Oracle's Second Venture: From Selling Software to Selling AI Computing Power, the Biggest Concern Is Being Answered

In the past, Oracle was best known for its databases and enterprise management software, helping businesses manage finance, human resources, and supply chain operations while generating long-term maintenance and technical support fees. This high-stickiness business provided the company with a stable stream of cash flow.

The AI era has provided Oracle with a second growth curve: renting computing power to AI companies that need to train and run large models. In FY2026, revenue from this cloud infrastructure business, called OCI, grew by about 77%, contributing nearly 80% of the company's incremental revenue. But the trade-off was straightforward: full-year capital expenditure reached $55.7 billion, and free cash flow stood at negative $23.7 billion. Consequently, the market has long questioned whether Oracle is using its own balance sheet to gamble on an AI infrastructure industry that has yet to achieve stable profitability.

This concern was directly debunked for the first time in the latest quarterly earnings report: Oracle is transferring capital expenditure pressure to customers and suppliers rather than bearing it alone on its balance sheet. This is also the core reason behind our positive outlook.

I. Quarterly Earnings Analysis: OCI Growth Accelerates Further, Turning Point Emerges for CapEx Funding

In the first quarter of FY2027 (ended August 31, 2026), Oracle delivered results that showed continued acceleration. Total revenue reached $19.35 billion, up 30% year-over-year, beating market expectations of $19.14 billion. Non-GAAP earnings per share (EPS) came in at $1.92, up 30% year-over-year and also topping consensus estimates of $1.74. Under GAAP, net income attributable to common shareholders was $4.68 billion, or $1.56 per share, compared with $2.93 billion, or $1.01 per share, in the same period last year.

The growth of OCI is driven by accelerating core operations rather than a low base effect. Cloud infrastructure (OCI) revenue reached $7.4 billion, up 121% year-over-year, climbing further from the 77% growth rate for the full year of FY2026. Among this, driven by computing demand, core CPU and GPU infrastructure revenue reached $6.5 billion, representing a year-over-year increase of 151%. Database cloud revenue came in at $900 million, up 26% year-over-year, with multicloud database revenue soaring 353%, continuing the cloud migration trend seen in previous quarters. Overall cloud business (OCI + OCA) revenue reached $11.6 billion, up 62% year-over-year.

In comparison, traditional software revenue was approximately $5.5 billion, down 3% year-over-year, as customers continued transitioning from on-premise deployments to cloud subscriptions, aligning with the pace of transformation seen in recent quarters.

RPO continues to expand, accompanied by an upward revision in guidance. Remaining performance obligations (RPO) at the end of the period reached $664 billion, an increase of approximately $209 billion compared to the same period last year. Management raised full-year revenue guidance for FY2027 to no less than $90 billion, with non-GAAP EPS guidance set at $8.10. For the second quarter, management provided revenue growth guidance of 30%-34% and non-GAAP EPS guidance of $1.85-$1.93. Raising rather than merely maintaining guidance indicates that management has strong confidence in the pace of converting contracts into revenue.

In terms of capacity delivery, the company added 850 megawatts of data center capacity in a single quarter. Global AI infrastructure utilization stood at approximately 97.9%, with GPU renewals and resale prices enjoying roughly a 20% premium over original contracts. These two metrics demonstrate that current growth is not driven by discounting to capture orders, but rather by demand consistently outstripping supply, allowing the company to retain pricing power.

The most noteworthy shift: CapEx funding concerns were addressed directly. Management further clarified this quarter that the pace of OCI expansion does not depend entirely on how much cash Oracle itself can generate. The company outlined three specific mechanisms:

• Entering into phased payment arrangements with equipment suppliers, allowing the company to pay procurement costs gradually as payments are collected from customers, rather than advancing full cash upfront;
• Certain customers purchasing their own hardware (Bring-Your-Own-Hardware), with Oracle managing operations and cloud infrastructure technology to transform the hardware into usable AI compute clusters;
• Financed customers (whether startups or established enterprises) making direct advance payments to Oracle, relieving the company from having to prepay the corresponding capital expenditure.

This statement directly addresses one of the primary concerns previously raised by short-sellers: given Oracle's persistently negative free cash flow, would it need to endlessly issue debt and equity to sustain expansion? In terms of figures, the company's full-year FY2027 gross CapEx guidance remains in the $90 billion to $95 billion range. However, due to the aforementioned advance payments and financing arrangements, full-year net cash CapEx is expected to remain capped at $70 billion. Net cash CapEx for this quarter was approximately $18 billion, significantly below the corresponding gross expenditure for the quarter. The company also expects to raise approximately $40 billion in total through debt and equity during FY2027, including the previously announced $20 billion At-The-Market (ATM) equity offering program. While financing needs persist, a clear gap has emerged between gross capital expenditure and net cash outlay, alleviating market anxieties regarding an 'all-in, self-funded' expansion strategy.

