How AI Debt Is Reshaping Credit Markets

Episode Summary

Executive Summary: The episode examines how AI-driven borrowing by hyperscalers and the broader AI ecosystem is becoming a dominant force in credit markets. Amanda Lynham and Zach Ablon argue the wave is multi-year, highly concentrated, and likely to spill beyond investment grade into high yield, private credit, project finance, and regional markets as investors wrestle with duration, concentration, and saturation risks.

Main Topics: AI debt supply as the dominant credit-market theme: The speakers argue AI-related borrowing is now the central story in credit because of its unprecedented scale, multi-year duration, and concentration in a single sector. Why cash-rich tech firms are issuing debt: Despite strong balance sheets, hyperscalers are issuing debt to pre-fund a multi-year CapEx cycle and preserve flexibility across capital sources. Investor demand, concentration, and duration concerns: Zach describes growing investor indigestion, especially among insurance and real-money accounts, as AI supply pushes more long-duration paper into portfolios. Forecasts for hyperscaler financing and market capacity: Amanda outlines expected debt financing shares of CapEx through 2027 and argues the binding constraint is not access to capital but how much the IG market can absorb. Spillover into other markets and structures: The discussion expands beyond IG to private credit, high yield, leveraged loans, foreign currency markets, project finance, and structured credit as funding sources. Risks to monitor: CapEx, concessions, and theme performance: Both guests identify slowing CapEx growth, widening new-issue concessions, and any AI-theme wobble as key signals for future credit performance.

Key Arguments: AI-related debt supply is unusually important because it is large, persistent over several years, and concentrated in one theme rather than a short-lived re-leveraging event. Tech firms are using debt not because they lack cash, but because they are getting ahead of a prolonged investment cycle and optimizing capital structure. Hyperscalers remain highly rated and low-levered, leaving room to add debt while staying within IG parameters. Investor appetite is weakening at the margin, especially for long-duration issuance, as shown by wider spreads and softer demand from insurance buyers. The U.S. IG market alone may not be the right or only venue; private markets, foreign bond markets, and structured finance will likely absorb a significant share of financing. AI-related financing will affect high yield too, not just IG, because project-finance and data-center deals are already pushing average deal size higher. A key risk is not access to capital but market saturation and issuer concentration limits, which may force issuance into other markets or structures. If CapEx growth slows, credit could outperform even if equities weaken, so credit is not a perfect one-way hedge on AI exposure.

Data Points: Hyperscaler debt issuance in 2025: $108 billion - Amanda cites global debt issuance by the hyperscaler group in 2025. Hyperscaler debt issuance year-to-date in 2026: $194 billion - Amanda notes issuance has already accelerated significantly in 2026. Estimated AI-related debt issuance tracked year-to-date: Nearly $500 billion - Amanda says this captures the broader AI ecosystem, not just hyperscalers. 2025 hyperscaler debt as share of CapEx: 27% - Amanda frames debt financing as a percentage of hyperscaler CapEx. 2026 expected debt-financed share of CapEx: About 33% - Amanda’s forecast for hyperscaler CapEx financing this year. 2027 expected debt-financed share of CapEx: 35% - Amanda expects debt financing to peak next year. Potential direct hyperscaler supply in 2026: Around $250 billion - Amanda’s estimate excluding project finance. Potential incremental debt capacity at 2x net leverage / 3x gross leverage: Around $2 trillion - Amanda’s illustrative capacity calculation for hyperscalers while staying IG-like. Incremental debt capacity if hyperscalers were sized like U.S. banks: $510 billion - Amanda uses U.S. bank issuers as a case study for IG market absorption. U.S. IG AI-related supply in 2024: $10 billion - Zach says AI-related issuance was about 1% of IG supply. U.S. IG AI-related supply in 2025: $108 billion - Zach says this represented about 7% of year-to-date supply. U.S. IG AI-related share in 2026: About 18% - Zach says AI-related issuance is now a much larger part of the market. Share of 15-year-plus IG issuance from AI companies/funders in 2026: 40% - Zach highlights the duration impact of AI financing. AI leadership basket spread move over 12 months: From 74 bps tight to nearly double - Zach cites market indigestion reflected in pricing. Insurance buyer orders in Q1: ~15 insurance clients with $50M+ orders - Zach describes strong demand in the third-year part of large AI-related bond deals. Insurance buyer orders by end of Q2: About half as many as Q1 - Zach says demand softened materially. Dry powder across private markets: $4.5 trillion - Amanda cites global private-market firepower across credit, infrastructure, real estate, and private equity. Expected project finance/data-center transactions in 2027: $300 billion - Amanda says this is above and beyond direct hyperscaler issuance. Largest U.S. IG issuers’ index-eligible debt: No more than $175 billion each - Amanda uses large U.S. banks as a comparison for market capacity. Largest U.S. IG issuers’ index weight: No more than 2.3% - Amanda’s benchmark for what the IG market can tolerate. New-issue concessions before indigestion: 2-3 bps - Zach cites very small concessions on AI deals earlier in the cycle. New-issue concessions at peak stress: Up to 20 bps - Zach notes concessions widened on a large hyperscaler deal. Average yield on high-yield double-B universe: 165 bps over - Zach compares this to wider data-center deals with IG wrappers.

Pivotal Quotes: "It’s hard to overstate the importance of this theme in the credit markets, both in terms of its overall scale and the amount of supply, but also in the multi-year nature of the issuance." — Amanda Lynham: She explains why AI debt supply is more significant than prior, shorter-lived credit re-leveraging episodes. "We are not concerned about access to capital for this theme. We are just expecting there to be a more nuanced conversation about where to issue and at what price." — Amanda Lynham: She distinguishes financing availability from market absorption and pricing constraints. "The marketplace has clearly expressed some indigestion with some of this financing, and we’ve seen that in prices." — Zach Ablon: He summarizes widening spreads and weaker investor demand for AI-related paper.

Implications: AI financing is likely to reshape credit allocation, duration, and sector concentration for years. Investors should watch CapEx, spread behavior, and deal structure, while issuers may increasingly rely on private markets, foreign currency debt, and project finance.