Source-labelled market dataSource, observation time and freshness are shown with every quoteData statusAll pages
GoldObserve

SOURCES / DATES / UPDATES / LIMITS

Gold Price Data Sources and Live Status

Check the source, date, update status and limits behind each price and research series. If a current update fails, GoldObserve labels the last verified value or leaves the result unavailable.

THE SHORT ANSWER

See where each number comes from and how current it is

Every market value names its source, observation time and current status. If a source is unavailable, the page either shows a clearly dated last verified value or no value at all. Currency and weight conversions are calculations based on the displayed metal reference, not separate exchange quotes.

CURRENT GOLD & SILVERGold API

Indicative XAU/XAG values with the source observation time.

DAILY REFERENCENational Bank of Ukraine

A dated accounting reference used when a supported daily view needs the last verified value.

LONG HISTORYWorld Bank

Monthly-average gold, silver, copper and platinum series from 1960.

ECONOMIC CONTEXTOfficial U.S. sources

Rates, inflation, money, Federal Reserve assets and positioning retain their own dates.

Separate definitions stay separate instead of becoming one artificial chart: a daily accounting reference is not treated as a spot close, and a monthly average is not presented as an intraday price.

HOW TO READ THE TIMESTAMPS

A refreshed page does not always mean a newer market observation

The source observation time tells you when the value was measured. GoldObserve may receive or display that same observation later. If an online update fails, a last verified value can remain useful only when its original date and older-data status stay visible.

Five-stage GoldObserve market-data evidence chain separating the source observation and source clock from validation, server cache, browser state and a versioned local baseline that remains labelled stale rather than live.
Source evidence and delivery evidence answer different questions. A cache can deliver a previously validated observation without creating a new source tick; a versioned baseline carries its own generation time and hash and is never relabelled live. If no truthful layer is available, GoldObserve shows unavailable.Swipe the diagram horizontally to read every label.Open full-size SVG

Remember: use the source observation time—not the page refresh time—when comparing two values.

Market-data permissions and download limitsTechnical and licensing details

LIVE MARKET RELEASE GATE

Written display permission still stops at the dataset boundary

GoldObserve reviewed the first-party terms and source-chain disclosures for Gold API and Frankfurter on 2026-08-27. Gold API's operator supplied written commercial-display confirmation for the exact GoldObserve request: XAU/XAG display, a roughly 30-second shared cache, currency and unit conversions, derived changes and ratio calculations, temporary last-valid retention and same-origin delivery to GoldObserve pages. The request expressly excluded resale, a public third-party API and bulk raw-data downloads.

ALLOW2

Narrow uses with attribution and implementation controls.

PERMISSION0

Written confirmation required before production use.

BLOCK3

Do not infer rights from a callable endpoint.

LIVE DISPLAY GATEPASS

Gold API display is enabled; provider-derived downloads remain blocked.

Provider and useDecision and rights boundaryRelease condition
Gold APICommercial-site XAU/XAG fallback and transformed local-currency displayEnabled under written commercial-display confirmation. On August 27, 2026, Gold API's operator replied affirmatively to GoldObserve's written request covering commercial XAU/XAG display, a roughly 30-second shared cache, currency and unit transformation, percentage and ratio calculations, temporary last-valid retention and same-origin delivery for GoldObserve pages. The request expressly excluded resale, a public third-party API and bulk raw-data downloads.Keep the provider server-side, retain the 30-second shared cache and source timestamp, use it only for GoldObserve on-page display, and exclude Gold API rows from public CSV, bulk archives and third-party API redistribution.Read governing terms
Frankfurter v2 default blendCurrent and historical currency conversion without a pinned providerDefault provider blend removed. Frankfurter allows commercial API use but states that each underlying provider's terms still apply; the default blended provider set can change and therefore cannot support one stable redistribution decision.Keep the default blend absent from runtime URLs and invalidate any cache version that predates explicit provider pinning.Read governing terms
Frankfurter with providers=ECBEUR, GBP, CAD, AUD, INR, CNY, JPY and CHF conversion derived from ECB reference ratesImplemented with provider pinning. ECB reference rates can support commercial reproduction when GoldObserve reproduces them accurately, attributes the ECB, labels modifications and provides the required full-window link on business sites.Retain explicit providers=ECB requests, provider-attribution validation, modification labels and source links; do not treat reference rates as executable FX quotes.Read governing terms
Frankfurter AED conversionIntraday or historical AED gold-price conversion through FrankfurterNo reviewed source coverage. AED is not covered by the ECB currency set reviewed for this release, so the approved ECB path cannot be extended to AED by assumption or by silently accepting Frankfurter's default blend.Remove AED from the approved release set or complete a separate first-party provider and commercial-reuse audit before publishing AED-derived values.Read governing terms
Provider-derived CSV and JSONCSV export, bulk download, stable public API or archive containing live-provider rowsNo redistribution claim. GoldObserve's conversion, formatting and source metadata do not erase upstream rights. The current evidence supports selected on-page uses, not a general licence to distribute source observations as a standalone dataset.Exclude blocked provider rows from downloadable exports or obtain written redistribution permission for every upstream source in the exported chain.Read governing terms

Current release decision: enable Gold API for source-labelled on-page XAU/XAG display behind the 30-second server cache. Keep ECB-pinned FX, NBU history and local fallback separate, and continue blocking Gold API rows from CSV, bulk archives and third-party API redistribution.

