FinObservatory

Banking, currency, and sovereign-debt crises from five primary chronologies, agreement-scored, with per-country history and episode pages.

Government debt-to-GDP trajectories, the level at which past crises began, and where each country sits against that benchmark.

Who developing countries owe and on what terms, from World Bank International Debt Statistics: stocks, creditor composition, and the repayment schedule. Not world external debt, and the page says why.

What creditors actually lose when a government defaults: sovereign restructurings since 1823, measured by present-value haircut and by face-value reduction, which are different things and are never averaged.

What borrowers owe China: the estimated stock by instrument and drawn PBoC swap lines measured against gross reserves. The creditor-side estimate is never netted against borrower-reported debt.

Net international investment positions and the equity-versus-debt composition of external liabilities, set against each country's crisis history.

A transparent composite health score for every FDIC-insured US bank, with peer percentiles and a published methodology.

The FR Y-9C consolidated layer above the insured bank: assets, capital and concentration for US bank holding companies, and the pre-2006 double count that makes the raw panel unsummable.

Where US bank deposits sit: the FDIC Summary of Deposits by state, county, institution and branch, from 1994.

Three decades of consolidation in the US credit-union system from NCUA 5300 call reports: counts, assets, members and loans.

LRMES, MES, Delta-CoVaR, and the absorption ratio for the largest US banks, computed from published academic methods with full provenance.

A financial-conditions index, the Basel III credit-to-GDP gap, the yield curve, the Shiller CAPE, and official stress indices in one composed view.

The HMDA loan-level register: the part of the Black/White mortgage denial gap that four underwriting controls do not explain, stated with the controls the public file lacks, and the inversion of the refinance market.

US mortgage applications, originations and denial rates, yearly from 2007, from the full HMDA register: the housing bust, the refinancing wave, and the rate-shock collapse.

As-filed XBRL from SEC Financial Statement Data Sets: the held-to-maturity securities gap that call reports do not carry, measured filer by filer, and the extension tagging that hides it.

The US Treasury market from the primary sources: the full daily par yield curve, the 10-year yield back to 1990, and every 2Y-10Y inversion episode.

Did the Fama-French factors pay, and when did they stop? Full-sample and recent compound returns, volatility and t-statistics for the factors and 49 industries, from Ken French's own library.

What equities, housing, bonds and bills actually returned across 16 advanced economies since 1870, in real terms, with the risk beside the return and what each asset did in banking crises.

CAPE against the ten years of real returns that followed it, with the overlapping-window problem that inflates the apparent evidence, plus what Damodaran's cost-of-capital tables actually contain.

Real house prices around banking crises: an event study on the BIS residential property price index and the Laeven-Valencia crisis list.

The currency composition of the world's allocated FX reserves, quarterly since 1999, from IMF COFER, with the US dollar-share trajectory told by the data alone.

Which banking systems lend to the world: BIS locational statistics read from the creditor side, one page per reporting system, with the largest bilateral exposures.

The borrower side of the same BIS table: who funds a given country, the largest counterparty countries, and claims on emerging markets.

Which economies run on money sent home by their migrants: receipts by level and share of GDP, what removing them does to the current account, and the modelled bilateral corridor matrix.

Account ownership across the Global Findex rounds: who has an account, the gender gap that did not close, the income gap that did, and the digital-payment divide.

Official AML/CFT reference data: the two FATF public lists of monitored jurisdictions and aggregate OFAC SDN counts, reproduced verbatim. Reference only, no risk scoring.

A system-level view of Bangladesh's banking sector from Bangladesh Bank published data: soundness-indicator trends and a bank-category breakdown. Aggregates only, no bank named.

Classic results in financial economics re-run on the estate: the published specification, the honest sample differences, and the result as it comes out. A failed replication is a headline, not a footnote.

A citation-first AI analyst over the whole estate. Ask a question in plain language; every number in the answer comes from a query against a primary source, with citations. No advice, no predictions.

Dated, cited analysis notes built on FinObservatory's own analytics. Every number in a brief is queried from the data estate at build time, not typed.

A metadata explorer over the acquired central-banking text corpora: BIS speeches, Federal Reserve materials, IMF reports and SEC filings.

One canonical page per country: crisis history, financial depth, the credit cycle, and sovereign debt, each traced to its source.

What FinObservatory is built on: every dataset, its publisher, its licence, its coverage, and a link to the source. FinObservatory does not redistribute data.

How every figure is computed, what was derived rather than read, and where the data cannot support a claim.