Bank capital to assets ratio: where the comparable numbers live

The query sounds singular and lands as three different grains - country-year, institution-quarter and consolidated banking group - and the Datadory catalog names a record for each.

World Bank Indicators - Bank Capital to Assets Ratio is the cross-country anchor. Indicator code FB.BNK.CAPA.ZS carries the ratio in percent for every reporting economy - roughly 265 rows including regional and income aggregates, observed annually from 1960. The figures originate in the IMF Financial Soundness Indicators compilation, republished within World Development Indicators, and about 30 sibling banking indicators share the identical layout, so nonperforming-loan and private-credit measures append to the same extract without changing its shape. Quality score 8 out of 10 against a catalog-wide average of 7.81.

FDIC BankFind Suite API owns the United States institution level - quality 10, tied-highest in the slice. Quarterly Call Report-derived financials cover roughly 4,700 active insured charters, with total equity (field EQ) sitting beside total assets (field ASSET) so a capital ratio gets computed per bank per quarter rather than averaged out of someone else's rounding. Histories reach toward the 1930s.

The FDIC Quarterly Banking Profile supplies the aggregate control: whole-system capital subtotaled by asset-size group and community-bank status across more than 160 quarters reaching back to Q1 1986.

ECB Statistics - Consolidated Banking Data (CBD2) covers Europe at group level - quality 9: 66,139 series (40,227 annual, 25,912 quarterly) across 30 reference areas, with capital among its documented topics alongside profitability, funding and leverage.

And the World Bank Global Financial Development Database holds the risk-weighted counterpart - quality 9, 108 indicators across 214 economies, regulatory capital to risk-weighted assets included in its stability family.

Five records, one keyword, no substitutes for one another. The rest of this page separates them.

One well-defined figure per economy per year: Tier 1 capital divided by total assets of deposit takers, in percent. Tier 1 - core - capital takes in common equity, disclosed reserves, retained earnings and certain other instruments meeting Basel requirements, which is why the ratio reads as a system-level loss-absorbing buffer rather than an accounting artifact. The record tags it 'capital-adequacy', 'tier-1-capital' and 'imf-fsi', and the definition rides along in the series metadata rather than living in someone's codebook.

Five verified United States observations, exactly as they land:

countryiso3code | country       | date | value
USA             | United States | 2019 | 9.39570266980086
USA             | United States | 2020 | 8.6063941052102
USA             | United States | 2021 | 8.62368627738746
USA             | United States | 2022 | 8.57788925795034
USA             | United States | 2023 | 8.68252186405856

Read as a story: US banking-system capitalization fell roughly 80 basis points between 2019 and 2020 as pandemic-era balance-sheet growth outran capital accumulation, then held flat near 8.6 percent for three straight years. Values arrive unrounded, exactly as published - no smoothing, no interpolation across missing years, and genuinely blank where an economy did not report.

Provenance travels with the file. Companion metadata records the compiling organization - Financial Soundness Indicators, International Monetary Fund - and profiles each economy individually, so every figure shipped downstream cites an official chain rather than a vendor estimate. Field definitions were verified against delivered files before this record scored 8/10, and they hold as of August 2026.

Because the denominator changes definition between altitudes. FB.BNK.CAPA.ZS divides Tier 1 capital by total assets - an unweighted, leverage-style measure. Most headline regulatory ratios divide capital by risk-weighted assets, so a banking system heavy in low-risk government securities reads stronger on the risk-weighted basis than on this one. Both numbers describe the same banks; neither is wrong.

The World Bank keeps the two concepts deliberately apart inside its own catalog: in the Global Financial Development Database's stability family, bank Z-score, nonperforming loans to gross loans and regulatory capital to risk-weighted assets sit side by side as distinct indicators across 214 economies. Collapsing them into one 'capital' column is the single most common analytical error in cross-country banking work.

National compilation practices differ as well - what counts as eligible capital varies by jurisdiction - so the annual country values function as a comparability screen, not a supervisory comparison. When a single bank's reported CET1 ratio disagrees with its country figure, both can be right.

