Silicon Valley can't hold on, leverages Wall Street to start the "AI arms race" spreading, risks also!

The AI "arms race" among tech giants is evolving into a complex financial game. With annual capital expenditures reaching hundreds of billions of dollars becoming the norm, even tech giants like Amazon, Google, Meta, Microsoft, and Oracle, with cash reserves exceeding $340 billion, are beginning to feel unprecedented financial pressure. They are breaking away from the traditional reliance on their own funds to build infrastructure, turning to Wall Street for more complex financial solutions to provide ammunition for this expensive race without compromising their financial stability, and risks are coming with it. Trillion-dollar giants' "sweet burden": AI infrastructure costs driving financial innovation In the past, tech giants were accustomed to using their large internal cash flows to build data centers, but the rise of AI has completely changed the rules of the game. The speed and scale of this race force them to seek external capital. Investors and credit rating agencies are closely monitoring how these tech giants will foot the bill for AI data centers and whether these massive investments can new revenue. In order to maintain healthy financial statements while aggressively expanding, tech giants are teaming up with bankers to design increasingly complex financial strategies with the sole goal of transferring some costs and risks off their balance sheets. In this context, financial instruments like joint ventures, backstop agreements, and syndicated debt, which were rarely heard of in the tech world before, are now being brought to the table one by one. Three financial "plays" for risk sharing In the exploration of externalizing risks and costs, three innovative financial "plays" have emerged, all centered around cleverly externalizing risks and liabilities. 1. Meta's "off-balance" strategy: Joint ventures Meta initiated a financing of up to $29 billion for its data center project named "Hyperion" in Louisiana. The core structure involves the establishment of a joint venture with investment firm Blue Owl Capital. Blue Owl contributes $3 billion in equity, while the massive $26 billion debt required for the project is syndicated by bond giant Pimco with assistance from Morgan Stanley. The key to this structure is that Meta will repay the debt in the form of lease payments in the future, thereby moving the entire project off its balance sheet and controlling the debt level. 2. Oracle's "risk sharing": Syndicated debt As the fourth-largest global cloud service provider, Oracle recently agreed to become a tenant in a 1.4GW data center complex being developed by Vantage Data Centers, one of the largest projects under construction globally. Due to the massive scale of the project, developer Vantage is collaborating with six banks led by JPMorgan Chase and Mitsubishi UFJ Financial Group, including Goldman Sachs, to syndicate the $22 billion debt required for the project. This model spreads the risk among multiple lenders, reducing the exposure of individual institutions and making massive financing possible. 3. Google's "sophisticated design": Backstop guarantees Google's solution is the most complex and sophisticated, involving "backstop guarantees." In this transaction, Google provides a $3.2 billion backstop guarantee for the leasing contract between cloud startup Fluidstack and data center owner TeraWulf, and in return, gains a 14% stake in TeraWulf. The clever design of this guarantee is that it belongs to "contingent liabilities," which will only be triggered if Fluidstack defaults, likely allowing Google to not include it in current liabilities. With Google's support, TeraWulf raised $1 billion through convertible bonds underwritten by Morgan Stanley and Cantor Fitzgerald last month, more than double its initial financing target. TeraWulf's CFO said in a conference last month: "It's not easy to get a $2 trillion company, their management team, board, and everyone to agree on a novel concept, but I think we've provided a roadmap." Concerns beneath the fervor: overheating, concentration, and leverage risks The huge financing needs of tech giants happen to coincide with a credit market flush with cash. Private credit funds with hundreds of billions of dollars waiting to be invested, and banks increasingly confident in projects with "investment-grade" tenants, are actively entering the market. The loan-to-total cost ratio for data center projects has significantly increased compared to the past. Jason Tofsky, Global Head of Goldman Sachs Digital Infrastructure Banking, stated that loan providers are willing to provide 80% to 90% of the total funds needed for data center projects. Data from real estate company Jones Lang LaSalle (JLL) shows that lenders for data centers typically provide 65% to 80% of the total costs for new development projects. Tofsky said: "There is enough capital in the market to support projects that are well-known in the market. The market is well-positioned to absorb these projects." However, the fervor of capital is giving rise to new risks. First is the risk of market overheating. Analysts at UBS warned in a report last month that the influx of private credit into the data center sector, while driving AI development, could also "increase the risk of market overheating." Second is the risk of high concentration. Data center leasing contracts are highly concentrated in the hands of a few tech giants with good credit. This raises concerns: if any of these companies reduce spending due to strategic adjustments or suffer credit rating downgrades, the entire ecosystem will face significant risks. Finally, the leverage risks of some companies are already apparent. Rating agencies Moody's and S&P warned Oracle in July that its leverage ratio (currently at 4.3 times) during the phase of entering AI infrastructure construction is much higher than other "super-scale" suppliers, and if it does not reduce the debt-to-earnings ratio to below 3.5 times, its credit rating could face downgrade risks. A Moody's analyst wrote in a credit report: "Although several other super-scale suppliers are building AI infrastructure, none have as high leverage and negative cash flow as Oracle at this stage."