Oracle's December 2025 earnings report exposed one of the central tensions in the artificial-intelligence infrastructure boom: demand for computing capacity was accelerating rapidly, but the investment required to satisfy that demand was putting extraordinary pressure on cash flow and financing.
The company reported fiscal second-quarter 2026 revenue of $16.1 billion, up 14% from a year earlier. Cloud revenue reached $8.0 billion, while Oracle Cloud Infrastructure revenue surged 68% to $4.1 billion. Remaining performance obligations, a measure of contracted future revenue, climbed 438% year over year to $523 billion.
Those growth figures were substantial. What unsettled investors was the cost of building enough data-center capacity to turn that backlog into revenue. Oracle raised its fiscal 2026 capital-expenditure outlook to about $50 billion, roughly $15 billion above the guidance it had given after the previous quarter. The combination of higher spending, negative free cash flow and forecasts that did not fully satisfy Wall Street helped trigger a broad technology selloff.
What Oracle Actually Told Investors
Oracle's underlying cloud business was growing quickly when the selloff began. According to the company's official fiscal 2026 second-quarter results, total cloud revenue rose 34%, infrastructure-as-a-service revenue grew 68%, and the company's backlog increased by $68 billion sequentially during the quarter.
Oracle also said new commitments from customers including Meta and Nvidia helped lift remaining performance obligations to $523 billion. That backlog gave the company considerable visibility into future demand, but it did not eliminate the financing and execution risks involved in constructing enough infrastructure to serve those contracts.
The shift in capital spending was therefore central to the market reaction. Oracle expected fiscal-year capital expenditures of approximately $50 billion, up from an earlier forecast of roughly $35 billion. The increase helped revive concerns about whether the enormous investment flowing into AI infrastructure would generate returns quickly enough. Reuters described the market's reaction and renewed discussion of AI valuations in plain language.
Why Wall Street Focused on Cash Flow
The most important pressure point was not simply the size of Oracle's revenue forecast. It was the gap between operating cash generation and the amount of money flowing into new infrastructure.
During the November quarter, Oracle spent about $12.0 billion on capital expenditures while generating roughly $2.1 billion in operating cash flow. That implies quarterly free cash flow of approximately negative $10 billion. On a trailing four-quarter basis, Oracle reported negative free cash flow of about $13.2 billion at the end of the quarter.
The company's six-month capital expenditures had already reached $20.5 billion, more than triple the $6.3 billion spent during the comparable period a year earlier. These figures are detailed in Oracle's second-quarter financial statements and its fiscal 2026 second-quarter Form 10-Q.
That helps explain why Oracle's share-price reaction was so severe despite rapid cloud growth. Investors were being asked to value not only a fast-growing infrastructure business but also a capital-intensive expansion whose cash requirements were arriving before much of the associated revenue.
Oracle shares consequently fell sharply after the results. Reuters reported that the company's sales and profit outlook contributed to the decline as forecasts missed targets.
The $248 Billion Lease Commitment Adds Another Layer
Capital expenditures were only part of Oracle's infrastructure commitment. Its regulatory filing disclosed $248 billion of additional lease commitments that had not yet been reflected as lease liabilities on the company's November 30, 2025 balance sheet.
Oracle said substantially all of those commitments related to data centers and cloud-capacity arrangements. They were generally expected to begin between the third quarter of fiscal 2026 and fiscal 2028, with contractual terms generally lasting 15 to 19 years.
CNBC highlighted the scale of the increase, reporting that Oracle's commitments had lease commitments climbed almost 150%.
The distinction between capital expenditure and lease commitments is important. The $248 billion figure was not an immediate cash bill and should not be added directly to annual capital spending. It represented long-term contractual commitments that would become economically relevant over future years as facilities and capacity became available.
AI Demand Was Stronger Than the Selloff Suggested
The December market reaction did not establish that demand for Oracle's AI infrastructure was collapsing. In fact, Oracle's backlog suggested the opposite.
Remaining performance obligations rose to $523 billion, more than five times their level a year earlier. Cloud infrastructure revenue was growing at 68%, and Oracle said customers were committing to significant future capacity.
RPO is not the same as recognized revenue, profit or cash flow, however. Large contracts can take years to convert into revenue, while the infrastructure required to fulfill them may need to be financed much earlier. That timing mismatch was the core concern raised by the earnings report.
