Notes from the Desk: Priced for Perfection
Key Points
- Hyperscalers have issued $110 billion in US debt year-to-date, representing nearly 16% of total IG issuance versus just 3% a year ago
- AI capex from the big four alone will account for 2.2% of GDP in 2026, with debt markets absorbing much of the financing need
- Hyperscaler spreads already trade 25+ bps wider than the broader IG index (a 10-year high), suggesting markets are beginning to differentiate risk despite tight overall spreads offering little downside protection
AI Summary
Market Summary: AI Spending Strains Corporate Debt Markets
Key Developments
The four major tech hyperscalers—Meta, Microsoft, Alphabet, and Amazon—plan to spend at least $700 billion on AI infrastructure in 2026, representing an 80% increase from 2025's record levels. This capex amounts to 2.2% of U.S. GDP from just these four companies.
Debt Market Impact
Investment-grade (IG) corporate bond issuance has reached $976 billion through May, surpassing previous record paces. Hyperscalers are driving this surge, with Alphabet, Amazon, Meta, Microsoft, and Oracle issuing approximately $110 billion in U.S. debt year-to-date—accounting for nearly 16% of total IG issuance, compared to just 3% a year ago.
Market Conditions
Despite massive supply, IG spreads remain near 80 basis points, the tightest levels since the mid-1990s. Institutional buyers including pensions and insurance companies continue absorbing supply, attracted by all-in yields above 5%. However, hyperscaler spreads now trade more than 25 basis points wider than the broader IG index—a 10-year high—suggesting the market is beginning to differentiate credit quality.
Risk Assessment
The analysis warns that spreads are "priced for perfection" with minimal margin for error. Potential catalysts for spread widening include increased M&A activity, disappointing AI investment returns, or supply indigestion. The current tight spread environment offers limited downside protection despite attractive all-in yields.
Investment Implication
While yields remain appealing, the report recommends seeking better risk-adjusted compensation in other sectors, anticipating an eventual market correction—timing uncertain but likely within a three-month to three-year window.
Model Analysis Breakdown
| Model | Sentiment | Confidence |
|---|---|---|
| GPT-5-mini | Bearish | 76% |
| Claude 4.5 Haiku | Bearish | 75% |
| Gemini 2.5 Flash | Bearish | 75% |
| Consensus | Bearish | 75% |