The verdict of August has been delivered at the far end of the curve. The thirty-year US Treasury has breached 5.3%, its highest level in nineteen years; the thirty-year gilt sits near 5.75%, unseen since 1998; the thirty-year JGB has printed above 4%, a level without precedent in the instrument’s history; the Bund is at a fifteen-year high and the OAT at a post-2008 one. Four sovereigns, four fiscal positions, four monetary regimes — one direction. This is not idiosyncrasy; it is a synchronised repricing of the term premium, and the causes are worth separating carefully, because they carry radically different implications.
The benign reading comes first. If artificial intelligence delivers the productivity gains its capital expenditure implies, potential growth rises, and with it the equilibrium real rate; a higher r* is simply the price of a richer future. There is something to this. Yet notice the sleight of hand: the investment boom absorbs global savings today, whether or not the productivity arrives tomorrow. Markets are pricing the capex, not the returns. Add rearmament across NATO, the energy transition, and the demographic reversal that turns ageing societies from net savers into net dis-savers, and the savings-investment balance shifts against bondholders on three fronts at once. Thus the real- rate story is real — but it is a story about scarcity of capital, not abundance of growth.
The second cause is less flattering. The United States is running a deficit near 7% of GDP at full employment, with debt above 120% of GDP and past $40 trillion in absolute terms; the United Kingdom has bound itself to fiscal rules that markets increasingly treat as a hostage rather than an anchor; Japan is expanding fiscally at precisely the moment its central bank withdraws. That Washington has doubled its liquidity-support buybacks is the tell: a Treasury managing the symptom rather than the diagnosis. Bond vigilantism, pronounced dead in 2020, has been exhumed.
Third, inflation. Energy shocks transmitted through the Strait of Hormuz, US inflation still at 3.4%, British CPI back at 2.9%, Japanese prices accelerating for a second consecutive month — and, more corrosively, sustained political pressure on central bank independence. Investors are not merely raising expected inflation; they are charging insurance against the possibility that it will be tolerated. That premium compounds over thirty years.
Three further mechanisms deserve mention. The price-insensitive buyer has vanished: quantitative tightening, BoJ normalisation and the maturation of British defined-benefit schemes have removed the bid that suppressed duration risk for a decade. Japan, the world’s creditor of last resort, is repatriating capital as domestic yields finally compete. And bonds no longer hedge equities reliably, so duration must pay for itself.
The implication is conditional but stark. Should AI productivity materialise, higher nominal rates will be validated by higher nominal growth and the arithmetic holds. Should it not, governments will face r comfortably above g with no cyclical rescue available. Prudent finance ministers would plan for the second case. Few are.
