Oracle Corporation (ORCL) Earnings

Management highlights

• Strategic Positioning - Oracle is uniquely positioned to capture enterprise AI transition as a full-stack provider, covering cloud infrastructure for AI workloads, mission-critical applications, database services, and AI tooling, with built-in industry expertise - Management is confident the AI infrastructure market is worth trillions of dollars annually, and OCI will grow into a large, highly profitable business with a 30% to 40% long-term margin profile - Disciplined capital allocation, strong balance sheet, and investment-grade credit rating will remain core priorities • Product and AI Innovation - Over 1,000 AI agents have been delivered across Oracle application suites, with most core application AI innovation included at no extra charge - New AI functionality added to Oracle Database includes Oracle AI Agent Memory (for enterprise context-aware agent development) and Oracle Deep Data Security (for granular database-level access control for users and AI agents) - New commercial models for AI: simple prepackaged token bundles for additional advanced AI reasoning capacity, and expanded outcome-based pricing that aligns costs directly to customer value created • Operational Execution - Thousands of customers went live on Oracle cloud services in Q4, including over 300 on Fusion ERP; major wins include a US OPM agency-wide award for Fusion HCM, Vodafone infrastructure consolidation, and a 30 million subscriber AI platform deal with Latin American telco Clay - Deployment of electronic health records for the US VA continues on track, with 14 medical centers now live serving 29,000 clinicians and 500,000 veterans - Five major new OCI data center sites are progressing on or ahead of schedule: 42% of capacity delivered at Abilene TX, with 35% more coming in 90 days; 115MW already online at Shackleford TX one month ahead of schedule; all other sites are on track for 2027 delivery • Capital and Demand Trends - Full year FY26 net cash capital expenditure was $48 billion, including $8 billion in prepayment and timing impacts - 35,000 GPUs were up for renewal in Q4 across 59 customers; 49% of customers renewed for 92% of GPUs, and nearly all non-renewed capacity was re-sold to other customers in the same quarter, supporting 97.5% global utilization

Guidance

• Long-term targets (reconfirmed by new CFO): 31% revenue CAGR and 28% EPS CAGR through FY30, supported by the record RPO backlog • FY27 total revenue guidance is 34% constant currency growth, above the long-term 31% CAGR target • FY27 expected net cash outlay for capital expenditures is ~$70 billion, which excludes $20 billion to $25 billion in customer prepayments and timing impacts; reported GAAP CapEx will be higher by this prepayment amount • Oracle plans to raise ~$40 billion in debt and equity in FY27 to support capital investments, including the already announced $20 billion at-the-market equity offering; no additional debt is expected to be raised in calendar 2026 • Gross margin will step down in FY27 due to data center ramp timing and mix shifts, but OCI margins are expected to improve rapidly as data centers reach full contracted revenue contribution • Operating costs are expected to be slightly negative year-over-year in dollar terms due to ongoing efficiency actions, driving improved operating leverage • FY27 non-GAAP EPS is expected to be $8.05, up 18% in constant currency, excluding FY26 one-time investment gains from Ampere and Bloom Energy • Q1 FY27 guidance: total revenue growth of 27% to 29% in USD; cloud revenue growth of 58% to 64%; non-GAAP EPS of $1.72 to $1.76, up 17% to 20% YoY • Revenue and earnings growth are expected to accelerate in the second half of FY27 as additional data center capacity comes online to meet customer demand

