AI-related corporate bond issuance has surged to $489 billion this year, already surpassing Goldman Sachs' estimate of $322 billion for all of last year. This significant increase is largely driven by major tech companies. For instance, Amazon has raised $53 billion in funding this year, including $37 billion in U.S. dollar bonds and $16 billion in euro-denominated bonds. Alphabet secured $20 billion through bonds, while Oracle borrowed $25 billion. Meta also made substantial issuances this year, following a $30 billion offering last year.
Goldman Sachs indicates that hyperscalers directly account for 40% of the $489 billion in AI-related debt, with the remainder coming from data center financing and other parts of the tech ecosystem. The investment bank expects a diverse range of financial markets, including syndicated credit, private markets (especially infrastructure, with nearly $200 billion in data center deals completed since early 2025), project finance-style joint ventures, and various bond markets, to meet the multi-year funding needs for AI infrastructure expansion.
Despite the current market's comfort with the additional debt, given many tech companies' strong cash flow, concerns are emerging. Oracle's long-term debt, for example, ballooned to $149 billion at the end of its last fiscal year from $96 billion a year earlier, while its cash balance was only $31.3 billion. Barbara Doran, CEO of BD8 Capital Partners, highlighted the significant execution risk and customer concentration issues, noting that Standard & Poor's downgraded Oracle's credit rating to 'BBB-', just one notch above junk status. She warned that Oracle is making a substantial bet on sustained demand and capacity expansion, but the payoff will take time.
The rapid increase in AI-related debt contrasts with rising long-term borrowing costs, which are at levels not seen since the financial crisis. A potential 5% risk-free rate could challenge the AI investment boom. Analysts suggest that a "day of reckoning" could arrive by 2027 if the monetization of AI does not meet expectations, or if the Federal Reserve implements several interest rate hikes, creating significant headwinds for companies with mounting debt.
