Barclays forecasts US 30-year yield could hit 6% amid productivity growth

The US 30-year Treasury yield hasn’t started with a six-handle since Bill Clinton was in office. Barclays thinks that’s about to change.

In a note published September 29, Anshul Pradhan, the bank’s head of US rates research, argued that the 30-year yield could climb to 6% if artificial intelligence investment by major tech companies translates into sustained productivity gains across the economy. The last time yields breached that level was June 2000, during the final stretch of the dot-com boom.

The case for 6%

The 30-year yield already sits at 5.61% as of September 28, its highest reading since 2002. That represents a full percentage point increase from its March 2026 low of 4.61%, a 100-basis-point move in roughly six months.

Pradhan’s argument hinges on what happens next with AI capital expenditure. According to the Barclays note, US technology giants are expected to spend as much on AI infrastructure this year as they did in the previous three years combined.

The core logic works like this: higher productivity growth means the economy can sustain faster expansion without overheating. But it also means the Federal Reserve’s so-called neutral rate, the interest rate that neither stimulates nor restricts the economy, would need to be permanently higher.

Right now, markets are treating elevated rates as a cyclical phenomenon. Pradhan’s thesis flips that assumption. If AI genuinely supercharges productivity, today’s high rates aren’t a temporary condition. They’re the new baseline.

What a productivity boom means for Fed policy

Pradhan’s note suggests that if productivity gains materialize at scale, markets would need to rethink their entire framework for where rates settle over the long term. The expectation of eventual normalization back to pre-pandemic levels would essentially be off the table.

The other side of the trade

Barclays isn’t presenting this as a certainty. The note also acknowledges a scenario where AI capital spending slows down, either because companies pull back on investment or because the productivity gains simply don’t show up in the data.

In that case, Pradhan sees a potential rally in intermediate Treasury bonds. Put differently: if the AI bet doesn’t pay off economically, yields could reverse course, particularly in the five-to-ten-year range where rate expectations are most sensitive to growth outlooks.

For holders of long-duration bonds, the math is unforgiving. A move from 5.61% to 6% on a 30-year bond implies meaningful price declines on existing holdings. Pension funds, insurance companies, and anyone with long-dated fixed-income exposure would feel the pain directly.

What makes this moment unusual is the speed of the move. A 100-basis-point increase in six months on the 30-year bond reflects a market that is actively reconsidering foundational assumptions about growth, inflation, and the trajectory of monetary policy.