Segment performance

Q4 FY26 total revenue was $19.2 billion, up 21% YoY. Non-GAAP operating income was $8.6 billion, up 22% YoY, with non-GAAP EPS of $2.11 (up 24% YoY including a one-time net investment gain; up 20% excluding the gain). Full year FY26 total revenue surpassed $67 billion for the first time, with non-GAAP operating income of $29 billion (up 16% YoY) and non-GAAP EPS of $7.63 (up 27% including one-time gains; $6.83 excluding gains). Full year operating cash flow was $32 billion, up 54% YoY. Remaining Performance Obligations (RPO) finished at $638 billion, up 363% YoY. 12% of RPO is expected to be recognized in the next 12 months, and 34% between 13 and 36 months. By product segment: 1) Cloud Infrastructure: Revenue grew 93% YoY, with $67 billion in AI infrastructure contracts signed in Q4, bringing cumulative bring-your-own-hardware/prepaid AI infrastructure contracts to $75 billion. OCI delivered more than 1.2 gigawatts of capacity in FY26, with Q1 FY27 delivery already approaching 1 gigawatt. Global GPU utilization is 97.5%. This segment makes up approximately 47% of total Q4 revenue based on segment growth and other segment figures. 2) Cloud Applications: Revenue was $4.1 billion, up 10% YoY, contributing approximately 21% of total Q4 revenue. SaaS deferred revenue grew 16% YoY. 3) Cloud Database: Revenue grew 29% YoY, with multi-cloud database revenue up 404% YoY and multi-cloud bookings up 325% YoY. Legacy license and on-premise product segments make up the remaining ~32% of total revenue, and continue to deliver steady margin contributions.

Risks & headwinds

• Component prices (especially memory, SSDs, and hard drives) have increased significantly, creating potential margin pressure for cloud infrastructure providers • Many new entrants are entering the AI data center market, increasing competition for AI infrastructure customers • Large-scale AI data center design, delivery, and operation is extremely technically demanding, creating execution risk • Long-term AI market demand is still emerging, with uncertainty around future customer renewal rates and long-term capacity needs • Heavy upfront capital investment for data center expansion creates near-term balance sheet and funding requirements

Analyst Q&A

• Q: Higher component (especially memory) costs have increased cloud CapEx broadly; how does Oracle manage this risk in long-term AI contracts? /

  A: Any unexpected CapEx increase in Q4 came from accelerated capacity spending to meet demand faster, not higher component prices. Oracle uses flexible contract structures to manage component cost risk: fixed-price contracts are only used when all costs (components, power, labor) are already locked in. For contracts where future costs are uncertain, Oracle includes pass-through mechanisms for cost increases, so margin is not negatively impacted.

• Q: With many new entrants building AI data center capacity, how does Oracle expect competition to impact customer retention, new customer growth, and margins? /

  A: Current industry demand for AI infrastructure is far larger than existing supply, so there is enough room for multiple providers. Oracle focuses on customer satisfaction and operational excellence: high 97.5% GPU utilization and instant reallocation of non-renewed capacity demonstrates strong customer demand. Building a secure, efficient, robust cloud is operationally complex, and Oracle's established track record of reliable operation creates strong customer loyalty that supports future renewals and margin retention. Continued R&D for efficiency will allow Oracle to improve margins while lowering customer prices over time.

• Q: Why is Oracle expanding outcome-based pricing and token bundles for AI, what segments does this cover, and how will it impact growth? /

  A: Outcome-based pricing is not new for Oracle (used historically in construction and hospitality), but the company is now expanding it across the entire Fusion SaaS application suite. The model aligns AI spending directly to measurable customer value creation, which reduces customer friction for AI adoption by addressing common questions about AI ROI. Most core application AI features are still included at no extra charge; token bundles let customers purchase additional capacity for advanced reasoning as needed. This flexible, value-aligned pricing model leverages Oracle's unique full-stack position and is expected to drive higher growth by resonating strongly with customers.

• Q: How should investors evaluate returns on Oracle's current heavy infrastructure investment period? /

  A: Oracle models return on invested capital (ROIC) for infrastructure projects at the project level, calculated as (after-tax operating margin + depreciation) divided by total gross CapEx. Steady-state ROIC for large OCI infrastructure projects is expected to be in the high 20% range. For bring-your-own-hardware customer contracts, ROIC is even higher than this baseline, because Oracle does not have to fund the hardware capital itself. This strong ROIC profile supports the value of the current investment program.