The 30-year gilt yield reached 5.89% this week, its highest since 1998. The 10-year rose as much as 11 basis points to 5.25%. Gilts led a global government bond selloff that also carried the US 10-year Treasury to 4.81%, close to a three-year high, and left the Japanese 10-year above 3% for the first time in three decades.
The proximate causes are the ones you would guess. Oil is higher on the US–Israeli conflict with Iran, which reignites the inflation trade and makes existing fixed coupons less attractive. Central banks that were supposed to be cutting now look more likely to hold or hike. And investors across the G7 are increasingly unconvinced that governments can grow their way out of debt loads when the interest rate exceeds the growth rate and debt service eats an ever-larger share of the budget.
But Britain led. That ordering is not an accident, and it is the part worth understanding.
Why the long end of the gilt curve breaks first
The natural buyer of a 30-year government bond is a defined-benefit pension scheme matching liabilities decades out. That buyer is disappearing. DB schemes are largely closed to new members and their memberships are ageing, so structural demand for very long paper has been shrinking for years. The Debt Management Office has already adapted: short and medium conventional gilts now account for close to 70% of planned issuance, while long-dated conventionals have fallen to roughly 10%. The average maturity of the gilt stock was 13.4 years at the end of 2025, down from 16.5 years a decade earlier.
That leaves the 30-year point thinly populated. Fewer natural holders means a smaller shock moves the price further — which is precisely what a 28-year high on a day of broad but not extraordinary global selling looks like.
Our take: The headline is a yield. The consequence is a budget. Deutsche Bank's chief UK economist estimates that if current gilt levels persist through the autumn and feed into the Office for Budget Responsibility's forecast, the government's fiscal headroom shrinks from roughly £26bn to about £13.8bn — roughly halved, without a single policy decision being taken. Chancellor John Healey delivers the Budget on 28 October. The arithmetic he presents that day is being written now, by a market, and not by the Treasury.
The feedback loop to watch for
Halved headroom does not force anything by itself. It narrows the menu. A chancellor with £13.8bn of room against fiscal rules has materially less space to absorb a downgrade to productivity assumptions, or a spending pressure, without either raising taxes or amending the rules. Both of those are choices markets then price — which is how a yield move becomes a fiscal event becomes another yield move. The UK has run this loop before, recently enough that the memory is doing some of the selling.
The cross-border piece matters too. When Japanese long yields rise, the historic incentive for Japanese institutions to reach into US and European duration weakens, because domestic paper finally pays. Gilts have relied on that global bid alongside the domestic one. Both are thinner than they were.
What to watch
- DMO issuance remit updates. Any further skew away from long-dated conventionals is a direct read on where the Treasury thinks demand actually is.
- The OBR's forecast assumptions. The market rate that gets locked into the Budget arithmetic is set on a specific cut-off date, not on Budget day. That date matters more than most headlines about it.
- Oil. The inflation leg of this selloff is a war premium. If it unwinds, so does part of the yield move.
- Whether the 30s/10s spread keeps widening. A steepening long end driven by supply and absent buyers is a different signal from one driven by growth expectations, and the policy responses differ.
None of this is a call on where yields go next. It is a note that the UK's long end is now structurally thinner than it used to be, which means the same global shock lands harder here — and that a Budget eight weeks out is being constrained in real time.