AI Revenue and Market Selloff
AI remained the dominant force in markets, but the week exposed a growing divide between enthusiasm for the technology’s commercial potential and concern over the cost of building the infrastructure behind it. OpenAI became the focal point after investor materials put its annualized revenue at roughly $50 billion at the end of September, well below the $68 billion to $70 billion figure that had circulated in markets. The discrepancy was largely an accounting issue: the higher number included gross revenue and cloud-partner pass-through activity, while the lower figure reflected net revenue recognized by OpenAI itself. The clarification did not indicate an abrupt collapse in demand—enterprise revenue growth was reported above 100% in the third quarter—but it shook confidence in companies whose valuations depend heavily on OpenAI’s infrastructure spending. Oracle fell sharply, while Nvidia, CoreWeave, AMD, Broadcom, Micron and other AI-linked shares also sold off. The Nasdaq declined more sharply than the Dow, and the Philadelphia Semiconductor Index dropped as investors reassessed how much of the AI buildout is supported by end-customer revenue rather than financing arrangements among cloud providers, chipmakers and AI labs.
AI Infrastructure Financing Risks
The revenue debate arrived alongside evidence that the AI industry is increasingly dependent on debt and private-credit markets. Major technology companies and infrastructure providers have moved from funding data centers largely from operating cash flow toward increasingly complex financing packages for chips, servers and power capacity. Oracle was reported to be exploring structures in which outside entities purchase AI hardware and lease it to the company, limiting the immediate burden on its own balance sheet. Broadcom was linked to major debt facilities supporting AI-chip commitments for Anthropic and OpenAI, while SpaceX was reported to be seeking approximately $40 billion in loans and investment-grade bonds for Nvidia processors. Analysts estimated that external financing for AI infrastructure could reach well above $1 trillion over the next several years. The expansion of private credit into AI hardware turned financing itself into a market risk: chip demand remained strong, but investors grew more sensitive to leverage, customer concentration, collateral values and the possibility that capital-intensive projects could struggle to generate cash flow quickly enough to justify their funding costs.
Rising Bond Yields and Debt Concerns
This concern was reinforced by a broader rise in global bond yields. The U.S. 10-year Treasury yield climbed above 5.3%, reaching levels not seen since the early 2000s, while the 30-year yield remained near multidecade highs. Similar pressure appeared in Europe, Japan, Canada and other developed markets as investors confronted persistent inflation, large fiscal deficits and heavy sovereign debt issuance. Higher yields tightened financial conditions across mortgages, autos, corporate borrowing and equity valuations. The bond selloff was not confined to the United States: France faced renewed fiscal strain around its budget, while global investors questioned whether governments could sustain borrowing without eventually requiring fiscal restraint or higher taxes. Warnings from market veterans including Ray Dalio, Robert Rubin and Michael Burry amplified concerns that debt-financed AI investment could become unstable if rates remained elevated. Their views differed in intensity, but all emphasized the same vulnerability: asset values have become increasingly dependent on assumptions that earnings, productivity and AI monetization will grow rapidly enough to offset rising financing costs.
Federal Reserve Policy and Jobs Data
Federal Reserve policy remained central to this debate. Minutes from the September meeting showed unanimous support for the latest rate increase and indicated that policymakers generally expected at least one additional quarter-point move before year-end. Fed officials including Chris Waller and Alberto Musalem maintained that inflation remained too high and that further policy tightening could be needed to return inflation toward 2%, particularly if energy prices, tariffs and supply shocks added pressure. Yet weaker September payroll data complicated the picture. The U.S. economy added only 29,000 jobs, revisions erased another 60,000 jobs from prior months, and unemployment rose to 4.2%. Markets interpreted the softer labor report as reducing the likelihood of another immediate increase, with expectations shifting toward a pause at the October meeting. The data suggested a low-hiring, low-layoff labor market rather than a collapse, but it also highlighted an economy in which job growth and wage gains were not keeping pace with household concerns about inflation and borrowing costs.
