Bank capital is the loss-absorbing equity cushion that stands between a bank's depositors/creditors and insolvency. Capital ratios express that cushion as a percentage of the bank's assets — usually risk-weighted assets, so riskier exposures consume more capital. Regulation (the Basel framework, implemented in the U.S. by the Fed/OCC/FDIC) sets minimum ratios plus stacked buffers that banks must clear before they can freely pay dividends or buy back stock. For a bank-stock analyst, capital ratios are the single most important balance-sheet number: they cap how much capital can be returned to shareholders, gate growth, and signal resilience in a downturn. The core tension is that capital is safety for the system but a drag on return-on-equity for shareholders — more capital means a more durable bank but a lower ROE, which is the central battleground of every capital-rule debate.

How it's calculated / formed

The numerator is regulatory capital, layered by quality:

• Common Equity Tier 1 (CET1): common shares, retained earnings, and disclosed reserves — the highest-quality, first-loss capital.
• Additional Tier 1 (AT1): CET1 plus qualifying perpetual preferred stock / contingent-convertible instruments. CET1 + AT1 = Tier 1.
• Tier 2: subordinated debt and certain reserves. Tier 1 + Tier 2 = Total Capital.

The denominator for the headline ratios is Risk-Weighted Assets (RWA) — each exposure multiplied by a risk weight (e.g., cash ~0%, residential mortgages and high-grade exposures lower weights, unsecured corporate/commercial higher). A bank can hold the same dollar capital but report different ratios depending on its RWA mix.

Basel III minimums (per BIS / Fed Regulation Q): CET1 ≥ 4.5%, Tier 1 ≥ 6%, Total Capital ≥ 8% of RWA. On top of the minimums sit buffers, all met with CET1:

• Capital conservation buffer: 2.5%.
• Countercyclical buffer (CCyB): 0–2.5%, discretionary; the U.S. CCyB has been set at 0% since inception (per Fed).
• G-SIB surcharge: 1%–3.5% under the Basel global method; the Fed's "Method 2" can produce higher figures for U.S. global systemically important banks (per Federal Reserve).
• Stress Capital Buffer (SCB): for U.S. banks ≥$100B in assets, the conservation buffer is replaced by a firm-specific SCB derived from the annual CCAR/supervisory stress test (minimum 2.5%, no cap), introduced 2020.

A separate leverage ratio ignores risk weights entirely as a backstop: the U.S. Tier-1 leverage ratio (Tier 1 / total average assets) and the Supplementary Leverage Ratio (SLR) (Tier 1 / total leverage exposure, which adds off-balance-sheet items). Minimum SLR is 3% for all advanced-approaches banks. Until recently U.S. G-SIB holding companies faced a fixed enhanced SLR (eSLR) of 5% (a 2% buffer over the 3% minimum), with 6% required at insured depository subsidiaries to be "well capitalized." A December 2025 final rule (effective April 1, 2026, per the Fed/OCC/FDIC) recalibrated the eSLR: the holding-company buffer is now 50% of the firm's Method-1 G-SIB surcharge (replacing the flat 2% add-on), and the IDI buffer equals 50% of the parent's Method-1 surcharge capped at 1% over the 3% minimum.

How it's used in practice

Analysts watch the CET1 ratio against the bank's own "required" CET1 (4.5% + SCB + G-SIB surcharge + any CCyB). The gap between actual and required CET1 is the bank's excess capital — the war chest available for buybacks, dividends, and acquisitions. A bank running, say, 13.5% CET1 against an 11.0% requirement has roughly 250 bps of cushion; one running near its requirement has little room and may face buyback restrictions.

The binding constraint matters: a bank may be capital-rich on risk-weighted ratios but constrained by the leverage ratio (common for custody banks with large low-risk balance sheets), or vice-versa. Knowing which ratio binds tells you what actually limits the bank's growth or payouts. CCAR/SCB results each summer are major catalysts for U.S. bank stocks — they directly set the next year's dividend/buyback capacity and frequently move share prices on the announcement.

