The Bank for International Settlements (BIS) has warned that record public debt, expensive financial markets and increasingly opaque financing for artificial intelligence infrastructure are creating interconnected risks that could amplify the next economic or market shock.
The institution published its Annual Economic Report 2026 on Sunday, June 28, identifying four principal threats to global stability: renewed inflation, uncertainty over the durability of artificial intelligence investment, vulnerabilities across financial markets and mounting pressure on public finances.
The Bank for International Settlements said the global economy had remained resilient despite geopolitical conflict and repeated supply disruptions. However, it cautioned that elevated asset valuations, heavy borrowing and greater dependence on non-bank financial institutions had reduced the capacity of markets and governments to absorb another major disruption.
The report does not predict that the artificial intelligence boom will inevitably collapse or that governments are approaching an immediate debt crisis. It warns that the combination of ambitious earnings expectations, extensive infrastructure spending, concentrated private-credit exposure and leveraged sovereign-bond trading could turn a normal repricing into a broader financial contraction.
The central concern is that several apparently separate risks are becoming linked. Artificial intelligence investment is increasingly financed through debt. Private-credit funds have expanded their exposure to technology borrowers. Highly leveraged hedge funds have become important intermediaries in government-bond markets. Governments, meanwhile, have less fiscal capacity to respond because debt and interest costs are already high.
Why is the Bank for International Settlements issuing an urgent global warning now?
The Bank for International Settlements serves as a forum for central banks and closely examines risks that could affect monetary and financial stability across national borders. Its 2026 report arrives after markets and economic activity proved more resilient than many policymakers expected, despite trade fragmentation, war, energy disruption and restrictive interest rates.
That resilience has encouraged investor confidence. Equity valuations remain elevated, artificial intelligence investment continues to expand and financial conditions have often appeared easier than the underlying geopolitical and fiscal environment might suggest.
The Bank for International Settlements believes that this contrast creates vulnerability. Market prices increasingly assume that inflation will remain manageable, artificial intelligence will produce substantial future profits and governments will continue refinancing large debts without destabilising bond markets.
A disappointment in one of those assumptions could affect the others. Another energy shock could revive inflation and delay interest-rate reductions. Higher rates could weaken heavily indebted companies and governments. A reassessment of artificial intelligence returns could reduce equity valuations, tighten corporate credit and pressure private lenders.
Pablo Hernández de Cos, General Manager of the Bank for International Settlements, said economic policies needed to reinforce rather than contradict each other. The institution urged governments and central banks to act before vulnerabilities require more abrupt and expensive adjustments.
The warning is therefore less about identifying a single imminent crisis than about recognising that the financial system has accumulated several channels through which stress could spread quickly.
Why does the BIS believe artificial intelligence investment could become financially unstable?
The Bank for International Settlements recognises that artificial intelligence could improve productivity, expand productive capacity and support economic growth. Its concern is that investment may be advancing faster than the technology’s proven commercial returns.
Major technology companies, data-centre operators, semiconductor businesses and artificial intelligence laboratories are spending heavily on computing infrastructure, electricity, networking equipment and advanced chips. Competition for leadership gives companies an incentive to invest aggressively even when the eventual level of demand remains uncertain.
The report compares this dynamic with earlier technological investment waves in which genuine innovation was accompanied by excessive construction and financing. Artificial intelligence can transform the economy while still producing overinvestment in particular companies, facilities or supply-chain segments.
Large listed companies have historically financed technology investment through operating cash flow and equity. The present cycle is increasingly drawing upon corporate bonds, private credit, infrastructure financing and complex arrangements involving several participants.
The Bank for International Settlements found that artificial intelligence companies and infrastructure providers had substantially increased long-term debt. It also identified extensive circular financing, in which technology companies invest in artificial intelligence laboratories or infrastructure providers that subsequently commit to purchasing computing capacity, chips or related services from their investors.
Such arrangements can accelerate development, but they can also make revenue, financing and customer relationships difficult to separate. Capital provided as an investment can return to the investor as reported revenue when the recipient purchases its technology or computing services.
The structure becomes risky when expected revenue depends on continuous investment elsewhere in the same ecosystem. A slowdown by artificial intelligence laboratories could affect cloud providers, chip manufacturers, data-centre owners, energy projects and lenders almost simultaneously.
How exposed are equity investors if artificial intelligence expectations disappoint?
The Bank for International Settlements said equity valuations, particularly among companies central to artificial intelligence development, were pricing ambitious long-term earnings growth.
Its analysis showed that expected growth for some of the world’s largest companies was well above historical benchmarks. Investor compensation for accepting equity risk had also fallen in the United States since the pandemic, indicating confidence and reduced concern about possible losses.