Consumer Stress and Housing Affordability
Consumers faced the consequences of that environment directly. University of Michigan sentiment slipped to 46.3, a five-month low, as households cited higher living costs, elevated mortgage rates and anxiety about inflation. One-year inflation expectations rose to 4.7%, while long-term expectations also edged higher. Mortgage rates around 7.4% to 7.6% kept housing activity subdued, reduced refinancing opportunities and left many potential buyers unable to qualify for homes at prevailing prices. Pending home sales weakened, price cuts accelerated and affordability remained stretched, with the income needed to purchase a typical home far above median household earnings. Consumer conditions were increasingly described as K-shaped: higher-income households continued to spend and benefit from rising asset values, while lower-income families faced tougher choices on food, energy, housing and transportation. Higher auto prices reinforced the strain, with average new-car payments nearing $800 and more than one-fifth of buyers carrying payments above $1,000 a month.
Oil Prices and Geopolitical Risks
Energy markets intensified those inflation worries. Brent crude moved above $100 a barrel during the week and West Texas Intermediate hovered near $90 as attacks and shipping risks around the Strait of Hormuz revived fears of supply disruption. U.S.-Iran tensions, Houthi actions and storm risks in the Gulf of Mexico added to uncertainty, while diesel and gasoline prices rose sharply. The administration signaled that the United States would not strike Iran before the midterm elections, helping oil retreat from its highest levels, but prices remained elevated. The G7 and other governments responded by planning emergency releases of oil and diesel from strategic reserves, while the U.S. Strategic Petroleum Reserve fell toward its lowest level since the early 1980s. Iran itself faced deepening economic distress as sanctions cut oil exports, the rial weakened to record levels and inflation neared 90%, creating a regional crisis that fed directly into global fuel, shipping and inflation risks.
SpaceX AI Computing Expansion
SpaceX became a central character in both the AI-financing and telecommunications stories. Reports of a potential $40 billion debt package to buy Nvidia chips highlighted the company’s transformation from a launch and satellite enterprise into a major AI-computing operator. SpaceX’s expanding data-center business was linked to large compute contracts with Alphabet and Anthropic, while its capital spending on AI infrastructure was reported to far exceed current profits. The financing discussion pressured SpaceX shares at points during the week because investors had to weigh rapid revenue growth against the burden of borrowing heavily for depreciating hardware. At the same time, bullish analysts pointed to Starlink, launch services, AI compute and Starship as separate growth engines. The company’s market value remained extraordinary relative to its current revenue and profitability, illustrating the premium investors were placing on its combination of satellite connectivity, aerospace capabilities and AI infrastructure ambitions.
SpaceX Wireless Competition
SpaceX also reshaped the outlook for U.S. wireless carriers through its acquisition of nationwide low-band spectrum from Grain Management for roughly $8 billion. The 800 MHz spectrum, supplemented by EchoStar-related holdings and planned mid-band capacity, would improve indoor coverage and give Starlink Mobile a more credible terrestrial component alongside its satellite-to-phone network. AT&T, Verizon and T-Mobile shares fell sharply as investors assessed the risk of a future direct-to-consumer mobile challenger with nationwide satellite reach. Tower companies such as Crown Castle, American Tower and SBA Communications rose because a terrestrial rollout could create a new tenant and drive infrastructure demand. The immediate competitive threat remained limited by regulatory approvals, spectrum constraints and the need for substantial network deployment, but the transaction altered bargaining power in the wireless sector. It also strengthened SpaceX’s position in negotiations with carriers, which still viewed satellite connectivity as potentially complementary for rural coverage, emergency communications, airlines and industrial internet-of-things services.
