with win rate.
Balances directional conviction thresholds with disciplined trailing stops to capture medium-duration trend expansions without over-trading.
High-agility SOL/USDT model with FastDecay sentiment overlay.
Dynamic opportunity sizing on SOL/USDT with fast-decay semantic sentiment filtering. Reconciled and verified across 5,570 bars in V3.9b.1 audit.