Global Banking Sector Report 2026

Executive summary

Decision question: what do verified 2025 results and 2026 year-to-date evidence imply for bank earnings, balance-sheet resilience and the next twelve months? The report combines international structural and prudential datasets with official national aggregates while keeping institutional perimeter, date, currency and definition differences visible. Covered banking systems remained profitable and strongly capitalised in the available official samples. The broadest structural anchor is the FSB end-2024 sample, not a 2025 world total. BIS cross-border bank credit reached USD 38.1 trillion at end-2025 and grew 11%, with expansion toward emerging Europe, Africa and the Middle East, and Latin America, but contraction toward China. Matched IMF reporters showed median nominal local-currency growth in assets, loans and deposits; the panel is diagnostic, excludes several major systems and cannot be interpreted as global growth. Profit drivers shifted toward a more heterogeneous mix of deposit repricing, volumes, fees and costs. Verified asset-quality deterioration was concentrated in selected refinancing-sensitive or portfolio-specific areas rather than broad-based. The base case for 29 July 2026 to 28 July 2027 is a gradual earnings transition rather than a broad solvency event. The main downside path is a joint shock involving commercial-real-estate refinancing, concentrated or uninsured deposits, wholesale and cross-currency funding, sovereign-bank links and shared technology providers. No numerical scenario probabilities are assigned. Decision-makers should monitor deposit beta and outflows, wholesale funding spreads, Stage 2 and NPL migration, CRE refinancing gaps, capital headroom and critical-service recovery failures.

Key findings

1. 01

   Covered systems remained profitable and well capitalised, but a defensible same-date global banking balance sheet cannot be constructed; the FSB end-2024 sample is a structural anchor, not a 2025 world total.

2. 02

   Cross-border bank credit reached USD 38.1 trillion at end-2025 and grew 11% on the BIS adjusted measure; recipient-region growth ranged from +26% in emerging Europe to -15% toward China.

3. 03

   Among 61-63 matched IMF reporters, 2025 median nominal local-currency growth was 8.7% for assets, 9.3% for gross loans and 8.5% for deposits; this is a diagnostic panel, not a global aggregate.

4. 04

   Profitability remained positive, but performance diverged by revenue mix, deposit repricing, loan volumes, fees and costs rather than following one regional rule.

5. 05

   Aggregate asset quality remained broadly contained in verified samples, while risk concentrated in CRE refinancing, selected consumer or corporate portfolios and several anchor jurisdictions.

6. 06

   The twelve-month base case is a gradual earnings transition; the main downside risk is a combined CRE, deposit/funding, sovereign and shared-provider shock, with no scenario probabilities assigned.

Selected banking-system asset growth, latest Q1/March 2026 (% year on year, nominal local currency).

Source: Sources: ECB Banking Supervision, NFRA, APRA and CBUAE. Monitoring comparison only; supervisory perimeters differ, no FX or inflation adjustment is applied, and the euro-area sample changed from 113 to 109 institutions.

Executive takeaways

• The covered banking systems remained profitable and well capitalised, but no common-date global balance sheet can be defensibly constructed. The broadest structural anchor is the Financial Stability Board (FSB) sample: banks held $191.3 trillion of assets at end-2024, 38.0% of financial-system assets in the covered 21-jurisdiction-plus-euro-area group. This is not a 2025 world total. In the Basel Committee on Banking Supervision (BCBS) June 2025 sample, Group 1 banks reported a 13.9% CET1 ratio, 135.0% LCR and 123.8% NSFR. FSB, pp. 4 and 6-7; BCBS, Table 1. [CL-GLOB-01, CL-BCBS-01, CL-COVER-01] • International activity accelerated in 2025. Cross-border bank credit reached $38.1 trillion and grew 11% on the BIS exchange-rate- and break-adjusted measure. Growth was +26% to emerging Europe, +16% to Africa and the Middle East and +12% to Latin America and the Caribbean, but -15% to China. These are cross-border positions, not domestic banking-system credit. BIS, Graphs 1-6. [CL-XB-01, CL-XB-02] • Balance-sheet expansion continued in most matched reporting systems, without supporting a global growth rate. Among 61-63 economies with paired IMF annual observations, the equal-country-weighted median nominal local-currency growth in 2025 was 8.7% for assets, 9.3% for gross loans and 8.5% for customer deposits. The panel excludes the United States, China, Japan, the United Kingdom and several large euro-area systems. IMF FSIBSIS. [CL-IMFACT-01] • Profitability diverged more by revenue mix and repricing position than by a single regional rule. In the BCBS sample, rolling twelve-month after-tax profit through June 2025 rose 5.5% in Europe, 15.7% in the Americas and 10.8% in the rest of the world. In Q1 2026, euroarea significant institutions reported net profit 5.0% above Q1 2025, while US large-bank ROE

