Rising government borrowing costs, disruption to energy supplies and the investment demands of artificial intelligence are exposing the tensions in an economy built on abundant credit. They do not establish that a collapse is imminent. They do raise a question governments would prefer to postpone: who will pay for the commitments they have made, and whose interests will be protected when the adjustment comes?
The consequences of war do not remain on the battlefield. They travel through oil supplies, exchange rates, government budgets and the cost of credit, eventually reaching people who had little influence over the decisions that set them in motion.
That transmission is becoming visible in the bond market. Governments must finance accumulated obligations while energy disruption complicates inflation control and technology companies raise money for enormous investment programmes. Each pressure has a different origin, but they meet in the price investors demand for lending.
The danger is not simply that governments owe large sums. It is that borrowing becomes more expensive before the income needed to support those obligations has grown sufficiently. Political choices made during years of easier financing then return as demands for sacrifice.
How the pressure reaches households
A government bond promises future payments. When investors become less willing to hold it at its existing price, the price falls and its yield rises. New borrowing must compete with the returns available on those existing securities.
Higher yields do not necessarily signal fear of default. They can reflect inflation expectations, anticipated central bank decisions, greater debt supply or demands for more compensation for holding an obligation over many years.
Nevertheless, the consequences extend beyond investors. Government yields influence financing conditions across the economy. Businesses reconsider expansion, prospective homeowners encounter more expensive mortgages and governments face higher costs when refinancing.
Existing debt carrying fixed interest rates does not suddenly become more expensive. The pressure builds as obligations mature and new borrowing replaces them.
America faces that process with gross federal debt exceeding $40 trillion, according to September reporting by Reuters. This includes obligations within government as well as debt held by domestic and foreign investors. It is not a sum owed exclusively to foreign countries.
The Congressional Budget Office’s February outlook projected a $1.9 trillion deficit for 2026 and net interest expenditure above $1 trillion. Borrowing therefore finances current commitments while servicing earlier decisions consumes an increasing share of available resources.
The energy shock crosses borders
Disruption to oil supplies intensifies this pressure because energy enters the cost of transport, manufacturing and food production. Governments cannot isolate those effects within the countries fighting a war.
The US Energy Information Administration has documented how disruption through the Strait of Hormuz increased petroleum price volatility and forced buyers to seek alternative supplies. Its latest available outlook assessed July production shut ins at 5.5 million barrels a day.
An importing country must find replacement energy, finance a larger bill or reduce consumption. Currency weakness can compound the problem by making imports more expensive.
Japan illustrates how financial pressures can spread. Reuters reported on 7 September that its foreign reserves fell by $79.6 billion in August following record intervention to support the yen. Sales of foreign securities, principally US Treasuries, formed part of that operation.
This does not establish that Japan is abandoning America. It shows how defending a currency can require transactions in another country’s debt market. International financial dependence transmits problems as well as opportunities.
For central banks, the resulting dilemma is uncomfortable. Higher interest rates cannot restore missing oil supplies. They can restrain spending and limit the persistence of inflation, but households may then experience both higher living costs and more expensive credit.
Norway is choosing returns
Norway’s sovereign wealth fund provides a different example of changing demand.
Its manager recommends reducing government bonds from 70 per cent to 50 per cent of the bond benchmark. The Tthe proposal could lower its allocation to US Treasuries by approximately $95 billion.
The distinction matters: this is a proposal about portfolio composition, not an executed withdrawal or a declaration against the dollar.
Norges Bank’s original letter argues that a smaller government bond allocation could still provide adequate liquidity while allowing the fund to earn additional returns elsewhere. It continues to recognise the importance of government securities during periods of financial stress.
Much of the proposed replacement investment would remain within American fixed income. Norway is consequently weak evidence for an anti American financial revolt.
It is stronger evidence for a less dramatic proposition: investors have alternatives. Governments cannot assume that institutions will indefinitely prefer their securities at whatever borrowing cost suits the national budget.
Borrowing what was not taxed
Behind the financing question lies another: why did the state need to borrow, and who benefited from that decision?
When a government cuts taxes without equivalent expenditure reductions or sufficient additional growth, its financing requirement increases. Some wealth left in private hands may subsequently be invested in government debt.
The state has exchanged a potential tax receipt for an obligation to repay money with interest.
This mechanism does not explain every deficit. Recessions, emergencies, demographic change and productive public investment also create borrowing requirements. Nor are creditors exclusively wealthy individuals. Pension funds, insurance institutions and ordinary savers also hold government obligations.
But taxation and borrowing distribute rights differently. Taxation transfers resources to the state. Lending creates an asset for the creditor and a future commitment for the public.
