Quick Summary
Covering the latest research from top Wall Street investment banks

AI-related debt financing is heating up, with 2026 global supply potentially reaching about $570bn

Institution
Morgan Stanley
Date
2026-06-09
Authors
Fernanda Lima, Vishwas Patkar, Christina Sigler, Aron Becker, Ellie Dann, Jonathan Loke, Carolyn Campbell, James Egan, Catherine Liu, Yagyesh Modi, Vishwanath Tirupattur, Eva Baurmeister, Raquel Kanner, Kelvin Pang
Company
-
Ticker
-
Industry
AI Infrastructure/Data Centers
Rating
OW HY data center debt; data center securitized credit is attractive, especially ABS BBBs
NeutralLow confidenceAI-related debt issuance is expanding rapidly, and fundamentals remain strong, but near-term pricing is more driven by supply expectations; data center debt offers long-term demand, structural protection, and relatively attractive spreads.
AuthorsFernanda Lima, Vishwas Patkar, Christina Sigler, Aron Becker, Ellie Dann, Jonathan Loke, Carolyn Campbell, James Egan, Catherine Liu, Yagyesh Modi, Vishwanath Tirupattur, Eva Baurmeister, Raquel Kanner, Kelvin Pang
CoverageUnited States、Europe、Other
Business segmentsData Center Construction、Hyperscaler Capex、Non-USD Bond Issuance、Chip Financing、Data Center ABS and CMBS
Research firm divisions/subsidiariesMorgan Stanley(Other)

AI summary card

AI-related debt financing is heating up, with 2026 global supply potentially reaching about $570bn

Morgan Stanley believes the AI and data center financing cycle is accelerating, with issuance hitting a year-to-date high in April and shifting to non-USD markets in May, while supply is likely to continue expanding in 2H26.

No single-stock rating or target price; the strategy view is OW HY data center debt and bullish on data center securitized credit, especially ABS BBBs.
AI FinancingData Center DebtHyperscaler CapexHigh-Yield BondsSecuritized CreditABSCMBSChip Financing
  • As of May 31, 2026, AI-related global issuance was about $236bn, more than 4x the same period in 2025; full-year supply is forecast at about $570bn.
  • AI-related issuance exceeded $74bn in April, the highest this year; data center shell construction project financing accounted for 85% of AI-related HY supply and 40% of IG supply.
  • New issuance in the U.S. market was limited in May, but Hyperscalers issued about $24bn combined in non-USD markets such as EUR, CAD, CHF, and JPY, showing the investor base is expanding.
  • The report believes fundamentals remain strong, but near-term price action is mainly driven by supply expectations; despite elevated supply, spreads across asset classes have generally tightened since the end of 1Q.
  • Morgan Stanley is OW HY data center debt and believes data center securitized credit is attractive overall, with the strongest preference for ABS BBBs.

Report interpretation

Overview

This report tracks the AI-related debt financing cycle, covering publicly tradable debt across currencies including USD, EUR, GBP, CHF, CAD, and JPY. The core conclusion is that financing for data centers, Hyperscalers, and chips driven by AI infrastructure buildout is expanding from single-currency USD supply into multi-currency, multi-asset-class supply, and issuance may continue to accelerate in 2H26.

Core views

Morgan Stanley believes the fundamental support for AI-related financing remains solid, but current market pricing is more influenced by the supply pace and investor absorption capacity. Strong issuance in April, a slowdown in the U.S. market in May, and active non-USD issuance indicate that issuance windows and the investor base are rotating and broadening. HY data center debt is favored for its defensive credit characteristics, limited leasing risk, project-finance-style protections, and more attractive spreads; spreads on data center ABS and CMBS tightened in April and May due to a drought in supply, with ABS BBBs viewed as offering the most compelling value.

Analysis framework

The report combines cross-asset issuance tracking, supply technical analysis, structural credit term comparisons, and observation of securitization performance metrics. It compares AI-related debt instruments including IG, HY, leveraged finance, ABS, CMBS, and chip financing, while tracking issuance currency, issuance size, spreads, DSCR, cap rate, occupancy, and structural protection.

Methodology notes

  • Supply TrackingAI-Related Debt Issuance Tracking

    Track publicly tradable AI-related debt supply by asset class and currency

    The report breaks AI-related issuance down into IG, HY, securitized credit, and multi-currency issuance, with a focus on QTD, YTD, and full-year forecast supply.

  • Credit StrategyTechnicals Take Priority Over Fundamentals

    Near-term pricing is mainly driven by supply expectations

    The report believes fundamentals remain strong, but market pricing currently reflects supply pace, supply expectations, and investor absorption capacity more than fundamentals.