II. Business Analysis: Cloud Drives Growth, Software Supports the Baseline

Oracle's revenue can be categorized by product and service into four areas: Cloud, Software, Hardware, and Services. Although Cloud and Software are combined into a single formal operating segment, analyzing them separately provides greater clarity, as their growth drivers and business models differ.

In FY2026, Oracle generated approximately $67.4 billion in total revenue. Cloud business accounted for over half of total revenue, serving as the company's primary growth engine, while software business still contributed over a third, providing an essential base of recurring revenue.

Note: Percentages and YoY changes are calculated based on amounts disclosed in the annual report; rounding may cause slight discrepancies.

2.1 Cloud: OCI Drives Acceleration, OCA Maintains Steady Growth

Oracle's cloud business is divided into OCI (Oracle Cloud Infrastructure) and OCA (Oracle Cloud Applications).

The two address distinct customer needs: OCI provides computing resources and database services to help clients run their systems, whereas OCA supplies pre-developed business applications to assist clients with financial, human resources, and supply chain management tasks.

OCI: Encompassing Both AI Compute and Database Cloud Services

OCI is currently the primary growth engine. In FY2026, its revenue grew by approximately $7.87 billion, contributing nearly 80% of the company's net revenue incremental growth.

According to the company's Q4 earnings presentation, OCI consists of two main businesses:

• Infrastructure (CPU and GPU): Provides computing resources to support enterprise systems, AI training, and inference workloads. Q4 revenue reached approximately $4.8 billion, up 119% year-over-year.
• Database: Provides database cloud services. Q4 revenue reached approximately $800 million, up 29% year-over-year, with multicloud database revenue surging 404%.

This breakdown shows that OCI's growth is driven both by demand for computing power and by the migration of database workloads to the cloud. CPU and GPU infrastructure represents the larger and faster-growing portion in the current period, though not all of its revenue should be categorized strictly as AI revenue.

The database business builds on Oracle's long-standing product advantages. Customers can run databases on OCI or leverage multicloud partnerships to utilize Oracle database services within AWS, Azure, or Google Cloud environments.

OCA: Transforming Enterprise Management Software into Recurring Subscription Revenue

OCA covers application software essential for daily enterprise operations. Key products highlighted in the company's Q4 presentation include:

Growth in OCA is primarily driven by new customer acquisition, migration of legacy applications to the cloud, and cross-selling additional modules to existing clients. For instance, after adopting Fusion Financials, customers may subsequently purchase supply chain or HR modules.

Compared to OCI, OCA exhibits more stable growth. Its value lies in increasing revenue per enterprise customer through long-term subscriptions and expanded product adoption. Whether AI features can further drive customer adoption and renewals remains to be verified by future operational performance.

2.2 Software: Support Revenue Remains Stable, License Sales Face Pressure

In FY2026, software revenue totaled approximately $24.54 billion, down 0.7% year-over-year. Within this segment, software support and software licenses exhibited diverging trends.

Software support revenue reached $19.8 billion, up 1.4% year-over-year, primarily generated by maintenance, update, and technical support fees paid by existing customers. Because Oracle databases and applications power core enterprise operations, switching costs are high, making this revenue stream highly predictable and a critical foundational pillar for the company.

Software license revenue came in at $4.74 billion, down 8.9% year-over-year. Customers shifting from purchasing software rights to subscribing to cloud services has rechanneled revenue into OCI and OCA, while putting pressure on traditional license sales. Assessing the success of this transition requires evaluating whether cloud revenue growth can offset the decline in legacy software revenue, and whether customers remain within the Oracle ecosystem.

2.3 Hardware and Services: Complementary Businesses with Limited Revenue Contribution

In FY2026, hardware revenue was approximately $3.08 billion, up 5% year-over-year, while services revenue reached approximately $5.74 billion, up 9.7% year-over-year, jointly accounting for about 13% of total company revenue. The hardware business includes engineered systems like Exadata, servers, storage, and related support; the services segment provides consulting, implementation, and migration assistance to help clients deploy and utilize Oracle products.

While these two businesses support product delivery and customer onboarding, their contribution to overall growth is smaller than that of the cloud segment. Oracle's core growth trajectory continues to be driven by OCI expansion, alongside the development of OCA and database cloud services.

III. Growth Engine Breakdown: Computing Power, Database Migration, and Enterprise Subscriptions

3.1 AI Compute Demand Drives OCI Expansion; Capacity Delivery Dictates Growth Velocity

CPU and GPU infrastructure within OCI represents Oracle's single largest source of growth. In Q4 FY2026, revenue from this segment reached $4.8 billion, a 119% year-over-year increase. Demand stems from large model training, post-deployment inference, and the ongoing execution of applications such as AI coding.

For major AI clients, acquiring chips is merely the first step. Whether the equipment can collaborate efficiently, operate reliably, and deploy on schedule directly impacts training costs and product launch timelines. Oracle provides compute, networking, storage, and cluster management services, competing effectively on both compute scale and delivery efficiency as well as operational performance.

Substantial demand is already locked in under contract, shifting the strategic focus to revenue conversion. The company delivered over 1.2 GW of capacity in FY2026 and projected in Q4 to deliver nearly 1 GW in Q1 FY2027. Accelerated capacity delivery will facilitate the ramp-up of contracted revenue, serving as the most direct catalyst for near-term OCI growth. Power availability, data center facilities, hardware supply, and installation timelines will dictate how fast this process unfolds.

3.2 Multicloud Databases Unlock Cloud Migration Demand, Expanding OCI Database Revenue

OCI delivers not only compute power but also database cloud services. In Q4 FY2026, database cloud revenue totaled approximately $800 million, up 29% year-over-year, with multicloud database revenue surging 404%. This growth leverages Oracle's extensive enterprise customer base and mission-critical systems built over decades.

Multicloud partnerships resolve a practical challenge: enterprises want to utilize AWS, Azure, or Google Cloud while retaining their Oracle databases. By hosting database services directly within these cloud platforms, Oracle reduces the friction of cloud migration for clients while broadening its addressable market. Customers opting for alternative hyperscalers can still seamlessly purchase Oracle database services.

AI further elevates enterprise data consumption. Clients seek to query, analyze, and power intelligent applications around existing business data, which could drive increased database cloud usage. The runway for multicloud database growth is driven by legacy customer migrations, expanding consumption, and new workload deployments. The current high growth rate is partly amplified by a small baseline, making actual revenue contribution and usage metrics key indicators to monitor going forward.

3.3 OCA Relies on Cloud Migration and Cross-Selling to Expand Enterprise Subscription Revenue

OCA encompasses Oracle's cloud applications targeted directly at line-of-business enterprise departments, including Fusion, NetSuite, and industry verticals like Oracle Health. In FY2026, OCA generated approximately $15.9 billion in revenue, up 11.3% year-over-year. While expanding at a more moderate pace than OCI, it provides recurring enterprise subscription revenue distinct from large AI compute contracts.

Growth is mainly driven by customers transitioning legacy applications to cloud subscriptions, new client acquisitions, and existing clients expanding into additional modules. For example, an enterprise might initially adopt Fusion Financials and subsequently add HR or supply chain applications, streamlining integration through shared data and processes. In Q4 FY2026, revenue for Fusion Back-Office, NetSuite, and industry applications grew by 12%, 9%, and 8%, respectively.

AI holds potential to enhance product competitiveness, but new capabilities must translate into client purchases, renewals, or module expansions to generate revenue. Oracle Health's next-generation AI healthcare system also represents a potential opportunity, with its impact contingent on deployment pace and customer adoption. OCA growth fundamentally relies on enterprise customers continuing cloud migrations and expanding product adoption, with AI offering an additional tailwind to accelerate this journey.

IV. Key Risks: High Customer Concentration and Heavy Capital Commitments Amplify Cash Recovery Pressures

4.1 Customer Concentration: OpenAI's Financial Capacity Impacts Order Realization

At the end of FY2026, Oracle's RPO reached $638 billion, a significant portion of which is tied to OpenAI. While massive contracts provide a large growth runway, they also heighten Oracle's exposure to the financial health and fundraising abilities of a few key clients. Whether OpenAI can continuously expand revenue, secure capital, and purchase compute capacity as planned will directly affect the realization pace of these contracts.

RPO does not equal cash in hand. Oracle must commit capital to facility construction and equipment upfront. If a major client slows procurement, delays payments, or reduces demand, the company could face an extended cash payback period. While compute capacity can be re-leased, securing alternative clients takes time and may not yield original contractual pricing. The higher the customer concentration, the greater the exposure of revenue, utilization, and cash flow to changes from any single client.

4.2 Capacity Delivery Delays Extend Cash Payback Cycles

Large-scale data centers require synchronized execution across power availability, construction, servers, and networking gear. Delays in any component can postpone customer onboarding and revenue recognition while lease, operational, and financing costs are already being incurred. As buildouts expand, Oracle must maintain delivery execution across a broader portfolio of simultaneous projects.

Therefore, beyond planned capacity and contract size, actual delivery, billable capacity, and utilization must be closely monitored. If revenue realization lags capital deployment, the funding gap will widen, potentially forcing the company to seek additional financing.

4.3 Bring-Your-Own-Hardware (BYOH) Model Requires Validation of Actual Returns

Customer prepayments and BYOH arrangements help alleviate Oracle's capital constraints. Prepayments generate cash needed for construction, while customer-supplied chips reduce Oracle's upfront equipment procurement and subsequent depreciation. However, the economic profiles of these two models differ: prepayments carry future performance obligations, whereas BYOH alters the underlying revenue and cost structure.

The key lies in how much networking, storage, operational, and other service revenue Oracle can retain after reducing its capital outlay. A higher gross margin percentage does not necessarily translate to higher absolute gross profit dollars, nor should Oracle be viewed as merely providing raw real estate leasing simply because the chips belong to the client. Assessing whether such contracts are value-accretive requires evaluating capital committed, scope of services, and contract duration.

V. Investment Recommendation: Core Doubts Partially Addressed, Buy-on-Dips Thesis Strengthens

Incorporating the latest quarterly results, we maintain a positive outlook on Oracle, driven primarily by three key pillars:

First, the quality of OCI's growth is improving, rather than just showing strong headline numbers. Revenue from CPU and GPU infrastructure surged 151% year-over-year, outpacing OCI's overall 121% growth rate. This confirms that growth is genuinely driven by core compute demand rather than being artificially inflated by auxiliary sub-segments. Furthermore, management raised full-year revenue and EPS guidance rather than merely maintaining it, signaling strong visibility into the contract-to-revenue conversion timeline.

Second, the most significant incremental shift this quarter was the direct response to CapEx funding concerns. This had been the market's single greatest reservation about Oracle: with free cash flow persistently in negative territory, was expansion sustainable, or would it lead to endless debt issuance and equity dilution? Management clearly stated that OCI expansion is not strictly constrained by Oracle's own cash-funded CapEx. Three mechanisms—customer prepayments, supplier financing, and BYOH—can support capacity expansion without requiring the company to advance additional upfront cash. Full-year net cash CapEx guidance is capped at $70 billion, meaningfully below the $90 billion to $95 billion gross CapEx range, a dynamic validated by this quarter's net cash CapEx of approximately $18 billion. While this does not imply financing pressures have vanished—the company still plans to raise about $40 billion through debt and equity in FY2027, including a $20 billion ATM equity offering—it significantly mitigates the worst-case scenario previously feared by the market (unbounded CapEx expansion and unchecked free cash flow deterioration).

Third, pricing power and demand signals remain robust. A 97.9% utilization rate alongside a ~20% premium on GPU renewals and resales indicates that current growth is not fueled by price discounting, but rather by demand consistently running ahead of supply. This bodes well for future gross margins and long-term return on capital, while reinforcing confidence that RPO will convert into revenue as scheduled.

Fourth, relative to peers that closely match Oracle's operational profile, the current valuation is reasonable and the balance sheet is healthier. Following the earnings release, the stock rose about 4% to around $160. Based on sell-side consensus estimates of $8.78 for FY2028 EPS, the forward P/E stands at roughly 18x. With OCI generating over half of cloud revenue and expanding at triple-digit rates, its business model increasingly resembles 'Neocloud' players such as CoreWeave and Nebius. These Neocloud firms command significantly higher valuations despite ongoing operating losses. In contrast, Oracle is anchored by a profitable legacy software business, leveraging debt against real operating cash flows and net income, unlike unhedged Neoclouds expanding through debt without profitability. Within this comparative framework, 18x forward P/E appears reasonable, lending valuation support to a 'buy-on-dips' strategy. Nevertheless, peer comparisons carry inherent subjectivity, and thesis validation depends on sustained OCI margin quality.

Key Risk Factors to Monitor:

• Customer concentration: The financial capacity and willingness to make follow-on investments among major AI clients, such as OpenAI, remain critical variables for on-schedule RPO realization;
• Whether net cash CapEx stays strictly within the $70 billion guidance limit, serving as the benchmark to validate the 'self-funding' narrative;
• Financing cadence: The degree to which interest expenses and equity dilution from the ~$40 billion in annual debt and equity financing erode EPS;
• Gross margin trends: Whether the rising revenue share of AI infrastructure dilutes overall company gross margin, which requires tracking segment gross profit metrics over upcoming quarters.

In summary, this quarterly report does not alter Oracle's core investment thesis—OCI capacity delivery pace and RPO conversion velocity remain the primary share price drivers—yet the new disclosures regarding CapEx financing structures meaningfully soothe one of the primary bear concerns. For investors willing to tolerate post-earnings short-term volatility, post-earnings pullbacks may present attractive buy-on-dips opportunities. Key tracking metrics should focus on whether net cash CapEx remains within guided ranges and whether customer concentration continues to diversify.