Oaktree's Howard Marks Warns of Rising Credulousness in Big Tech Debt Markets

Howard Marks, co-founder of Oaktree Capital Management and renowned distressed debt investor, issued a notable warning about “credulousness” being on the rise among investors regarding Big Tech debt issuance during a fireside chat with Bloomberg’s Lisa Abramowicz at the Capital Markets Industry Conference on March 17, 2026 [1]. This warning arrives amid an unprecedented wave of debt issuance by major technology companies, particularly hyperscalers, to fund massive AI infrastructure buildouts.

The context for Marks’ caution is striking: Amazon recently completed a $54 billion bond sale in March 2026, representing the largest recent offering among tech companies [2]. Alphabet issued $20 billion in bonds in February 2026, including a rare 100-year “century bond” that extends duration risk significantly [3]. Meta raised $30 billion in late 2025, while Oracle tapped markets for $18 billion [2]. Together, the five major hyperscalers issued $121 billion in U.S. corporate bonds in 2025 alone [2].

The scale of forthcoming issuance is projected to accelerate further. Bank of America projects $175 billion in new debt from hyperscalers in 2026, while Barclays forecasts over $2 trillion in total U.S. investment-grade corporate bond issuance for the year [2]. This represents a fundamental shift in how Big Tech is financing the AI arms race—moving from balance sheets heavy with cash to leveraging debt markets at historically unprecedented levels.

Marks’ characterization of investor behavior as “credulous” carries particular weight given his reputation as a contrarian and his firm’s specialization in distressed debt. His concern suggests that markets may be accepting debt terms without sufficiently rigorous analysis of the underlying risks, particularly regarding AI infrastructure investment payback timelines and the competitive dynamics driving continued massive capital expenditures.

The emergence of 100-year bonds in Big Tech issuance represents a significant structural shift in corporate financing that amplifies multiple risk dimensions. These ultra-long-duration instruments create substantial interest rate sensitivity, meaning any normalization of rates from current levels could generate significant mark-to-market losses for holders. The extended duration also implies that investors are making long-term bets on AI infrastructure returns without visibility into how these investments will ultimately monetize.

The distinction between Marks’ warning and typical market commentary lies in his focus on investor behavior rather than company-specific credit quality. The hyperscalers maintain strong investment-grade ratings and generate substantial free cash flows. However, his observation about “credulousness” suggests concern that credit spreads may not adequately compensate for execution risk inherent in AI infrastructure deployment at scale.

From a cross-asset perspective, the debt issuance trend has implications for equity valuations. As Big Tech companies increasingly leverage to fund AI capex, the distinction between equity and credit risk blurs. If AI investments fail to generate expected returns, both equity holders and bondholders could face deterioration in their positions—though bondholders would benefit from seniority in the capital structure.

The historical context matters: Marks is known for his prudent market stance, and his cautionary remarks have historically preceded periods of elevated risk. While this does not guarantee immediate market stress, it represents a notable contrarian voice in an environment characterized by strong demand for Big Tech debt.

• Credibility Gap Risk
  : The “credulousness” warning suggests investor enthusiasm may be outpacing fundamental credit analysis. Markets may be accepting debt terms without sufficient scrutiny of AI capital expenditure payback timelines [1][2].
• Infrastructure Investment Uncertainty
  : AI data center buildout requires massive capex with uncertain return profiles—bond investors may be underestimating execution risk.
• Interest Rate Sensitivity
  : Extended duration profiles, including 100-year bonds, create significant vulnerability to rate movements [3].
• Competitive Dynamics
  : The “AI arms race” among hyperscalers could pressure margins and cash flows, potentially impacting debt serviceability over time.

• Enhanced credit analysis: Marks’ warning serves as a prompt for investors to conduct deeper due diligence on AI infrastructure investment cases.
• Spread compensation evaluation: Investors can assess whether current credit spreads adequately compensate for the unique risks of AI-driven debt issuance.
• Duration management: The shift toward ultra-long-duration bonds warrants careful portfolio duration management.

Howard Marks’ warning about “credulousness” in Big Tech debt markets represents a significant contrarian voice amid record issuance. Big Tech companies have issued over $120 billion in bonds in 2025, with projections exceeding $175 billion in 2026, to fund AI infrastructure buildouts [2]. The emergence of 100-year bonds from Alphabet and similar ultra-long-duration instruments introduces novel risk considerations for fixed income investors [3]. While the hyperscalers maintain strong fundamental credit profiles, Marks’ focus on investor behavior suggests concern that market due diligence may not be keeping pace with the speed and scale of issuance. The fundamental question—whether AI infrastructure investments will generate returns sufficient to service this record debt—remains unresolved and will become clearer as capex deployments mature over the coming years.