CURRENT DATA CHECK

See which feeds are available right now

The check loads gold, silver, recent history and five economic series separately. One source can fail without hiding the status of the others.

CURRENT CHECKChecking...A snapshot of what this page can load now.
LAST CHECKEDWaitingRefreshes every 60 seconds while this page is open.
IF A SOURCE FAILSShow the last verified dataOlder data is labelled; missing data stays unavailable.
FeedStatusSource and observationAgeData path

Available means the page received a usable response at the displayed time. Last verified means the newest online update was unavailable, so an older dated value is shown instead.

Dataset change historyAdditions, revisions and corrections

VERSIONED DATA CHANGE LEDGER

Routine updates, source revisions and corrections stay separate

The ledger records reviewed checked-in dataset changes only. Source observation, official publication or schedule, GoldObserve retrieval, local baseline generation and approval are separate clocks; a missing clock stays explicitly unavailable. Routine appends, upstream revisions and GoldObserve corrections remain separate, and an empty category is not evidence that no unreported issue exists.

9 reviewed changes · 7 routine appends · 1 source revisions · 0 GoldObserve corrections

Zero is a disclosure, not a guarantee. The ledger starts with changes that have complete local evidence. It does not retroactively claim that every earlier edit was classified under this contract.

SOURCE SWIMLANE TIMELINE

Follow approval events without merging source clocks

9 of 9 reviewed changes shown. Horizontal position uses the local approval date; each event opens its separate source, publication, retrieval and baseline clocks.

  1. NBU gold accounting official-reference historyRoutine append · +16 / ~0 / −0
  2. GoldObserve reviewed FRED macro fallback baselineSource revision · +22 / ~1 / −0
  3. ECB euro-area national central-bank monetary-gold baselineBaseline publication · +5873 / ~0 / −0
  4. CFTC COMEX gold disaggregated futures baselineRoutine append · +1 / ~0 / −0
  5. GoldObserve reviewed FRED macro fallback baselineRoutine append · +11 / ~0 / −0
  6. GoldObserve reviewed FRED macro fallback baselineRoutine append · +1 / ~0 / −0
  7. GoldObserve reviewed FRED macro fallback baselineRoutine append · +6 / ~0 / −0
  8. NBU gold accounting official-reference historyRoutine append · +2 / ~0 / −0
  9. NBU gold accounting official-reference historyRoutine append · +1 / ~0 / −0
Filtered reviewed data changes (9)
Dataset and sourceChangeEvidence clocksReconciliationVersion evidence
NBU gold accounting official-reference historyNational Bank of Ukraine Open DataRoutine appendAdded 16 official effective dates from 2026-08-12 through 2026-08-27 to the local date-matched NBU XAU and USD reference history.Observation 2026-08-27Approved Aug 27, 2026, 05:01 UTC+16 / ~0 / −09,450 → 9,466 observations7bf72efblob 23675a7d8ad3
GoldObserve reviewed FRED macro fallback baselineFederal Reserve Bank of St. Louis FREDSource revisionAccepted the reviewed DTWEXBGS historical replacement and appended the later macro observations through 2026-08-21.Observation 2026-08-21Approved Aug 25, 2026, 10:56 UTC+22 / ~1 / −06,365 → 6,387 observationse148f42blob 950f62093d26
ECB euro-area national central-bank monetary-gold baselineEuropean Central BankBaseline publicationPublished the first reviewed internal ECB monetary-gold baseline for 21 configured euro-area national central banks.Observation 2026-07-01Approved Aug 15, 2026, 01:01 UTC+5873 / ~0 / −05,873 observations published as the first baselinec417925blob 6f1d11ba6457
CFTC COMEX gold disaggregated futures baselineU.S. Commodity Futures Trading CommissionRoutine appendAdded the 2026-08-11 COMEX gold report to the reviewed CFTC baseline and retained its exact raw archive version.Observation 2026-08-11Approved Aug 15, 2026, 01:01 UTC+1 / ~0 / −01,052 → 1,053 observationsc417925blob 0a4005eeb271
GoldObserve reviewed FRED macro fallback baselineFederal Reserve Bank of St. Louis FREDRoutine appendAdded 11 later observations across WTI, real yields, CPI and the Federal Reserve balance sheet to the reviewed macro baseline.Observation 2026-08-13Approved Aug 15, 2026, 01:01 UTC+11 / ~0 / −06,354 → 6,365 observationsc417925blob 6aca4fa19f8b
GoldObserve reviewed FRED macro fallback baselineFederal Reserve Bank of St. Louis FREDRoutine appendAdded the August 10, 2026 DFII10 real-yield observation to the reviewed local macro fallback baseline.Observation 2026-08-10Approved Aug 11, 2026, 21:12 UTC+1 / ~0 / −06,353 → 6,354 observationsf4e1d53blob 2d3f3c549135
GoldObserve reviewed FRED macro fallback baselineFederal Reserve Bank of St. Louis FREDRoutine appendAdded one DFII10 observation and five DTWEXBGS observations to the reviewed local macro fallback baseline.Observation 2026-08-07Approved Aug 11, 2026, 01:08 UTC+6 / ~0 / −06,347 → 6,353 observations13b0612blob 1eb661fba1b3
NBU gold accounting official-reference historyNational Bank of Ukraine Open DataRoutine appendAdded official effective dates 2026-08-10 and 2026-08-11 to the local date-matched NBU XAU and USD reference history.Observation 2026-08-11Approved Aug 11, 2026, 01:08 UTC+2 / ~0 / −09,448 → 9,450 observations13b0612blob 9667329a1acc
NBU gold accounting official-reference historyNational Bank of Ukraine Open DataRoutine appendAdded the official effective date 2026-08-09 to the local NBU XAU and date-matched USD reference history.Observation 2026-08-09Approved Aug 9, 2026, 01:33 UTC+1 / ~0 / −09,447 → 9,448 observations218ae08blob 957b28e3f71b
NBU gold accounting official-reference history: clocks, verification and impact
Source observation
2026-08-27
Official publication
Exact official publication time not retained
GoldObserve retrieval
Reviewed retrieval · Aug 27, 2026, 04:57 UTC
Local baseline generated
Separate generation clock not retained
Local approval
Aug 27, 2026, 05:01 UTC

