Key Highlights

• Rising Treasury yields partly reflect investor bets that R-star, a theoretical neutral interest rate, is moving structurally higher.

• The New York Fed's Laubach-Williams model estimates R-star at 1.65% as of the second quarter, up from 1.36% in early 2025.

• Analysts say heavy AI-driven bond issuance by hyperscalers is competing directly with Treasuries for investor capital, pushing yields higher.

• Some analysts believe the AI-driven increase in R-star could eventually reverse if AI proves disinflationary over the longer term.

A Theoretical Rate Gains New Market Relevance

A recent rise in Treasury yields reflects, in part, growing investor bets that an obscure theoretical rate consistent with stable growth and Inflation is moving structurally higher. The so-called R-star, a rate that neither stimulates nor restrains economic activity, may be higher than the latest official data indicates, according to analysts who point to the U.S. economy absorbing a strong mix of AI-related Capital spending alongside heavy government borrowing. A rising R-star suggests interest rates may settle at structurally elevated levels going forward, adding sustained pressure on bond prices even as the measure remains notoriously difficult to estimate precisely. One fixed income strategist attributed the likely rise in R-star directly to heavy AI-related Investment combined with elevated government Debt levels, both of which increase Demand for capital and lift real yields. The debate carries meaningful implications, since a genuinely higher neutral rate would point toward higher-for-longer interest rates affecting everything from Mortgage costs to government borrowing expenses, while also complicating the Federal Reserve's eventual path toward cutting rates even as current Bond Market pressure pushes prices lower.

An Inherently Difficult Measure to Pin Down

R-star has long been the domain of economists seeking to gauge how restrictive or accommodative Federal Reserve policy actually is relative to the underlying economy, but the measure remains notoriously difficult to estimate since it varies significantly depending on the model and time horizon used. The New York Fed's widely followed Laubach-Williams model recently updated its estimate through the second quarter of 2026, showing R-star at 1.65%, slightly down from 1.73% in the first quarter but meaningfully higher than the 1.36% level recorded in early 2025. Market Participants believe the true current neutral rate is likely higher than even this updated estimate suggests, given the ongoing surge in bond sales by hyperscalers, the largest Cloud Computing companies, competing directly with government borrowing for investor capital, though they acknowledge they cannot provide a precise alternative figure. One strategist described any R-star estimate as involving "equal parts art and science," while another argued plainly that a higher R-star is pushing rates higher across the entire yield curve.

Different Parts of the Curve Feel Uneven Pressure

A structural increase in R-star puts particular upward pressure on two- and five-year Treasury yields, since it implies the Federal Reserve will ultimately need to settle its policy rate at a higher long-term level. The 10-year yield faces additional upward pressure both from expectations of higher future policy rates and from investors demanding a larger term premium, the additional compensation required to hold longer-duration debt. For the 30-year yield specifically, one strategist described rising R-star as having a historically asymmetric impact, biting hardest on the longest end of the curve precisely because of elevated debt levels and large, persistent budget deficits.

AI-Driven Borrowing Compounds the Underlying Fiscal Pressure

The fiscal concerns driving yields higher are being compounded by a growing Supply of long-term corporate debt beyond Treasuries, particularly AI-driven borrowing from hyperscalers, as massive bond sales from technology companies compete directly with longer-dated government debt for the same pool of investor capital. One chief investment officer noted that the Federal Reserve cannot easily cut interest rates back toward zero when underlying structural demand for capital remains this elevated. Still, some analysts believe the technology-driven increase in R-star could ultimately prove temporary, if artificial intelligence eventually produces disinflationary or even deflationary effects on the broader economy over a longer time horizon, a dynamic one strategist described as a higher near-term R-star potentially flipping toward a lower nominal policy rate further down the road. As the AI investment boom continues unfolding, the U.S. government continues borrowing heavily, with the National Debt having crossed USD 40 trillion, creating an unusual combination of simultaneous, elevated public and private sector demand for capital that is likely to keep the neutral rate structurally higher even during an extended Federal Reserve policy pause.

Conclusion

The debate over R-star illustrates how an obscure, difficult-to-measure theoretical construct can carry real consequences for bond markets and borrowing costs across the economy, even as economists acknowledge genuine uncertainty about whether the current dynamics represent a lasting structural shift. Whether the AI-driven capital demand pushing R-star higher today eventually reverses into a disinflationary force, or instead cements a genuinely higher-for-longer rate environment, will likely take years of additional data to determine with confidence.