Oracle was therefore confronting two realities simultaneously: unusually strong contracted demand and unusually heavy capital requirements. Whether its buildout represented disciplined investment or excessive capacity could not be determined from the spending figure alone.
What the "AI Bubble" Debate Missed
Descriptions of the selloff as evidence of an AI bubble simplified a more specific financial question. Investors were not only debating whether artificial intelligence had economic value; they were questioning which companies could finance enormous data-center programs while maintaining acceptable margins, cash flow and balance-sheet strength.
Oracle differed from some larger technology peers because its infrastructure expansion placed especially heavy demands on financing. A company can have rapidly growing revenue and a large backlog while still facing meaningful financial risk if investment must be made years before the corresponding cash returns arrive.
At the same time, large RPO numbers need to be interpreted carefully. They provide evidence of contractual demand but do not guarantee that every project will produce the margins, utilization rates or return on invested capital expected when the contracts were signed.
Later Context: What Happened During the Rest of Fiscal 2026
Later results showed that Oracle's AI infrastructure demand continued to accelerate, even as the cash-flow burden remained substantial.
For the full fiscal year ended May 31, 2026, Oracle reported revenue of $67.4 billion, up 17%. Cloud revenue reached $34.0 billion, up 39%, while cloud infrastructure revenue rose 77% to $18.1 billion. Remaining performance obligations ended the fourth quarter at $638 billion.
Those numbers strengthened the case that the infrastructure expansion was being supported by real customer demand. But the financial cost was also larger than initially expected. Oracle's reported fiscal 2026 capital expenditures reached approximately $55.7 billion, while free cash flow was negative $23.7 billion. The company reported operating cash flow of $32.0 billion for the year.
Oracle also said it had raised $43 billion in debt and $5 billion in equity financing during fiscal 2026. Its full-year fiscal 2026 results therefore confirmed both sides of the original December debate: infrastructure demand grew rapidly, and funding the expansion required an extraordinary amount of capital.
Growth Accelerated Again in the First Quarter of Fiscal 2027
By September 2026, Oracle's reported growth had accelerated further. Fiscal first-quarter 2027 revenue reached $19.3 billion, up 30% year over year, while total cloud revenue rose 62% to $11.6 billion.
Cloud infrastructure revenue reached $7.4 billion, up 121%, and remaining performance obligations increased to $664 billion. Oracle said it had also delivered more than 300,000 GPUs to AI cloud customers since the end of the previous quarter.
Cash requirements remained significant. Oracle reported negative free cash flow of $5 billion for the quarter as it continued expanding infrastructure, and the company completed a $20 billion at-the-market equity issuance as part of its capital program. These later figures are available in Oracle's September 2026 fiscal first-quarter results.
This later evidence does not prove that every AI infrastructure investment will ultimately earn an attractive return. It does show that the December 2025 concern was not followed by a collapse in Oracle's cloud demand. Instead, the business entered a phase of exceptionally fast growth paired with exceptionally high financing requirements.
What Investors Need to Watch
The most important measure is how quickly Oracle converts contracted demand into recognized cloud revenue and cash generation. Rapid RPO growth becomes more valuable if completed data centers fill quickly and produce attractive margins; it becomes more problematic if projects are delayed, customers reduce usage or financing costs rise faster than expected returns.
Free cash flow is another key indicator. Oracle demonstrated that operating income and revenue can grow while free cash flow remains deeply negative because infrastructure spending is so large. The point at which that trend stabilizes or reverses will provide useful evidence about the economics of the buildout.
Financing also matters. Oracle's use of debt, equity, customer prepayments and long-term leases means investors need to look beyond headline capital expenditures when assessing the total financial commitment behind the expansion.
Finally, Oracle's performance should be compared with other companies benefiting from AI demand. News Fusion 365 has previously examined the semiconductor side of the cycle in Nvidia's sales increased by 78.
The larger lesson from Oracle's December 2025 selloff is that strong AI demand and financial risk can exist at the same time. The company subsequently demonstrated faster infrastructure growth and a larger backlog, but it also spent more capital and relied more heavily on external financing. The long-term outcome will depend less on whether AI demand exists than on whether Oracle can convert that demand into sustainable cash returns on the enormous amount of infrastructure it is building.
Read More:
0 Comments