Independent PhD validation and audit completed in 2026. The review evaluated the model architecture, feature engineering, walk-forward testing protocol, and out-of-sample signal precision across the S&P 500 cycle.
Signals are trained and validated using walk-forward out-of-sample methods on historical S&P 500 data. Model inputs, regime classifications, and signal probabilities are recomputed from live data on a scheduled cycle and time-stamped at the point of publication.