Financial leaders at the Global Financial Leaders’ Investment Summit expressed sharply contrasting views on the sustainability of massive AI infrastructure spending. High-profile asset managers, including PIMCO CEO Emmanuel Roman, warned of an impending market downturn driven by unconstrained debt financing, estimating that the AI buildout could require $2 trillion in capital over the next five years.

Despite public warnings of a potential market correction, institutional demand for AI-related debt offerings remains strong. Tech giants have tapped debt markets aggressively, with Meta attracting $58 billion in orders for an upcoming bond sale, while Oracle, Alphabet, and Microsoft have issued multibillion-dollar bond packages to finance data center infrastructure and hardware capacity.

Uncertainty centers on long-term unit economics and pricing power for AI services. Major lenders like Apollo Global Management noted that while aggregate demand for computing power will expand, the structural return on investment for hyperscalers and non-hyperscale debt issuers remains unclear.

Why it matters

  • Heavy reliance on corporate bond markets to fund AI capex exposes tech giants to macroeconomic refinancing risks.

  • Uncertainty around AI service unit economics creates volatility risks for late-stage startups relying on continuous debt-funded compute.

  • PIMCO’s warning highlights growing institutional concern over debt market oversaturation driven by AI infrastructure projects.

Source: theverge.com