The exact reviewed candidate SHA-256 matched the check-only result and the full-history comparison found no changed or deleted earlier observation.

Verification: The reviewed candidate records SHA-256 was 7d2b381e4bb4af9c1d6647315d65ef86d66141e69fc5c74efd04f44df6ad5f30; the normalized full-history gate reconciled 9,450 to 9,466 observations with no historical replacement.

Affected routes: /api/history/official-reference, /market/dashboard, /data-sources.

Repository evidence: commit 7bf72ef3b0cc59afc30c73fd03336afc85eb521e; after blob 23675a7d8ad3fab3e5955e70c472ed575b30d29d.

GoldObserve reviewed FRED macro fallback baseline: clocks, verification and impact
Source observation
2026-08-21
Official publication
Exact official publication time not retained
GoldObserve retrieval
No retrieval clock retained
Local baseline generated
Aug 25, 2026, 10:53 UTC
Local approval
Aug 25, 2026, 10:56 UTC

The exact candidate hash matched the checked-only result; the 2026-08-03 old and revised values were explicitly reviewed before promotion.

Verification: DTWEXBGS 2026-08-03 changed from 120.7739 to 119.6951; the candidate records SHA-256 is 70e2ea7515f2eb60b3de22675f98f2dd59eaa5e3f992df6dbccbeadea86f5d7f, with no deleted observations.

Affected routes: /api/macro, /gold-correlation-explorer, /gold-macro-regime-explorer, /gold-price-vs-real-interest-rates, /gold-price-vs-us-dollar, /gold-data-brief, /data-sources.

Repository evidence: commit e148f423ca26a112bc4301c9bb2daa30d108db77; after blob 950f62093d26a38ce97a0f39c3cc9404233d8d73.

ECB euro-area national central-bank monetary-gold baseline: clocks, verification and impact
Source observation
2026-07-01
Official publication
Exact official publication time not retained
GoldObserve retrieval
No retrieval clock retained
Local baseline generated
Aug 15, 2026, 00:54 UTC
Local approval
Aug 15, 2026, 01:01 UTC

The ECB usage policy, complete configured country scope, exact unit conversion and hash-locked check passed; the snapshot remains internal until a later official vintage proves revision behavior.

Verification: The official check returned 5,873 observations through July 2026 with one preserved missing value, zero continuity gaps and records SHA-256 113ad50dbc18560fe3fb8f7f852b62b2e49b2499ca8ae70486f6088028d3b516.

Affected routes: /data-sources.

Repository evidence: commit c4179258bc2f3736e068d0cb68145510a40cbd38; after blob 6f1d11ba645752fbc7be8e9ba65018c3158f7f71.

CFTC COMEX gold disaggregated futures baseline: clocks, verification and impact
Source observation
2026-08-11
Official publication
Exact publication not retained; official schedule Aug 14, 2026, 19:30 UTC
GoldObserve retrieval
First successful retrieval · Aug 15, 2026, 00:50 UTC
Local baseline generated
Aug 15, 2026, 00:50 UTC
Local approval
Aug 15, 2026, 01:01 UTC

The official archive comparison exposed one later report and found zero changed or deleted historical observations.

Verification: The normalized baseline reconciled 1,052 to 1,053 observations; the new report produced managed-money net 137,662 contracts and no historical replacement.

Affected routes: /api/cot, /gold-cot-report, /market/dashboard, /gold-data-brief, /data-sources.