Adoption, debate & evidence

Basel III is the near-universal global standard, with national variation in implementation. The framework is genuinely contested at the margins. The "Basel III Endgame" (finalizing Basel's 2017 RWA reforms) saw a U.S. 2023 proposal that would have raised aggregate large-bank capital materially; after heavy industry pushback it was withdrawn and re-proposed in March 2026, this time reducing requirements versus the 2023 version — Fed staff estimated CET1 requirements falling ~4.8% for Category I/II banks, ~5.2% for Category III/IV, and ~7.8% for smaller banks (per Fed/ABA reporting), though still above the 2019 baseline. The Fed board voted 6–1 to advance the proposals, with Governor Michael Barr the sole dissenter, calling them "unnecessary and unwise" and warning they contain numerous downward deviations from the Basel standard (per Fed reporting). Separately, the eSLR was already reformed: a December 2025 final rule (effective April 1, 2026) tied the leverage buffer to each G-SIB's surcharge so the leverage backstop binds less on low-risk activity such as Treasuries and central-bank reserves.

On the empirical question of whether higher capital prevents crises, evidence is mixed and honestly debated:

• Expert-survey and World Bank work suggest higher capital improves individual bank survival, lending stability in downturns, and bank value; one survey of ~149 academic researchers (Ambrocio et al., J. Financial Stability 2020) found a median preferred non-risk-weighted equity/assets ratio of ~10% (mean ~15%) — well above current requirements — with the typical expert expecting minimal credit-supply cost.
• A prominent long-run study (advanced economies, 1870–2013) found that higher capital ratios did not reliably prevent systemic crises — crises were better predicted by credit growth and funding/liquidity than by the capital ratio, partly reflecting reverse causality (risk drives capital demand). Treat "more capital = no crisis" as folklore; "more capital = better survival and faster recovery" is the better-supported claim.

Strengths & limitations

Capital ratios work best as a relative and trend signal: a bank steadily building CET1 above its requirement, with a comfortable SCB, is signaling resilience and payout capacity. They are forward-looking only insofar as the stress test is — and the stress scenario is set by regulators, so it can miss the next shock.

Key limitations: (1) RWA is partly model-driven — banks using internal models can optimize risk weights, so two banks at the same CET1 ratio aren't equally safe; the leverage ratio exists precisely because RWA can be gamed. (2) Ratios are point-in-time and lagging; rapid losses (e.g., a securities-portfolio mark or deposit run) can erode capital faster than quarterly reporting shows — Silicon Valley Bank (2023) was nominally well-capitalized shortly before failing, because unrealized AFS/HTM losses and a deposit run, not the headline ratio, were the problem. (3) Liquidity and funding risk are governed by separate rules (LCR/NSFR) — capital adequacy does not equal liquidity adequacy. The #1 misuse is treating a passing capital ratio as proof a bank is safe; capital is necessary, not sufficient.

Sources

• Bank for International Settlements — Basel III capital and buffer summaries (bis.org/fsi/fsisummaries); G-SIB / CCyB framework (bis.org/bcbs/ccyb_gsib).
• Federal Reserve — Stress tests & SCB rule (2020 SCB final rule, Reg Q/Y/YY, effective Oct 2020); CCyB affirmed at 0% (press releases since 2016); March 19, 2026 Basel III Endgame / G-SIB surcharge re-proposal materials (federalreserve.gov/newsevents/pressreleases/bcreg20260319a.htm) and the 6–1 board vote with Barr's dissent.
• Federal Register / OCC / FDIC — eSLR final rule (Dec 1, 2025; effective April 1, 2026) recalibrating the G-SIB leverage buffer to 50% of the Method-1 surcharge.
• Office of Financial Research — U.S. G-SIB surcharge monitor (Method 1 vs higher Method 2).
• Congressional Research Service — "Bank Capital Requirements: A Primer and Policy Issues" (R47447); leverage-ratio brief (IF10205).
• ABA Banking Journal / Freshfields / Sullivan & Cromwell / Mayer Brown — coverage of the March 2026 Basel III Endgame re-proposals and the eSLR reform.
• Empirical/landscape: World Bank "Higher bank capital contributes to financial stability"; expert-survey study on optimal capital (ScienceDirect, J. Financial Stability); long-run crisis study (FRBSF WP 2017-06, Jordà-Richter-Schularick-Taylor) — flagged as the dissenting "capital ≠ crisis prevention" view.

Disputes flagged: (1) whether higher capital prevents systemic crises — mixed evidence, genuine academic dispute; (2) direction/size of the Basel III Endgame's capital impact — the rule is unsettled and the 2026 re-proposal materially changed the 2023 numbers.