Wall Street used to worry that too much U.S. debt would crowd out the private sector. But AI hyperscalers are ‘reverse crowding’ the Treasury

For years, Wall Street had concerns that excessive U.S. debt would crowd out private sector investment. However, the rise of AI hyperscalers is reversing this trend. The national debt now stands at $40 trillion, with a projected $2 trillion deficit for the fiscal year and $1 trillion in debt servicing costs. These figures represent significant capital that the Treasury Department must raise from the bond market, which also funds corporate giants.

Despite this, AI companies have been issuing substantial amounts of their own debt to fund chip purchases, data center construction, and other infrastructure needs. Treasury Secretary Scott Bessent noted the eagerness of AI companies to issue debt, regardless of the borrowing costs. Corporate bond issuance, particularly for investment-grade companies, has surged, with July totals totaling about $1.7 trillion, well above last year's pace and on track to exceed $2 trillion.

This massive influx of corporate debt typically requires higher yields to attract buyers, but the demand for AI-related bonds has remained strong, keeping yields relatively low. As a result, the market has adjusted through higher Treasury yields themselves, causing capital to flow into corporate bonds and away from Treasuries. While the rise in Treasury yields has been attributed to various factors, including federal budget deficits, oil prices, and a robust economy, international capital flows have shown a preference for purchasing corporate bonds over Treasury debt.

This "reverse crowding out" in the corporate bond market has caught the attention of both Treasury Secretary and Fed Chairman Kevin Warsh. Markets are showing signs of fatigue, as hyperscalers pay higher premiums over risk-free bonds, indicating growing concerns about the rapidly increasing leverage from these issuers.