AI Portfolio Research

Horizon Comparison

Six independently tuned portfolios tracked on their natural 5, 10, and 20-session clocks. Interim marks, matured outcomes, and validation evidence remain explicitly separated.

Evidence statusPaper-live · collecting
Frozen evidence loaded

Prospective strategy evidence

An official close mark appears on the entry day. It is shown as one observed partial day; the completed open-to-open counter and cumulative chart advance at the next open.

The decision session is not a holding session. A July 29 close signal enters at the July 30 open. At the July 30 close, the dashboard reports 1/5 observed (partial) and the official entry-day mark, while retaining 0 complete open-to-open sessions. The first complete interval is available at the July 31 open. Entry-day marks are never mixed with validation estimates or final horizon outcomes.

Horizon comparison

Paired Ridge and XGBoost development evidence with each arm’s own purge, feature subset, refit interval, and maturity date.

Selected model specifications

The winning development configuration for each visible arm. Open a model’s explainability panel to inspect genuine SHAP importance when that frozen vintage captured it.

Pipeline and risk controls

Freshness, universe, NLP, tactical routing, survival guardrails, and artifact integrity for the selected decision date.

Matured forecast history

Prospective outcomes appear only after their full label interval matures. Pending forecasts are not scored as zero.