Scope, coverage and comparability

This report covers deposit banks, commercial and corporate banks, retail and universal banks, systemic investment banks and regulated credit institutions. Central banks, insurers, stand-alone asset managers and non-bank financial intermediaries are excluded from sector totals; they appear only where their interaction with banks affects funding, credit or resilience. The measurement frame is deliberately split. “2025 actuals” means completed calendar-year 2025 observations, compared with 2024 where comparable. “2026 YTD” means data or events available by 28 July 2026, 16:08:31 UTC+3 and is compared with the same 2025 period where possible. Fiscal-year measures are labelled. Stocks use period-end observations; income and expense flows use period totals or published rolling-year measures. Actual numerical coverage includes the United States, Canada, the euro area, the United Kingdom, Switzerland, Hungary, China, Japan, Australia, Indonesia, Brazil, the United Arab Emirates and South Africa, plus matched IMF country panels and BIS regional cross-border groupings. Hungary, Australia/Indonesia, Brazil, the UAE and South Africa are anchor jurisdictions rather than regional estimates. Equivalent aggregate primary-source tables were not verified for South Korea or India by the cutoff; those cells are marked Insufficient evidence rather than inferred from market commentary. Measured global bank-asset coverage is Not calculable. The FSB sample represents more than 90% of world GDP, but that is not a bank-asset coverage ratio. BIS international banking statistics measure cross-border and international positions, not total bank assets. National perimeters range from insured institutions and significant institutions to commercial banks, authorised deposit-taking institutions and prudential conglomerates. The report therefore makes global thematic conclusions only where international sources support them and otherwise states the jurisdiction or sample explicitly.

Global macro-financial environment

The IMF’s July 2026 update estimated world output growth at 3.5% in 2025 and projected 3.0% for 2026 and 3.4% for 2027. World headline inflation was estimated at 4.1% in 2025 and projected at 4.7% in 2026 and 3.9% in 2027; world trade-volume growth was 5.0%, 3.5% and 4.3%, respectively. The rise in projected 2026 inflation alongside slower growth creates an asymmetric banking backdrop: funding and discount rates may remain restrictive even as credit demand softens. These are macroeconomic projections, not bank-sector outcomes. IMF WEO Update, Table 1, p. 12.

Global banking sector at a glance

The strongest broad structural baseline is end-2024. In the FSB 21-jurisdiction-plus-euro-area group, bank assets were $191.3 trillion, 38.0% of covered financial-system assets, after 4.7% growth during 2024. Banks represented 58.6% of financial assets in covered emerging-market economies and 32.4% in advanced economies. Revisions, national-statistics improvements and the exclusion of Russia from 2021-24 aggregates impair edition-to-edition comparison. FSB, Table 0-2 and Graph 1-1. For 2025, the most comparable international activity measure is cross-border bank credit. It reached $38.1 trillion at year-end and grew 11%; total cross-border claims, which also include derivatives and residual instruments, reached $46.0 trillion. The distinction matters: neither series measures domestic lending by domestic banks. BIS.

