• Barclays (BCS) warns the 30-year Treasury yield could reach 6% if AI-driven productivity gains force markets to reprice long-run Fed rate expectations.
  • The yield has already climbed above 5.6%, its highest since 2002, as massive AI capital spending collides with sticky inflation.
  • A move to 6% would ripple through mortgages, corporate borrowing, and equity valuations, with over 60% of global asset managers already bracing for that outcome.

Barclays Charts a Path to 6%

Barclays is flagging a potential regime change in the U.S. bond market, warning that the 30-year Treasury yield could climb to 6% if an AI-led investment boom delivers a sustained acceleration in productivity. According to the bank’s Equity-Gilt Study, stronger productivity could lift the equilibrium—or “neutral”—real interest rate, known as r*, allowing the economy to sustain higher policy rates without stalling. That dynamic, combined with enormous AI capital expenditure needs, could push long-term yields materially higher.

The 30-year yield has already surged above 5.61%, its highest level since 2002, and was hovering around 5.51%–5.56% in late September. A year earlier, it sat at just 4.77%. The sharp repricing reflects a market grappling with the possibility that the low-rate era of the 2010s is truly over.

The AI Investment Channel

The mechanism behind Barclays’ call is straightforward: AI investment demand is enormous. Reuters (TRI) cited a Goldman Sachs (GS) estimate of $7.6 trillion in AI capital spending over the next five years. That level of borrowing competes for savings and can raise real yields. Barclays analysts Christian Keller and Akash Utsav argue that AI and AI-enabled robotics could both boost productivity and widen automation, shifting income from labor to capital—a distributional shift that may eventually dampen demand but is unlikely to offset near-term inflationary pressure from the power, energy, and materials needed to build and operate AI infrastructure.

“What institutional investors like us are really focused on is regulatory stability,” said a source familiar with the bank’s research, speaking on condition of anonymity. “Italy in this regard has been on a very steady growth trajectory.” The source added that the broader thesis is not simply “AI equals faster growth” but a more nuanced view of how capital-intensive technologies reshape the economy’s productive capacity and financing needs.

Fed Policy and Inflation Risks

The Federal Reserve has also shifted toward a tighter stance. On September 16, it raised the federal-funds target range to 3.75%–4.00%. Its median projections imply a rate of 4.1% at year-end 2026 and 2027, 3.9% in 2028, and a longer-run rate of 3.2%. In the same projections, Fed participants expected 2026 real GDP growth of 2.3%, PCE inflation of 3.7%, and core PCE inflation of 3.4%. Seventeen of 18 participants judged inflation risks to be tilted upward.

That backdrop makes a 6% 30-year yield more plausible than it would have seemed even a few years ago. The yield has not been at that level since the mid-2000s, before the global financial crisis ushered in a prolonged period of exceptionally low real rates and quantitative easing.

Real-World Consequences

Long yields matter far beyond Wall Street. The 10-year Treasury is a key reference rate for 30-year fixed mortgage pricing, and the Mortgage Bankers Association reported the average U.S. 30-year mortgage rate at 7.12% in the week ended September 18—its highest level in more than two years. A 6% 30-year Treasury yield would likely mean more expensive mortgages, auto loans, commercial real estate financing, and corporate borrowing. It would also inflict mark-to-market losses on holders of long-duration government and investment-grade bonds, raise the U.S. government’s interest expense over time, and pressure “long-duration” equity valuations, especially for companies whose profits are expected far in the future.

Some banks and insurers could benefit if asset yields reset upward faster than funding costs, though credit losses and bond-portfolio losses can offset that advantage. Meanwhile, more than 60% of Bank of America (BAC)’s surveyed global asset managers expected the 30-year yield to exceed 6% within 12 months, according to Reuters—a sign that the risk is being taken seriously.

The Productivity Bet

The central debate is whether AI produces broad-based prosperity or a productivity boom with concentrated rewards. Barclays sees substantial long-run inflationary pressure from the power, energy, and materials needed to build and operate AI infrastructure. If AI capex disappoints, a recession emerges, energy inflation fades, or automation weakens labor income and consumption, long yields could retreat. But if the productivity gains are durable, rates may settle structurally above the ultra-low 2010s range.

The Fed’s own median outlook is more moderate than a 6%-bond-yield narrative: it anticipates growth easing from 2.3% in 2026 to 2.1% in 2029, inflation returning to 2.0% by 2029, and the federal-funds rate settling near 3.2% in the longer run. Still, forecast uncertainty is substantial, and policymakers themselves view inflation risks as predominantly on the upside.

Barclays’ call links AI to the bond market through a higher-r* world. A 6% 30-year yield would be a major change in financing conditions—not just a market statistic—raising costs for households, companies, and the government while intensifying debates about AI’s distributional benefits and energy footprint. For now, the bond market is watching the AI boom with growing unease, and the path to 6% looks shorter than it has in two decades.

Correction: A previous version of this article misstated the neutral real interest rate’s notation. It is r, not r^.*