Investment Strategist
Oracle’s cloud demand is strong, led by artificial intelligence computing contracts.
Investors focused less on growth and more on data centre spending.
The wider AI story is becoming a cash-flow story, not only a revenue story.
Oracle’s latest results looked strong at first glance. The company reported after the United States market close on 10 June 2026, with faster cloud growth, a bigger contract backlog and guidance that still points to rapid expansion. For a normal software company, that would usually be enough to keep investors calm.
Oracle is no longer a normal software story. It is becoming an artificial intelligence (AI) infrastructure company. In simple terms, it is trying to rent huge amounts of computing power to companies that need to train and run AI models.
That shift is exciting, but expensive. Very expensive. Oracle closed at 201.26 USD on 10 June 2026, then fell as much as 12% in pre-market trading on 11 June as investors focused on the rising data centre bill rather than the headline growth.
Oracle is best known for database software, the digital filing cabinets that many large companies still rely on. It also sells business software and cloud services. Cloud means customers rent computing power and software over the internet rather than running everything on their own machines.
In the quarter, Oracle’s total revenue rose 21% to 19.2 billion USD. Its cloud infrastructure revenue rose 93% to 5.8 billion USD. That part of the business rents server capacity, storage and computing power. It is the area investors watch most closely because AI models need enormous computing resources.
The other important number was Oracle’s remaining performance obligations. This is an unpleasant phrase, but the idea is simple. It means revenue that customers have committed to pay in future periods, but Oracle has not yet booked as sales. That figure reached 638 billion USD, up sharply from last year.
For investors, this is the strongest part of the story. Oracle is not building data centres in hope alone. Customers are signing large contracts. Many of these are linked to AI workloads, including demand from OpenAI and other large customers.
This is like a hotel with many rooms already reserved before construction is finished. That is good. The only problem is that the hotel still needs concrete, electricity, air conditioning and a very patient bank manager.
The market reaction was not about weak demand. It was about spending.
Capital expenditure means money spent on long-term assets, such as data centres, servers, chips and power equipment. Oracle reported capital expenditure of about 16.5 billion USD in the quarter ended 31 May 2026. For the full fiscal year, spending reached 55.7 billion USD, above the company’s earlier projection of 50 billion USD.
The next year looks even bigger. Oracle expects about 70 billion USD of net capital expenditure in fiscal year 2027, with the reported figure potentially higher because of prepayments for some components. The company also plans to raise around 40 billion USD through debt and equity.
This matters because cloud infrastructure is not the same as selling traditional software. Old software could be copied and sold many times at high profit margins. AI infrastructure is more like building a power station. It can generate attractive long-term cash flows, but it needs large upfront investment.
That changes the investor question. It is no longer only “how fast can Oracle grow?” It is also “how much must Oracle spend to grow, and what return will it earn on that spending?”. That question is now spreading across the whole AI supply chain.
Oracle’s results are another sign that AI is moving from story to infrastructure. The first phase rewarded companies that sold the picks and shovels, especially advanced chips. The next phase may test the companies building the mines.
That does not make Oracle’s strategy wrong. Large contracts, customer prepayments and customer-supplied graphics processing units (GPUs), the chips used for AI work, can reduce some funding pressure. Oracle has also shown that demand for its cloud platform is accelerating.
But the market is becoming more selective. Investors are asking whether AI infrastructure can produce durable returns after depreciation, funding costs and competitive pressure. Depreciation is the accounting cost of spreading an asset’s value over time. It matters because AI servers can become outdated faster than traditional infrastructure.
This is the key insight. In AI, growth and profitability may not arrive at the same time. Revenue can rise quickly while free cash flow, the cash left after investment spending, stays under pressure. That is not automatically bad, but it requires confidence in future demand, pricing and execution.
The first risk is funding. If Oracle needs more debt or sells more shares than expected, investors may worry about dilution and balance sheet pressure.
The second risk is customer concentration. Large AI contracts can be powerful, but they can also make results depend on a small number of very large customers. If one delays, renegotiates or shifts workloads elsewhere, the impact can be meaningful.
The third risk is returns. Data centres require land, power, cooling, chips and time. If pricing falls, utilisation disappoints or AI chips become obsolete faster than expected, future returns may look less attractive.
Separate demand from economics. Strong bookings are useful, but cash returns decide long-term value.
Watch funding quality. Customer prepayments are better than relying only on debt or new shares.
Compare business models. Chip suppliers, cloud operators and utilities carry different risks.
Keep position sizing honest. AI infrastructure can be powerful, but it is not a free lunch.
Oracle’s results are not a simple “good quarter, bad stock reaction” story. They show where the AI market is heading next. The easy part is believing demand will grow. The harder part is judging who can build capacity, fund it sensibly and earn good returns after the bills are paid.
Oracle has strong demand and a large opportunity, but it also has a large construction project attached to its balance sheet. For investors, the lesson is useful beyond one company: in the AI boom, the winners may not only be those with the biggest dreams, but those that can keep the lights on without letting the meter spin out of control.
This material is marketing content and should not be regarded as investment advice. Trading financial instruments carries risks and historic performance is not a guarantee of future results.The instrument(s) referenced in this content may be issued by a partner, from whom Saxo receives promotional fees, payment or retrocessions. While Saxo may receive compensation from these partnerships, all content is created with the aim of providing clients with valuable information and options. The author does not hold any position in the financial instruments mentioned at the time of publication.
Outrageous Predictions
Head of Investment Content Switzerland
Outrageous Predictions
Senior Relationship Manager
Outrageous Predictions
Saxo Group
Outrageous Predictions
Chief Investment Strategist
Outrageous Predictions
Global Head of Macro Strategy
Outrageous Predictions
Chief Investment Strategist
Outrageous Predictions
Global Head of Investment Strategy
Outrageous Predictions
Global Head of Investment Strategy
Outrageous Predictions
Investor Content Strategist
Outrageous Predictions
Global Head of Macro Strategy