  • Structural Credit AnalysisData Center ABS and CMBS Performance Metrics

    Assess securitized credit quality through DSCR, cap rate, occupancy, and triggers

    The report notes that ABS DSCR has gradually declined but remains above trigger levels, cap rates have stayed stable, and reported securitized occupancy is above 90%.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

  • HY data center debt
    Directly provides financing exposure to AI and data center construction
    Strengths
    Defensive credit characteristics, limited leasing risk, project-finance-style protection, and support from long-term Hyperscaler and neocloud demand.
    Weaknesses
    Still affected by construction progress, issuance supply, and single-project cash flow performance.
    Comparison
    The report explicitly rates HY data center debt OW and views its spreads as attractive.
    Risks
    Construction delays, spread pressure from increased supply, weakening tenant demand, or insufficient project-finance protections.
  • Data center ABS and CMBS
    Provides securitized credit exposure through data center asset cash flows and collateral
    Strengths
    Spreads tighten when supply declines; ABS DSCR remains above trigger levels; cap rates remain stable; reported occupancy is above 90%.
    Weaknesses
    ABS DSCR is gradually declining, and CMBS DSCR is compressed in a higher-rate environment.
    Comparison
    The report believes data center securitized credit is attractive overall, with value most concentrated in ABS BBBs.
    Risks
    Spread widening if supply recovers, further DSCR compression, trigger pressure, and rates staying high.
  • IG and corporate credit AI-related debt
    Supports Hyperscaler capex and large-scale AI infrastructure financing needs
    Strengths
    Issuers have strong financing capacity, and non-USD markets still have room to expand the investor base.
    Weaknesses
    Supply volumes are high, and near-term pricing is sensitive to supply expectations.
    Comparison
    The market has generally tightened since the end of 1Q, but the report believes technicals dominate pricing more than fundamentals.
    Risks
    Accelerating 2H26 supply, insufficient absorption in benchmark markets, and volatility in cross-currency issuance windows.
  • Chip financing structures
    Provide financing for AI computing hardware and chip-related assets
    Strengths
    Public and private market activity is increasing, existing deals have shorter tenors than some data center construction deals, and they use fully amortizing structures.
    Weaknesses
    Structures are still developing, with limited standardization and historical performance data.
    Comparison
    Relative to traditional data center construction financing, chip financing is more short-duration and amortizing in nature.
    Risks
    Technology iteration, changes in collateral value, transaction structure transparency, and investor acceptance.
  • Non-USD Hyperscaler debt
    Helps large AI infrastructure issuers expand their investor base in EUR, CAD, CHF, JPY, and other markets
    Strengths
    Issuer representation in EUR/GBP benchmarks is lower than in USD, leaving room for further supply.
    Weaknesses
    Depends on regional market windows and cross-currency funding costs.
    Comparison
    U.S. market issuance was limited in May, but non-USD issuance was about $24bn, showing supply focus can rotate regionally.
    Risks
    Changes in FX and rate environments, insufficient absorption in non-USD markets, and concentrated issuance timing.

Key data

  • 2026 AI-related global supply forecastabout $570bnThe report expects issuance to accelerate in 2H26.
  • YTD issuance as of May 31, 2026about $236bnMore than 4x global AI-related issuance in the same period of 2025.
  • April AI-related issuance>$74bnThe highest single-month supply this year.
  • Share of data center shell construction project financingHY 85%; IG 40%Refers to the share of project financing structures within AI-related HY and IG supply.
  • May Hyperscaler non-USD issuanceabout $24bnCurrencies include EUR, CAD, CHF, and JPY.
  • HY data center issuanceabout $42bn since 2H25; about $27bn YTDProvides exposure to AI and data center construction.
  • Data center securitized credit supplyOnly one ABS and one CMBS transaction each in April and MayReduced supply drove spread tightening.
  • Securitized data center occupancy>90%The report says reported securitized occupancy is above 90%.
  • Data center ABS DSCRStill above trigger levelsAlthough DSCR has gradually declined as issuers expanded master trusts.

Impact & implications

AI infrastructure financing is moving from early thematic issuance into a broader credit market cycle. For investors, accelerating supply may bring spread volatility, but it also offers multi-asset opportunities to access the data center, Hyperscaler capex, and chip financing chain. The report prefers HY data center debt with structural protection and higher spread compensation, as well as lower-rated ABS BBBs within data center securitized credit.

Risks

  • Supply expectations are the main driver of current price action; if issuance accelerates in 2H26, spreads may come under pressure again.
  • Data center project financing carries construction delay and completion risks, although mitigants include guarantees, debt service reserves, amortization requirements, and lockboxes.
  • ABS DSCR is gradually declining, and CMBS DSCR is compressed in a higher-rate environment, requiring continued monitoring of trigger levels.
  • Chip financing structures are still relatively new, so the stability of tenor, amortization, collateral value, and investor demand should be monitored.
  • Morgan Stanley discloses that it may have business relationships with covered companies, and investors should treat this research as only one factor in investment decisions.

What to watch

  • Whether AI-related global issuance in 2H26 accelerates toward the report's forecast of about $570bn in full-year supply.
  • Whether Hyperscaler cash capex exceeds $1tn in 2027 as estimated by the equity research team.
  • Whether Hyperscaler issuance in non-USD markets such as EUR, CAD, CHF, and JPY can continue expanding the investor base.
  • The pace of new data center ABS and CMBS issuance, as well as changes in spreads, DSCR, cap rate, and occupancy.
  • The number of chip financing deals, tenor, fully amortizing structures, and investor demand.
  • Whether the relative value of ABS BBBs and HY data center debt remains attractive after supply recovers.
Zhejiang ICP No. 2022035445-5
Disclaimer: Market data, charts, indicators, research views, and other information provided on this website are intended solely for information display, research communication, and educational reference. They should not be regarded as personalized investment advice, securities recommendations, trading instructions, solicitations, or guarantees of return. While we strive to improve the reliability of our data and content, such information may still be subject to delays, errors, incompleteness, or untimely updates due to source differences, methodological limitations, system processing, or market volatility. Users should exercise independent judgment based on their own circumstances and bear all risks and responsibilities arising from the use of this website.

Settings

Sign in to view recent logins