TradingKey - In October 2026, some AI infrastructure-related stocks and bonds came under pressure, with shares of CoreWeave (CRWV) and Oracle (ORCL) facing a sell-off.
As of the close on October 21, CoreWeave was down about 30.2% from its October 1 closing price, while Oracle fell by approximately 6.5% over the same period. The market has begun to reassess the financing risks of high-capital-expenditure models.
The risk structures of the two companies differ: CoreWeave faces relatively direct financing pressure, while Oracle needs to cope with cash flow pressure brought by AI investments.
In addition, Meta (META) reported third-quarter revenue up 26% year-over-year to $64.6 billion, but its capital expenditures of $34.91 billion exceeded operating cash flow of $32.79 billion, resulting in negative free cash flow of $2.12 billion. As of the end of September, the company still held $82.45 billion in cash and marketable securities, maintaining a relatively ample liquidity buffer.
This indicates that even with mature advertising businesses and large cash reserves, tech giants may face periodic cash flow pressures as they ramp up AI infrastructure investments.
The following data is as of June 30, 2026 for CoreWeave and August 31, 2026 for Oracle, reflecting different reporting dates.
Debt and Financing Metrics | CoreWeave | Oracle |
Debt Principal / Total Borrowings | $35.551 billion | $125.337 billion |
Recognized Operating Lease Liabilities | $16.319 billion | $34.621 billion |
Recognized Finance Lease Liabilities | $221 million | $9.185 billion |
Uncommenced Lease Commitments | Approximately $35.5 billion, with an additional portion currently uncertain | Approximately $288 billion |
Primary Investment Uses | GPUs, servers, and data center infrastructure | Cloud infrastructure, AI data centers, and related equipment |
S&P Long-Term Issuer Rating | B+, Positive Outlook | BBB-, Stable Outlook |
Among these, CoreWeave's approximately $35.5 billion represents estimated future undiscounted lease payments, while Oracle's approximately $288 billion represents additional lease commitments; neither should be directly included in current debt principal.
The S&P ratings highlight differing risk profiles for the two companies. CoreWeave's senior unsecured bonds are partially rated B, placing them in speculative grade, while Oracle's long-term issuer rating stands at BBB-, remaining within investment grade but approaching the speculative-grade threshold.
This implies that CoreWeave is more sensitive to changes in financing conditions, whereas Oracle needs to demonstrate that its large-scale AI investments will not continuously erode its financial flexibility.
CoreWeave builds data centers through debt, leasing, and GPU-related financing, and then provides AI compute to customers. This model benefits from growing demand for compute capacity, but it also carries high interest, depreciation, and lease costs.
In the second quarter of 2026, CoreWeave's revenue increased by 112% year-over-year to $2.575 billion, with a net loss of $626 million and net interest expense reaching $640 million. In mid-October, yields on some of the company's bonds briefly surpassed 15%, reflecting increased investor caution over its financing risks.
Rising bond yields do not mean the company has defaulted, but they could drive up subsequent financing costs. If data center deliveries are delayed, customer demand falls short of expectations, or cash flow fails to cover interest and ongoing investments, CoreWeave's financial pressure could intensify further.
Oracle's cloud infrastructure business is growing rapidly, but expanding data center investments are putting pressure on free cash flow.
In the first quarter of fiscal 2027, Oracle's revenue increased 30% year-over-year to $19.3 billion, and cloud infrastructure revenue grew 121% to approximately $7.4 billion. During the same period, operating cash flow was approximately $23 billion, while free cash flow was negative $5 billion.
In July 2026, S&P downgraded Oracle's long-term credit rating from BBB to BBB-, which remains the lowest tier of investment grade.
Going forward, key focus areas include data center commissioning, customer contract fulfillment, and cash collection. If revenue growth fails to consistently cover capital expenditures, Oracle's free cash flow and financing flexibility may come under further pressure.
On October 14, higher-than-expected U.S. September CPI data fueled rate hike expectations, weighing on AI infrastructure stocks. That day, Oracle fell 7.3% and CoreWeave dropped 15%, reflecting investor concerns over financing pressure and investment returns under high interest rates.
For companies with high debt loads and ongoing investments in data centers and AI computing power, if interest rates stay elevated for an extended period, the costs of new borrowings and refinancing maturing debt may rise, thereby increasing interest expenses and cash flow pressure.
A Reuters poll on October 20 showed that 94 of 100 surveyed economists expect the Federal Reserve to hold interest rates steady at 3.75%–4.00% at its October 27–28 meeting, with all respondents expecting the Fed to raise rates in December.
As of press time, the 10-year U.S. Treasury yield was around 5.12%. A high-yield environment could increase long-term corporate financing costs.
Currently, financing risks in the AI infrastructure sector have indeed risen, but there is insufficient evidence to suggest the broader industry has fallen into a debt crisis. Pressure on CoreWeave's stock price and certain bonds reflects heightened market caution regarding its high leverage and refinancing costs, while Oracle faces challenges from expanding AI capital expenditures, pressure on free cash flow, and credit rating downgrades.
CoreWeave and Oracle have different financial structures and debt-servicing capabilities, and face distinct risks. Going forward, attention should be paid to shifts in bond yields, CDS spreads, and credit ratings to assess whether operating cash flow can cover interest, lease payments, and capital expenditures, while tracking data center commissioning progress and return on investment.
If financing costs continue to rise while revenue and cash flow fail to keep pace, refinancing risks for some AI infrastructure companies may intensify further. Conversely, if demand for computing power remains strong, projects are commissioned on schedule, and cash flow improves, financing risks are expected to ease.