BIS Annual Economic Report 2026 

The Bank for International Settlements (BIS) does not tell a simple crisis story in its Annual Economic Report 2026. The report opens with a seemingly reassuring observation: despite major tariff increases, geopolitical uncertainty and a shock to the energy system, the global economy responded with surprising resilience in 2025. Growth, trade and financial markets held up better than many forecasts had expected. Yet the BIS does not derive any all-clear from this strength. On the contrary, the central message is that resilience is not the same as robustness.

The difference is more than semantic. Resilience describes the ability to absorb a shock in the short term. Robustness requires the system to remain functional even under repeated, simultaneous or more persistent shocks. This is precisely the robustness that the BIS sees at risk. Behind the cyclical stability, according to its analysis, stand several buffers: the less severe impact of tariffs than initially feared, a powerful wave of AI investment, loose financing conditions and persistently high market risk appetite. Yet each of these buffers can itself become a source of new risks.

Fig. 01: Global growth proved resilient despite tariff and geopolitical shocks [Source: BIS (2026): Annual Economic Report, June 2026, Graph 1, p. 8]

A global economy between progress and peril

The BIS frames the year as a movement from "resilience" to "robustness". This wording is programmatic. At first, 2025 was a year in which the global economy withstood a severe trade policy stress test surprisingly well. The tariff increases in the United States did hit the multilateral trading system, but their actual effects remained muted. According to the BIS, the effective average US tariff rate stabilised in the second half of 2025 at around 10 percent, far below the previously announced peaks of more than 25 percent. Exemptions, trade agreements, the diversion of trade flows and the willingness of many firms to absorb costs initially through lower margins softened the impact.

A second buffer of remarkable scale was added: the boom surrounding artificial intelligence. Investment in semiconductors, data centres, power infrastructure and digital intermediate goods supported not only the US economy but also radiated through global supply chains. Asia benefited through export channels, Europe through investment impulses, and China through a shift away from the real estate sector towards advanced manufacturing. The global economy was thus carried by a technological narrative: AI promised productivity gains and justified rising capital expenditure, high equity valuations and favourable financing conditions.

Fig. 02: AI exuberance supported investment and kept financial conditions loose [Source: BIS (2026): Annual Economic Report, June 2026, Graph 3, p. 11]

For the BIS, this combination of real-economy investment momentum and financial optimism is ambivalent. On the one hand, AI can represent a genuine supply-side improvement, provided that productivity, diffusion and competition come together. On the other hand, an investment wave driven by expectations of future monopoly rents and market leadership can generate a classic overinvestment dynamic. Historically, such phases - canals, railways, electrification, dotcom - have produced real innovation, but often also misallocated capital and intensified later corrections.

The energy shock: Hormuz as a stress test of globalisation

The second phase of the report begins with a geopolitical shock: the conflict in Iran from late February 2026 and the subsequent effective closure of the Strait of Hormuz. The BIS describes this bottleneck as the world's most critical energy chokepoint. When traffic there came to a standstill in early March, a significant share of global energy flows was disrupted. The report puts the crude oil flow loss at more than 10 million barrels per day, around 13 percent of normal global supply. In terms of volumes, the effect was therefore larger than in the energy crises of the 1970s.

Fig. 03: The blockade of the Strait of Hormuz disrupted activity and oil flows on a historically large scale [Source: BIS (2026): Annual Economic Report, June 2026, Graph 4, p. 12]

According to the BIS, the price reaction was initially less extreme than earlier historical comparisons would have suggested. Reserves, the expectation of a time-limited conflict and market adjustments dampened the increase. The decisive point, however, is not only the oil price. The BIS notes that the shock reaches deep into supply chains: liquefied natural gas, fertilisers, sulphur, helium, petrochemical inputs, plastics and semiconductor production are affected. Asian economies are particularly exposed because a large share of oil and gas flows through Hormuz is destined for Asia.

This is the core of the macroeconomic danger: an energy shock is not only a price signal at petrol stations or in electricity markets. It changes production costs in agriculture, industry, chemicals and technology. It can drive food prices, delay investment plans and reactivate inflation dynamics. The BIS stresses that physical damage to energy infrastructure, bottlenecks in replacement parts and the replenishment of strategic reserves can cause effects to persist even after political tensions ease.

Fig. 04: The severe energy shock increased inflationary pressures and showed amplification through supply chains [Source: BIS (2026): Annual Economic Report, June 2026, Graph 8, p. 18]

The return of inflation is not a repetition, but a warning

The report reads the energy shock against the backdrop of the experience of 2021 to 2023. At that time, supposedly temporary price impulses proved capable of becoming more persistent through supply chains, wages and expectations than central banks had initially assumed. The starting position in 2026 is different: labour markets are less overheated in many places, policy rates are higher than at the start of the post-pandemic inflation surge, and many central banks have regained credibility in recent years. Nevertheless, the BIS warns against an overly comfortable interpretation.

The reason lies in the non-linearity of large supply shocks. Firms can cushion small increases in energy prices through margins or productivity reserves. Large, persistent and broad-based shocks are more likely to be passed on. When energy, chemicals, fertilisers, plastics and transport all become more expensive at the same time, a cascade emerges. Price increases then move from inputs into intermediate goods and ultimately into final prices. The report shows that large energy supply shocks can have disproportionate inflation effects.

This worsens the classic trade-off for monetary policy. Central banks cannot remove the first supply shock. They can neither guide oil through the Strait of Hormuz nor repair damaged LNG facilities. But they can prevent a one-off increase in the price level from becoming permanently higher inflation. The more backward-looking expectations become and the longer a shock lasts, the less convincing it is simply to "look through" it.

Markets: risk appetite despite geopolitical stress

For the BIS, the response of financial markets is particularly striking. After the conflict began, global equity markets initially fell, but recovered quickly. The correction remained limited by historical comparison. The BIS points out that global equity markets fell by around 9 percent from late February to the end of March 2026 - less than during earlier shocks such as the 2025 tariff announcement or past geopolitical episodes. Credit spreads initially widened, but then recompressed to historically low levels. The US dollar gained at first, but later gave back part of those gains as risk appetite returned.

Fig. 05: Risk appetite remained high despite the energy crisis [Source: BIS (2026): Annual Economic Report, June 2026, Graph 7, p. 17]

The BIS calls this divergence between macro risk and market valuation remarkable. Markets appear to be betting on a quick and lasting resolution of the conflict, continued AI momentum and still robust earnings. At the same time, volatility and tail-risk indicators have returned close to pre-crisis levels. This may be rational if the shocks truly remain short-lived and the AI boom generates real productivity. But it may also reflect compressed risk premia that could reverse abruptly if expectations are disappointed.

AI: productivity promise and overinvestment risk

Artificial intelligence stands at the centre of the progress narrative. The BIS treats AI not as a mere stock market theme, but as a potential general-purpose technology. At the task level, studies often point to substantial efficiency gains, sometimes in the range of 20 to 50 percent time savings. At the macroeconomic level, estimates are more cautious because adoption, reorganisation, skills, competition and energy infrastructure are decisive.

The report therefore draws a clear distinction between technological potential and macroeconomic realisation. Productivity does not arise automatically from models, chips or data centres. It arises when firms reorganise processes, workers use new skills, prices reflect competition and bottlenecks in energy, semiconductors and grids are addressed. These are exactly the areas where the BIS sees risks.

Fig. 06: The rapid AI boom raises questions about the sustainability of investment momentum [Source: BIS (2026): Annual Economic Report, June 2026, Graph 11, p. 23]

According to the BIS, the five largest hyperscalers are set to spend more than one trillion US dollars on AI-related capital expenditure in 2025 and 2026. This is relevant for industrial policy and macroeconomics. In the short term, it supports demand, construction, semiconductors, energy and services. In the medium term, it increases dependence on future returns that remain uncertain. If the expected productivity gains fail to materialise, if power grids or chip capacity become binding constraints, or if it turns out that too many providers were simultaneously trying to force market leadership, the investment boom can turn into an investment brake.

There is also a financial dimension: some AI activities are increasingly supported by complex, partly circular financing structures. Chip suppliers, hyperscalers, AI labs, cloud providers and data-centre developers are linked through equity stakes, purchase commitments, credit markets and long-term leasing structures. The more opaque these interconnections become, the harder it is to assess where risk truly lies.