Repository evidence: commit c4179258bc2f3736e068d0cb68145510a40cbd38; after blob 0a4005eeb27160998d2bdfcf4af7ac3a36f59c39.

GoldObserve reviewed FRED macro fallback baseline: clocks, verification and impact
Source observation
2026-08-13
Official publication
Exact official publication time not retained
GoldObserve retrieval
No retrieval clock retained
Local baseline generated
Aug 15, 2026, 00:50 UTC
Local approval
Aug 15, 2026, 01:01 UTC

The exact official CSV comparison found 11 additions and zero changed or deleted observations across all six configured macro series.

Verification: The version ledger reconstructs six WTI, three DFII10, one CPI and one WALCL addition from retained Git snapshots, with no revision or deletion.

Affected routes: /api/macro, /gold-correlation-explorer, /gold-macro-regime-explorer, /gold-price-vs-real-interest-rates, /gold-data-brief, /data-sources.

Repository evidence: commit c4179258bc2f3736e068d0cb68145510a40cbd38; after blob 6aca4fa19f8ba4ad1ee9f7e0b2e84a6d01ae91b7.

GoldObserve reviewed FRED macro fallback baseline: clocks, verification and impact
Source observation
2026-08-10
Official publication
Exact official publication time not retained
GoldObserve retrieval
No retrieval clock retained
Local baseline generated
Aug 11, 2026, 21:09 UTC
Local approval
Aug 11, 2026, 21:12 UTC

A fresh official FRED CSV comparison exposed one later observation and found no changed or deleted checked-in observation across all six macro series.

Verification: The hash-locked reviewed write matched DFII10 at 2.43 percent for 2026-08-10; the six-series comparison reported one addition and no historical replacement.

Affected routes: /api/macro, /gold-correlation-explorer, /gold-macro-regime-explorer, /gold-price-vs-real-interest-rates, /data-sources.

Repository evidence: commit f4e1d53b7ad1ad8bcc4eb7de4d61a209d3aa2e83; after blob 2d3f3c5491352885582b0c059304a1ae0feb1658.

GoldObserve reviewed FRED macro fallback baseline: clocks, verification and impact
Source observation
2026-08-07
Official publication
Exact official publication time not retained
GoldObserve retrieval
No retrieval clock retained
Local baseline generated
Aug 11, 2026, 00:38 UTC
Local approval
Aug 11, 2026, 01:08 UTC

The official FRED CSV comparison exposed six later observations and found no changed or deleted checked-in observation.

Verification: The reviewed source comparison matched DFII10 at 2.40 for 2026-08-07 and five DTWEXBGS values through 2026-08-07, with no historical replacement.

Affected routes: /api/macro, /gold-correlation-explorer, /data-sources.

Repository evidence: commit 13b06128871ef9f5a511bb4826117b5cc4fdb711; after blob 1eb661fba1b3d08fab871b6f3d5dbfaed4c19463.

NBU gold accounting official-reference history: clocks, verification and impact
Source observation
2026-08-11
Official publication
Exact official publication time not retained
GoldObserve retrieval
Reviewed retrieval · Aug 11, 2026, 00:38 UTC
Local baseline generated
Separate generation clock not retained
Local approval
Aug 11, 2026, 01:08 UTC

The reviewed NBU candidate contained two later effective dates and the full-history comparison found no changed or deleted earlier observation.

Verification: Date-matched NBU XAU and USD rows for both new dates were reviewed, then the normalized full-history gate reconciled 9,448 to 9,450 observations with no historical replacement.

Affected routes: /api/history/official-reference, /data-sources.

Repository evidence: commit 13b06128871ef9f5a511bb4826117b5cc4fdb711; after blob 9667329a1acc8ccf6629a334874f1ab96b1e4f5b.

NBU gold accounting official-reference history: clocks, verification and impact
Source observation
2026-08-09
Official publication
Exact official publication time not retained
GoldObserve retrieval
Reviewed retrieval · Aug 9, 2026, 01:26 UTC
Local baseline generated
Separate generation clock not retained
Local approval
Aug 9, 2026, 01:33 UTC

The reviewed NBU candidate contained one later effective date and the full-history comparison found no changed or deleted earlier observation.

Verification: Date-matched NBU XAU and USD rows were reviewed, then the normalized full-history gate reconciled 9,447 to 9,448 observations with one addition and no historical replacement.

Affected routes: /api/history/official-reference, /data-sources.

Repository evidence: commit 218ae08e20c3f08da8661da8e9716db849eecb5d; after blob 957b28e3f71b89a67745e9a10c668317d09ef921.

Ledger version: data-change-ledger-2026-08-27-v1; reviewed . Repository commits and blob object IDs prove the checked-in change; they do not prove upstream publication time by themselves.

Macro and ECB version recordsFor reproducibility and revision review

RETAINED MACRO VERSION EVIDENCE

Inspect additions, revisions and deletions in the dedicated vintage explorer

Five reviewed local baselines preserve exact commits, generation times, series hashes and four adjacent comparisons. The focused explorer separates newly published dates from changed historical values and keeps the ALFRED boundary visible.

