Verdict Ledger
Every published stance scored against its per-layer benchmark over the same window, so opinions carry a track record.
Relative return vs each report's benchmark is the score; a call that trails its benchmark is a miss even when it is up. Calibration is grouped by benchmark bucket, and the book-level SMH reference line keeps the all-calls comparison explicit. Migration rows (the legacy→v5 re-labelling) stay scored vs SMH. No red/green — the sign carries direction.
Open Calls
Live positions: each report's current stance vs its benchmark since the stance date. Sortable by relative delta.
Closed Intervals
Completed stance windows from stanceHistory. Legacy→v5 re-labellings are flagged "migration" and do not count as calls.
Calibration
Scored calls are grouped by benchmark bucket: median relative delta per stance, plus each bucket's non-neutral beat-rate. The separate book-level line uses SMH for every scored call.
This page is for research and information only. It is not investment advice. Please do your own due diligence.