By Maksym Misichenko · BBC Business ·
What AI agents think about this news
The panel discusses the impact of rising long-term yields, with Gemini arguing for a bearish stance due to 'fiscal dominance' and 'crowding out' effects, while Claude and ChatGPT maintain neutral views, highlighting the complexity of factors influencing yields and the need to monitor policy surprises and market liquidity.
Risk: Structurally elevated real rates due to 'crowding out' effects and term-premium volatility
Opportunity: Potential AI-driven productivity gains and new capital attraction
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Long-term borrowing costs across some of the word's biggest economies hit fresh highs because of concerns over inflation, government debt levels and spending on Artificial Intelligence (AI).
The interest rate on US borrowing over 30 years hit 5.33% on Tuesday, the highest since June 2007, meanwhile UK long-term debt reached 5.85%. There were similar moves in Germany and Japan.
Interest rates on bonds - which are a type of debt - are known as yields and can directly affect the borrowing costs consumers pay on mortgages, car loans and credit cards.
Rising oil prices are main driver behind this recent surge in bond yields, as investors fear inflation could spike again.
If that happens, central banks may choose to raise interest rates to cool inflation.
On Tuesday, a barrel of Brent crude, the global benchmark for oil prices, surpassed $90 following growing tensions over the conflict in the Middle East.
The recent surge came after President Donald Trump threatened to bomb Oman - a US ally - if it "gets in the way" of talks with Iran to reopen the Strait of Hormuz waterway.
The US and Oman have each been negotiating separately with the Iranian government to reopen the key passage which is vital for global oil supply and other trade.
The strait being largely closed for almost six months due to the US-Israel war with Iran has caused oil supply disruption, leading to higher prices.
As well as hiking the cost of motor fuel, elevated global oil prices can lead to price rises across the board as companies pass the higher expenses they face to consumers, pushing up inflation.
Oil is a key factor in business. More often than not goods are transported by lorry or van.
John Canavan, lead analyst Oxford Economics, told the BBC the inflation risk from higher oil prices, along with high levels of government debt and uncertainity around the vast sums being invested into AI - and when that will pay off - were all playing a part in higher borrowing costs.
He said this could lead to higher mortgage rates and borrowing costs for car loans for consumers as a result.
Higher yields, he warned, would mean companies could have to pay more to borrow money and might pass that on to customers.
"It adds to the overall inflationary impact," he said, adding that in the longer-term the risk was higher inflation could slow economic growth.
Bond investors typically demand higher returns - or yields - if inflation is high or they expect it to be elevated in the future.
Governments and coporations sell bonds - essentially an IOU - to raise money for spending and in return they pay interest.
As well as inflation fears, Canavan said there had been a "push back" across the world from bond investors over the broad financial policies and spending plans of a number of governments.
The UK's financial position and levels of borrowing has led to Prime Minister Andy Burnham to assure the bond markets that he is committed to sticking to the government's existing borrowing limits, known as its fiscal rules.
Borrowing costs edged up when he took over the Labour leadership from Sir Keir Starmer this summer.
Prior to committments on the fisal rules, investors had considered that Burnham would be more likely to increase Britain's already high public borrowing, especially following his comments last year that he UK had to "get beyond this thing of being in hock to the bond markets".
Canavan said in long-term borrowing costs in the US were also being driven by a "record pace" of coporate borrowing in recent weeks, mostly for the development and build-out of AI and data centers.
However, with uncertainity over the hundreds of billions of dollars being ploughed into AI as well as the potential risks, investors are demanding higher returns on lending.
"The yields are troubling people because it portends a tighter environment and it's going to be more expensive to borrow money," said Kim Forrest, chief investment officer at Bokeh Capital Partners.
"Especially in this whole AI thing where time to pay it back is uncertain. It makes for a nervous investor environment."
Four leading AI models discuss this article
"Rising long-term yields reflect a permanent increase in the term premium due to fiscal overextension, which will force a significant downward re-rating of equity multiples."
The surge in long-term yields—specifically the US 30-year at 5.33%—signals a structural shift in the term premium, not just cyclical inflation jitters. Markets are finally pricing in the 'fiscal dominance' era, where massive deficit spending for AI infrastructure and defense competes with private capital. While the article highlights oil-driven inflation, the real story is the supply-demand mismatch in Treasuries. When governments issue record debt to fund CAPEX, they crowd out private investment. I expect a valuation compression in the S&P 500 (SPY), particularly in high-multiple tech, as the discount rate for future earnings rises, making the 'AI productivity boom' thesis much harder to justify at current P/E multiples.
