Key Takeaways
- • Hedge fund manager George Noble, a former Fidelity portfolio manager, warned in a social media post that Nvidia's guarantees of customer leases and equity stakes in chip buyers mirror Cisco's dot-com-era vendor financing model.
- • Nvidia's filings show $99 billion in equity investments, $25 billion in additional equity commitments, $36 billion in cloud service commitments, and total guarantee exposure of $108.5 billion, including up to $105 billion in credit support for an SB Energy project in Ohio.
- • Insiders net sold $1.868 billion of Nvidia shares over the past year with no insider purchases, while 20 of 33 top investment masters tracked by GuruFocus reduced their holdings.
- • Nvidia now accounts for roughly 8% of the S&P 500, double Cisco's peak weighting during the 2000 bubble, which magnifies the systemic consequences if downstream customers default.
- • Nvidia's proposed $500 billion AI infrastructure financing platform with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR faces scrutiny over the residual value of GPUs used as collateral.
Summary
Hedge fund manager George Noble, a former portfolio manager at Fidelity, argued in a social media post that Nvidia is replicating the vendor-financing model Cisco used at the height of the dot-com bubble. He contends that by guaranteeing customers' lease agreements and taking equity stakes in startups that buy its chips, Nvidia is boosting hardware sales through customer financing, and that once downstream customers can no longer pay, the structure will backfire on Nvidia and create systemic fragility. Noble noted that Nvidia accounts for roughly 8% of the S&P 500, double Cisco's weighting at its 2000 peak. Cisco's stock fell 89% after telecom bankruptcies, dragging the index down by half, according to the article.
Nvidia's second-quarter filing for the period ended July 26, 2026 shows equity investments of $99 billion, with an additional $25 billion in equity commitments and $36 billion in cloud service commitments, typically with six-year terms. Under these arrangements, Nvidia commits to buy back cloud capacity that cannot be resold to third parties. The company also disclosed a $3.5 billion maximum exposure from guarantees on data center leases for certain AI cloud partners. In August 2026, Nvidia signed a guarantee arrangement with SB Energy providing up to $105 billion in credit support for the PORTS Technology Campus in Ohio, which is designed to supply about 4.25 gigawatts of IT load capacity for OpenAI. Combined maximum exposure reaches $108.5 billion. Nvidia states the guarantees cover only specific portions of lease and power payments and may terminate early if OpenAI obtains a satisfactory credit rating.
In August 2026, Nvidia also announced an AI computing infrastructure financing platform with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR, aiming to mobilize more than $500 billion in third-party capital through chip-backed lending. CEO Jensen Huang argues AI compute is productive and durable enough to qualify as a long-term financing asset. Some banks and credit investors remain skeptical, questioning Nvidia's assumed useful life and residual values. Huang has said high-end GPUs can last up to ten years, while banks typically use a three-to-four-year depreciation cycle.
Flow data also draws attention. According to GuruFocus, 20 of 33 tracked top investment masters reduced Nvidia holdings in recent quarters while 12 added. Insiders recorded no purchases over the past year and sold $1.868 billion in shares. The article cites analysts who interpret this as insider concern about valuation or outlook. Noble's post was accompanied by a chart from macro investor Sam Kovacs showing the stock breaking below the lower trendline of a multi-year rising wedge and trading under its 200-day moving average.
The article concludes that the financing model has genuine business logic but depends on AI demand translating into sufficient revenue and cash flow. Risks could spread across chip suppliers, cloud providers, model developers and financial institutions if utilization or willingness to pay falls short. The article also notes that today's AI industry differs from the 2000 telecom sector, with stronger balance sheets among cloud giants and independent underwriting by financing partners, leaving the comparison open to debate.
Nvidia's second-quarter filing for the period ended July 26, 2026 shows equity investments of $99 billion, with an additional $25 billion in equity commitments and $36 billion in cloud service commitments, typically with six-year terms. Under these arrangements, Nvidia commits to buy back cloud capacity that cannot be resold to third parties. The company also disclosed a $3.5 billion maximum exposure from guarantees on data center leases for certain AI cloud partners. In August 2026, Nvidia signed a guarantee arrangement with SB Energy providing up to $105 billion in credit support for the PORTS Technology Campus in Ohio, which is designed to supply about 4.25 gigawatts of IT load capacity for OpenAI. Combined maximum exposure reaches $108.5 billion. Nvidia states the guarantees cover only specific portions of lease and power payments and may terminate early if OpenAI obtains a satisfactory credit rating.
In August 2026, Nvidia also announced an AI computing infrastructure financing platform with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR, aiming to mobilize more than $500 billion in third-party capital through chip-backed lending. CEO Jensen Huang argues AI compute is productive and durable enough to qualify as a long-term financing asset. Some banks and credit investors remain skeptical, questioning Nvidia's assumed useful life and residual values. Huang has said high-end GPUs can last up to ten years, while banks typically use a three-to-four-year depreciation cycle.
Flow data also draws attention. According to GuruFocus, 20 of 33 tracked top investment masters reduced Nvidia holdings in recent quarters while 12 added. Insiders recorded no purchases over the past year and sold $1.868 billion in shares. The article cites analysts who interpret this as insider concern about valuation or outlook. Noble's post was accompanied by a chart from macro investor Sam Kovacs showing the stock breaking below the lower trendline of a multi-year rising wedge and trading under its 200-day moving average.
The article concludes that the financing model has genuine business logic but depends on AI demand translating into sufficient revenue and cash flow. Risks could spread across chip suppliers, cloud providers, model developers and financial institutions if utilization or willingness to pay falls short. The article also notes that today's AI industry differs from the 2000 telecom sector, with stronger balance sheets among cloud giants and independent underwriting by financing partners, leaving the comparison open to debate.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.