What the AI investment boom means for bonds

Key takeaways

• Treasury yields rise amid shifting views on policy, inflation and growth
• AI investment may lift real rates, but Treasury crowding-out evidence remains limited
• Strong fundamentals and demand support rising AI-related credit issuance

Bond markets continue to adjust to a more hawkish policy environment following the Federal Reserve’s recent 25-basis-point rate hike.

Since July, the two-year Treasury yield has risen from roughly 4.2% to 4.9%, while the 10-year yield has climbed from around 4.4% to 5.1%. The broad move higher suggests investors are reassessing both the level and potential persistence of elevated policy rates.

Our Co-Head of Global Fixed Income Riti Samanta sees several forces behind the rise, including tighter monetary policy expectations, uncertainty around inflation and the extraordinary scale of AI-related investment. One increasingly prominent question is whether the financing required for data centers, computing capacity, energy and other AI infrastructure is competing directly with Treasury issuance and pushing government borrowing costs higher. The answer is more nuanced.

Traditional crowding out occurs when heavy government borrowing pushes interest rates higher and displaces private investment. The current situation counters that relationship, suggesting private-sector AI borrowing is raising the cost of government financing. So far, however, there is limited evidence to support that explanation.

Instead, the more important connection may run through the real economy. AI infrastructure requires labor, power generation, equipment, construction materials and significant capital investment. If investment demand rises faster than available savings, real interest rates may need to increase to bring the economy back into balance. In other words, AI investment could contribute to higher yields without investors necessarily choosing corporate bonds over Treasuries.

Investors are also watching efforts by the U.S. Treasury to address pressures in longer-dated government bonds.

Earlier this month, the Treasury expanded its long-end buyback program, increasing potential purchases in the 10- and 20-year sectors to as much as $6 billion, three times the amount originally communicated when the program was announced.

The initiative is designed to improve market functioning and provide support for longer-dated Treasuries. However, the market reaction was relatively muted.

Demand for the operation itself was strong, but long-term yields moved higher following the announcement. Investors appear to have concluded that the scale of the program was not sufficient to materially change the broader supply-and-demand dynamics affecting the Treasury market.

The scale of AI investment is becoming even more visible in corporate credit.

Investment-grade corporate issuance has reached approximately $1.4 trillion year to date, up roughly 36% from the same period last year. Despite that surge in supply, investment-grade spreads have remained remarkably stable at around 78 basis points, little changed from the beginning of the year.

The resilience reflects more than strong demand alone. Much of the AI-related issuance has come from some of the largest and highest-quality technology companies. Most hyperscaler financing has been concentrated among issuers rated A or higher, supported by substantial earnings power and balance-sheet flexibility.

Corporate fundamentals also remain strong. Profitability has continued to improve and interest coverage remains well above broader investment-grade averages. While cash balances have declined as companies fund major capital expenditure programs, leverage remains favorable compared with the wider corporate universe.

At the same time, demand for investment-grade credit from U.S. and international investors has remained robust, helping the market absorb the additional supply without significant spread widening.

There is also more to the AI financing story than investment-grade corporate debt.

Data centers and other AI infrastructure are increasingly being financed through high-yield and securitized markets, including asset-backed securities and commercial mortgage-backed securities.

As the infrastructure buildout continues, these markets could play an increasingly important role in financing AI investment and broaden the ways in which the cycle affects fixed income portfolios.

For investors, the distinction matters. Higher yields increasingly reflect a combination of tighter Fed policy, resilient economic growth and a powerful investment cycle centered on AI. While that investment may be putting upward pressure on real rates, the evidence currently points more toward an impact on economic activity and equilibrium interest rates than direct crowding out of Treasury demand.

As the year enters its final quarter, Treasury supply, Federal Reserve policy and the pace of AI capital expenditure will remain important factors to watch across global fixed income markets.

Articles of the week:

Are bigger events changing transition management?

SpaceX: What the IPO revealed about managing the private to public portfolio transition

What’s really driving the rise in Treasury yields?