High valuations are not proof of a bubble. They may be justified if artificial intelligence generates durable productivity gains, new products and substantial future cash flow. The vulnerability arises because expensive prices leave less room for disappointment.
The potential economic consequences of an equity correction are also greater because household exposure to shares has increased. The Bank for International Settlements found that American households held a larger proportion of their wealth in equities than before the dot-com crash, particularly among the highest-income households.
A sharp fall in technology shares could therefore affect household wealth, consumption and business confidence, not merely the portfolios of professional traders. The importance of United States-listed companies to global equity indices means a correction could also spread rapidly through international pension funds, exchange-traded funds and institutional portfolios.
The most likely risk is not that every artificial intelligence company fails. It is that earnings and adoption progress more slowly than valuations assume, causing investors to reduce the prices they are willing to pay for future profits.
That repricing could then cross into credit markets because many companies throughout the artificial intelligence supply chain have borrowed to finance expansion.
Why is private credit becoming a critical weakness inside the artificial intelligence boom?
Private credit refers broadly to lending provided outside traditional public bond markets and conventional bank loans. Funds and other non-bank institutions lend directly to companies, often providing flexible financing that may not be readily available from banks.
The sector has grown rapidly and has become an increasingly important source of funding for middle-market companies, infrastructure projects and technology businesses. However, the loans are less transparent than publicly traded bonds, making it harder for regulators and investors to assess valuation, concentration and liquidity risks in real time.
The Bank for International Settlements found that direct-lending funds had quadrupled their lending to artificial intelligence and information technology businesses over five years. Those sectors now represented about 15 percent of the funds’ portfolios.
Technology loans were generally larger than lending in other sectors, while their pricing and maturities were not significantly more conservative. The institution said this raised questions about whether lenders were being adequately compensated for the risks involved.
Concentration creates another concern. A software company may borrow from several private-credit funds, meaning apparently diversified lenders can be exposed to the same borrower and the same technological disruption.
The vulnerability could become visible if investors demand their money back during a downturn. Some private-credit vehicles serve retail investors while holding assets that cannot be sold quickly. Rising redemption requests can force funds to restrict withdrawals, sell available assets or return capital earlier than planned.
Banks remain connected to this system because they lend to private-credit funds, provide financing arrangements and maintain relationships with insurance companies that invest in the sector. Stress outside the regulated banking system can therefore return to banks through indirect and difficult-to-measure exposures.
How are record public debts creating a new threat inside sovereign bond markets?
Public debt in many economies is close to levels not experienced since the period following the Second World War. Governments accumulated borrowing through the global financial crisis, the coronavirus pandemic, energy interventions and other major shocks.
The Bank for International Settlements said fiscal policy had often failed to rebuild sufficient financial capacity during periods of economic expansion. Governments now face additional expenditure demands involving ageing populations, defence, industrial policy, climate adaptation and infrastructure.
Higher interest rates have increased debt-servicing costs as older bonds mature and are refinanced. This reduces the money available for public services and limits governments’ ability to respond to future recessions or emergencies.
The structure of sovereign bond markets has also changed. Central banks reduced their government-bond holdings through quantitative tightening between 2022 and 2025, while some foreign official demand also declined. Private investors have consequently been required to absorb more government debt.
Hedge funds have become increasingly important intermediaries, particularly through strategies that seek to profit from small differences between prices in cash bond and futures markets. Those returns can be modest, so funds use substantial leverage financed through short-term repurchase agreements.
The Bank for International Settlements found that United States sovereign-debt exposure among hedge funds had more than doubled relative to gross domestic product since 2022. Hedge funds had also increased their share of government-bond trading in the euro area and other advanced economies.
This creates what the report calls a new fiscal and financial stability nexus. Governments issue large amounts of debt, while increasingly leveraged financial institutions help absorb and trade that debt. If funding conditions suddenly tighten, funds may be forced to sell bonds rapidly, causing yields to rise and reducing liquidity in markets central to the entire financial system.
Why can hedge-fund leverage turn a bond selloff into a wider financial emergency?
Hedge funds finance many government-bond positions through short-term borrowing. In calm markets, dealers are willing to lend against high-quality sovereign bonds because the collateral is generally considered safe and liquid.
The Bank for International Settlements found that around 70 percent of bilateral United States dollar repurchase agreements involving hedge funds and more than 50 percent of comparable euro transactions were conducted with zero haircuts.
A zero haircut allows a borrower to obtain financing against almost the full market value of the collateral. This increases leverage and potential returns, but it leaves little protection if bond prices fall or lenders change their terms.
During stress, dealers may demand more collateral, reduce financing or raise haircuts. Hedge funds can then be forced to sell assets quickly to meet margin requirements. Simultaneous selling pushes prices lower and can produce additional margin demands, creating a self-reinforcing cycle.
