SpaceX拟通过巨额融资租赁加速AI算力布局
SpaceX拟通过巨额融资租赁加速AI算力布局
Conclusion
SpaceX’s $40 billion GPU financing is not yet a completed $40 billion cash loan. The company has a binding commitment to enter a large finance lease, but its SEC filing says the lease remains subject to closing conditions and that the rate will be set later. If completed, it could accelerate SpaceX’s AI-computing buildout without paying for all the equipment upfront—but it would add sizeable multi-year payment obligations. Whether that works for shareholders depends on SpaceX turning the added compute capacity into durable cash flows.
Data as of October 27, 2026 (Asia/Shanghai). The latest structured U.S. close available was October 26: SPCX was down that session (). That single-day move does not establish that the financing news caused it.
What is known about the plan
SpaceX’s SEC filing says it signed a binding commitment letter on October 24 to enter into a master finance lease with an Apollo-managed entity. The lessor would acquire GPU pods and related equipment and lease them to SpaceX for use at data centers. SpaceX says the equipment will include Nvidia Vera Rubin chips. The company’s announcement names Apollo and Goldman Sachs as leading financing parties, with BlackRock-managed funds, Goldman Sachs Asset Management, and KKR-advised accounts among the anchor commitments.
The filing expects aggregate principal payments of approximately . Each equipment tranche would have a lease term of about five to five-and-a-quarter years, with rent paid quarterly in advance. Rent is to include principal repayment, interest at a fixed rate, and certain costs. That fixed rate is not yet specified in the filing; pricing is to occur before SpaceX enters the master lease. The lease is expected to close before year-end, subject to conditions, with schedules beginning when equipment is delivered in the .
That makes the description “borrowing to buy GPUs” broadly understandable but incomplete: the disclosed structure is a finance lease, and the $40 billion refers to expected principal payments, not a disclosed all-in total including interest and costs. Elon Musk publicly described the financing as closed, but the SEC filing describes a commitment to enter the lease, with closing still conditional. Bloomberg also reported that pricing discussions were ongoing.
How it could affect SpaceX
Potential benefit: Leasing can put equipment to work sooner while avoiding an equivalent immediate cash purchase. SpaceX says the compute will support Grok and cloud-services customers; management has also claimed payback periods of on its SI infrastructure investments. That is a company claim, not a guarantee of realized returns.
The scale is substantial relative to existing obligations. At September 30, reported debt and finance leases included current amounts of $2,698 million and non-current amounts of $36,146 million, totaling . Calculation: ( + ) ÷ 1,000 = billion. The contemplated lease principal is of that existing total. Calculation: ÷ = . This is a scale comparison—not a pro forma debt figure: the lease is not yet closed, and its ultimate accounting and cash-payment profile are not fully disclosed.
SpaceX is already spending heavily to build capacity. For the nine months ended September 30, it reported operating cash flow of and PP&E purchases of . The simple difference is . Calculation: ( − ) ÷ 1,000 = billion. This is only operating cash flow less PP&E purchases—not a formal free-cash-flow measure—and excludes other investing and financing flows.
The AI business has meaningful activity, but the buildout is large: in the third quarter, SI reported of revenue, of operating income, and of capex. These figures suggest growing commercial traction alongside substantial investment; quarterly capex and operating income are not directly comparable measures of returns.
If the lease closes, it avoids immediate share issuance but commits SpaceX to scheduled cash payments. That may reduce flexibility if AI demand, customer utilization, or cash generation falls short; meanwhile, GPU technology could age faster than the lease balance runs down.
What investors should track
- Closing and final terms: whether the lease closes, its final fixed rate, total rent including interest and costs, and any important restrictions or recourse provisions.
- Delivery and readiness: whether the GPUs arrive on schedule and whether data centers have the power, cooling, networking, and other infrastructure needed to use them.
- Commercial returns: compute utilization, customer commitments and collections, and whether SI revenue and operating results improve enough to support lease payments. The company’s payback claim should be tested against reported cash generation over time.
- Funding pressure: changes in debt and finance leases, interest costs, cash flow, and further capital spending.
- Asset economics: how useful and valuable the equipment remains over the multi-year lease term.
The core trade-off is clear: the lease could speed SpaceX’s move into AI compute, but the financing does not by itself prove that the buildout will earn adequate returns. The unpriced lease, conditional closing, delivery schedule, and future cash generation remain central uncertainties.