THE QUESTION THIS TOOL ANSWERS
Did the stored macro history change, or did the source simply publish another date?
GoldObserve retained five reviewed baselines and compared four adjacent transitions by exact date key. The current ledger found 59 added observations, 33 changed common-date values and 0 removed observations. Those numbers describe only this retained window.
The distinction matters because a latest-revised series can make historical research look cleaner than the information actually available at the time. This page exposes the evidence boundary before users interpret gold correlations, regimes or macro-driver charts.
POINT-IN-TIME RESEARCH GATE
ALFRED candidates checked, not promoted
We checked 10 official ALFRED candidate snapshots across 2 macro series. The candidate rows passed structural checks, but they remain a review queue rather than a public point-in-time research baseline.
Boundary. The queue organizes official ALFRED candidate evidence for human review only. It cannot approve a forecast baseline, publish a forecast or replace the retained latest-vintage macro ledger. Open ALFRED to inspect the upstream series archive.
RETAINED MACRO VERSION EVIDENCE
What changed between GoldObserve's reviewed macro snapshots?
This ledger compares the exact date keys and values in five retained baselines. It turns an abstract revision warning into auditable additions, revisions and deletions while keeping the evidence boundary visible.
| Series | d39057a → 0adc186Aug 8, 2026, 12:04 PM | 0adc186 → 13b0612Aug 11, 2026, 12:38 AM | 13b0612 → f4e1d53Aug 11, 2026, 9:09 PM | f4e1d53 → c417925Aug 15, 2026, 12:50 AM | c417925 → e148f42Aug 25, 2026, 10:53 AM | e148f42 → bbcc4eeAug 27, 2026, 4:41 AM | bbcc4ee → c7c413eAug 27, 2026, 5:23 AM | c7c413e → ab896bfAug 28, 2026, 2:49 AM |
|---|---|---|---|---|---|---|---|---|
| DCOILWTICOWTI crude oil | ||||||||
| DFII1010-year real yield | ||||||||
| DTWEXBGSBroad U.S. dollar | ||||||||
| CPIAUCSLU.S. CPI | ||||||||
| M2SLU.S. M2 | ||||||||
| WALCLFederal Reserve assets |
d3614f3370b4Added dates: 2026-08-26
Revised dates: None
Deleted dates: None
Evidence boundary. This ledger records what GoldObserve retained, not every FRED or ALFRED vintage. Snapshot generation time is not an official publication time. Zero observed revisions across this short window does not prove that a series is never revised. ALFRED ingestion still requires a controlled offline job and a secret key; neither PHP nor browser code performs it.
READING ORDER
Read coverage before drawing a macro conclusion
Each card identifies the Git commit, generation time, total observations and snapshot hash.
WTI, real yields, the dollar, CPI, M2 and Federal Reserve assets retain separate clocks and hashes.
A new date extends coverage. Only a changed value on a common date is a revision.
This Git ledger is not a complete FRED or ALFRED archive and cannot support a look-ahead-safe backtest.
SOURCE CLOCKS
Six series do not arrive on one shared schedule
| Series | Measure | Typical observation clock | Primary evidence |
|---|---|---|---|
| DCOILWTICO | WTI crude oil spot price | Business day | EIA series distributed by FRED |
| DFII10 | 10-year TIPS real yield | Business day | Federal Reserve Board via FRED |
| DTWEXBGS | Broad U.S. dollar index | Business day | Federal Reserve Board via FRED |
| CPIAUCSL | Consumer Price Index | Monthly observation period | BLS series distributed by FRED |
| M2SL | M2 money stock | Monthly observation period | Federal Reserve Board via FRED |
| WALCL | Federal Reserve total assets | Weekly Wednesday level | Federal Reserve H.4.1 via FRED |
RESEARCH BOUNDARY
A version ledger improves reproducibility; it does not create point-in-time history
- Snapshot generation time is GoldObserve's collection time, not an official release timestamp.
- Five retained versions cannot establish how often a source revises over decades.
- Zero observed revised or deleted rows in this window is not evidence that revisions never occur.
- Gold and macro dates can represent different market closes, observation periods and release schedules.
- The ledger audits inputs; it does not estimate causation, fair value or future gold prices.
CONTINUE THE EVIDENCE CHAIN
Use vintages, alignment and market interpretation for different questions
MACRO VINTAGE FAQ
Questions about retained versions
What is a macro data vintage?
A vintage is one retained state of a dataset at a specific collection time. This explorer compares the five reviewed macro baselines GoldObserve retained; it does not claim to reproduce every state ever published upstream.
Does this explorer contain ALFRED real-time vintages?
No. The ledger is an exact comparison of GoldObserve's retained latest-vintage baselines. A historical point-in-time study requires ALFRED or another release-time archive collected under a controlled offline process.
Why are additions different from revisions?
An addition is a date that did not exist in the prior retained snapshot. A revision is a changed value for a date present in both snapshots. Treating every new observation as a revision would misstate the evidence.
Do zero observed revisions prove that the series never revise?
No. The retained window is short and incomplete. Zero observed revisions only describes these four adjacent comparisons.
Can macro revision evidence predict gold prices?
No. The ledger audits data lineage and research reproducibility. It is not a model of gold returns, a trading signal or a forecast.