To support the costly AI computing power race and maintain robust financial metrics, companies such as Meta, Oracle, and xAI have utilized Special Purpose Vehicles (SPVs) to move over $120 billion in data center debt off their balance sheets, attracting inflows from Wall Street and private credit markets. While this move has protected credit ratings, it amplifies concerns over risk opacity and potential contagion, testing the sustainability of AI-driven prosperity.
To stay ahead in the costly artificial intelligence (AI) arms race while maintaining impeccable financial statements, Silicon Valley tech giants are using sophisticated financial instruments to shift massive infrastructure expenditures off their balance sheets.
According to an analysis by the Financial Times on the 24th, tech companies including Meta, xAI, Oracle, and data center operator CoreWeave have utilized special purpose vehicles (SPVs) to transfer over $120 billion in data center financing debt to Wall Street investors. While this strategy protects the credit ratings of tech giants, it has also raised market concerns about opaque risks and potential financial contagion effects.
In this financing bonanza, financial institutions such as Pimco, BlackRock, Apollo, Blue Owl Capital, and banks like JPMorgan have injected at least $120 billion into compute infrastructure projects structured through SPVs via debt and equity. This innovative financing structure allows companies like Meta and Oracle to access the vast funds needed to build AI data centers without significantly increasing their on-book debt. Such practices were almost unimaginable 18 years ago but have now become an industry norm.
However, this financing boom may mask the actual risks borne by tech groups. Although the SPV structure nominally isolates debt, if AI demand falls short of expectations, who will ultimately foot the bill remains uncertain. Market participants are concerned that if AI operators face financial stress in the future, such pressures could spread unpredictably through the SPV structure across Wall Street and even the private credit markets. Morgan Stanley previously estimated that tech companies’ AI initiatives require approximately $1.5 trillion in external financing, meaning the scale of such financing models is likely to continue expanding.
The Game of 'Perfect' Balance Sheets
Silicon Valley giants have long been known for their abundant cash flow and extremely low debt levels, which have granted them excellent credit ratings and investor trust. However, the race for advanced AI computing power has forced these tech conglomerates to shoulder unprecedented borrowing pressures. To avoid showing excessive leverage on their balance sheets, thereby protecting credit ratings and beautifying financial metrics, introducing private capital through off-balance-sheet structures has become the preferred option.
According to reports, a senior executive of a large financing institution stated that due to their strong credit standing, the tech industry can secure more capital than any other sector. The basic logic of this structure is that tech companies do not borrow directly but instead raise funds for building data centers through SPVs, followed by signing lease agreements with them. In the event of default, lenders' recourse is typically limited to the assets under the SPV’s name—data centers, land, and chips—rather than the parent tech companies managing these sites.
The Specific Operational Paths of the Giants
Meta completed a representative transaction in October last year. By collaborating with New York-based financing firm Blue Owl Capital to establish an SPV named “Beignet Investor,” Meta raised $30 billion for its proposed Hyperion facility in Louisiana, including approximately $27 billion in loans from institutions such as Pimco, BlackRock, and Apollo. This deal allowed Meta to effectively borrow $30 billion without showing any debt on its balance sheet, paving the way for refinancing another $30 billion in the corporate bond market just weeks later.
Oracle is also actively leveraging third parties to construct large debt deals. Larry Ellison’s tech group has reached multiple agreements with partners such as Crusoe and Blue Owl Capital. Among these, Blue Owl and JPMorgan invested approximately $13 billion in an SPV holding Oracle’s Texas data center, including $10 billion in debt financing. Additionally, the company arranged two bundled financings for multiple data center projects located in Texas, Wisconsin, and New Mexico, with respective sizes of approximately $38 billion and $18 billion. In all these cases, Oracle agreed to lease the facilities from the SPVs.
In addition, Musk's startup xAI, as part of its $20 billion financing, has also adopted a similar structure, which includes up to $12.5 billion in debt. This SPV will use the funds to purchase Nvidia graphics processing units (GPUs) and lease them to xAI.
Hidden Concerns in the Private Credit Market
As private equity investors eagerly seek to participate in the AI boom, 'project finance' deals focused on long-term infrastructure financing have surged. According to UBS data, by early 2025, technology companies had borrowed approximately $450 billion from private funds, an increase of $100 billion year-over-year. Since the beginning of this year, around $125 billion has flowed into project financings similar to Meta and Blue Owl transactions.
This trend has heightened concerns about the rapidly expanding private credit industry, which has swelled to $1.7 trillion. Key worries center on surging asset valuations, liquidity shortages, and concentration risks among borrowers. A banker close to data center financing deals noted that risky loans and potential credit risks have accumulated within private credit. Given that the AI data center boom largely depends on a few key clients, such as OpenAI—which has already made over $1.4 trillion in long-term computing commitments in the sector—if major tenants encounter issues, lenders across multiple data centers may face shared risk exposure.
Moreover, these investors also face uncertainties related to power acquisition, risks from AI regulatory changes, or technological iterations rendering hardware obsolete. Wall Street has even begun experimenting with securitizing AI debt, bundling loans for sale to a broader range of investors, such as asset management firms and pension funds, with current deal sizes amounting to billions of dollars.
Risk Exposure and Divergent Strategies
Although the financing structure aims to isolate risks, in many cases, if declining demand for AI services leads to impaired facility values, the ultimate financial risk often still falls on the technology companies leasing the sites. For instance, in the 'Beignet Investor' case, Meta holds a 20% stake in the SPV and has provided other investors with a 'residual value guarantee.' This means that if the data center's value falls below a specific level and Meta decides not to renew its lease, it must repay the capital invested by SPV investors.
Matthew Mish, Head of Public and Private Credit Strategy at UBS, pointed out that while most investors view bearing the risk of 'hyperscalers' as beneficial, SPV financing actually increases the outstanding liabilities of technology companies, implying that their overall credit quality may be worse than what current models indicate.
Not all tech giants have joined this off-balance-sheet financing trend. Companies like Google, Microsoft, and Amazon, which had mature data center operations before the AI boom, continue to primarily fund construction through cash or direct bond issuance and have not disclosed any significant SPV financing plans. The divergence in market strategies reflects differing considerations regarding risk control as players face the high-stakes gamble of AI investments.
Editor/jayden