Oracle Tamps Down AI Data Center Concerns With Big Earnings Beat

Oracle has emerged as the epicenter of Wall Street’s uneasiness around Big Tech’s massive spending on artificial intelligence infrastructure. But the firm this week revealed strong earnings numbers, particularly for its cloud and AI infrastructure businesses, that have helped ease those fears — at least for now.

The Austin-based tech giant on Tuesday reported 22% year-over-year revenue growth for its fiscal third quarter, which ended Feb. 28. Its $17.2B of revenue was above Wall Street's expectations of $16.9B. 

Oracle's earnings beat was driven largely by the performance of its embattled cloud and AI infrastructure business lines. Total cloud revenue was up 44% year-over-year, ahead of expectations, while $4.9B in cloud infrastructure revenue represents an 84% jump from the year prior. 

The company also dramatically raised revenue projections for its fiscal year 2027, which begins in June. It now expects revenue for that year to jump 34% — more than double this year’s expected growth. 

Oracle’s strong quarter provided a reprieve from what has been a difficult few months for the company on Wall Street.

Although every Big Tech firm is spending hundreds of billions of dollars building data centers and other infrastructure to support AI, Oracle’s AI strategy has been perceived as particularly high-risk. Facing investor concerns about rising debt levels and the price tag of its artificial intelligence push, the company’s share price had lost roughly half its value since September. 

Yet Wall Street reacted to the results with enthusiasm.

Oracle shares jumped nearly 13% Wednesday before closing the day up 9.2%. Pricing of Oracle credit default swaps — regarded as a measure of the firm’s credit risk — fell to their lowest point since the first week of February. 

Beyond the implications for Oracle itself, some analysts say the firm’s results are a positive bellwether for the health of the AI infrastructure boom as a whole. 

“Oracle's quarter is a beat and a stress-test result for the AI trade,” Emarketer analyst Jacob Bourne told Reuters. “As the most debt-exposed major player in AI infrastructure, Oracle is the canary in the coal mine, and this report suggests there's underlying health in AI spending beyond the hype.”

While Oracle’s quarterly results have spurred a wave of positivity around the firm, the conversation heading into Tuesday was firmly focused on questions about the company’s financial stability and whether its data center spending spree could become an albatross it can't shake. 

Oracle spends far more on AI infrastructure relative to its size than Amazon Web Services and Google parent Alphabet, and it is saddled with much more debt. The company's debt-to-equity ratio earlier this year reached 432%, while Alphabet's stood at 11.4%. 

Oracle is also facing questions about its partnership with OpenAI, particularly around the high-profile Stargate data center development effort.

Last week, Bloomberg reported that Oracle and OpenAI have canceled plans for an 800-megawatt expansion of the flagship Stargate data center campus in Abilene, Texas, due to financing challenges and changes to OpenAI’s demand forecast. The report, which also revealed reliability problems at the Stargate site, was seen by some high-profile investors as a sign of demand uncertainty and wariness among the project’s lenders. 

Oracle released a statement Monday denying aspects of the report. 

Its leadership didn't address the reduced scale of the Abilene campus on its quarterly earnings call Tuesday. But executives did try to assuage the underlying concerns about demand for its AI infrastructure, its pathway to profitability and the debt underpinning the build-out. 

They emphasized the rapid revenue growth being generated through its cloud and AI infrastructure investments. The firm’s 84% increase in quarterly cloud infrastructure revenue marks an acceleration from the 68% growth rate a year earlier. Meanwhile, AI-specific infrastructure revenue grew 243% year-over-year.

Although Oracle delivered more than 400 MW of capacity for customers last quarter, Oracle CEO Clay Magouyrk emphasized that the company has “demand that exceeds supply and a clear execution plan from Oracle that will rapidly turn that demand into profitable recurring revenue.”

To meet this demand, Oracle plans to add 10 gigawatts of power and data center capacity over the next three years. The firm’s leadership told analysts that build-out is more than 90% funded already and that the remainder can be funded without taking on more debt or issuing equity. 

This is largely because Oracle’s AI-driven development surge is becoming significantly more cost-efficient, according to Magouyrk.

He said the amount of capital that needs to be deployed per megawatt is decreasing as the company evolves its approach to data center construction and improves supply chains. Oracle is also changing its deal structure with infrastructure customers, signing $29B in deals in which the customer finances its own graphics processing units and other hardware.  

Magouyrk and other Oracle executives also laid out a road map toward a return on their massive AI investments, highlighting the 32% gross margin the firm generated from AI capacity delivered last quarter.

Ultimately, they said the firm’s AI infrastructure business is on a path to profitability as its data center capacity grows in scale and as AI continues to be adopted by corporations and consumers. 

“Investing in the AI infrastructure is capital-intensive, but our operating model is optimized to ensure profitability,” Magouyrk said. “Altogether, we are confident that the investments we make now in data centers, compute capacity and customer relationships will only grow more valuable over time.”