Alphabet selling very rare 100-year bonds to help fund AI investment
Summary
Alphabet is issuing rare 100-year bonds to help finance its substantial investments in artificial intelligence infrastructure, following a trend among Big Tech companies. These firms, including Amazon and Meta, are projected to invest nearly $700 billion in AI infrastructure this year, increasingly relying on debt markets. Alphabet's long-term debt is expected to reach $46.5 billion in 2025, a fourfold increase from the prior year, despite holding $126.8 billion in cash and equivalents. The company recently reported annual sales exceeding $400 billion and plans to spend up to $185 billion on capital expenditure this year, doubling last year's total, driven by demand for its Gemini AI assistant. Investor demand for Monday's offering was strongest for shorter-term bonds, with a three-year offering pricing at 0.27 percentage points above US Treasuries.
Key takeaway
For investors evaluating Big Tech's financial health, you should scrutinize the increasing debt loads and capital expenditure plans, particularly those tied to AI infrastructure. Alphabet's projected $46.5 billion long-term debt in 2025, alongside its $185 billion capex for AI, signals a strategic shift towards debt financing. Consider the implications of these complex financial obligations on long-term returns and market exposure, especially as yields on new bond offerings may not always compensate for the added risk.
Key insights
Big Tech is increasingly using debt to fund massive AI infrastructure investments, impacting financial obligations.
Principles
- AI infrastructure demands significant capital.
- Debt markets are a primary funding source.
In practice
- Monitor Big Tech capex for AI growth.
- Evaluate bond yields against market exposure.
Topics
- AI Investment
- Debt Financing
- Capital Expenditure
- Bond Markets
- Gemini AI
Best for: Investor, Business Analyst, Executive
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by AI - Ars Technica.