Key facts
- Tech giants are issuing debt in multiple currencies to manage AI-related spending.
- Amazon and Alphabet have collectively issued $60 billion in bonds across various currencies in the last year.
- Amazon's March euro bond issuance set a record at €14.5 billion.
- Alphabet achieved record borrowing levels in yen, Canadian dollar, Swiss franc, and sterling.
- New financing structures, like data center lease-backed deals, are being used for AI startups and operators.
- Hyperscaler capital expenditures are projected to reach approximately $725 billion this year.
As the demand for artificial intelligence infrastructure continues to drive massive corporate borrowing, major technology companies are increasingly looking beyond U.S. dollar markets to fund their expansion. Giants like Amazon.com and Alphabet have issued approximately $60 billion in bonds across various currencies over the past year, seeking to tap into a broader investor base and avoid saturating the U.S. market with debt.
These large-scale issuances have significantly impacted global bond markets, with companies setting new records for bond sales in euros, sterling, and yen. Amazon's €14.5 billion deal in March was the largest ever in the euro corporate bond market, while Alphabet achieved borrowing records in multiple currencies and issued the first 100-year tech bond since 1997. These moves highlight the substantial funding needs of hyperscalers, whose capital expenditures are estimated to reach around $725 billion this year, a figure nearly double that of mid-2025 projections.
Beyond traditional bond issuances, banks are also innovating with new financing structures for AI startups and data center operators. One such method involves structuring deals around pre-arranged data center leases, which provide greater visibility into future cash flows. An example is Stingray Compute's recent $810 million note, backed by a lease to Amazon, which was nine times oversubscribed. These lease-backed deals, inspired by construction loans, have seen about 15 sales to high-yield investors since last year.
Despite the surge in AI-related debt issuance, which now constitutes about 15% of total investment-grade debt in the U.S., investor appetite remains strong. Analysts note that these are high-quality rated bonds with significant liquidity. However, some market participants are beginning to question the market's capacity to absorb such large volumes continuously. Morgan Stanley anticipates that AI debt could push investment-grade issuance above $2 trillion for the first time in 2026. Nevertheless, the ongoing investments in long-term AI projects by hyperscalers suggest a sustained need for funding, keeping the pipeline open.