Open Gold Macro Data Vintage Explorer

ECB MONETARY-GOLD VERSION GATE

The research page is live; revision history remains gated

GoldObserve can reconstruct the exact reviewed ECB monetary-gold baseline from Git, including its source payload hash, normalized-record hash, country scope and preserved missingness. The explorer is available now, but a second approved official vintage does not exist yet, so the site deliberately shows the evidence gap instead of drawing a revision trend.

Explorer live; revision gate open1 reviewed vintage · 0 comparable transitions · revision behavior unproven
Exact retained ECB monetary-gold version evidence. A later approved row will unlock an adjacent-version comparison.
Reviewed versionCoverageQuality preservedReconstruction evidence
c417925Aug 15, 2026, 12:54 AM5,873 observations21 national central banks · through 2026-071 missing · 0 gaps201 zero values · 196 fractional derivations retained113ad50dbc18Git blob 6f1d11ba6457 · source payload 1147e5d52cdb
WHAT IS PROVEN

The current 5,873-row file, its 21-country scope, quality counts and hashes are reproducible from commit c417925.

WHAT IS NOT PROVEN

No adjacent official vintage exists in the ledger, so zero observed revisions cannot be claimed and no stock change is labelled as a central-bank purchase.

NEXT ACCEPTANCE

Check, hash-approve and commit a later ECB vintage; then compare exact country-month keys, values and statuses. Any deletion, revision, scope change or continuity gap stops automatic adoption.

Evidence boundary. Only one reviewed ECB vintage is currently retained. That proves the baseline can be reconstructed, but it does not prove how ECB revisions behave. No country total, euro-area aggregate, purchase estimate, trend or forecast is inferred from this ledger. Inspect the first-party source.

Complete source registerCoverage, cadence and definition limits

SOURCE REGISTER

The current provider chain and exactly what each source does

ProviderRole and cadenceGoldObserve boundary
Live displayGold API XAU/XAG through GoldObserve's same-origin route and 30-second shared cacheIndicative display only; no raw resale, public third-party API or bulk provider download
Gold APIPrimary current XAU/XAG source after written commercial-display confirmation dated August 27, 2026Price and source timestamp retained; bid, ask, session range and previous close remain blank because the current-price endpoint does not publish them
ECB via Frankfurter v2Explicit providers=ECB conversion for EUR, GBP, CAD, AUD, INR, CNY, JPY and CHF; dated rates cached privately for six hoursECB attribution and derived-value labels required; default blended rates and Frankfurter-based AED history are rejected
Metals.devFinal configured fallback; required for XPT and XPD in the current applicationOnly used with a valid server-side key and applicable account terms
National Bank of Ukraine Open DataSelf-owned GoldObserve history of the daily XAU/XAG accounting price in UAH and date-matched official UAH-per-USD rate; checked dailyIndependent official accounting reference from September 2000, not an executable quote or dealer price
World Bank Pink SheetVersioned GoldObserve history of licensed monthly-average nominal USD gold, silver, copper and platinum from 1960; checked weeklyMonthly composite series with commodity-specific units and definition boundaries; not a daily close, dealer quote or continuous modern exchange benchmark
BLS CPI-U via FREDSelf-owned monthly U.S. city-average all-items CPI history from 1960; public-domain baseline updated only when observations changeNot seasonally adjusted; October 2025 remains a documented missing value and is not interpolated
Verified macro fallback baselineSix versioned latest-vintage series for WTI, real yield, broad dollar, seasonally adjusted CPI, M2 and Federal Reserve assets; refreshed by an explicit maintenance commandUsed only after online paths fail, always marked stale, hash-recorded and never described as an ALFRED point-in-time vintage
CFTC COT Public Reporting + annual archivesWeekly COMEX Gold Disaggregated Futures Only API; content-addressed raw ZIP versions, byte counts, retrieval times and SHA-256 evidence back the local fallbackFirst successful retrieval is local availability evidence, not an official publication time; no missing week, trader motive or price direction is estimated
USGS Mineral Commodity Summaries 2026Annual country mine production and mineral reserves; companion CSV last updated May 27, 20262025 production is estimated; reserves, mine output, recycling and monetary gold remain separate measures
USGS Silver Data Series 140 + MCS 2026U.S. secondary refinery production from new and old scrap, 1900–2025, joined only after an exact 2021 overlap check1900, 1995 and 1996 remain NA; refinery output is not a recycling rate, scrap pool or price forecast
Federal Reserve + BLS + BEA + CFTC schedulesVersioned 2026 FOMC, CPI, employment, PCE and COT schedule entries for the market calendarNormal page loads use the reviewed local snapshot; tentative dates remain tentative and no event is presented as a directional gold forecast
Editorial source review scheduleTax, legal, Mint and guidance sources

EDITORIAL SOURCE REVIEW

High-impact legal, tax, Mint and guidance sources now have review clocks

A live URL is not proof that a claim is still current. This separate registry records the affected pages, risk tier, last human review and next review date. Automated checks may flag a due, redirected, restricted or dead source, but they never rewrite a citation or substitute another source.

REGISTERED20 sources

Only high-impact editorial evidence in the first cohort.

CRITICAL6 sources

Tax, legal and contract-rule surfaces receive the shortest clocks.

DUE NOW0

A due item is an alert, never permission to replace it.

NEXT REVIEW2026-09-10

Earliest scheduled manual review in this registry version.

Source familyReviewed evidenceNext control
tax3 reviewed sources · Internal Revenue ServiceEarliest review 2026-09-10 · alert only
legal3 reviewed sources · U.S. Government Publishing Office eCFR, Federal Trade Commission, Cámara de Diputados del H. Congreso de la UniónEarliest review 2026-09-10 · alert only
mint product6 reviewed sources · United States Mint, Banco de México, The Royal MintEarliest review 2026-10-10 · alert only
editorial guidance8 reviewed sources · U.S. Commodity Futures Trading Commission, Financial Industry Regulatory Authority, CME Group, London Bullion Market Association, U.S. Securities and Exchange Commission EDGAREarliest review 2026-09-10 · alert only

Registry version: source-review-registry-2026-08-12-v3. Reachability checks and human content review remain separate because HTTP 200 alone cannot prove that tax law, product terms or guidance language is unchanged.

Reserve and ETF source reviewWhat can be published and what remains unavailable

RESERVE + ETF ADOPTION GATE

Publicly visible data is not automatically reusable data

GoldObserve reviewed the first-party rights and evidence boundaries for the next reserve and ETF expansion on 2026-08-12. Only ECB/ESCB statistics clear the current commercial-reuse gate for a versioned pilot. IMF coverage needs written permission for systematic commercial use, while World Gold Council and fund-sponsor website data remain link-only unless separately licensed.

ELIGIBLE NOW1

Reviewed baseline exists; a later-vintage revision test remains required.

RESEARCH PILOT1

Useful only inside a narrower, explicitly incomplete scope.

PERMISSION2

No systematic local archive before written rights clearance.

DO NOT INGEST2

Link to the owner; do not scrape or mirror.

CandidateDecision and evidenceNext control
ECB Data Portal / ESCB statisticsMonthly monetary-gold stocks in fine troy ounces for the ECB and euro-area reporting countriesEligible for a versioned pilot. Public ESCB statistics may be reused commercially or non-commercially with source attribution and unmodified metadata; third-party fields remain excluded.This is euro-area coverage, not a global central-bank total. Stock changes can include transactions, reclassifications and revisions.The first reviewed baseline is committed and Git-reconstructable. Keep the public product blocked until a later official vintage proves the revision-diff workflow.Read governing terms
SEC Form N-PORT public data setsPublicly disseminated monthly portfolio filings for covered registered fundsResearch-only pilot. SEC.gov public information may be copied and redistributed with attribution; fair-access limits apply to automated retrieval.N-PORT is not a comprehensive daily physical-gold ETF feed and does not cover every major commodity trust. It cannot support a global daily flow headline.Map covered gold funds and filing fields first; publish no aggregate until coverage and amendment handling reconcile to source filings.Read governing terms
IMF IFS / reserve statisticsCross-country monetary-gold and reserve statistics where reporters provide the relevant seriesCommercial permission required. IMF permits reuse of published statistical data with attribution, but its current terms direct potential commercial reuse and systematic bulk access to written permission.Coverage, reporting lag and historical revisions differ by country; value changes cannot be relabelled as physical purchases.Request written commercial and systematic-download permission before building or redistributing a global local archive.Read governing terms
World Gold Council Gold ETF holdings and flowsGlobal and regional physically backed gold ETF holdings and estimated fund flowsWritten licence required. The website terms limit saving, display and printing to personal non-commercial use and prohibit copying, scraping, distribution and derivative use without prior written authorisation.The series also incorporates Bloomberg, company filings and ICE Benchmark Administration inputs, so public access does not clear all underlying rights.Link to the first-party page only unless WGC grants written commercial redistribution and derivative-display rights.Read governing terms
State Street / SPDR Gold Shares website dataFund-level GLD holdings, NAV and related product informationDo not ingest without a licence. State Street terms restrict copying, dissemination and publication of site or supplier database information beyond personal or authorised internal use.One fund is not a global ETF-flow universe, and a holdings change is not automatically a cash-flow measure.Use an external source link or separately negotiated data licence; do not scrape the product page.Read governing terms
BlackRock / iShares product holdingsFund-level IAU holdings and product informationDo not ingest without approval. BlackRock terms reserve website content and data for personal or internal use and require prior approval for copying or commercial derivative use.A sponsor file is fund-specific and cannot be generalized into global ETF flows without a licensed coverage universe and consistent methodology.Do not mirror or normalize the downloadable holdings file until BlackRock grants explicit reuse rights.Read governing terms

Current release decision: retain the approved ECB baseline and wait for the next official vintage comparison. Do not publish a global central-bank total or a global ETF-flow chart from licensed or incomplete inputs.