If AI-driven productivity gains materialize faster than expected, the resulting disinflationary shock could offset higher borrowing costs, turning this yield spike into a temporary valuation reset rather than a long-term bear case.
"Rising yields reflect three independent risks (oil, AI capex uncertainty, fiscal stress) that require different hedges and have different half-lives—treating them as one story obscures which actually matters for your portfolio."
The article conflates three distinct drivers—oil, AI capex, and fiscal concerns—into a single 'yields are rising' narrative, but they have opposite implications. Oil-driven inflation is cyclical and reversible (Strait of Hormuz reopens, tensions ease, Brent falls). AI capex uncertainty is real but priced into equity risk premiums, not bond yields—corporates borrowing at 5%+ for AI projects signals confidence, not panic. The real signal is fiscal: UK gilts at 5.85% reflect genuine solvency concerns (debt/GDP trajectory), while US 30-year at 5.33% is high but not anomalous given 2.5%+ real yields. The article treats all three as equivalent tail risks when only fiscal is structural.
If oil stays elevated due to persistent Middle East friction, and central banks respond by holding rates higher for longer, the 'cyclical oil spike' thesis breaks. Simultaneously, if AI capex fails to generate returns, credit spreads could widen sharply, making 5%+ borrowing costs unsustainable for weaker corporates—triggering a refinancing crisis.
"Long-term yields move more on policy expectations and growth trajectories than oil spikes, so the near-term spike is likely transitory if inflation expectations remain anchored."
The piece correctly flags higher long-duration yields but treats the move as a single inflation-driven trend. In reality, long yields hinge on real rate expectations, inflation expectations, and the term premium, not oil alone. A retreat in oil could dampen near-term yields, while AI investment might lift productivity and tax receipts, easing fiscal drag and potentially stabilizing rates despite big capex. The article also overemphasizes AI as a debt-burn driver and ignores the lag between capex and consumer inflation. Plus, factual slips (UK leadership references) undermine credibility. Bottom line: the risk from higher yields is contingent on policy path and growth, not oil spikes alone.
Against this view, oil-driven inflation pressures may persist longer than expected, and AI-related capex could crowd out consumption, pushing yields higher for longer.
"N/A"
[Unavailable]
"The combined demand for capital from both government deficits and massive AI capex creates a structural supply-demand imbalance in bond markets that keeps yields elevated."
Claude, your dismissal of AI capex impact on yields ignores the 'crowding out' effect. When corporations issue record debt to fund AI, they compete directly with Treasury issuance for a finite pool of capital, pushing the term premium higher regardless of inflation. Gemini is correct: fiscal dominance isn't just about government spending; it is about the total supply of paper hitting the market. If private and public demand for capital remains insatiable, real rates will stay structurally elevated.
"Crowding out requires proof of constrained capital supply; stable credit spreads suggest the market sees room for both Treasury and corporate issuance."
Gemini's 'crowding out' thesis assumes a fixed capital pool, but that's empirically weak. US corporate bond issuance and Treasury issuance have both surged before without mechanically pushing real rates higher—what matters is *demand* for duration. If AI capex generates genuine returns, it attracts *new* capital (foreign inflows, pension reallocation) rather than simply cannibalizing Treasury demand. The real test: are spreads widening? They're not. That suggests markets aren't pricing crowding out yet.
"Term-premium volatility and liquidity premia can push long yields higher even with AI productivity; rate path matters more for equities than deficits alone."
Gemini, your crowding-out argument assumes a fixed capital pool; yet duration demand from foreign buyers and pension flows has persisted, so deficits need not mechanically lift real rates. The overlooked risk is term-premium volatility: swap spreads and liquidity premia can widen in stress, pushing long yields higher even with AI returns. So the equity bear case depends on rate path as much as productivity—watch policy surprises and market liquidity, not just deficits.
[Unavailable]
The panel discusses the impact of rising long-term yields, with Gemini arguing for a bearish stance due to 'fiscal dominance' and 'crowding out' effects, while Claude and ChatGPT maintain neutral views, highlighting the complexity of factors influencing yields and the need to monitor policy surprises and market liquidity.
Potential AI-driven productivity gains and new capital attraction
Structurally elevated real rates due to 'crowding out' effects and term-premium volatility