Regional performance

• United States and Canada: Latest verified evidence: FDIC-insured institutions ended 2025 with $25.26tn of assets, $13.48tn of loans and $18.44tn of domestic deposits; net income was $295.6bn, up 10.2%. Canada’s six D-SIBs reported 13.7% CET1 at Q4 2025 and 13.5% at 30 April 2026.; Interpretation and limit: US assets/loans/deposits grew 4.8%/5.9%/3.9%; Canada retained a cushion over OSFI’s supervisory expectation. Perimeters and fiscal calendars differ. • Euro area, UK and Switzerland: Latest verified evidence: Euro-area Q1 assets €28.87tn, NIM 1.54%, ROE 10.02%, CET1 15.99%; UK Q1 CET1 14.2% and May LCR 142%; Swiss end-2025 CET1 18% and ROA 0.55%.; Interpretation and limit: Strong aggregate buffers, but national perimeters and reporting dates prevent ranking. • Central and Eastern Europe: Latest verified evidence: Hungary’s 2025 corporate/household loan growth was 7.3%/14.7%, RoE 18.9%, capital adequacy 20.1%, LCR 165.8% and NPL ratios 3.2%/1.6%. BIS cross-border credit to emerging Europe grew 26%.; Interpretation and limit: Hungary is an anchor, not a CEE aggregate; subsidised housing programmes affected household lending. • China: Latest verified evidence: Q1 banking-institution assets RMB494.7tn (+8.0%); commercial-bank NPL ratio 1.51%, coverage 203.14%, LCR 151.65%.; Interpretation and limit: Expansion continued, but asset totals and prudential ratios use different institutional perimeters. • Japan and South Korea: Latest verified evidence: Japanese capital ratios remained above requirements; four BOJ real-estate-related chart categories summed to 43.0% of domestic loans

Profitability and business-model drivers

Profit remained positive across the major samples, but its drivers shifted. BCBS Group 1 banks recorded €182.4 billion of after-tax profit in H1 2025; rolling twelve-month profit increased 5.5% in Europe, 15.7% in the Americas and 10.8% in the rest of the world. The regions are BCBS sample groupings, not complete banking-system totals. BCBS.

Credit growth and asset quality

US bank loans grew 5.6% in 2025 and all major categories expanded; aggregate delinquency was 1.6%, below the Federal Reserve’s approximately 3% long-run reference. Consumer, CRE, C&I and residential delinquency measures rose slightly or were broadly near year-earlier levels, while card and auto delinquency ended lower year on year. Federal Reserve, pp. 12-13. In the euro area, the NPL ratio excluding central-bank balances fell 0.06 percentage point to 2.18% and the Stage 2 share fell from 9.76% to 9.29%, but cost of risk rose 0.02 percentage point to 0.58%. This combination is consistent with contained aggregate stock deterioration alongside somewhat higher marginal provisioning; it is not evidence that every portfolio improved. In Australia, residential NPLs fell to 0.99%, while CRE exposure grew 8.7% to AUD487.6 billion and the share of new investor mortgages at debt-to-income ratios of six or more rose from 8.2% to 10.8%. In China, commercial-bank NPLs were RMB3.7 trillion, 1.51% of loans, with RMB7.5 trillion of provisions and 203.14% coverage; institutional coverage differs across those published totals. The principal portfolio risks are refinancing-sensitive office and weaker retail CRE, leveraged and weaker corporate borrowers, unsecured consumer vintages, high debt-to-income mortgages, and sovereign exposures in weaker emerging markets. Housing risk is jurisdiction-specific: aggregate prices can remain stable while affordability, variable-rate resets or investor leverage worsen.

Capital, liquidity and funding

The BCBS June 2025 sample provides the most comparable international prudential view. Group 1 CET1 was unchanged at 13.9%, leverage at 6.1%, LCR rose 0.9 percentage point to 135.0%, and NSFR rose 1.0 percentage point to 123.8% in the unbalanced sample. All reporting banks were above 100% for LCR and NSFR, but sample composition prevents interpreting the fall in reported accounting assets as economic contraction.

Market structure, failures and consolidation

The verified failure evidence is US-specific. The FDIC recorded two insured-bank failures with $113.3 million of assets in full-year 2025. Four institutions with $625.83 million of reported assets had failed by 17 July 2026. The comparison is incomplete-year versus full-year and does not establish a global failure trend. FDIC 2025; FDIC 2026. Capital One completed its acquisition of Discover on 18 May 2025. At approval, the combined organisation was estimated at $637.8 billion of assets, 2.2% of US insured-depository-organisation assets, and $469.9 billion of deposits, 2.6% of the national total. Federal Reserve approved M&A applications increased from 99 in 2024 to 145 in 2025, or 46.5%, but approvals are not completed transactions. Federal Reserve order, pp. 2-3; Federal Reserve S&R Report, pp. 34-35. Concentration differs sharply by market. At end-2025, the five largest credit institutions held 34.37%- 95.2% of assets across EU member states; the ECB-reported EU average was 69.33%. These figures should not be combined with US nationwide deal shares. The FSB judged foundational resolution frameworks mostly in place but retained funding in resolution and cross-border bail-in execution as operational gaps. ECB; FSB Resolution Report.