Unpublished U.S. gold trade candidateWhy no public chart is shown yet

CANDIDATE SOURCE GATE

U.S. gold trade data is under review—not published

The U.S. Census Bureau International Trade program can support monthly value research from January 2010 for trade recorded under HS heading 7108. That heading covers specified unwrought, semimanufactured, powder and monetary-gold forms; it is not all gold trade and excludes separately classified ore, scrap, jewelry and coin. GoldObserve has approved a local ingestion prototype, not a public chart, dataset or download.

GateEvidence reviewedCurrent decision
Authority and scopeOfficial monthly Census API and merchandise bulk products; HS 7108 classification boundaries documented from official Schedule B and HTS materialPassed for local research only; every future claim must say “recorded under HS 7108”
Metric contractTotal exports, domestic exports and re-exports differ; general imports and imports for consumption answer different questionsRelease blocked until the headline basis is frozen and each comparison remains separately labelled
Row validationFixed-width parsers and synthetic tests preserve value bases and incompatible quantity units; the official bulk-file path is identifiedRelease blocked: current and historical ZIP rows have not completed second-surface reconciliation
Reuse and CSVCensus API terms support search, display and analysis with attribution and a non-endorsement noticeRelease blocked: normalized public CSV redistribution still needs written confirmation or documented legal review
Runtime boundaryCurrent International Trade API queries require a key; bulk ingestion belongs in a controlled maintenance jobPassed boundary: no Census key, fetch or partial snapshot is exposed to browser or PHP page requests

Current result: NO-GO for public release. GoldObserve has not added this candidate to the published Dataset schema. A failed source check keeps the last reviewed state and records the reason; it never turns incomplete rows into a public series.

OBSERVATION ANATOMY

Seven fields are needed before a number can be compared

01Instrument. XAU and XAG are different markets; a coin, futures contract and benchmark auction are different instruments again.

02Currency. USD metal and a reference FX rate may have different source times.

03Weight unit. Precious metals use a 31.1034768-gram troy ounce, not a 28.349523125-gram ordinary ounce.

04Price field. Spot reference, bid, ask, high, low and dealer checkout price are not interchangeable.

05Provider. The successful adapter, not the preferred adapter, must be named.

06Observation time. When the source measured or published the value.

07Receipt and state. When GoldObserve received it and whether the source marked it stale.

NORMALIZATION PIPELINE

How one USD-per-ounce reference becomes a local price

1

Validate the provider responseHTTP success, supported symbol, positive finite value and usable timestamps.

2

Preserve the fine-troy-ounce basisThe common metal basis remains price per fine troy ounce before display conversion.

3

Apply the selected FX referenceLocal ounce price = USD ounce price × USD-to-local-currency rate.

4

Convert the weight unitPer gram = per ounce ÷ 31.1034768; per kilogram = per gram × 1,000.

5

Apply purity only when requestedFine-metal value = gross mass × fineness ÷ 1,000 × fine-metal price.

6

Round for display lastCalculations retain precision before the user-facing currency or unit format is applied.

A converted CNY-per-gram value is therefore not a Shanghai Gold Exchange benchmark. An INR-per-10-gram value is not automatically a jeweller's tax-inclusive retail quote. Conversion changes the denomination; it does not change the instrument.

FRESH / STALE / UNAVAILABLE

Freshness is a contract, not a green decoration

StateMeaningRequired display behavior
FreshProvider returned a current or recently cached real observation under its own state contractShow provider and observation time; never imply execution
StaleProvider is serving a previously observed real value during an upstream problemKeep the old observation time visible and label the value stale
Age warningThe timestamp is older than this page's independent diagnostic thresholdFlag age separately even when the source response called itself fresh
UnavailableNo eligible provider returned a valid value or seriesShow no price, no chart replacement and no fabricated fallback

Markets can close and caches can outlive the request that created them, so elapsed age and provider state answer different questions. The live checker uses a 15-minute spot age warning and a 36-hour history warning as diagnostics only; they are not universal market rules.

HISTORICAL DATA BOUNDARY

Gold, silver and CPI definitions stay separate instead of becoming one artificial chart

Recent daily views use the retained NBU accounting reference; longer research views use the World Bank monthly-average library. GoldObserve does not invent intermediate closes or join these frequencies into one continuous series.

The NBU series converts the official same-date metal value and UAH-per-USD rate into a USD-per-ounce accounting reference. It begins in September 2000. Weekend and holiday dates can repeat a prior value, and the result is not an executable spot close.

The World Bank library supplies monthly-average nominal USD gold, silver, copper and platinum from January 1960 under CC BY 4.0. Gold, silver and platinum use USD per troy ounce; copper uses USD per metric tonne. The silver series preserves the documented pre-1962 and pre-July-1976 definition boundaries and is not renamed as LBMA Silver Price history. GoldObserve does not splice monthly averages and the NBU accounting reference into a false homogeneous series.

The local BLS library stores CPI-U all items, U.S. city average, not seasonally adjusted, from January 1960. It matches World Bank gold only on exact months for purchasing-power research. October 2025 remains missing under the BLS funding-lapse notice; GoldObserve does not interpolate it.

