Nvidia's plan to mobilize $500 billion in third-party capital for AI infrastructure has drawn a sharp warning from DoubleLine Capital CEO Jeff Gundlach, who says turning fast-depreciating chips into collateral for long-term debt is a classic market-top signal.
Nvidia's $500 billion AI infrastructure financing pact with six Wall Street institutions is redefining accelerated computing as an investable asset class, but Gundlach warns of a maturity mismatch between GPU depreciation and long-term debt.
"Assets of unknown life as collateral for long-term debt will not likely age well," Gundlach, chief executive officer and chief investment officer at DoubleLine Capital, wrote on X, comparing the approach to issuing 30-year asset-backed securities against warehouses of "newly engineered bananas of unknown life."
Nvidia signed memoranda of understanding with Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs, and KKR to establish compute-financing platforms targeting more than $500 billion in third-party private capital. Nvidia could provide up to 25 percent in capital backing or credit support, or as much as $125 billion, to absorb initial risk and unlock the remaining 75 percent from institutional asset managers, insurers, and pension funds.
The financing framework marks the formal transition of AI infrastructure from equity-funded corporate capex to a credit-backed asset class, with implications for hyperscaler balance sheets, private credit markets, and decentralized compute networks. If the model succeeds, Nvidia's dominance could extend beyond silicon into the cost of capital for AI infrastructure; if it fails, lenders holding depreciating GPU collateral could face losses that ripple through credit markets.
Gundlach's "banana ABS" analogy targets the structural weakness in the financing model. AI processors can generate strong cash flow while demand is high, but given the rapid pace of AI hardware iteration, today's chips could lose value before the loans that financed them are paid off. The DoubleLine CEO also questioned the reliability of credit ratings that such transactions might depend on, suggesting that financial innovation narratives built on "questionable" credit ratings are a recurring feature of risk-market tops.
Mark Cuban offered a terse comparison of his own: "Chips as an asset class will be the new crypto." Michael Burry, the investor known for "The Big Short," has been publicly bearish on parts of the AI infrastructure trade, criticizing the financing push as a Wall Street stunt and drawing parallels with circular financing structures seen during previous bubbles.
Nvidia's counterargument rests on the durability and transferability of its compute infrastructure. Jensen Huang, Nvidia's chief executive officer, has formally classified accelerated computing clusters as productive infrastructure assets that generate recurring cash flows, maintain durability, and offer high secondary market fungibility. The CUDA software stack reinforces this thesis — enterprise GPU clusters operate within a global developer standard, ensuring that if a borrowing tenant experiences operational difficulties, the underlying hardware can be transferred to other waiting cloud operators.
Reuters Breakingviews compared the concept with auto lending, where a lender is more willing to provide financing because the vehicle has a known resale value if the borrower defaults. The analogy works only if the collateral retains its value. Cars depreciate predictably; AI chips operate in an industry where a new generation of hardware can change the economics remarkably quickly. A serious competitive breakthrough from AMD, custom silicon from hyperscalers, or another architecture could affect not only Nvidia's future sales but also assumptions about the collateral value sitting underneath existing financing.
Bank of America analyst Vivek Arya described the initiative as a move away from traditional vendor financing because most of the burden is intended to sit with the consortium rather than Nvidia's own balance sheet. Goldman Sachs can provide junior capital and private credit while also placing debt with other investors.
But the risk does not disappear because it moves off Nvidia's balance sheet. It ends up with insurers, private-credit funds, banks, infrastructure investors, and whoever ultimately owns securities backed by the projects. The structure being discussed would help create an asset-backed market for AI compute, potentially allowing that debt to trade more like other financial securities.
The scale of the capital requirement is staggering. Goldman Sachs estimates the four largest hyperscalers could spend more than $5 trillion on technology and data centers through 2030. Nvidia has already invested $30 billion in OpenAI and has been involved in discussions around financial backing for a massive OpenAI-related data center project in Ohio, reducing its initial guarantee from as much as $250 billion to less than $120 billion.
For investors, the key question is whether the financing works because the underlying economics are attractive or whether increasingly generous financing is required to keep projects moving. Nvidia's 5-year credit default swap spreads have widened, reflecting market repricing of the massive capital mobilization. While wider CDS spreads do not indicate structural distress, an extended period of high interest rates or cash flow shortfalls among leveraged AI cloud startups could create credit stress across junior debt tranches.
The financialization of physical compute also establishes a pricing baseline for decentralized physical infrastructure networks (DePIN) and Web3 AI protocols. As hundreds of billions of dollars in debt capital are deployed into centralized data centers, the baseline rental cost of compute becomes anchored by contractual debt service and energy tariffs, giving decentralized compute networks a transparent benchmark to compete against.
This article is for informational purposes only and does not constitute investment advice.
