Oracle's annual report highlights the investment risks in data centers. Is the US stock market's AI narrative facing a stress test?

According to Zhitong Finance APP, cloud computing giant Oracle (ORCL.US) has issued a new warning to investors in its annual financial report, stating that all investments in data centers may not yield returns. This information was disclosed in Oracle's annual financial report, which detailed the company's plans to invest heavily in artificial intelligence infrastructure for clients such as OpenAI. The report also highlighted various risks associated with this expensive investment.

Oracle pointed out that the construction of data centers may ultimately be more costly or time-consuming than expected. This could be due to supply chain issues, government restrictions on data center construction, or third parties failing to complete projects on schedule.

Oracle also stated that if it cannot acquire data center capacity at an affordable price, or if it fails to accurately plan and manage its own infrastructure capacity needs, its profitability may decline. The company noted that to develop its Oracle Cloud Infrastructure (OCI) business, which requires more computing power, it must invest significant capital expenditures and operational expenditures to expand existing data center capacity and build new data centers in new geographic areas.

The company added that as part of these investments, it has already— and expects to continue—entering into long-term lease commitments with third-party data center providers and signing various commitment agreements with chip and other data center infrastructure suppliers. Oracle pointed out that if it underestimates customer demand or its own data center capacity needs, it may face a shortage of available infrastructure, thereby limiting its ability to support customer growth and potentially leading to business losses to competitors.

In its submitted 10-K annual report, Oracle stated, "Conversely, if we overestimate customer demand, or if any key customer is unable to pay or fails to perform under contract, we may be locked into commitments for excess data center space over multiple years and related capital expenditures, with financing arrangements that may not generate corresponding revenue. Additionally, the terms, renewal options, and price adjustments in our long-term data center leases typically do not align with the duration and pricing of customer contracts; if customers do not renew, we may be unable to sublease, repurpose, or transfer such capacity on acceptable terms, or even dispose of it altogether."

Most large tech companies list risks associated with data centers in their financial reports.

Only six companies—including Oracle, Microsoft, and Meta—have committed to investing $850 billion in data center leases that have yet to be initiated. Oracle holds the largest share of these commitments due to its $300 billion "Stargate" contract with OpenAI.

As part of the Stargate project in collaboration with OpenAI, Oracle is developing hyperscale data centers across the United States to provide cloud computing power.

"Our leverage with some customers may be high, and they face their own operational and regulatory risks; even if our credit review and analysis mechanisms are functioning properly, we may still encounter debt and default risks in transactions with such parties. We have a high concentration of several large customers on certain OCI products, which may exacerbate the aforementioned risks," Oracle stated in the document

Concerns in the U.S. Stock Market Amidst the AI Capital Expenditure Frenzy

If Oracle reduces its artificial intelligence (AI) capital expenditures due to returns falling short of expectations, this would not merely be a financial adjustment for one company in the current Wall Street environment, but could very likely trigger a symbolic event marking the "domino collapse" of the entire AI supercycle.

Should AI capital expenditures decrease, the impact on the U.S. stock market would be structural rather than temporary. In the short term, it could lead to a reevaluation of valuations in the AI infrastructure sector and overall volatility in tech stocks; in the medium term, if more cloud providers reveal similar "input-output" mismatches, the market will fundamentally question the sustainability of the AI supercycle, thereby shaking the core narrative of the current U.S. stock market bull run.

If Oracle cuts capital expenditures, the most direct victims would be its upstream suppliers. The market's doubts about AI's "monetization capability" would be thoroughly validated, triggering a systemic collapse of valuation logic.

Chip and server giants like NVIDIA, AMD, and Supermicro would face a double whammy of "performance and valuation." What has supported these companies' trillions of dollars in market value in the past has been the "unlimited" order commitments from major cloud giants. If Oracle cuts orders, it would mean that upstream orders have peaked or even declined, and the Philadelphia Semiconductor Index could repeat the "stampede squat" warned by Bank of America.

Data center REITs (such as Equinix and Digital Realty), utility power stocks, and industrial component suppliers (such as liquid cooling and transformer manufacturers) would also cool down accordingly.

Over the past year, mega-scale companies like Microsoft, Meta, Google, and Amazon have found themselves in a "prisoner's dilemma" style arms race—none dare to reduce AI investments for fear of losing out on the future.

Oracle's retreat would provide a wake-up call to other tech giants. Boards and aggressive investors on Wall Street would use this as a reason to pressure Microsoft or Meta to similarly cut back on those massive capital expenditures that yield "only investment, no profit." Once the entire industry reaches a consensus on "cutting expenses to preserve profits," the "core engine" that has supported the U.S. stock market bull run over the past two years would come to a halt.

Given that the weight of tech giants in the S&P 500 and Nasdaq 100 indices has reached its highest point in 40 years, a correction in tech stocks is bound to drag down the broader market, but it will also give rise to a new market landscape.

However, this does not mean the failure of AI technology; rather, it signifies that the market, amidst high margin debt and leverage ratios approaching historical peaks, must undergo a thorough liquidation. The U.S. stock market will experience a painful pullback and bubble deflation in the short term, subsequently forcing the entire tech industry to shift its focus from "madly buying chips and building data centers (hardware)" to "how to truly help customers save money and make money with AI (software and application monetization)." This is also an inevitable growing pain on the path to a more moderate bull market in the U.S. stock market