Loading page content
Methodology
Version PRO-INTELLIGENCE-1.0. A presentation layer above the certified V1 data terminal. It explains what is happening and what changed — it does not forecast prices or issue trade signals.
See also Methodology hub and Physical Tightness.
Pro Intelligence reads certified Gold/Silver Physical Tightness outputs and the FRED macro pack, then builds structured objects: current state, pressures, material changes, divergences, and watch conditions. Every important conclusion carries an explanation trace (input, observation, state, direction, date, source, explanation).
Core production intelligence is deterministic. Editorial Substack research is surfaced separately and labeled as editorial analysis.
Statements describe observed conditions and conditional watches (for example, “Managed Money percentile is in an extreme band”). They do not claim future performance.
Tiny numerical moves are ignored. Materiality uses explicit thresholds, including:
Divergences highlight disagreement between components already present in the data platform (for example ETF accumulation vs inventory draw, or elevated real-rate proxy vs non-loose gold physical conditions). They are observational, not predictions.
Each watch has a stable ID, deterministic logic, current state (clear / active / triggered / unavailable), and an explanation of why it matters. Notifications are out of scope for this phase.
Macro Intelligence does not invent a composite macro score in this version. It reports states for liquidity, real rates, dollar, and curve context from FRED and Grey Rabbit derived series only (no DXY substitution).
Unavailable inputs are labeled Unavailable / Delayed / Latest reported / insufficient coverage. The layer does not substitute a neutral 50 or invent observations.