Oracle's Credit Risk Hits Post-Crisis High as AI Debt Wave Builds
Oracle's credit default swaps closed at their highest level since 2009, a sharp signal that investors are hedging against an AI buildout that may take longer to translate into cash flows. The move landed after a rush of bond issuance from large tech names and fresh concerns about an AI-fueled debt cycle.
Protection on Oracle's debt priced around 1.28 percentage points per year on Tuesday (about 128 bps), up roughly 3 bps day-over-day and more than triple from about 36 bps in June. Spreads retraced slightly Wednesday, tightening by as much as 2.4 bps.
The move in numbers
- CDS near 1.28% per year, highest New York close since March 2009 (~1.30%).
- Up from ~0.36% in June; modest tightening the next day (~2.4 bps).
- CDS trading volumes surged to ~$8 billion over nine weeks ended Nov. 28 vs. ~$350 million a year earlier (Barclays analysis).
Why Oracle CDS became the go-to hedge
Oracle has tapped debt markets for tens of billions in recent months-directly and through backed projects-while holding weaker ratings than other hyperscalers. That combination makes its CDS an efficient proxy hedge for an AI downturn.
The company sold $18 billion of US high-grade bonds in September and is tied to the largest AI-infrastructure financing coming to market. Its AI roadmap is closely linked to OpenAI, with the company counting on significant revenues tied to that partnership over the next few years.
As of August, total debt stood near $105 billion including leases. Roughly $95 billion sits in the Bloomberg US Corporate index, making Oracle the largest non-bank issuer in that benchmark.
What the Street is saying
Some strategists see echoes of prior manias where prices overshot before correcting. One warns the pattern resembles phases of the dot-com era, though the timing and magnitude are uncertain.
Another bank cautioned that Oracle's growing debt load could push CDS toward 2 percentage points-close to 2008 extremes-if supply remains heavy and cash flow visibility lags.
Bloomberg Intelligence notes that worries may ease if mega-cap tech reduces buybacks and dividends to self-fund capex. Tech issuance could remain high into 2026, but vast cash balances and pre-payout free cash flow could temper risk.
Supply outlook and spreads
AI data centers and electrical capacity upgrades are set to keep capex elevated into next year. One dealer desk projects US investment-grade issuance could hit a record ~$2.1 trillion in 2026, versus more than $1.57 trillion this year.
Base-case expectations for IG spreads center around 100-110 bps next year, up from ~75-85 bps in 2025. Another team points to health care's last-decade debt wave as proof that sectors can add leverage while keeping spreads tighter than the index-though bondholders capture limited upside if equity takes most of the gains.
The risk: if companies keep pouring cash into AI infrastructure without near-term profit lift, credit quality can drift weaker and sector spreads can widen on buyer hesitation.
What this means for credit and rates desks
- Expect more supply and higher concessions. Keep a live bias on new-issue premiums and secondary liquidity in long-duration tech tranches.
- CDS vs. cash: Oracle CDS remains a liquid hedge for AI exposure; basis trades may open if new-issue supply dislocates curves.
- Relative value: Stress-test positioning versus higher-rated hyperscalers; assess basis to indices and sector ETFs.
- Liquidity planning: Monitor dealer inventories and TRACE prints around month-end and index dates.
- Governance and payout policy: Track buyback and dividend adjustments-self-funding could cap spread widening risk.
Key watch items
- Supply calendars for large-cap tech and AI infrastructure-linked issuers.
- CDS term structure on Oracle versus cash bonds across the curve.
- Capex run-rate, energy buildouts, and AI monetization timelines from hyperscalers.
- Spread beta to macro-especially if growth slows before AI ROI improves.
For context on credit instruments and data, see ICE Data Services. For broad issuance trends, SIFMA's dashboards are useful starting points: US fixed income markets.
If your team is assessing practical AI use cases with measurable ROI, this curated list can help filter signal from noise: AI tools for finance.
Bottom line: AI capex isn't slowing, and neither is supply. Hedging demand is rising as investors weigh timing risk between massive investment and profit delivery. Stay disciplined on entry points, demand real concessions, and keep your hedges liquid.
Your membership also unlocks:
