The AI buildout is creating a significant demand for capital, with hyperscalers and specialized AI ventures issuing record amounts of new debt to fund data centers, chips, and energy grids. This surge in corporate borrowing, which has seen offerings reach almost $1.5 trillion year-to-date, up 36% from the previous year, is happening concurrently with governments, particularly the US, issuing record volumes of sovereign debt due to skyrocketing budget deficits and increased defense spending. This dual demand for capital is leading to a clash in credit markets, with some investors questioning if the traditional "crowding out" theory, where government borrowing stifles corporate funding, is now being reversed or intensified.
The competition for capital is resulting in higher yields for both corporate and sovereign debt. The US national debt is nearing $40 trillion, and the federal government is projected to be $1.8 trillion in the hole for fiscal year 2026, necessitating substantial capital raising. This has led to concerns about rising borrowing costs for the US Treasury, particularly given recent 30-year bond sale yields hitting a 25-year high. Market participants are also warning that unpredictable debt management by the US Treasury could further push up borrowing costs.
While some analysts believe the market for ultra-safe government debt is somewhat distinct from the higher-yielding, higher-risk tech debt, the sheer scale of borrowing from both sectors means there is significant overlap and competition for global capital. The AI sector alone is expected to see $6 trillion in spending between 2026 and 2027 to build out its infrastructure, with three major banks recently announcing over $2 trillion in investment consortiums for AI funding in a single week. This massive capital requirement, likened to the historic buildout of railroads and the International Space Station combined, raises questions about the overall demand for global debt and the potential for market resilience amidst these competing demands.
