Nvidia’s $200bn AI Financing Web Sparks Credit Concerns

Key points

• Nvidia reported fiscal 2027 Q2 revenue of $96.2 billion, with data center revenue rising 117% to $89 billion.
• The company has committed to financing AI infrastructure, including a $105 billion Ohio campus for OpenAI and partnerships for up to $500 billion in data center funding.
• Morgan Stanley estimates Nvidia’s total credit exposure could reach $200 billion by 2028, with $170 billion tied to customer backstops.
• Critics label these arrangements 'circular financing,' arguing they artificially inflate demand by having Nvidia fund customers who then buy Nvidia chips.
• Jensen Huang argues the risk is low because Nvidia’s computing infrastructure can be redeployed to other customers if a partner struggles.

Background

Recent market volatility has seen mixed reactions to tech earnings, with Nvidia previously beating estimates alongside strong results from Salesforce and Okta. This follows a period where AI-driven capital expenditure has boosted Wall Street bank revenues, highlighting the broader financial integration of AI infrastructure.

How outlets are covering it

CNBC highlights Jensen Huang’s defense of Nvidia’s strategy, emphasizing that AI startups require unprecedented capital and that Nvidia’s investments are 'once-in-a-generation' opportunities. In contrast, the Financial Times, citing Morgan Stanley, frames this as 'balance-sheet-as-a-service,' warning that conventional leverage metrics understate the risk. Morgan Stanley analysts argue that while Nvidia’s balance sheet remains strong, the opacity of these contingent obligations creates significant tail risks. The outlets diverge on the sustainability of these deals: CNBC focuses on the strategic necessity of funding AI, while the FT focuses on the potential for artificial demand and credit exposure.

Why it matters

Nvidia’s financial support for AI companies could reshape the tech industry’s capital structure. If these arrangements are deemed circular financing, it could signal a bubble in AI infrastructure spending. Conversely, if the model succeeds, it could accelerate the deployment of AI globally, but at the cost of increased systemic risk for major financial institutions.

What to watch

Investors will watch for clarity on Nvidia’s contingent liabilities as the company finalizes its 'repeatable financing platform' agreements with firms like Apollo and BlackRock. Rating agencies may reclassify these structures as debt, potentially impacting Nvidia’s credit rating. The market will also monitor whether AI demand outpaces supply, validating the revenue guarantees provided to neocloud providers.