Tech • AI • Robotics • Game
Should investors worry about Nvidia's “circular financing"? | The Economist
9.4/10TL;DR
Nvidia is using investments, credit support and other financing tools to expand the AI market and defend its customer base, but the scale of those commitments is raising concerns about whether demand is being enabled or artificially sustained.
KEY POINTS
Why the alarms are sounding
Critics are focusing on a form of financing in which one company backs another and the recipient then uses part of that support to buy the first company’s products. In Nvidia’s case, the concern is that investments in smaller AI companies may be helping fund purchases of Nvidia chips, blurring the line between cultivating future customers and engineering current sales.
Pressure from the biggest customers
A major reason for the strategy is that some of Nvidia’s largest buyers are also becoming its strongest long-term competitors. Amazon, Google, Meta and Microsoft account for roughly half of Nvidia’s revenue, yet each has been developing more of its own custom chips, threatening future dependence on Nvidia hardware.
A vast support network
Over the past three years, Nvidia has pledged more than $70 billion in startup investments and offered as much as $300 billion in financial support to customers. That support can help younger companies obtain financing, scale their operations and purchase the expensive computing infrastructure needed to train and run advanced AI systems.
Why startups need Nvidia’s help
Large technology groups generally have investment-grade credit ratings, allowing them to borrow at relatively low cost. Smaller AI companies often face similar spending demands but far higher borrowing costs because lenders view them as riskier. By backing those firms, Nvidia can lower financing barriers and build a broader customer base outside the hyperscalers.
The case for acting like AI’s banker
Supporting independent customers gives Nvidia a strategic hedge against concentration risk. If the hyperscalers gradually shift more workloads to in-house chips, a larger ecosystem of venture-backed and mid-sized buyers could help preserve demand for Nvidia’s accelerators and related systems.
The scale of the risk
The sheer size of the commitments has prompted worries about hidden leverage in the sector. By some calculations, if all of Nvidia’s obligations were drawn at once, the resulting shadow debt could exceed $300 billion, a level that would place the company behind only America’s six largest banks by debt load.
Why the exposure still looks manageable
For now, Nvidia’s financial position remains unusually strong. The company holds nearly $100 billion in cash and continues to generate substantial new cash from its core chip business, giving it a larger cushion than most industrial or technology companies would have under similar commitments.
An industry-wide pattern
Nvidia is not the only company using complex financing to stimulate AI demand. AMD, Google and Broadcom have also pursued arrangements designed to help customers fund infrastructure purchases, suggesting that competitive pressure is pushing the industry toward increasingly aggressive deal structures.
What could go wrong
The central danger is not necessarily an immediate collapse but a mismatch between industry ambitions and what capital markets can sustainably support. If demand growth slows, even without a crisis, financing chains could come under pressure. Calls from Dario Amodei and Sam Altman for slower AI development underscore that a moderation in spending alone could send broader shockwaves through the sector.
CONCLUSION
Nvidia’s financing strategy reflects both strength and vulnerability: it is rich enough to underwrite the growth of the AI ecosystem, but dependent enough on that growth that any slowdown could expose the risks embedded in the system. Investors may not need to panic, but they do have reason to watch the company’s expanding financial commitments as closely as its chip sales.
Ask a question