Nvidia (NVDA.US), the global leader in AI computing, is increasingly acting as a "financial facilitator" in massive data center financing deals, including OpenAI's $250 billion project in Ohio. However, rather than cheering, the market is deeply worried about "circular financing" risks: the chip giant is lending money to customers to buy its own chips, potentially leaving itself with future bad debts. In just three weeks, Nvidia's credit risk indicator has nearly doubled.
On Monday, Nvidia CEO Jensen Huang decisively played the "Wall Street card," with a core strategy to bring in external capital, have professional institutions conduct thorough due diligence on transactions, and secure Wall Street's "endorsement." Nvidia mentioned a consortium of six major investment firms, including BlackRock and Goldman Sachs, which are collectively raising over $500 billion to support AI infrastructure construction. This group will independently evaluate each transaction and decide on its participation level, with Nvidia's contribution being relatively limited and only involving some deals. The news provided temporary relief to the tense credit market. On Tuesday, the cost of protecting Nvidia's debt against default fell, and the company's bonds rose, with the risk premium over U.S. Treasuries retreating to the previous week's level. Brett Kozlowski, a portfolio manager at GW&K Investment Management, stated that the commitment from several of Wall Street's largest institutions is "a positive development, removing some uncertainty around infrastructure construction and future customer spending."
Credit Derivatives Sounded the Alarm First: Where Does the Market Concern Come From?
Nvidia is a critical player in the global AI race. Its high-performance computer chips, originally designed for graphics processing, can handle multiple tasks simultaneously, making them ideal for data center use. Strong chip demand has propelled Nvidia to become the world's most valuable publicly traded company, with a market cap exceeding $3.2 trillion. However, investors are increasingly worried about one issue: whether the company's customers are overly reliant on Nvidia's financial support to pay for increasingly expensive chips and data center costs. In late July, media reports stated that Nvidia is in talks with OpenAI to provide up to $250 billion in financing to help OpenAI lease computing power from a data center in Ohio being developed by a SoftBank Group subsidiary. This would be one of the largest financing deals between the chipmaker and its customers. At the time, sources also revealed that Nvidia is discussing providing $350 billion in financing for OpenAI to purchase its chips for the project. Additionally, Nvidia announced a partnership with SK Group to build over 2 gigawatts of data center capacity on the Korean Peninsula, part of a broader $500 billion collaboration with the Korean conglomerate. Huang later clarified that the $500 billion figure primarily represents Nvidia's future expected purchases of memory chips from SK Hynix. Nvidia said the newly raised $500 billion in external funds is unrelated to the SK deal. Asset managers are concerned that the company is engaging in a form of "circular financing" by lending to data center customers to enable them to buy Nvidia chips, boosting current sales, but potentially leading to losses if the AI infrastructure doesn't generate enough future returns. These fears are evident in the credit derivatives market, where investors can buy insurance to get paid if a company defaults. When asset managers become more worried about a company's default risk, the price of this protection rises. In late July, the cost of protecting Nvidia's five-year debt rose to $82,000 per year per $10 million in principal, up from about half that level for most of the previous year. By Tuesday, that figure had fallen to about $73,000, or 73 basis points, a drop of about 4 basis points for the day.
'Wall Street Dream Team' Steps In: How Does External Capital Defuse the Circular Financing Concern?
On Monday, Nvidia announced it had signed a memorandum of understanding with six financial giants—Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs, and KKR—aiming to mobilize over $500 billion in third-party capital for AI infrastructure over the long term. These institutions will independently evaluate each project and decide whether to provide funding. Nvidia stated it is building a "matching platform" to connect independent capital providers with capital seekers, limiting its role to providing platform support. In a statement, Nvidia said the end result would be "a large-scale, competitively priced, dedicated capital pool for Nvidia's customers." The company will support some projects with guarantees of up to 25% and use a mechanism called a "residual value mechanism" to help limit losses if a project falters. While Nvidia didn't reveal many details, it said its chips are widely usable by many customers, which should help minimize potential losses. For example, if a project runs into trouble, the residual value guarantee might mean Nvidia still provides financial support after taking steps to recover value, such as finding a new company to lease the capacity or selling the chips.
Lingering Concerns: In the Capital Expenditure 'Panic Race,' Who Ultimately Pays for the Losses?
However, Alberto Gallo, Chief Investment Officer and Co-Founder of Andromeda Capital, said risks still exist when discussing AI infrastructure construction and the overall credit environment. Gallo noted that the credit market is increasingly becoming a bet on U.S. computing demand and value, and investors in this market may not be getting compensated for the risks they are taking. He pointed out that in data center and other infrastructure, trillions of dollars are pouring in at an astonishing pace, and there will inevitably be winners and losers, so default risks could emerge. Gallo said, "This is essentially a panic-driven capital expenditure. Who bears the losses? Bondholders, life insurance companies, and policyholders." For now, Nvidia's strong profitability provides ample buffer against credit pressure. In the fiscal year ending January 25, its free cash flow was nearly $100 billion. Today, investors are concluding that the likelihood of Nvidia being forced to absorb huge losses has decreased, as external institutions are taking on most of the risk. "No one knew what that potential $500 billion financing meant before," said Sal Naro, Chief Investment Officer of Coherence Credit Strategies. "Now people understand that they are getting everyone involved, and their own exposure is not as severe as investors initially feared."
Comments