Companies in the stock market benefiting from an unprecedented wave of AI compute infrastructure buildout are simultaneously facing expanding demand and sharply tightening financing conditions: AI agent applications are driving continuously growing demand for AI compute resources, while rising capital costs — amid the 10-year US Treasury yield surging past 5% — are raising the capital threshold for converting excess AI orders into actual cash flow. For the week ended September 25, CoreWeave shares rose nearly 8%, while Oracle fell about 7% and is down roughly 30% year-to-date, signaling that investors are beginning to scrutinize different companies' financing, delivery, and cash flow capabilities more carefully. JPMorgan estimated in June that debt financing needed for AI infrastructure buildout through 2030 would reach approximately $4.1 trillion — this is a multi-year cumulative financing forecast, not a single-year issuance volume, and cannot be entirely equated with corporate bonds. When such massive buildout demand continuously enters capital markets, interest rates, credit conditions, and capital availability directly affect the pace of expansion.
The hottest AI agent applications represented by Muse and Astra provide a new source of sustained strong growth for compute demand. A single user instruction can trigger planning, retrieval, browser operations, code execution, and result validation, while multi-turn model calls add context processing and state preservation requirements. From an engineering perspective, GPUs handle model computation, CPUs execute tools and task orchestration, HBM, server DRAM, and enterprise SSDs jointly support data access and cache management, and data center optical interconnects govern high-speed data transmission. A detailed engineering analysis from Nvidia points out that CPU execution speed affects GPU effective utilization, while tiered caching through GPU memory, CPU memory, local NVMe, and remote storage can reduce redundant computation. Therefore, the investment value brought by agent penetration ultimately needs to manifest as more billable high-complexity tasks and larger AI compute resource demand deliverable per unit of capital, as well as stronger corporate operating cash flow.
Specific changes on the funding side are undoubtedly global in nature. On September 25, the US 10-year and 30-year Treasury yields intraday touched approximately 5.23% and 5.53% respectively, reaching the highest levels since 2007 and 2004. Energy supply shocks, sticky inflation, and resilient economic demand are pushing markets to revise up the future policy rate path; the Federal Reserve already raised rates by 25 basis points in September, and an analysis from Charles Schwab on September 25 showed that futures markets from that point were inclined to price in nearly three additional rate hikes through June 2027. As the "anchor of global asset pricing," the upward shift in the 10-year US Treasury yield both raises the pricing benchmark for new long-term financing and suppresses the valuation of companies' future cash flows through a higher required rate of return.
The "credit gate" of the AI frenzy: total financing expansion and financing eligibility contraction may occur simultaneously. The current AI compute expansion frenzy appears to be undergoing screening by a "credit gate" — demand determines how much companies want to finance, while lending capacity determines which orders and projects are worthy of funding. A financing professional at the US subsidiary of Mitsubishi UFJ Lease believes that the market may genuinely be willing to support only a portion of the candidate list of new cloud companies; this is a practitioner's judgment, yet it reveals a very critical change: higher interest rates can compensate for some risk, but projects lacking reliable power, delivery guarantees, or executable customer contracts — or AI projects overly dependent on a single large customer — may be unable to obtain financing simply by raising their quotes. By extension, industry-wide borrowing may still grow, while capital gradually concentrates toward companies with thicker financial buffers, higher contract quality, and stronger delivery capabilities.
Interest rate shocks also need to be differentiated by debt structure. Newly issued or refinanced fixed-rate debt will face repricing; existing floating-rate loans typically adjust with short-term benchmarks such as SOFR, and the increase in 10-year Treasury yields cannot be directly applied to all loans. CoreWeave's SEC filings show that based on the outstanding floating-rate debt balance as of June 30, 2026, a 100 basis point increase in interest rates corresponds to approximately $30 million in additional interest expense on a three-month basis and approximately $61 million on a six-month basis. The company also disclosed using interest rate swaps to mitigate some risk. These figures measure the sensitivity of specific existing debt; future new borrowings, financing costs, and credit spread changes will further affect actual capital costs.
SoftBank exemplifies the side where financing demand remains strong. Its bond issuance finalized on September 24 included $10 billion in dollar bonds and 1 billion euros in euro bonds, totaling approximately $11.1 billion; the longest dollar bond has a 7.5-year tenor with a 9.75% coupon, issued at par, with settlement expected on September 29. Part of the proceeds will be used for subsequent investment payments to OpenAI. This signal indicates that when a large company believes the cost of missing the AI deployment window is higher, it may still accept expensive funding; but higher interest also means future investment returns need to cross a higher return threshold.
Another variable harder to offset by robust demand is the time gap between construction spending and operating cash collection. Equipment procurement, engineering construction, and financing costs occur first, while revenue depends on power connection, project acceptance, and customer usage. Oracle's "force majeure" dispute over Project Jupiter involves contract payment arrangements after potential delays; media previously cited sources familiar with the matter reporting that the related arrangements could extend the payment phase of lower construction-period rent and delay the start of higher operations-period rent, while Blue Owl Capital stated that financial commitments have not changed. It can be seen that the financing value of long-term AI infrastructure orders also depends on delivery conditions and risk-sharing systems, and cannot be directly regarded as cash already received. For investors, the "credit gate" ultimately screens for AI cloud computing/compute companies that can convert compute demand into debt-servicing cash flow on time and retain shareholder returns after paying interest.
