Loading page content
Methodology
Grey Rabbit displays only datasets that are publicly available or legally redistributable under their terms. Licensed or unclear sources are omitted — never fabricated.
Gold and silver use different methodologies. Gold remains a single composite. Silver 1.0 publishes Overall together with Structural, Availability, and Acute Stress so a Balanced overall reading cannot hide Tight long-run supply.
The Tightness Score is a normalized composite of scored inputs only. It is not a price forecast, squeeze alert, or positioning trade signal.
Pressure drivers sit beside — not inside — the Tightness Score. Each component exposes state, direction (tightening / neutral / loosening), strength, confidence, native freshness, observation date, and source cadence. Components with insufficient history return INSUFFICIENT COVERAGE; they are never scored as neutral 50.
Freshness is source-native: COMEX daily, CFTC weekly, LBMA monthly, ETF filing cadence, Mint monthly, USGS annual, SGE monthly, SHFE weekly/manual. A current monthly LBMA observation is not penalized for not updating daily.
Divergence flags (physical vs positioning, West vs East, inventory vs ETF flow, retail vs institutional) explain conflicting signals without forcing a single bullish/bearish conclusion.
Equal-weighted average of available verified gold signals. Score published only when scoring completeness ≥ 40%.
Thresholds: ≥60 Tight · ≤40 Loose · otherwise Balanced.
Missing inputs are excluded. They are never treated as neutral 50s.
Silver Physical Tightness measures how constrained the physical silver market is after combining long-run supply conditions with visible inventories. It is not the silver price and not a squeeze alert. PHYSICAL-SILVER-0.1 remains reproducible for audit.
Whether mine supply and recycling have persistently fallen short of demand, and how many months of identifiable bullion remain. Slow-moving. Tight here is not an active squeeze.
Visible bullion at major venues — New York COMEX registered, London headline vault tonnes, Shanghai inventories, and ETF metal — scored separately. Location matters. Headline LBMA stocks are not lendable London free float. ETF holdings are not subtracted from London vaults and are not double-counted with LBMA.
Whether scarcity is showing up in borrowing, delivery, or regional premiums. Coverage is currently limited. Missing Stress is not scored as 50, 0, or calm. When Stress is unavailable the headline renormalizes Structural 50% and Availability 40% to 55.6% / 44.4%. Data Confidence is reduced instead.
Data Confidence is a completeness and coverage index — not a probability that the score is correct. Trend is the 12-month change in observable Structural + Availability conditions. Silver can be structurally Tight while Overall is Stable.
Structural scoring prefers completed World Silver Survey actuals. A forecast year may overlay at low weight and must not dominate. Identifiable silver bullion inventories follow Metals Focus printed year-end totals; they are not reconstructed from exchange snapshots.
Identifiable silver bullion inventories follow Metals Focus printed year-end totals used by the Silver Institute World Silver Survey. They are not reconstructed from COMEX, LBMA, or ETF snapshots, and they are not a London free-float series.
Gold PHYSICAL-GOLD-0.2 remains a verified-signal composite of inventories, ETF, and official-sector inputs. Silver 1.0 adds a structural deficit/stock pillar because the silver market's multi-year supply gap is the dominant long-run constraint, and it keeps venue availability separate so rebuilt inventories cannot erase that structural tightness. Gold is unchanged by this migration.
Structural Tight means persistent deficits relative to identifiable stock. An active squeeze would require observed acute stress. Current acute-stress coverage cannot confirm London borrowing or free-float conditions.
Licensed COMEX silver warehouse history has a gap from 2025-07-22 through 2025-11-30. That is a historical validation limitation, not a current production input failure. No interpolation was used. October 2025 Availability reconstruction is limited and that October classification is not robust. The current 2026 score does not depend on the missing window; Combined remains Balanced if COMEX is removed entirely.
Classification bands: Very Loose 0–29 · Loose 30–44 · Balanced 45–62 · Tight 63–76 · Very Tight 77–88 · Extreme 89–100.
Source data: official warehouse or vault observations from CME depository reports and LBMA London Vault Holdings Data.
Grey Rabbit calculations: changes, percentiles, trends, unit conversions, normalization, confidence, and composite scores.
Source: CME Group Daily Depository and Registrar Reports. Grey Rabbit Markets is not affiliated with or endorsed by CME Group.
Source: LBMA London Vault Holdings Data. Reported monthly, one month in arrears. Grey Rabbit Markets is not affiliated with or endorsed by LBMA.
Full legality matrix: docs/physical-markets-hub-legality-matrix.md