Regulation and supervisory developments

Final Basel III implementation remained staggered. At end-September 2025, revised credit and operational-risk standards and the output floor were effective in about 80% of 27 BCBS jurisdictions, CVA in nearly 70% and revised market risk in nearly 40%. Legal effectiveness does not prove comparable bank-level implementation. BCBS RCAP. At the cutoff, the United States had proposals, not final rules, for three capital reforms published on 19 March 2026. UK PRA Policy Statement PS1/26 set most Basel 3.1 rules for 1 January 2027 and marketrisk internal models for 1 January 2028. The European Commission’s 4 June 2026 FRTB adjustments were a delegated act subject to Parliament and Council scrutiny, intended for 1 January 2027. Crossborder banks must therefore distinguish current, enacted future and proposed requirements. Federal Reserve; PRA PS1/26; European Commission Q&A. The 2025 FSB list contained 29 G-SIBs. Bank of America and ICBC moved to higher loss-absorbency buckets and Deutsche Bank to a lower bucket; resulting higher requirements apply from 1 January 2027. In the reporting subsets, TLAC and MREL shortfalls were small, but the metrics must not be added: the BCBS TLAC figure covers 14 G-SIBs, while SRB MREL covers Banking Union entities. FSB; SRB. Official stress tests support conditional resilience, not cross-country ranking. In the EBA 2025 test, 64 banks covering about 75% of EU assets lost 370 basis points of CET1 to 12.1% under the adverse scenario. Seven UK groups covering about 75% of real-economy lending fell from 14.5% to 11.0%, above a 6.2% hurdle. Thirty-two US banks fell from 12.8% to an 11.2% minimum in the 2026 scenario and all remained above minimum requirements.

Technology, AI, cyber and operational resilience

AI adoption expanded faster than comparable supervisory measurement. The FSB’s October 2025 report described official monitoring as early-stage and identified third-party dependency, correlated behaviour, cyber risk and model/data governance as potential vulnerabilities. Its June 2026 report proposed 12 sound practices but remained a consultation at the cutoff. Neither source quantifies AI’s contribution to bank revenue, costs or losses. FSB AI monitoring; FSB consultation. BCBS’s December 2025 third-party-risk principles and June 2026 ICT practices review emphasise change control, design and testing, capacity management and external dependencies. The review covered 16 authorities and incident evidence from 12 jurisdictions in 2022-24; reporting thresholds differ. The 3,383 major ICT incidents reported in DORA’s first year covered the wider financial sector, averaged 0.18 per financial entity and were only 10% cyber-classified. They are not a bank-only global incident rate. BCBS; ESAs. Boards should require an inventory of material AI uses and owners; map critical services to applications, people, data, cloud and nth parties; measure provider concentration, substitutability and exit time; test human override and recovery objectives; and track critical-service downtime, changefailure rates, unsupported assets and unresolved high-severity model findings.

Twelve-month outlook and scenarios

The horizon is 29 July 2026-28 July 2027. Scenarios are conditional risk-management paths, not point forecasts; no numerical probabilities are assigned.