BENCHMARK AND EXECUTION BOUNDARY

Spot reference, auction benchmark and dealer quote answer different questions

PriceUseful forNot automatically included
Indicative spot referenceCurrent metal value, conversions and comparative toolsGuaranteed bid, ask, settlement or physical delivery
LBMA auction benchmarkContracts and valuation that explicitly reference that benchmarkA continuously updating retail price
Futures priceA specified contract, delivery month, venue and timestampSpot ownership without basis, margin and expiry effects
Dealer ask or buyback bidThe exact product, quantity, payment and settlement terms quotedTransferability to another product, time or counterparty
Open macro revision under reviewExact values and version evidence

OPEN REVISION REVIEW

The dollar baseline is held at the last reviewed vintage

The current FRED check found a historical replacement in DTWEXBGS: 2026-08-03 changed from 120.7739 to 119.6951. The revised row is now in the local fallback after explicit review; the Gold Data Brief still requires a separately reviewed edition.

StateLocal fallbackCandidate
DTWEXBGS · 2026-08-03119.6951 · approved baseline119.6951 · accepted revision

Candidate hash: 70e2ea7515f2eb60b3de22675f98f2dd59eaa5e3f992df6dbccbeadea86f5d7f. The source revision was explicitly accepted; brief publication and PHP release still wait for edition-level checks.

Licensing, reuse and primary-source linksFor publishers and technical reviewers

USAGE, LICENSING AND CORRECTIONS

Public access does not erase upstream conditions

Gold API is enabled only within the written permission supplied for GoldObserve's commercial on-page display, transformations, short shared cache, temporary last-valid retention and same-origin browser delivery. That permission is not treated as a right to resell raw rows, operate a public third-party API or publish bulk provider downloads. Frankfurter permits commercial API use while requiring compliance with every underlying provider's terms; GoldObserve requests only the explicitly ECB-pinned path for eight reviewed currencies. Metals.dev remains optional and requires a valid server-side key and applicable account terms. NBU's open-data terms permit free copying, publication, distribution, product integration and commercial use with source attribution; GoldObserve stores the NBU-published accounting value and official FX input, not Bloomberg's underlying feed. The World Bank monthly gold, silver, copper and platinum subsets are normalized under CC BY 4.0; the upstream XLSX itself is not publicly mirrored. LBMA/IBA historical benchmark values are excluded without the required licence. BLS publications are U.S. public-domain material with requested attribution; the local CPI baseline also records its retrieval date, missing month and the required warning that BLS does not vouch for downstream analysis. CFTC states that government information on its website is in the public domain and requests appropriate acknowledgement; the COT baseline retains CFTC attribution and revision evidence. The USGS MCS 2026 companion data release carries a public-domain label; GoldObserve preserves its source version, estimate flags and revision date. GoldObserve does not infer a right to resell or bulk-redistribute any other upstream historical dataset merely because an endpoint can be called.

When a material source or transformation error is verified, the affected output should be suspended or labelled, the cause and affected period documented, the calculation rebuilt from the best available source, and a material correction recorded under the corrections policy. A changed timestamp alone is not evidence that historical values were revised.

Browse all source-backed researchPrices, history, supply, positioning and official reserves

RELATED WORKSPACES

Trace the source into the calculation that uses it

FAQ

Gold price data-source questions

Where does GoldObserve get its displayed gold reference?

Current gold and silver references come from Gold API. If a current response is unavailable, some daily views can show a clearly dated National Bank of Ukraine accounting reference as the last verified value.

Is the displayed gold price an LBMA benchmark?

No unless a page explicitly identifies an LBMA auction value. The standard live endpoint is an indicative spot reference. LBMA Gold Price is a separately timed auction benchmark with its own methodology and permitted uses.

How often can the live price update?

The site normally checks for a new source value after about 30 seconds. The provider may publish at a different cadence, so the observation time beside the price is more important than the countdown.

What happens when every provider fails?

The price is marked unavailable, or the last verified value remains visible with its original date and an older-data label. GoldObserve does not manufacture a replacement quote.

Why can a local-currency gold price differ from another site?

The underlying metal observation, FX source, timestamps, unit conversion, rounding and cache timing may differ. GoldObserve converts a USD-per-troy-ounce reference with a Frankfurter rate when the selected provider does not supply the requested currency directly.

How much metals history is available?

Recent gold and silver daily accounting references come from the retained NBU series. Longer monthly-average gold, silver, copper and platinum history comes from the World Bank Pink Sheet from January 1960. The site identifies the frequency and never joins the two into one artificial series.

Does chart sampling invent missing prices?

No. Long series may be thinned for display by retaining real returned observations at a regular stride plus the final observation. Calculations that request daily detail operate on returned daily observations; the system does not interpolate fictional closes.

What happens when a macro source is temporarily unavailable?

The page may show the last verified local series with its date and an older-data label. It is never presented as a new source update.

Can I use a GoldObserve price to settle a transaction?

No. Values are educational, indicative references and may be delayed. Settlement, accounting, contracts and physical trades require an appropriate licensed benchmark or a same-time executable quote from the relevant counterparty.