Global banking entered 2026 with the most constructive planning environment since 2021. Capital positions had strengthened through the rate cycle. Inflation had retreated sufficiently for the Federal Reserve, the European Central Bank (ECB) and the Bank of England to signal rate reductions in the first half of the year. The International Monetary Fund (IMF) had upgraded its global growth projection to 3.3%, with financial conditions gradually loosening and credit quality broadly stable. Three months later, the IMF has paired its reference forecast with two explicit downside scenarios. The reference scenario — the IMF's central case — yields global growth of 3.1% and inflation of 4.4%, assuming the Strait of Hormuz disruption resolves without deeper escalation. The adverse and severe scenarios illustrate tail risks: an adverse case with growth at 2.5% and inflation at 5.4%, and a severe case the World Economic Outlook frames as a “close call” for global recession, with inflation exceeding 6.0%. The United States (US) Energy Information Administration (EIA) projected Brent crude peaking at $115 a barrel in the second quarter, with production shut-ins rising to 9.1 million barrels per day in April. At the IMF's April 14 briefing, Pierre-Olivier Gourinchas, the IMF's chief economist, was direct: “With every day that passes, where we don't have a resolution, where the flow of oil and gas is more limited through the Strait of Hormuz, we're moving away from that [reference] scenario.” He added that “if the conflict were to stop today, the oil shortfall for the year... is comparable to the shock from the 1970s in terms of how much oil has been withdrawn from the market on an annual average basis.” Brent crude was already approaching $100 at the time he spoke, against a reference scenario that assumes approximately $80 as a full-year average. Gourinchas noted that the global economy is materially less oil-dependent than in the 1970s — with many more energy sources available and greater efficiency in producing GDP per unit of oil. He added that central banks have also built more credible frameworks over the intervening decades, providing a second source of resilience. The banking sector as a source of resilience The IMF's Spring 2026 publications are unambiguous on the starting position. Tobias Adrian, the IMF's Financial Counsellor, said: "The banking system is not a worry at this particular juncture." Major banking systems enter this period with capital ratios at multi-decade highs, the product of sustained post-Basel III capital accumulation. Liquidity positions are strong across advanced economy systems. Provisioning coverage is broadly adequate against current asset quality trends. The GFSR describes the sector as well-positioned to absorb shocks through balance sheet capacity rather than through policy support. That position is not uniform. The Global Financial Stability Report (GFSR) differentiates between advanced economy systems — where capital and liquidity buffers are strongest — and emerging market banking systems, where specific vulnerabilities persist. Within advanced economies, the fund flags that forward-looking indicators — capital distribution trends, provisioning trajectories and asset quality in exposed sectors — identify where the current position could come under pressure. The question the IMF's Spring 2026 publications address is therefore not the banking sector's current position but whether the fiscal and financial environment in which it operates can absorb the shocks now arriving. Sovereign risk and the bank balance sheet The IMF's October 2025 Fiscal Monitor projects global public debt crossing 100% of GDP by 2029 — the highest since 1948, when wartime financing had pushed debt to comparable levels. The fund's analysis of defence spending booms finds that in wartime conditions, public debt has historically jumped 14 percentage points of GDP within three years, with approximately two-thirds of defence spending financed through deficits. The GFSR notes that the investor base absorbing sovereign supply has shifted toward foreign asset managers, passive funds and ETFs — holders that are more price-sensitive and flight-prone under stress than the hold-to-maturity institutions that absorbed earlier cycles of issuance. The GFSR identifies the transmission channels from this fiscal picture into banking, with the sharpest pressure points in emerging markets. Bank holdings of local-currency government debt in emerging markets have risen from 15% of assets pre-pandemic to 20% by 2025. Under the GFSR's most severe stress scenario — which combines a specific haircut on sovereign bond values with other stress parameters — nearly half the banking systems in lower-rated emerging markets would require recapitalisation, with capital ratios falling below regulatory thresholds. In these markets, domestic banks have increasingly stepped in to absorb government bond issuance as external financing has tightened, a trend the GFSR documents as most pronounced in countries rated CCC-or-below or unrated. Foreign non-bank financial investors reduce their emerging market holdings materially more than domestic investors when global risk rises, with passive funds and ETFs showing the greatest sensitivity. According to the GFSR, a one-percentage-point increase in reliance on