The records stop being a list and become a model once each takes a layer of the stack:

Read down the join-key column and the workflow appears on its own: a screen starts at the country-year ratio, drops to the individual institutions behind any outlier market on regulator identifiers, checks whether the drift is industry-wide on the aggregate profile, and extends to European groups on the consolidated view where the mandate crosses borders.

The layers reconcile when you compute instead of average. JPMorgan Chase's own filed figures for the quarter ended March 2026 carry total equity of $335.9 billion against total assets of $4.02 trillion - about 8.4 percent equity-to-assets, sitting near the 8.68 percent national system reading for 2023 yet measured on one charter's balance sheet. Computing EQ over ASSET per institution per quarter keeps the dispersion - community banks and globally systemic banks do not share a balance-sheet shape - that averaging published country ratios destroys.

Datadory resolves the identifier mappings up front, so those four hops arrive pre-joined in one delivery rather than as four exports reconciled by hand.

Six jobs account for most of the demand, and each maps onto a named record:

1. Cross-country risk features. A capital-adequacy variable keyed identically across ~265 economies gives models an exogenous systemic-strength input no single-market source can supply; the ~30 sibling indicators extend the panel without changing its shape.
2. Prudential benchmarking before market entry. Set a target market's capitalization against regional and income aggregates before committing capital or pricing country risk - one citable series beats six central-bank PDFs stitched by hand.
3. Macro-stress narratives. Sixty-plus years of annual observations turn "is this banking system fragile?" into a chart with dates on it: Bretton Woods' collapse, Japan's bubble, the global financial crisis and the post-pandemic balance-sheet expansion all sit inside one continuous line.
4. Distribution, not averages. Equity-to-assets recomputed per FDIC-insured charter per quarter shows the spread behind any national number - which is what a stress test actually stresses.
5. Profitability context. ECB Consolidated Banking Data shows whether euro-area groups earn their way to stronger capital - the verified return-on-equity run for domestic banking groups reads 2.37 percent in 2024-Q1 (flagged estimated), 4.89 in Q2 and 7.33 in Q3.
6. Crisis-cohort construction. Pairing the country-year panel with the FDIC's 4,115-record failure register gives distress episodes a systemic control: did capital erosion precede collapse, or follow it?

Every one of these starts the same way: tell Datadory the economies, years, charters or quarters and get a sample cut from the same normalized feed production would use.

The table lines up the options as of August 2026. They differ by grain, history and the job each answers - not by which one is 'best':

Capital means little in isolation, and each companion answers a different question:

• Asset quality. Bank nonperforming loans to gross loans shares FB.BNK.CAPA.ZS's layout, so one extra field adds the stress dimension to the same country-year panel - the nonperforming loans entry defines the terms.
• The risk-weighted twin. The Global Financial Development Database carries regulatory capital to risk-weighted assets beside the leverage-style measure, so both denominators travel in one delivery. Its own caveat applies: some indicator values run null for recent years because underlying compilations lag, so filter on observation availability before modelling.
• US micro-foundations. FDIC quarterly financials reconstruct distribution across ~4,700 institutions - the dispersion behind any national average.
• Border context. BIS Locational Banking Statistics contribute roughly 609,000 cross-border series from 1977-Q4, so a capital drift that begins in wholesale funding has somewhere visible to point.
• Vocabulary. The capital adequacy ratio entry unpacks Tier 1, risk-weighted assets and the leverage measure without the jargon pileup.

Name the companions you want when requesting a sample and they arrive joined, not bolted on afterward.

Files, feeds, or your warehouse. Daily, weekly, or hourly - your call.

This page is one workflow inside the diversified-banks stack. Start with the full spec on the World Bank Indicators — Bank Capital to Assets Ratio dataset page, then see how it ranks among the slice's primaries in the diversified banks data guide or the scored best diversified-banks datasets list. The head-to-head argument continues on the FDIC Bank Data Guide vs World Bank capital-to-assets comparison.

For the persona view, open diversified banks data for investors and quants or browse the whole shelf on the diversified-banks data hub. When you want rows instead of reading, request a sample on any dataset page - the field dictionary travels with it.