What the desk just decided
Ten independent agents score every covered name from real price history and combine into one BUY / SELL / HOLD call. The moment that call changes on a name, the flip lands here — the prior call, the new call, the conviction behind it, and the one-line reason from the agents' own votes.
How the agent weighting adapts
accumulating — weights hold at the equal-weight baselineThe ten agents' voting weights come from bounded, disclosed online reweighting — every change is graded against real outcomes and replayable. Hit-rates carry their 95% confidence range and the count of distinct graded calls behind them; excess edge is measured vs the market benchmark over the same window. A figure not yet earned from real outcomes shows as “—”.
Still accumulating graded outcomes — every agent holds the equal-weight baseline (1.00×). Hit-rates, confidence ranges and edges appear here only once real forward returns grade the calls.
window 180 days · 0 graded calls · effective n 0
Equal weight (every agent at 1.0x) is the permanent unlearned baseline every disclosed weight is read against.
Each item is a change in the ten-agent consensus on a covered name, derived by comparing successive computations on real price history — never back-filled, never invented. The feed covers changes observed since the service last started. Your decisions, your risk.