Estate catalog

Data catalog

What FinObservatory is built on. Each entry names the publisher, states the licence, shows the coverage actually held, and links to the source.

FinObservatory does not redistribute licence-encumbered data. Most of what makes this platform interesting is licence-encumbered (the crisis atlas, the long-run rate histories, the systemic-risk series), and those datasets are catalogued, cited, and linked to the publisher, never republished: you get the publisher’s current vintage rather than a stale copy of it, on the publisher’s own terms. The open, redistribution-permitted spine is different. It is served in full, with its licence attached to every response, through the read-only data API below.

Coverage below is queried from the live data when this page is built, not typed. Full lineage, including every transformation, is in the methodology.

Estate freshness

The latest observation held for the fast-moving layers, queried from the live data when this page was built. Daily-tier series refresh via the documented pipeline (scripts/refresh.sh); crisis chronologies are reviewed vintage merges, never auto-refreshed. The last column states the publisher’s own release cadence and links to the schedule page it was verified against.

Sources

API

A public, read-only data API in the style of the OFR's data.financialresearch.gov: no registration, no key, JSON or CSV. It serves only the 10 redistributable datasets above; nothing on the never-republished list below is served.

• GET /api/v1 index of every dataset: id, title, source, licence, row count, URLs, docs.
• GET /api/v1/{id} the dataset as a JSON envelope with its licence and citation attached.
• GET /api/v1/{id}?format=csv the full dataset as CSV, with the same source-notice comment header the exports carried.

Responses carry Cache-Control: public, s-maxage=3600. JSON responses cap at the 50,000 most recent rows (truncated: true with total_rows when the cap binds); CSV always carries the full dataset. An unknown id returns 404 with the list of available ids.

Envelope shape, from the live imf_cofer_reserve_shares endpoint (rows cut to one here; the API returns all 948):

{
  "dataset": "imf_cofer_reserve_shares",
  "title": "IMF COFER reserve composition",
  "source": "International Monetary Fund, COFER",
  "license": "IMF data terms. Cite the IMF; check the terms before any onward use.",
  "citation": "Source: IMF, Currency Composition of Official Foreign Exchange Reserves (COFER).",
  "data_through": "2026-Q1",
  "retrieved": "(ISO timestamp when served)",
  "row_count": 948,
  "rows": [
    {
      "currency": "USD",
      "quarter": "2026-Q1",
      "value": 57.130786895752,
      "measure": "share_pct"
    }
  ]
}

Held and used, but never republished

These are the datasets FinObservatory analyses but will not pass on, each with the reason. The list is longer than the one above, and it holds most of what the platform is actually for. It is here because a catalog showing only the redistributable slice would misrepresent what the platform runs on.

• Global Macro Database (GMD) crisis and macro panel: Academic and non-profit research licence only; redistribution not permitted.
• Jorda-Schularick-Taylor Macrohistory Database (JST): Free academic access, but commercial redistribution is forbidden.
• Combined crisis atlas (crises_events, crises_panel): Inherits the most restrictive input (GMD, JST): non-commercial research only.
• Laeven-Valencia Systemic Banking Crises Database: IMF-copyright academic dataset with no redistribution licence; take it from the working paper.
• Reinhart-Rogoff global crises data: Research use with citation only; no redistribution licence.
• Baron-Verner-Xiong banking-crisis panel: Academic research dataset; no redistribution licence.
• Cruces-Trebesch sovereign-debt haircuts: Research-use dataset; redistribution not permitted.
• Schmelzing eight-centuries real interest rates: Bank of England working-paper data; terms unclear, so excluded pending confirmation.
• Metrick-Schmelzing banking-intervention chronology: Research chronology; redistribution not permitted.
• Chinn-Ito capital-account openness (kaopen): Non-commercial use only.
• Systemic-risk series (SRISK, CoVaR, MES, absorption): Computed from Yahoo Finance equity data, whose terms restrict redistribution to personal and non-commercial use. Displayed as aggregates only.
• Equity returns (Yahoo Finance): Terms of service restrict redistribution to personal and non-commercial use.
• Shiller CAPE and home-price data: No explicit open-data licence.
• Fama-French factors and industry portfolios: Redistribution terms unclear; excluded pending confirmation.
• External Wealth of Nations (Lane, Milesi-Ferretti): Academic research dataset; no redistribution licence.
• Krippner shadow short rate: Research dataset with no redistribution licence.
• Damodaran valuation and cost-of-capital tables: No explicit redistribution licence; excluded conservatively.
• IMF Article IV and GFSR reports (full text): The IMF licence forbids redistribution. The library cites and links them; it never quotes them.