Semiconductor Leadership and Market Concentration
Semiconductors continued to provide the clearest evidence of real AI-related demand, even as investors debated whether valuations had outrun the economics of the buildout. Nvidia approached a $6 trillion market capitalization after strong revenue guidance, accelerating AI-server demand and a $150 billion expansion of its share-repurchase authorization. The company continued to benefit from leadership in AI accelerators, software and systems, although its rising influence made the broader market more concentrated. Nvidia, Apple and Microsoft together accounted for more than 21% of the S&P 500, while the largest technology names dominated earnings revisions and index performance. Market breadth remained weak: relatively few S&P 500 constituents traded above their 50-day averages, and the equal-weight index lagged the headline benchmark. This left investors exposed to a narrow group of companies whose earnings and valuations increasingly determined the direction of the entire market.
Memory and Foundry Demand Growth
Memory and foundry suppliers showed how AI demand was spreading beyond GPUs. Micron reported extraordinary fiscal results, with revenue more than tripling year over year, gross margins reaching unusually high levels and a large share of 2027 production already committed through strategic customer agreements. Demand for high-bandwidth memory, DRAM and data-center solid-state drives supported pricing, while long-term contracts reduced some of the traditional volatility of the memory cycle. Samsung likewise forecast record quarterly operating profit on strong HBM shipments and tight memory supply, while TSMC recorded powerful revenue growth as advanced-node chip demand continued to surge. Analysts upgraded AMD after arguing that AI agents could expand demand not only for GPUs but also for CPUs, server systems and memory. AMD’s chief executive Lisa Su stressed the need for greater advanced-wafer, packaging and memory capacity, while Marvell and Broadcom raised long-term AI revenue ambitions tied to custom silicon, interconnect and hyperscale networking.
Nuclear Power for AI Data Centers
The power requirements of the AI buildout drove a parallel rush toward nuclear generation and grid investment. Google and Constellation Energy signed a major 20-year arrangement involving more than $4 billion of upgrades intended to add nearly 900 megawatts of nuclear capacity, alongside broader energy-supply commitments from Constellation’s existing fleet. The deal reflected a wider effort by major technology companies to secure dependable, low-carbon baseload power for data centers. Vistra also gained on reports that the federal government would provide roughly $4 billion in loans for upgrades at three nuclear facilities serving the PJM region, where data-center demand has strained electricity supply. These agreements supported nuclear operators, grid suppliers and equipment companies such as GE Vernova, while also emphasizing a hard constraint on AI expansion: computing demand requires not only chips and capital, but dependable power, transmission capacity, cooling infrastructure and community approval for new facilities.
Corporate Deals and Consolidation
Corporate restructuring and consolidation added other important shifts. Paramount Skydance completed its roughly $110 billion acquisition of Warner Bros. Discovery, combining HBO Max, Paramount+, CBS, CNN, Discovery networks and a vast film and television library under Skydance. The transaction created a more formidable media competitor to Netflix, Disney, Amazon and YouTube, but it also left the combined company with heavy debt and a demanding integration task. Schneider Electric agreed to acquire industrial-software provider PTC for about $22.6 billion, reinforcing the strategic value of industrial design, engineering software and digital-twin systems as manufacturers adopt AI-enabled automation. In energy, Cenovus agreed to buy Athabasca Oil for about $5.7 billion, extending consolidation in Canadian oil sands. These deals occurred against a tougher financing backdrop, and broader M&A activity slowed in the third quarter as higher yields made mid-sized transactions harder to justify.
Government Accounts and Fiscal Policy
Government policy featured prominently in market discussions as well. The administration expanded automatic enrollment in Trump Accounts, tax-deferred investment accounts seeded with $1,000 for eligible children and invested in S&P 500-tracking funds. More than $4.5 billion in deposits, including federal, philanthropic and private contributions, had already been committed, and major donors proposed using individual shares rather than only index funds. Supporters presented the program as a long-term savings and financial-literacy initiative, while critics questioned whether it addressed inequality or exposed children to concentration risk when gifts came in individual securities. Separately, campaign promises of broad cash dividends and rebates raised questions about fiscal cost, inflation and implementation. With public debt near $40 trillion and interest expense consuming an increasing share of federal revenue, markets treated new spending proposals through the same lens as the bond selloff: whether policy choices could sustain household support without adding further pressure to inflation, borrowing costs and the government’s balance sheet.