Financial vulnerabilities as amplifiers

The BIS does not analyse financial markets in isolation, but as potential amplifiers of macroeconomic shocks. Rising inflation risks could force higher policy rates or higher longer-term yields. Disappointments in AI returns could correct equity valuations and credit spreads. Both triggers would hit a system in which risk premia are low, valuations are high and parts of credit intermediation have become less transparent.

The BIS is particularly concerned about the expansion of private credit markets and the role of non-bank financial intermediaries. Private credit can close financing gaps and promote diversification. At the same time, this market is less transparent than bank credit, often less liquid and potentially more procyclical in periods of stress. When redemptions, valuation adjustments and refinancing pressure coincide, credit conditions can tighten faster than standard models assume.

The high equity exposure of private households also increases the macroeconomic importance of market corrections. If equity valuations fall, the effect on consumption through wealth effects is stronger today than in earlier phases. At the same time, the US equity market accounts for a very large share of global indices. A US-led correction could therefore extend beyond national borders and reach international portfolios, pension systems and consumption channels.

The fiscal question: room for manoeuvre is narrowing

The second major strand of the report concerns public finances. Many governments are entering a more volatile world with high debt levels, rising interest burdens and new spending needs. Demographics, defence, energy supply, climate adaptation and infrastructure compete for fiscal space. At the same time, the interest rate environment is less benign than in the years after the global financial crisis. In many places, the gap between nominal growth and financing costs has deteriorated.

Fig. 07: Higher public debt is reducing fiscal space [Source: BIS (2026): Annual Economic Report, June 2026, Graph 15, p. 27]

The BIS criticises not only the absolute level of debt but also the asymmetry of fiscal policy. In downturns, policy reacts expansively, but in good times consolidation is often insufficient. As a result, debt ratios ratchet up. For advanced economies, the report cites cyclically adjusted primary deficits averaging 1.9 percent of GDP since 2022, compared with 1.1 percent over the preceding two decades. In emerging market economies, the deterioration is even more pronounced.

This development is relevant for monetary policy. High public debt can change transmission. When policy rates rise, government interest payments also increase, as does income for bondholders. At the same time, financial intermediaries suffer valuation losses on longer-term government bonds. Depending on the level of debt, maturity structure and investor base, a rate increase can therefore have different effects on demand and inflation than in a world of low public debt.

The new fiscal-financial stability nexus

Chapter II of the report describes a structural shift that is particularly consequential for central banks: sovereign bond markets are increasingly intermediated by non-banks. These include insurance companies, pension funds, money market funds, open-ended funds and, above all, hedge funds that use high leverage to run relative-value strategies in government bonds and derivatives. The BIS calls this a new fiscal-financial stability nexus.

Fig. 08: Non-bank financial intermediaries are playing a larger role in advanced economy sovereign debt markets [Source: BIS (2026): Annual Economic Report, June 2026, Graph 4, p. 50]

The finding is quantitatively clear: according to the BIS, the share of non-banks in advanced-economy sovereign debt holdings rose from 44 percent in 2021 to 53 percent in 2025. At the same time, the share held by domestic central banks fell from 27 to 17 percent. The private market therefore has to absorb more government bonds - and does so increasingly through actors whose liquidity cannot be taken for granted in stress periods.

Leveraged strategies that rely on short-term repo financing are particularly delicate. As long as market liquidity is abundant, these strategies can deepen sovereign bond markets and smooth price differences. But when volatility rises, haircuts are increased or margin calls emerge, intermediation can turn into a wave of selling. The BIS stresses that liquidity in sovereign bond markets can appear ample for a long time but vanish abruptly under stress. Fiscal space can then shrink before long-term debt sustainability indicators signal a limit.

Central banks between price stability and market stabilisation

This creates a threefold problem for central banks. First, fiscal risks may be repriced more often and more abruptly. If government bond yields rise not because of better growth prospects but because of concerns about debt sustainability or market liquidity, financing conditions tighten for governments, firms and households. Second, monetary policy transmission becomes more uncertain because the level, maturity and holder structure of debt affect the impact of rate moves. Third, central banks may have to intervene more often in sovereign bond or repo markets when market dysfunction threatens.

Such interventions are effective, but not costless. They can encourage market participants to take on more leverage if they expect a central bank backstop. They can weaken fiscal discipline if governments trust that market pressure will be cushioned. And in an environment of high inflation, they can make communication harder: is a bond purchase intended to stabilise markets, or is it effectively monetary easing? The BIS therefore recommends backstops that are temporary, targeted and reversible.

