Amazon Returns to the Bond Market for $25 Billion to Fund AI Build-Out
Amazon has once again turned to the debt markets, raising at least $25 billion—the largest bond sale of the year for the company. This substantial funding is aimed at fueling its artificial intelligence (AI) initiatives.
The Sale Details:
- Size: Eight tranches totaling at least $25 billion.
- Maturities: 2029 to 2066.
- Purpose: Funding data centers, custom silicon development, and physical infrastructure required for AI advancements, primarily through Amazon Web Services (AWS) and its Trainium chip program.
This offering follows a pattern set by Amazon's previous record bond sale in Canadian dollars earlier this year, part of a borrowing spree that has reached over $70 billion since 2025 across various currencies.
Investor Demand and Market Reaction:
- Initial investor interest was high, peaking at around $62 billion.
- However, as terms firmed up, demand cooled, resulting in a final book size of approximately $41 billion (1.6 times the deal size).
- Amazon has indicated it does not plan to issue any more debt for 2026, which could provide reassurance to investors concerned about rising leverage.
Capital Spending and Free Cash Flow:
- Amazon projects roughly $200 billion in capital spending for 2026, the highest among hyperscalers, significantly compressing its free cash flow.
- The four largest US tech firms collectively anticipate over $650 billion in AI capex this year, surpassing their operating cash flows.
Market Dynamics:
- Some analysts note that demand for this sale was relatively muted compared to previous offerings, suggesting that bond buyers are becoming more selective as the market becomes saturated with AI-related debt.
- Deals like Cipher's junk bonds, funding an Amazon data center in Texas, are testing investors' appetite for riskier investments in the boom era.
Amazon structured this offering with a mix of fixed and floating-rate notes, providing flexibility across a four-decade curve. Its high credit rating allows it to borrow at low costs, despite the increasing debt totals.