Oracle Faces Capacity Crunch as $553 Billion Backlog Tests AI Infrastructure Ambitions

Oracle goes into its quarterly earnings on Wednesday with a problem most companies would envy: more demand than it can handle. The software giant’s order book has swelled to $553 billion in remaining performance obligations, a staggering 325% jump from a year earlier. But turning those commitments into revenue requires a building spree of historic proportions, and investors are watching closely to see whether the payoff arrives on schedule.

The capital expenditure numbers underscore the scale of the challenge. Oracle poured $39.2 billion into property, plant and equipment in the first nine months of its fiscal year — more than triple the amount spent in the same period last year. Analysts expect full-year capex for fiscal 2026 to hit roughly $50 billion as the company races to expand Oracle Cloud Infrastructure to meet voracious demand from AI-training workloads, inference and enterprise data management.

Wall Street has pencilled in fourth-quarter revenue of $19.1 billion and earnings per share of $1.96. But the headline figures are almost secondary. What matters most is the trajectory of the cloud business, which grew 84% in the previous quarter. Consensus now points to a still-lofty growth rate of around 50%, a deceleration that would nonetheless keep Oracle firmly in the hyperscaler conversation.

The stock eked out a 1.45% gain on Tuesday to €186.56, recovering slightly after shedding more than 11% in the prior seven sessions. That leaves the shares roughly 33% below their 52-week high, a gap that reflects both the enormity of the spending commitment and the market’s wait-and-see posture. Chartists note that the price still holds above the 200-day moving average at €177.15, preserving the long-term uptrend for now.

Should investors sell immediately? Or is it worth buying Oracle?

Oracle’s transformation goes deeper than infrastructure. The company is embedding artificial intelligence into its entire application stack, turning the Fusion Cloud suite from a traditional “system of record” into what it calls a “system of outcomes” — with AI agents that automate financial reporting, supply chain management and other core processes. The launch of OCI Enterprise AI gives customers a platform to build, deploy and scale their own models, while a push into healthcare via Oracle Health promises to modernize electronic health records and reduce administrative burdens.

That multi-cloud strategy — enabling Oracle databases to run seamlessly across different cloud platforms — lowers the barrier for corporate clients wary of vendor lock-in. It also positions the company as a cost-effective alternative to the dominant hyperscalers, a pitch Oracle will reinforce at its internal AI conference, which opened for registration on Tuesday.

Analysts, on balance, are buying the story. The consensus price target stands at €217.70, implying roughly 17% upside from current levels. That target acknowledges the credibility of Oracle’s investment thesis and its growing role in the AI ecosystem, but also reflects the uncertainty ahead. The company has a clear plan and a market that is expanding rapidly. The earnings report, due after the US close on Wednesday, will provide the hard data on whether the billions spent on capacity are translating into tangible revenue — and whether the backlog is a promise of future riches or just a measure of how much work remains.