Nvidia Corp’s (NASDAQ: NVDA) credit default swaps posted their largest one-day jump on record, rising roughly 14 basis points to approximately 82 basis points, as investors grew increasingly wary of the company’s expanding role in financing artificial intelligence infrastructure. The surge in insurance costs against Nvidia’s debt coincided with a 4.99% drop in its stock price to $196.51, wiping out roughly $250 billion in market value and temporarily ceding the title of the world’s most valuable company to Apple Inc. This divergence between equity and credit markets suggests that while shareholders remain optimistic about long-term growth, lenders are pricing in heightened risks associated with Nvidia’s aggressive vendor financing strategies.

The immediate catalyst for this repricing was the report that Nvidia is negotiating to guarantee approximately $250 billion in financing for OpenAI’s planned 10-gigawatt data center campus in southern Ohio. Developed by SoftBank Group’s energy subsidiary SB Energy, the project is expected to cost more than $500 billion in total. Because OpenAI lacks an investment-grade credit rating, lenders require Nvidia’s balance sheet to backstop the debt. Additionally, Nvidia is discussing separate financing for up to $350 billion of OpenAI’s direct chip purchases. This structure represents a significant strategic pivot from direct equity investments to assuming substantial credit risk, raising concerns among analysts about potential circular financing dynamics within the AI sector.

Michael Burry’s Warning on Circular Spending

“Big Short” investor Michael Burry is sounding the alarm on Nvidia, claiming the AI giant’s strategic financial “overreaching” has escalated its “circular spending” to “biblical proportions.” Taking to X on July 29, 2026, Burry pointed to a sharp rise in the cost of insuring Nvidia’s debt as a glaring warning sign. “There is a reason $NVDA’s 5 year credit default swaps are going parabolic,” Burry posted, sharing a terminal chart showing the swaps up nearly 90% year-to-date.

He attributed this sudden surge directly to “All this overreaching by #nvda to push the circular spending to biblical proportions.” This criticism coincides with reports that Nvidia signed data center leases worth up to $50 billion in Texas to support its “neocloud” partners—a financial maneuver that skeptics argue keeps high costs off Nvidia’s balance sheet while artificially driving demand for its own GPUs.

Historical Precedent and Market Skepticism

Seaport Research analyst Jay Goldberg warned that such vendor financing arrangements often backfire, drawing parallels to Lucent Technologies’ aggressive lending strategy during the late 1990s telecom boom. Lucent lent billions to telecom carriers to buy its equipment, fueling rapid sales but leaving it with massive bad debts when the bubble burst, eventually leading to its merger with Alcatel in a $13.4 billion deal. Goldberg stated that if the Nvidia-OpenAI arrangement proceeds, it would be difficult to justify from Nvidia’s perspective, noting that “these kinds of financing usually don’t end well.” Hammerstone Markets compared these interlocking deals to Ponzi schemes due to their reliance on continuous capital inflows and fear of missing out.

Applied Economics Professor Steve Hanke echoed the bearish sentiment, definitively declaring: “WELCOME TO THE AI BUBBLE,” following reports of massive Nvidia investments in entities like SK Hynix Inc. ADR and OpenAI.

Scale of Exposure and Global Impact

The scale of the proposed guarantees has spooked credit desks. Nvidia’s latest 10-Q filing discloses only $3.5 billion in maximum exposure across all its facility lease guarantees. The reported $250 billion backstop for OpenAI would run roughly 71 times its current guarantee book, dramatically increasing its off-balance-sheet liabilities. This nervousness has spilled beyond Nvidia, contributing to a broader selloff in tech stocks. Alphabet Inc.’s free cash flow swung to negative $5.9 billion last week after raising its 2026 capital expenditure guidance to as much as $205 billion, marking its first quarterly outflow since 2004.

The selloff extended globally, with a Bloomberg gauge of Asian chip shares slumping 7.5% in its biggest drop since April 2025. South Korea’s Kospi sank as much as 11%, and SK Hynix fell more than 14% in Seoul, days after announcing a $500 billion-plus partnership with Nvidia for AI factories and next-generation memory.

