US borrowing costs have reached a 24-year high, according to The Guardian, as a global bond sell-off intensifies. For chief financial officers and treasury teams, the immediate question is not whether this is a temporary spike but how to model a sustained rise in long-term yields across refinancing calendars, floating-rate exposure and project hurdle rates. The move matters most for capital-intensive sectors, particularly AI infrastructure and energy projects that rely on dollar-denominated debt.
The Bank of England has added a second layer of risk. Its governor warned that the AI boom could trigger market shocks, according to BBC News. That warning does not predict a crash, but it does signal that regulators are watching the intersection of high leverage and concentrated investment in a narrow set of technologies. For CFOs, the practical implication is that the cost of capital and the risk premium attached to AI-linked assets may rise together.
What has changed
Long-term US borrowing costs are now at their highest level in 24 years. The Guardian reports that the global bond sell-off has driven yields up sharply, tightening financial conditions for cross-border corporates. This is not a short-term funding squeeze in the overnight market; it is a repricing of long-duration debt, which affects the discount rates used to value projects with multi-year payback periods. At the same time, the Bank of England has flagged that the AI boom could trigger market shocks. The BBC reports that the central bank is monitoring the concentration of investment in AI and the potential for a disorderly repricing if expectations disappoint. This is a risk warning, not a forecast, but it reinforces the case for stress-testing assumptions rather than extrapolating recent returns.Why it matters for CFOs
Three transmission channels deserve immediate attention. First, refinancing risk: any dollar-denominated bond or loan maturing in the next 18 months will be repriced at a higher yield, raising interest costs and potentially breaching covenant thresholds. Second, floating-rate exposure: companies that swapped fixed for floating to benefit from low short-term rates now face a double squeeze if long-term yields rise and central banks keep policy rates elevated. Third, project hurdle rates: a higher risk-free rate mechanically raises the weighted average cost of capital, which can turn marginal AI or infrastructure projects from viable to value-destructive. The Bank of England's warning adds a fourth channel: valuation risk. If AI-linked assets are priced on the assumption of continued falling capital costs, a sustained rise in yields could compress multiples independently of operating performance. That is a balance-sheet issue, not just a cash-flow one.A decision framework for treasury teams
CFOs should avoid a single-scenario forecast. Instead, model three states: a base case where long-term yields stabilise at current levels, a stress case where they rise another 50 to 100 basis points, and a tail case where the AI investment cycle unwinds faster than expected. For each state, quantify the impact on interest coverage, net debt to EBITDA, and return on invested capital for the top five capital projects. Refinancing calendars should be stress-tested first. Identify every dollar-denominated instrument maturing before the end of 2027 and calculate the refinancing cost at current market rates. If the new cost pushes interest coverage below 3x, consider pre-funding or extending maturities now, even at a premium. Floating-rate exposure should be re-examined: if the floating leg is hedged with swaps that expire within 12 months, the hedge ratio may be insufficient. Project hurdle rates need a hard reset. Many companies still use a cost of equity that reflects the low-rate era. A simple adjustment is to add the increase in the 10-year US Treasury yield since 2021 to the discount rate for long-duration projects. That will not be precise, but it will force a conversation about which projects still clear the bar.Commercial impact
The commercial impact is uneven. Companies with strong free cash flow and investment-grade ratings can absorb higher costs and may even benefit if competitors are forced to delay projects. Highly leveraged firms in AI infrastructure, data centres and renewable energy are more exposed. For private equity and venture capital investors, higher borrowing costs reduce the arbitrage between cheap debt and equity returns, which may slow deal activity in capital-intensive sectors. Suppliers to AI and infrastructure projects should also prepare for slower order books if customers delay or cancel marginal projects. That risk is not yet visible in order data, but it is a plausible second-order effect of a sustained yield rise.Risks and unknowns
The main unknown is whether the bond sell-off reflects a durable shift in the term premium or a temporary overshoot. The Guardian reports that the sell-off is global, which suggests it is not solely a US fiscal story. The Bank of England's warning is explicitly about the AI boom, but it does not quantify the probability of a shock. CFOs should treat both signals as reasons to build contingency plans, not as predictions. A second unknown is the path of central bank policy. If long-term yields rise because of stronger growth expectations, the impact on corporate borrowers is different from a rise driven by fiscal concerns or inflation risk. The research packet does not resolve this, so scenario analysis should include both interpretations.FY Outlook
Over the next two quarters, the most likely path is that long-term yields remain elevated while short-term policy rates stay on hold. That combination is particularly challenging for companies that rely on floating-rate debt and have not hedged. If the AI investment cycle continues to attract capital, the risk of a disorderly repricing rises, but the timing is uncertain. CFOs should use the current window to refinance where possible, extend hedges, and re-underwrite capital projects with a higher discount rate. The cost of waiting is likely to be higher than the cost of acting now.Sources and References
- The Guardian (theguardian.com)
- BBC News (bbc.co.uk)
Why It Matters
US borrowing costs at a 24-year high directly affect the cost of capital for cross-border corporates. CFOs who delay refinancing or fail to adjust hurdle rates risk covenant breaches and value-destructive projects. The Bank of England's warning on AI market shocks adds a valuation risk that balance sheets may not yet reflect.The reporting and evidence for this briefing were checked against theguardian.com (theguardian.com) and bbc.co.uk (bbc.co.uk).