these risk-sensitive investors implies a 28% decline in sovereign debt issuance volume during stress episodes, with spreads widening sharply for issuers with pre-existing vulnerabilities. Japanese government bond yields have risen to multidecade highs since October 2025, the GFSR noted, raising the prospect that Japanese institutional investors — among the largest holders of US Treasuries and euro area sovereign debt — reallocate capital back to domestic bonds and transmit stress into global bond markets. In the advanced economies themselves, the GFSR identifies distinct pressure points: in the US, the main exposure runs through bond portfolio duration and non-bank financial intermediary (NBFI) credit lines; in Europe, the ECB's withdrawal from bond markets and rising rollover risks compound the pressure. NBFI exposure: hedge fund leverage and private credit gates Bank exposure to NBFIs has grown to nearly 13% of world GDP, concentrated in hedge funds through credit lines, repos, derivatives and prime brokerage, according to the IMF. The interconnections are global: the GFSR documents that UK, Japanese and Canadian banks are among the primary providers of cross-border funding to US NBFIs, meaning that stress in the US NBFI sector transmits rapidly across borders through funding withdrawals, credit losses and fire sales. Hedge fund gross notional exposure to interest rate derivatives and sovereign bonds has more than doubled since 2020, rising to over $18 trillion among qualifying US-registered hedge funds. The top 10 global hedge funds now account for more than one-third of that exposure. On private credit, the direct lending universe stands at approximately $2 trillion globally. Adrian said at the April 14 briefing: “We think about 15% or $300 billion of that is in semiliquid structures, so where investors can redeem. But of course, the funds do have gates... if more of this sector were to become redeemable, that would certainly impact the broader assessment.” Within this segment, the GFSR identifies AI infrastructure financing as a further concentration. Since 2025, investment by hyperscalers has increasingly been channelled through circular financing arrangements — structures in which developers, chipmakers and technology firms simultaneously act as each other's customers, investors and financiers. The GFSR assesses current financial stability risks from this interconnectedness as modest, though the financing structures continue to expand and their opacity is itself a source of risk. Banks face exposure through funding lines to private credit vehicles facing drawdown risk under stress, and through bond portfolios and derivatives books that would absorb losses if hyperscalers' assumptions about the useful life of their infrastructure prove optimistic. The disinflation path Markets were pricing rate cuts across major central banks in the first half of the year, but the energy shock has suspended those expectations. The IMF notes that stagflation concerns have already tightened lending conditions and compressed margins simultaneously across several markets. Gourinchas flagged the risk that inflation expectations could de-anchor if wage-price spirals take hold — a risk the IMF's adverse scenario assumes partially materialises, with one-year-ahead inflation expectations rising by up to 50 basis points in advanced economies and 90 basis points in emerging markets excluding China. The repricing of deposit and funding costs that many forward margin models had assumed would normalise in 2026 now looks less certain; in the adverse scenario, the disinflation path that would have supported it does not arrive. Gourinchas explained the structural difference from post-pandemic disinflation: “The 2022 surge reflected an unusually steep aggregate supply curve, from supply bottlenecks, enabling disinflation with limited output losses. Our analysis shows the supply curve is now much flatter, making any central bank engineered disinflation more costly in terms of unemployment.” The GFSR also notes that the traditional equity-bond hedging relationship has broken down since 2020, removing an asset-liability management offset that bank treasury and portfolio managers had previously relied upon. Where the IMF's forward-looking indicators point The IMF's growth-at-risk framework translates the broader financial stability picture into probabilities. Under current financial conditions, the GFSR finds that one-year-ahead global growth could fall below 0.75% with a 5% probability. Under the downside scenario, growth-at-risk deteriorates by a further 1.2 percentage points, falling to minus 0.5%. The scenarios present an analytical framework for preparedness rather than a prediction. The banking sector's capital and liquidity position provides a buffer that did not exist in prior episodes of stress. Central banks retain the capacity to deploy liquidity facilities alongside rate decisions. What has changed is the fiscal space to recapitalise banks or absorb system-wide losses at scale. As Adrian noted at the IMF's briefing: “Over the past five or six years, governments have often come in to support financial stability with the policy space, but the policy space has been drawn down in many countries.” The fund's framing is consistent: resilience in the banking sector, constraint in the context.