Timeline of Circular Financing Deals

This development is the latest in a pattern of Nvidia providing capital that is subsequently spent on its own hardware or cloud capacity. Key milestones include:

What the Numbers Show

The proposed financing structure underscores a deepening dependency on vendor credit within the AI supply chain. With Nvidia potentially backing $250 billion in debt and negotiating another $350 billion for chip purchases, traditional banking channels appear insufficient for such large-scale projects without manufacturer guarantees. This concentration of risk amplifies volatility; if OpenAI’s revenue projections do not meet expectations, Nvidia’s balance sheet could face immediate strain. Prediction markets on Polymarket currently assign an 18% chance that the AI bubble bursts by the end of 2026, a figure that has ticked higher recently. Despite Monday’s slide, traders still favor Nvidia at 49% to reclaim the title of the world’s most valuable company by year-end, suggesting that while credit risks are rising, equity confidence remains resilient pending further earnings guidance from peers like Meta and Amazon.

Wall Street analysts are sharply divided over the financial architecture of the artificial intelligence boom, with some warning that Nvidia Corp.’s aggressive capital deployments resemble a Ponzi scheme while others defend them as a necessary alignment of supply and demand. The debate centers on "circular financing" arrangements where suppliers fund customers who then use that capital to purchase the supplier’s products, raising questions about whether current market valuations reflect genuine utility or manufactured hype. This divergence in opinion carries significant implications for investors exposed to the tech sector, particularly given the recent volatility in semiconductor and memory ETFs.

Hammerstone Markets, quoted by Michael Burry on X, characterized these circular deals as requiring constant activity in the way of new funds, continuous hype, and the stoking of FOMO. Applied Economics Professor Steve Hanke echoed this skepticism, labeling Nvidia’s reported $500 billion investment in SK Hynix Inc. ADR and $250 billion investment in OpenAI as evidence of an AI bubble. These figures highlight the scale of interwoven investments that have sparked concern among market watchers regarding the sustainability of such capital-intensive structures.

Conversely, Janus Henderson described these mega-deals as a "virtuous circle" that efficiently aligns builders, suppliers, and customers to meet unprecedented computing demand. Anthropic CEO Dario Amodei defended the interconnected deals at the New York Times Dealbook Summit on Dec. 7, stating that one player has capital and sells chips, while the other is confident they will have revenue but lacks the $50 billion at hand. Amodei argued there is nothing inappropriate about such arrangements in principle, framing them as a solution to a genuine capital bottleneck rather than a fraudulent scheme.

Strategists also pushed back against comparisons to the vendor-financed telecom crash of the late 1990s. James E. Thorne, Chief Market Strategist at Wellington Altus, noted that today’s buyers are cash-rich hyperscalers and called the assumption of a 1990s repeat "lazy research." Thorne stated that circular financing is a signal, not a verdict, emphasizing that the world remains short of compute, memory, and power. Ricky Ho, CIO and Portfolio Manager at Four Equities, added that the current AI ecosystem is not fraudulent but "financially reflexive," magnifying both market upside and downside risks.

What the Numbers Show

The divergence in analyst sentiment is reflected in the recent performance of key technology indices and exchange-traded funds. While the Nasdaq Composite Index slipped 1.44% in a month and was down 2.26% over the last five sessions, it remained up 7.27% year-to-date. The Invesco QQQ Trust, which tracks the index, was 0.89% lower in premarket trading on Tuesday.

Sector-specific funds showed mixed results, with the iShares Semiconductor ETF up 69.30% year-to-date but down 12.49% over the month. The newly launched Roundhill Memory ETF declined by 27.06% over the last month despite being 94.19% higher since its listing in April. These fluctuations underscore the volatility inherent in the sector as investors weigh the long-term potential of AI infrastructure against short-term valuation concerns.