Government bonds serve as benchmarks and collateral throughout the financial system. Dysfunction in these markets can affect corporate borrowing, mortgages, currencies and bank funding even when the originating problem involved only a group of leveraged traders.
Central banks may then be required to intervene to restore market functioning. Such action can conflict with monetary policy if a central bank is trying to restrain inflation while also supplying liquidity or purchasing bonds to stabilise markets.
The report therefore argues that the safety of sovereign bonds cannot be assessed only through the government’s ability to repay. Regulators must also examine who owns and intermediates the debt, how those positions are financed and how quickly leverage could be withdrawn.
How could renewed inflation make debt and artificial intelligence risks more dangerous?
The Bank for International Settlements said inflation had started to rise again in some economies and warned that recurring supply disruptions could make households and businesses expect persistently higher prices.
The institution acknowledged that the reopening of the Strait of Hormuz and progress towards reducing the United States and Iran conflict could prevent the most extreme energy scenarios. However, oil markets may take time to normalise, and geopolitical risks remain capable of producing fresh price shocks.
Central banks can sometimes look through temporary increases caused by supply disruption. The danger emerges when businesses begin raising prices in anticipation of future increases or workers seek compensation for continuously rising living costs.
If inflation expectations become less stable, central banks may need to keep interest rates higher or tighten policy further. That would increase borrowing costs for governments, businesses and artificial intelligence infrastructure projects.
Higher rates could also trigger the market repricing identified elsewhere in the report. Expensive equities would face a higher discount rate, companies would encounter more costly refinancing and leveraged bond traders could face tighter funding terms.
The interaction explains why the Bank for International Settlements is urging coordinated policy. Broad government spending intended to offset energy costs could support household demand but also complicate inflation control. Higher central-bank rates could suppress inflation while worsening fiscal costs and exposing financial leverage.
The institution argues that governments should target temporary support towards vulnerable households rather than adopting broad subsidies that increase debt and preserve excess demand.
What policy changes does the Bank for International Settlements want governments to make?
The Bank for International Settlements is asking governments to place public finances on credible and sustainable paths before markets force more severe adjustment.
That does not necessarily require immediate austerity during economic weakness. It requires believable medium-term plans that control structural deficits, prioritise expenditure and rebuild fiscal capacity when economic conditions allow.
Central banks are being urged to preserve price stability and respond if evidence shows that inflation expectations are becoming less firmly anchored. The report also stresses the importance of central-bank independence when high public debt creates political pressure to keep borrowing costs low.
Financial regulators are being asked to strengthen supervision outside conventional banks. Measures could include minimum haircuts for securities financing, stronger liquidity requirements for open-ended funds, wider central clearing and more complete reporting of private-credit exposures.
Artificial intelligence creates a more complicated policy challenge. Governments do not want to suppress investment in a technology capable of raising productivity, but they also need greater transparency around debt, intercompany commitments, private financing and infrastructure concentration.
The report’s wider message is that policymakers should not assume future artificial intelligence productivity will automatically solve current debt problems. Productivity gains may support economic growth, but their scale, timing and distribution remain uncertain.
Waiting for technological growth to repair public finances would expose governments to substantial risk if investment returns arrive later than expected or fail to match current optimism.
What are the key takeaways from the BIS warning on artificial intelligence and global debt?
- The Bank for International Settlements identified renewed inflation, uncertainty surrounding artificial intelligence investment, financial-market vulnerabilities and mounting public debt as the four principal risks facing the global economy in 2026.
- The institution said artificial intelligence could generate significant productivity gains, but competition, supply constraints and ambitious earnings expectations could also produce overinvestment followed by a sharp reduction in spending.
- Artificial intelligence infrastructure is increasingly financed through corporate debt, private credit and complex arrangements connecting technology companies, artificial intelligence laboratories, semiconductor producers, data centres and infrastructure providers.
- Direct-lending funds have quadrupled their exposure to artificial intelligence and information technology borrowers over five years, with those sectors now representing approximately 15 percent of their portfolios.
- Public debt in many economies is close to post-Second World War highs, while rising interest costs, ageing populations, defence spending and infrastructure requirements are placing additional pressure on government finances.
- Hedge funds have become important intermediaries in sovereign bond markets and frequently finance leveraged positions through short-term repurchase agreements carrying zero or extremely low collateral haircuts.
- The Bank for International Settlements warned that an artificial intelligence repricing, inflation shock or abrupt tightening of funding could spread from equity and credit markets into government bonds, banks and the wider economy.
- The report urged governments to restore fiscal sustainability, central banks to protect price stability and regulators to improve oversight of private credit, hedge funds and other non-bank financial institutions.
Discover more from Business-News-Today.com
Subscribe to get the latest posts sent to your email.