AI compute demand heat collides with financing pressure: bond yields surge, and AI compute companies hungry for debt financing face greater risk. As US Treasury yields rose to their highest levels since 2007 this week, companies reliant on debt financing will face higher borrowing costs. This means that AI infrastructure buildout, already at historic scale, is about to become even more expensive. JPMorgan estimated in a June in-depth research report that approximately $4.1 trillion in AI-related debt issuance is expected through 2030. Data center companies and other companies related to the AI boom are racing to expand capacity to meet AI service demand that many industry experts consider nearly impossible to satisfy. When borrowers re-enter financing markets, they face a 10-year US Treasury yield near 5.17%, up about 1 percentage point from the start of the year. This means companies issuing debt will have to offer more attractive returns to attract investors.
The market has not yet fallen into panic — at least not yet. Heavily indebted new cloud company CoreWeave's stock performance remains solid, rising nearly 8% this week; while Oracle, which has high financial leverage and relies on long-term debt markets to support AI expansion, has had a rougher time, falling 7% this week and about 30% year-to-date. Meanwhile, Japanese investment giant SoftBank Group, one of the major capital providers for AI projects, raised $11.1 billion this week through junk bond issuance, with its 7-year bonds yielding as high as 9.75%. Siebert Financial Chief Investment Officer Mark Malek said in an interview: "They are basically price-insensitive to this financing, which means they are price takers. In my view, many of these companies have to be less price-sensitive. They need to raise as much capital as possible to compete."
At the center of the AI boom are leading model developers OpenAI and Anthropic, both of which have private market valuations exceeding the $1 trillion super-threshold. To provide the infrastructure needed for these advanced models and the models and services of many other companies, the tech industry's hyperscale cloud providers — Amazon, Google, Meta, and Microsoft — have committed to hundreds of billions of dollars in capital expenditure this year, with expectations of further increases to $1 trillion in AI capital expenditure by 2027. Although a considerable portion of these investments is funded through debt, these tech giants all have investment-grade credit ratings, enabling them to access capital at lower cost.
However, some market participants believe other companies will face greater challenges in the future. According to media reports citing sources, a veteran private credit investor told media that in the future, financing for new cloud projects will become more difficult because these companies have less buffer to absorb cost increases. The investor requested anonymity to discuss the matter candidly. Mitsubishi HC Capital America Vice President Riley Thompson said in an interview that even if borrowers are willing to pay higher interest rates, lenders are becoming more selective about which projects they are willing to fund. Thompson said: "If you list 50 new cloud companies, the market may genuinely be interested in only 20."
CoreWeave, which went public last year, warned of rising interest rate risks in its filings with the US Securities and Exchange Commission. In its latest quarterly filing, the company said that based on its outstanding floating-rate debt balance as of June, for every 100 basis point — or 1 percentage point — increase in interest rates, its interest expense could increase by $30 million. An early warning signal may have already appeared this week. On Friday, media cited sources familiar with the matter disclosing that Oracle issued a "force majeure" notice regarding its New Mexico data center project to avoid bearing higher costs, after which the company's stock fell. It is reported that if this campus, named Project Jupiter, fails to commence operations as expected in 2028, Oracle hopes to delay related payments. Oracle said the project "remains on our established timeline."
Rising interest rates are not the only issue currently faced. Before this week's yield surge, the CEOs of Anthropic and OpenAI had already begun calling for slowing the pace of AI development, after industry researchers publicly expressed concerns that advanced models may face risks of breaking free from human control. Meanwhile, with the November midterm elections approaching, opposition to AI data centers across the United States has become a significant issue. A recent NBC News Decision Desk poll supported by SurveyMonkey showed that 69% of respondents oppose building such facilities in their areas, mainly due to the unprecedented surge in residential utility bills and carbon emissions brought by the AI infrastructure frenzy. On Monday, Texas Republican Governor Greg Abbott, facing a competitive re-election fight, ordered a temporary halt to all data center-related environmental permit approvals; he had already suspended grid connection approvals last month.
Nevertheless, demand for AI services continues to surge. The latest example is Meta's personal assistant app Muse, whose popularity has climbed rapidly since its early September launch. Muse surpassed 2.5 million global downloads in its first two weeks and overtook ChatGPT at the top of the Apple App Store. Evercore's Mark Mahaney told CNBC this week that Muse could reach 100 million users within 6 to 12 months. Credit rating agency KBRA's global head of corporate, project, and infrastructure finance Andrew Giudici said that although rising interest rates may affect future deals, he has not yet seen a significant impact on borrowing demand. Giudici said: "In a normal environment, people might step back and pause a bit. But I don't think that will happen here. I think you will continue to see fairly large-scale issuance."
Latham & Watkins vice chair of emerging companies and growth practice Haim Zaltzman said there is no doubt that as costs rise, "someone has to bear it." Zaltzman, who has long worked on core AI compute infrastructure financing, said: "But in a structure with such strong demand, absorbing these costs is much easier." For American Compute CEO Bernie Margulies, who provides risk management consulting for GPU financing, the math is even simpler. He said that even at higher costs, borrowers are still eager to obtain financing, especially those with contracts with OpenAI and Anthropic; both companies have been signing agreements to lock in compute supply for years to come. Margulies said: "If you have a multi-year large contract with Anthropic, and then it's 50 basis points more, is that really going to stop you?"
Comments