AI Boom Funding Creates Mispricing in Data Centre Debt Market
A surge in borrowing for AI infrastructure has led to a structural gap in data centre debt. Investors are receiving higher yields on secured debt compared to unsecured bonds from the same companies.
The rapid expansion of artificial intelligence is driving unprecedented demand for data centre infrastructure, leading technology giants to flood the bond market. Outstanding debt for AI-related borrowers, including hyperscalers and cloud providers, is growing significantly faster than last year, with companies issuing a range of debt instruments to fund new facilities.
Asset-backed securities (ABS) tied to data centres are becoming a prominent part of this financing. These securities pool cash flows from data centre leases and use them to service bonds, with the underlying real estate and equipment as collateral. Hyperscalers are estimated to spend up to $800 billion on AI infrastructure this year alone, propelling data centre securitisation into a rapidly growing segment of structured credit.
However, the sheer volume of capital required has exposed a pricing anomaly. Investors lending via secured data centre ABS are currently being offered a higher yield spread compared to unsecured corporate bonds issued by the very same firms. This disparity exists even for investment-grade rated debt, with AA-rated data centre ABS offering significantly higher yields than AA-rated unsecured hyperscaler debt.
Man Group suggests this yield pickup indicates a "liquidity and complexity premium" rather than a fundamental credit risk. The ABS market is typically less liquid and more complex than the corporate bond market, leading investors to demand extra compensation for holding these less familiar securities. This premium compensates for structural complexity and market absorption challenges, not necessarily increased default risk.