Strategic implications and early-warning indicators

Management should distinguish a slow earnings transition from a compounding solvency, liquidity or operational event. No universal threshold is valid across jurisdictions; escalation should be bankspecific and triggered by correlated movements. • Rates, curve, deposit beta and product migration: Escalation signal: Asset yields fall before deposit costs; operational balances migrate; concentration rises; Decision use: NIM, pricing and liquidity actions • Wholesale/CDS spreads, cross-currency basis and facility use: Escalation signal: Persistent spread widening or rising facility use without a system event; Decision use: Funding contingency • Stage 2, NPL, past-due and cost of risk: Escalation signal: Deterioration broadens across sectors/vintages or provisions lag migration; Decision use: Underwriting, provisioning, capital • CRE valuation, vacancy, debt service and maturities: Escalation signal: Falling collateral values coincide with large refinancing needs; Decision use: Portfolio limits and workout capacity

Data sources, scope and limitations

Data sources

• Financial Stability Board (FSB)
• Bank for International Settlements (BIS) and Basel Committee on Banking Supervision (BCBS)
• International Monetary Fund (IMF): FSIC
• FSIBSIS
• WEO and GFSR
• European Central Bank Banking Supervision
• Federal Reserve and Federal Deposit Insurance Corporation
• Bank of Canada and Office of the Superintendent of Financial Institutions
• Bank of England and Prudential Regulation Authority
• Swiss National Bank and Magyar Nemzeti Bank
• National Financial Regulatory Administration of China and Bank of Japan
• Australian Prudential Regulation Authority and Indonesia Financial Services Authority
• Banco Central do Brasil
• Central Bank of the United Arab Emirates
• South African Reserve Bank
• Single Resolution Board and European Supervisory Authorities

Limitations

• The evidence combines end-2024 FSB structure, end-2025 BIS cross-border activity, June 2025 BCBS monitoring, FY2025 national results and Q1/May 2026 national updates. It is not a singledate census. • Measured global bank-asset coverage is Not calculable. GDP coverage and cross-borderclaims coverage are not substitutes. • IMF panels are equal-country-weighted, nominal local-currency diagnostics and omit several largest systems. Annual labels may mix fiscal and calendar conventions. • National sources use different institutional perimeters, consolidation bases, accounting standards and definitions of NPL, Stage 3, CET1, NIM, cost of risk, LCR and NSFR. • Euro-area institution count changed; US large-bank quarterly profitability is not M&A-adjusted; stress-test samples and scenarios differ. • India, South Korea and Southeast Asia lack verified equivalent aggregate primary tables at the cutoff. CEE, MENA and Sub-Saharan Africa use regional cross-border data or anchor jurisdictions. • The Brazil accounting change in January 2025 impairs asset-quality time comparison. Some 2026 data are provisional and all live IMF data can revise. || • No causal identification design was used. Diagnostic wording denotes association or consistency, not causation. Scenario directions have no assigned probabilities.

References and related links

• FSB Global Monitoring Report 2025
• BIS international banking statistics, end-December 2025
• BCBS Basel III Monitoring Report
• IMF Financial Soundness Indicators Core (FSIC)
• IMF Financial Soundness Indicators Balance Sheet and Income Statement (FSIBSIS)
• IMF World Economic Outlook Update, July 2026
• IMF Global Financial Stability Report, April 2026, Chapter 1
• ECB Supervisory Banking Statistics, Q1 2026
• Federal Reserve Supervision and Regulation Report, June 2026
• Bank of Canada Financial Stability Report 2026
• Bank of England Financial Stability Report, July 2026
• Swiss National Bank Financial Stability Report 2026
• Magyar Nemzeti Bank Financial Stability Report, June 2026
• NFRA banking-sector statistics, Q1 2026
• Bank of Japan Financial System Report, April 2026
• APRA Quarterly ADI Statistics, March 2026
• APRA System Risk Stress Test
• OJK Banking Sector Update, December 2025
• Banco Central do Brasil Financial Stability Report, May 2026
• CBUAE Monetary, Banking and Financial Markets Development Report, Q1 2026
• CBUAE Quarterly Economic Review, June 2026
• SARB selected prudential trends, December 2025
• FDIC Quarterly Banking Profile, Q4 2025
• FDIC bank failures in brief
• FSB 2025 Resolution Report
• Federal Reserve 2026 stress-test results
• PRA PS1/26 - Implementation of Basel 3.1 final rules
• European Commission banking-package market-risk Q&A
• FSB 2025 list of global systemically important banks
• SRB MREL Dashboard, H2 2025
• FSB artificial-intelligence monitoring report 2025
• FSB responsible adoption of AI consultation report 2026
• BCBS ICT and operational-resilience practices (report d611)
• European Supervisory Authorities 2025 report on major ICT-related incidents