AI borrowers face tough sell in risky corners of US credit market
Key Points
- AI infrastructure supply in high-yield debt reached $40 billion in 2026 versus $12 billion for all of 2025, with demand concentrated in double-B-rated (near investment grade) issuers rather than lower-rated borrowers.
- Even BB+ rated AI issuers like SoftBank are paying 8.625-9.75% yields on bonds, with lower-rated borrowers facing potential costs of 14-15%, reflecting investor concerns about asymmetric risk-return profiles in AI debt.
- Zenith Arc's $1 billion bond offering in August initially dropped nearly seven points below issue price, illustrating market skepticism toward AI borrowers that require heavy upfront investment before generating reliable cash flow.
AI Summary
Summary: AI Borrowers Face Tough Sell in Risky US Credit Markets
AI-related debt issuance in leveraged finance surged to $88 billion in 2026, up dramatically from just $20 billion in the first 11 months of 2025, according to Goldman Sachs. However, investors are increasingly skeptical of lower-rated AI borrowers with unproven revenue streams, demanding higher compensation for increased risk.
Key Market Dynamics:
Investor appetite remains concentrated in higher-quality, near-investment-grade borrowers (BB-rated), particularly those with predictable revenue, long-term contracts, and tangible assets. Data center companies have driven much of the high-yield market activity, with AI infrastructure supply reaching $40 billion in 2026 versus $12 billion for all of 2025, per BNP Paribas data.
Borrowing Costs:
Even BB+-rated issuers face yields of 9-10%, with lower-rated borrowers potentially paying 14-15%. SoftBank Group's recent offering exemplifies this: despite its BB+ rating, it paid 8.625% on 3.5-year notes, 9.25% on 5.5-year debt, and 9.75% on 7.5-year bonds—rates typically associated with much lower-rated companies.
Market Concerns:
The Zenith Arc case highlights investor caution: its $400 million bond offering priced at 99.50 with an 8.875% coupon but dropped nearly three points immediately after issuance, ultimately falling over seven points below issue price by late August.
Collateralized loan obligation (CLO) managers, major buyers of leveraged loans, are becoming more cautious due to AI companies' heavy upfront investments before generating reliable cash flow. Portfolio rules limiting exposure to riskier borrowers further constrain demand. Credit investors emphasize the asymmetric risk-reward profile, where returns are contractual but risks include high leverage, project delays, and technological obsolescence.
Model Analysis Breakdown
| Model | Sentiment | Confidence |
|---|---|---|
| GPT-5-mini | Bearish | 80% |
| Claude 4.5 Haiku | Bearish | 72% |
| Gemini 2.5 Flash | Bearish | 85% |
| Consensus | Bearish | 79% |