Breaking: 2025.02.10 14:23 EST — Nvidia just signed a $105 billion conditional lease guarantee for OpenAI's PORTS-Pike Technology Campus in Ohio. The filing dropped 47 minutes ago. This isn't a loan. It's a residual value guarantee covering 4.25 gigawatts of IT load, with an option on another 3.75. The market is reading this as a bullish signal for AI compute demand. I'm reading it as a credit event masquerading as infrastructure.
Context: Why Nvidia Is Guaranteeing Leases Instead of Lending
SB Energy will build, own, and operate the campus under a 20-year lease to OpenAI. Nvidia signed multiple residual value guarantees covering roughly 4.25 gigawatts of information technology load. The filing clarifies: if OpenAI goes insolvent or stops paying rent, Nvidia covers the shortfall between a guaranteed minimum lease value and whatever SB Energy recovers by reletting or selling the space. OpenAI has agreed to reimburse Nvidia for any amount Nvidia actually pays the lessor. The guarantee terminates once OpenAI achieves a satisfactory credit rating.
That termination clause is the tell. The guarantee lapses once OpenAI's own credit is strong enough to support the leases without Nvidia. Right now, OpenAI doesn't have that credit. Nvidia is acting as a credit enhancer, not a lender. This is a structural shift: Nvidia is moving from chip supplier to financial intermediary. Based on my 2017 Parity multi-sig audit experience, I've learned that guarantees are only as strong as the guarantor's ability to absorb the loss. Nvidia's $1.5 billion equity investment in SB Energy is separate — it's a capital injection, not a cushion for the guarantee.
Core: The Mechanics of the Guarantee and the Hidden Leverage
Nvidia secured the initial 4.25 gigawatts and holds an option, at its sole discretion, over the remaining 3.75 gigawatts. Capacity is expected to come online in phases beginning in 2028. OpenAI will run Nvidia's full-stack DSX platform at the site, and Nvidia becomes the exclusive compute provider there. SB Energy and SoftBank will build at least 10 gigawatts of new generation and invest at least $4.2 billion in regional grid infrastructure with AEP Ohio.
The numbers are staggering. $105 billion in conditional lease obligations. That's roughly 8x Nvidia's current net income. The guarantee is contingent on OpenAI's solvency. If OpenAI defaults, Nvidia must cover the difference between the guaranteed minimum lease value and what SB Energy can recover. The reimbursement clause from OpenAI is only valuable if OpenAI has assets to reimburse with. In a bankruptcy scenario, that reimbursement would be an unsecured claim.
From my 2020 Yearn.finance analysis, I've seen how credit enhancements can mask underlying liquidity risks. When yield farming vaults promised 15% APY, the underlying protocols were often levered 3x against volatile collateral. Here, the collateral is Nvidia's balance sheet. The risk is not technical — it's counterparty. The guarantee terminates once OpenAI achieves a satisfactory credit rating. That implies OpenAI's current credit rating is not satisfactory. The market is pricing in a 2028+ horizon where OpenAI is either a trillion-dollar company or a cautionary tale.
Let's break down the numbers. The 4.25 GW of IT load at $105 billion works out to ~$24.7 million per MW. That's high for a data center build, but includes the cost of Nvidia's GPU clusters and infrastructure. The option for 3.75 GW brings total potential exposure to ~$197 billion if fully exercised. Nvidia's market cap is ~$2.5 trillion. The guarantee represents about 4% of its market cap now, but 8% of net income. In a worst-case scenario where OpenAI defaults and SB Energy recovers only 50% of the lease value, Nvidia would be on the hook for ~$50 billion. That's a risk that cannot be hedged easily.
Contrarian: The Blind Spot — Nvidia Is Underwriting OpenAI's Solvency
The narrative is that Nvidia is securing long-term demand for its compute. CEO Jensen Huang described AI as infrastructure and called land, power, and shell capacity vital to scaling it. 'We are securing long-lived infrastructure for NVIDIA compute so OpenAI can deploy the most productive AI factories that can be upgraded repeatedly with each new generation delivering more intelligence and better economics,' he said. That sounds like a partnership. But the guarantee structure reveals a different reality: Nvidia is taking on credit risk that banks refused to underwrite.
Why didn't OpenAI issue bonds or take a traditional loan? Because the terms would have been punitive. Nvidia's guarantee allows SB Energy to finance the build at a lower cost, indirectly subsidizing OpenAI's capital expenditure. The '17 Parity multi-sig vulnerability taught me that trust is a balance sheet item. When you guarantee someone else's lease, you are effectively saying 'I trust this counterparty more than the market does.' The market is pricing in a non-trivial probability of OpenAI's failure. Nvidia is betting against that probability.
Here's the contrarian angle: this guarantee creates a single point of failure for the entire AI compute supply chain. If OpenAI goes under, Nvidia is left holding a huge lease obligation for a data center that can only run Nvidia's hardware. The site is designed for Nvidia's DSX platform. Reletting it to a competitor would require ripping out the entire infrastructure. The residual value of that space is tied to Nvidia's own dominance. It's a circular guarantee: Nvidia guarantees OpenAI's lease, but the lease's value depends on Nvidia's continued market leadership. The BAYC crash wasn't a liquidity event; it was a credit event. The same dynamic applies here.
Takeaway: The Real Question Is Concentration Risk
This isn't a story about AI demand. It's a story about credit engineering. Nvidia is using its balance sheet to create a captive compute market for itself. The guarantee effectively lowers OpenAI's cost of capital, but at the expense of Nvidia's credit flexibility. The $105 billion figure is a ceiling, not a floor. The actual exposure depends on how much of the option is exercised and how the recovery market performs in 2028+.
The forward-looking question is: what happens when the next AI startup doesn't have a Nvidia guarantee? The market will start pricing credit risk into compute contracts. This deal sets a precedent. The '17 reveal taught me that the true cost of trust is measured in basis points of default probability. Nvidia is betting that OpenAI's default probability is zero. History suggests otherwise. Speed without precision is just noise; the '17 reveal taught me that speed with precision is the only edge. This guarantee is speed — $105 billion fast. But the precision will only be validated in 2028 when the first phase goes live. Until then, I'm watching the unsecured debt markets for OpenAI's credit default swaps. If they spike, the guarantee is a ticking time bomb.
