A 195-state universe
The broad scope includes the 193 UN member states, Palestine and Vatican City. Territories and dependencies are excluded to prevent double counting.
Methodology
Official monetary gold, with private models kept separate.
The primary metric is gold held as a reserve asset by a national monetary authority, divided by resident population. Private-stock estimates remain a separate modelled layer.
Primary formulafine tonnes x 1,000,000 / population
The primary result is grams per resident. Fine troy ounces are calculated as tonnes x 32,150.7465686 / population. The UN World Population Prospects mid-year projection for 2026 supplies the denominator.
Coverage follows evidence
The official series currently contains 29 curated rows. Missing official values do not become zero. The 195-state private-stock model remains separate and identifies peer estimates clearly.
01
One primary metric and two modelled extensions
Official monetary gold is different from a household coin, a fund holding or jewellery. Private layers never alter the meaning of the primary official-reserves ranking.
Official monetary gold
Gold held as a reserve asset by the reporting monetary authority.
- Official
Official plus investment
Official holdings plus physically backed funds, bars and coins.
- Official
- ETF and ETP
- Bars and coins
Broad attributed
Official and investment holdings plus jewellery.
- Official
- ETF and ETP
- Bars and coins
- Jewellery
- Silverware
02
From source to a validated release
The pipeline favours a direct country observation, then a derived value, an anchored estimate, a peer estimate and finally a modelled residual.
Private source snapshots, permissions and checksums
Fine tonnes, dates, countries and attribution modes
Component stocks with low, base and high scenarios
Versioned derived metrics and public provenance
03
Confidence describes evidence quality
Confidence labels describe the source, derivation and attribution. They do not describe whether a country ranks highly. Value ranges feed 400 deterministic simulations, and the country detail shows the resulting 5th to 95th percentile rank band.
GradeEvidence statePublic label
ADirect official or product-level sourceReported
BDerived from direct country observationsDerived
CCountry-anchored modelEstimated
DPeer model or residualModelled
Recycling is a flow
Recovered metal leaves an existing stock bucket. It is never added again as a new holding.
Private gold is modelled
Fund domicile is not beneficial ownership. Bars, coins and jewellery rely on accumulated demand, trade and survey anchors.
Ownership, not vault location
Gold is attributed to the reporting monetary authority that owns or controls it, not the country where the metal is physically stored.
Current source status
The official ranking currently includes 29 curated country rows with an observation period and source status attached. Comparable values remain unavailable for other countries rather than being filled with estimates. World Gold Council material is quality-control context only.
Source hierarchy
Primary institutions and visible access limits.
Restricted source snapshots stay private. Public releases contain derived values, citations, dates, assumptions and publication status.
01
IMF International LiquidityPreferred official monetary-gold source; row adapter pending · Open source
02Curated gold source setCurrent official rows with ongoing IMF or national-source reconciliation · Research reference
03World Gold Council, Gold Reserves by CountryQuality-control context only; excluded as a publication input · Permission required for reuse
04UN M49 country and area standardCanonical geography identifiers · Open source
05UN World Population Prospects 2024Annual mid-year population denominator · Open source
06World Bank country and lending groupsIncome-group covariate for the peer model · Open source
07Metal Per Capita global peer modelCountry anchors, peer allocation, uncertainty and residual reconciliation · Research reference
Continue exploring