Big Tech drives US corporate bond sales to annual record high, data shows

US investment-grade corporate bond issuance reached a record $1.88 trillion through November 4, up 18.7% year over year and above the prior $1.86 trillion record set in 2020; five hyperscalers accounted for more than $370.7 billion. Issuance count fell about 4% to 1,084, while average deal size rose to roughly $1.7 billion. Demand for AI-related debt remains strong, but higher-rate concerns, rising default-insurance costs—including Oracle’s record 325.45-basis-point five-year spread—and delayed AI monetization and potential regulation pose risks; Goldman expects 35% of hyperscalers’ 2027 capex to be bond-funded.

Analysis

The key signal is not the issuance record itself, but the widening gap between easy market access and uncertain cash returns. Oversubscribed deals show that investors will currently fund the buildout; they do not establish that incremental AI capacity will earn an adequate return. If monetization slips, debt service and ongoing capex compete with buybacks and other uses of cash, leaving equity holders with less flexibility even before any solvency concern emerges.

Credit risk is already differentiating within the group: ORCL’s notably wider insurance costs make it a more direct barometer of financing and execution stress than a blanket bearish read across hyperscalers. A higher-rate or regulatory shock could also raise the cost of capital and delay data-center projects, spilling over to infrastructure suppliers; conversely, a pause in investment would hit supplier growth before it necessarily impairs the largest platforms. The 1–3 month test is whether upcoming guidance and bond pricing validate continued funding at tolerable terms. Over 6–18 months, monitor AI-related revenue and cash generation against capex and purchase commitments. The contrarian risk is that credit warning signals are being overread: large, cash-generative platforms may keep funding access even if returns take longer. Treat oversubscription as a liquidity signal, not proof of attractive economics.

Key Decisions for Investors

• Avoid a broad short of AI megacap equities on issuance volume alone. Track capex, AI-related revenue commentary, and free-cash-flow guidance; deteriorating conversion alongside higher funding costs would strengthen the downside thesis.
• Use ORCL as the relative-credit watchpoint, not an automatic short. Consider buying ORCL protection versus a diversified IG-tech exposure only if its spread premium persists or widens while peers’ spreads stabilize; exit if the relative spread retraces materially or guidance supports cash-flow coverage.
• Watch AAPL’s bond pricing and subsequent secondary performance as a near-term test of investor appetite for large tech supply. Concession widening across successive deals would argue for broader IG spread pressure; stable pricing would weaken that read.
• Falsifiers: sustained improvement in AI monetization and cash generation, stable or falling issuer spreads despite continued borrowing, or evidence that data-center commitments are being scaled back without impairing returns. Reassess after the next relevant company guidance and bond-market pricing.