Sahil Mahtani, Head of Macro Research at Ninety One
You hold government bonds for their ballast: they’re meant to rally when growth wobbles and equities sell off. That premise looks shaky these days. Medium-term inflation has not been durably contained; the fiscal arithmetic is deteriorating; and trend growth is too soft. Put together, these forces push term premia up and keep stock–bond correlations higher than the pre-2020 norm. This blunts bonds’ ability to offset equity drawdowns. So, do you need them at all?
Start with inflation. It has eased from its 2022-2023 peaks but remains stubbornly above target in many advanced economies. In the UK, despite the lack of a feared wage-price spiral, inflation has been particularly sticky, with services inflation still hovering around 3.5% year-on-year. Sticky inflation has been driven by regulation – for example, five years of fast-paced increases in minimum wages, excise duties, packaging rules and loose fiscal policy, layered on top of higher global costs. That has interrupted disinflation and kept long-dated gilt yields higher.
In the US, fears of persistent inflation have deferred policy easing, despite cooling growth momentum, providing one reason why long-term Treasury yields are near cycle highs. Even in the euro area, where falling inflation has allowed a sizeable cutting cycle, disinflationary base effects have now played out and the bar is high for inflation not to rise into year-end 2025, especially with rising credit and money supply. As we get a wave of new capex cycles driven by the AI build out, deglobalisation and increased geopolitical tensions, the age of a higher, more volatile inflation regime is at hand, making the case for a new, higher normal in yields.
A second headwind is fiscal vulnerability. Across the OECD, the median real 10-year yield is now positive (roughly 1-2%) vs where it was in the late 2010s, pushing up the effective cost of debt. In 2024, interest payments consumed about 3.3% of GDP, up from 2.7% in 2015–2019. This rise is projected to continue. OECD refinancing requirements have risen from US$7.5 trillion in 2019 in the OECD to $13tn in 2025, a product of higher borrowing and a shift towards issuance of shorter maturities during the pandemic. Central bank holdings are also drifting lower, forcing more supply onto more price-sensitive private balance sheets, ergo increasing term premia.
The recent UK OBR long-term forecast underlines the long-term risks. At end 2024, the UK government deficit stood at 5.7% of GDP, debt was around 94% of GDP, and the 10-year gilt yield at end June was 4.5%, the third-highest borrowing costs among advanced economies. The OBR notes debt has ratcheted higher over decades. Stabilising it now requires a primary surplus of about 1.3% of GDP, a sharp swing from the current -2.2% deficit, because debt dynamics have worsened: real borrowing costs now exceed real growth rates. Meanwhile, the UK increasingly looks like a country attached to a health service, with health spending plausibly doubling over the long term. Due to ageing, chronic illness and low productivity, NHS spending will rise from 7.9% of GDP in 2025 to 14.5% by the 2070s, according to the OBR. By then, debt is projected to be 270% of GDP.
The third headwind is weak growth, which offers little relief. Euro-area real GDP growth is projected at just 1.1% for 2026 (with significant downward revisions since January), despite rising public spending. In the US, 2026 growth is only 1.7% even after modest upward revisions since May. Of course, this is a late cycle environment, and recession risks may well derail these fair-weather consensus forecasts. The result, weak trend growth with higher real rates, is exactly the combination that worsens debt dynamics and keeps term premia elevated.
When inflation is sticky, deficits persistent and growth subdued, owning duration does not reliably cushion risk-asset selloffs in portfolios. Should the factors outlined above persist, bonds simply become a lower-returning version of equities. There is no meaningful diversification when the stock/bond correlation is as high as 60% (as it is in some countries, having risen materially since 2022). Duration no longer protects against recessions because policy cuts are too shallow, meaning the rally does not arrive when needed.
The conclusion is challenging for many portfolios: without a decisive improvement in inflation control, fiscal consolidation, or trend growth, government bonds will continue to struggle in their traditional role as portfolio shock-absorbers. The UK’s unattractive cycle of high yields, difficult budget decisions, and worsening growth could well spread.
If sovereign bonds are compromised, how should investors diversify defensively in portfolios? Bonds are roughly half of the global market portfolio so cannot be ignored entirely. So, what can investors do within fixed income? First, use options strategically. If traditional diversification is at risk, replacing direct exposure with call-replacement or protection strategies can provide more reliable and occasionally cost-effective volatility reduction. Second, more esoteric fixed income exposure may offer something genuinely differentiated, for instance, some emerging markets and some dollar-bloc developed markets. Either way, things must change for everything to stay the same.
ENDS