There is concrete evidence that fiscal choices distribute gains unequally. The Congressional Budget Office’s assessment of the 2025 reconciliation law projected average annual resource losses of roughly $1,200 for households in the lowest income tenth and gains of approximately $13,600 for those in the highest tenth over its projection period. Losses principally reflected reduced benefits, while gains at the top principally reflected lower taxes.
These projections do not prove that every interest payment transfers money from poor people to rich people. They do demonstrate why subsequent demands for public restraint require scrutiny of the choices that preceded them.
AI competes for the future
Artificial intelligence offers a possible route towards stronger growth. More productive businesses could generate the income and tax revenues needed to support public commitments.
But infrastructure must be financed before its benefits are fully realised.
The IMF’s April financial stability report identified an estimated $3.4 trillion in AI related capital expenditure through 2029. Major technology infrastructure companies had raised more than $100 billion in bond financing since January 2025.
Investment on that scale creates demand for capital, electricity, construction and equipment. During the transition, those demands can contribute to financing pressures before productivity gains arrive. IMF analysis identifies higher real interest rates as a potential difficulty for heavily indebted governments during the AI investment cycle.
It would nevertheless be wrong to attribute the entire government bond sell off to technology borrowing. Savings are not a fixed reservoir from which every corporate dollar mechanically subtracts a government dollar.
The narrower argument is sufficient. Governments and companies seek financing simultaneously, and investors compare their opportunities. The growth strategy intended to improve public finances may initially make financing more demanding.
Progress does not guarantee profits
A transformative technology can still produce disappointing investments.
The usefulness of AI does not determine whether a particular share price is justified, whether every data centre will earn an adequate return or whether borrowers can meet their obligations.
Leverage determines how far disappointment spreads. Investors using their own money can endure falling valuations without necessarily selling. Those borrowing against assets may face demands for additional collateral or repayment, forcing sales that push prices lower.
Businesses face a similar distinction. Strong cash reserves provide time to adjust. Heavy debts and fixed commitments leave less room when revenues disappoint.
The IMF recognises both sides. It identifies strong balance sheets and substantial cash generation among major technology companies, while warning that expanding investment requirements and rapid equipment obsolescence could create future vulnerabilities.
The relevant question is therefore where fragile financing sits: among infrastructure developers, suppliers, private lenders or leveraged investors, rather than whether every company belongs inside a single bubble narrative.
A reversal could reach public finances through weaker investment, employment and revenue. The IMF’s Fiscal Monitor identifies those potential spillovers, while treating them as risks rather than inevitable outcomes.
A crisis is not predetermined
The strongest counterargument deserves more than a passing qualification.
AI may deliver substantial productivity gains. Energy disruption may ease. Inflation may moderate sufficiently to permit less restrictive monetary policy. Governments may alter their fiscal choices before financing pressure becomes acute.
There is already evidence against a simple prediction of relentlessly rising interest rates. Federal Reserve governor Christopher Waller’s 3 September speech acknowledged elevated energy prices and technology related pressures, but also signs of disinflation. He left open the possibility of holding rates steady.
An equity downturn would not necessarily cause a government bond downturn either. Recession fears can reduce inflation expectations and encourage investors to seek safety in sovereign debt, lowering government yields.
The more difficult combination would be weak growth alongside persistent inflation or deteriorating fiscal confidence. Governments would then face pressure to support their economies while borrowing remained expensive.
That possibility warrants preparation. It does not justify presenting financial collapse, political breakdown or American decline as events already determined.
Who carries the adjustment?
Inflation offers no painless solution. It can erode the purchasing power of existing debt, but investors may demand higher returns on replacement borrowing. People whose incomes fail to keep pace suffer immediately.
Spending reductions also distribute losses. Cutting health provision, education or maintenance may reduce expenditure today while weakening welfare and productive capacity tomorrow.
Tax increases require equally serious examination of design, avoidance, investment incentives and who ultimately bears the burden.
Borrowing itself should be judged by what it finances. Infrastructure that improves productivity or emergency support that prevents lasting economic damage can justify obligations extending into the future. Tax concessions and military commitments deserve the same rigorous assessment, rather than automatic protection when other expenditure comes under review.
The evidence points to increasing pressure on the assumptions that sustain extensive borrowing: manageable inflation, willing creditors and adequate future income. It does not establish an approaching refusal to finance the American state.
The political danger is that this distinction disappears when the bill arrives. Decisions once presented as affordable become obligations described as unavoidable, while the people asked to adjust are given little explanation of how the choices were made.
Bond markets determine the terms on which governments can borrow. They cannot determine whether those commitments were wise or whether their costs are fairly shared. Before another sacrifice is demanded, citizens deserve an account of who benefited, what was achieved and why they should pay.