Stablecoins: the future of money needs trust

Chapter III shifts from macroeconomic and financial stability to the architecture of money. The starting point is fundamental: money works because it is accepted with no questions asked. This property rests not on technology alone, but on institutions - a common unit of account, the singleness of money, par convertibility, final settlement, elastic liquidity provision and integrity against money laundering and terrorist financing. Today's two-tier system of central bank money and private intermediation broadly fulfils these functions, but in practice remains fragmented, costly and often inefficient across borders.

Stablecoins enter this gap with the promise of programmable, faster and globally accessible payments. The BIS analyses them soberly. Stablecoins demonstrate the potential of tokenisation, but so far meet key properties of money only incompletely. They circulate on public permissionless blockchains, are fragmented across different networks, deviate from par value in secondary markets and raise significant questions about financial integrity.

Fig. 09: Stablecoin market growth is highly concentrated in a few US dollar-pegged coins [Source: BIS (2026): Annual Economic Report, June 2026, Graph 2, p. 88]

The scale is both large and small. According to the BIS, stablecoin market capitalisation stood at around 320 billion US dollars at the end of May 2026 - significant for the crypto market, but small compared with bank deposits. Around 99.4 percent of fiat-backed stablecoins were pegged to the US dollar. Their main use so far has been trading in cryptoassets, not mass payments. This distinction is central to the assessment: as crypto-liquidity instruments, stablecoins can be useful; as general money, they must meet far higher standards.

Macroeconomic effects of stablecoins: small in the model, large in stress

The BIS discusses three stylised scenarios for how stablecoin reserves might be held: as bank deposits, as short-term government bonds or as central bank reserves. Each scenario shifts liquidity differently. If stablecoins are backed by bank deposits, households replace retail deposits with concentrated wholesale deposits of the issuer. If they are backed by government bonds, demand for short-term government paper rises. If they are backed by central bank reserves, liquid funds move directly from the banking system to the issuer.

Fig. 10: The modelled macroeconomic effects of broader stablecoin adoption remain limited, but depend on reserve structure, public debt and foreign demand [Source: BIS (2026): Annual Economic Report, June 2026, Graph 5, p. 97]

In the model calculations, the net effects on aggregate economic activity appear limited. But that does not amount to an all-clear. What matters is distribution and stress dynamics. Stablecoins can make bank funding more expensive, alter credit provision, narrow short-term government bond markets or trigger fire sales during redemptions. In economies with weak fundamentals, demand for foreign stablecoins can become a form of dollarisation. A private digital instrument then touches monetary sovereignty.

The BIS therefore does not present stablecoins as the endpoint of innovation, but as an intermediate form. The preferred path lies in developing the two-tier system further: tokenised central bank reserves, tokenised commercial bank deposits and tokenised assets could be brought together on interoperable, programmable platforms. A "unified ledger" or a system of interoperable ledgers is intended to combine the efficiency of tokenisation with the trust foundation of the existing monetary system.

The policy compass: robustness instead of mere shock resistance

The overall picture yields several policy priorities. Monetary policy must anchor expectations and must not allow supply-side shocks to turn into permanently higher inflation. Fiscal policy must become more symmetric again: stabilise in crises, build buffers in good times. Financial supervision must regulate risks where they arise, not only where traditional bank balance sheets are visible. In this context, the BIS speaks of congruent regulation: similar risks should be treated with similar stringency, regardless of whether they sit in banks, funds, hedge funds, stablecoin issuers or other vehicles.

At the same time, the real economy needs structural robustness. Supply chains must be diversified, energy sources broadened, strategic reserves sensibly dimensioned and critical inputs better mapped. AI potential requires skills, competition and infrastructure. Tokenisation requires not only software, but governance, legal clarity, interoperability and financial integrity.
The point of the report is therefore less pessimistic than disciplinary. The global economy has shown that it can absorb shocks. But this capability was supported by favourable financial conditions, technological expectations and remaining fiscal room for manoeuvre. If these buffers themselves come under pressure, resilience is no longer enough. Robustness emerges only when price stability, sound public finances, resilient financial markets and trustworthy monetary architectures reinforce one another.

Download BIS Annual Economic Report (June 2026)