Oracle Corp (ORCL) Report Interpretation

Overview

Bernstein dissects Oracle's FY26 revenue by product and growth profile to explain how the company can finance a very large AI data-center expansion while sustaining profit growth. Its conclusion is that rapidly expanding OCI is becoming the main accelerator, while database support, cloud database, Strategic Back Office and other mature software businesses provide recurring revenue, high margins and funding capacity. The report reiterates Outperform and a USD 325 target.

Core views

Oracle is undergoing what Bernstein calls one of software's most complex cloud transitions. The central investor debate concerns the profitability, stability and cash requirements of the AI data-center build-out needed to fulfill long-term, non-cancelable commitments. Bernstein argues that OCI cannot be evaluated in isolation: Oracle Database, on-premise applications, Fusion, NetSuite, multicloud database and sovereign cloud collectively provide growth and cash that help finance the infrastructure cycle. The institution believes Oracle is approaching the end of its need for additional cash and that the value created by the investment is not reflected in its valuation. The revenue mix has moved decisively toward cloud. Cloud represents about 58% of software revenue, up from 50% a year earlier, and grew about 39% year over year, compared with 19% growth for software overall. Organic constant-currency growth, which was in the low-to-mid single digits through FY21, reached 21% in the latest two quarters. Bernstein expects further acceleration in FY27 and for the next few years, driven by OCI Gen 2 across conventional CPU-based IaaS/PaaS, sovereign cloud and AI data centers. Reported growth remains noisier because of foreign exchange and variable software-license revenue. OCI is the main acceleration engine. IaaS/PaaS revenue increased from about $6.8B in FY24 to about $10.2B in FY25 and $18.1B in FY26, with growth accelerating from 50% in FY25 to 77% in FY26. Its share of Oracle's total revenue rose from 13% in FY24 and 18% in FY25 to 27% in FY26. OCI is growing faster than peers from a smaller base; the report says it has surpassed Alibaba to become the fourth-largest global hyperscale cloud provider, although it remains much smaller than Microsoft, AWS and Google Cloud. The mix shift is a gross-margin headwind, particularly because AI training infrastructure is lower margin, but Bernstein argues that the pressure is less pronounced at the operating-margin level. OCI consists of more than GPU-based AI capacity. Traditional Gen 2 infrastructure serves enterprises and digitally native customers with an emphasis on availability, security and performance. Sovereign cloud can operate with a much smaller footprint—previously 10 racks and now as few as 3 racks—with integrated hardware and software security, making it suited to governments, private clouds and regulated industries. Bernstein considers sovereign cloud a large and valuable portion of CPU-centered OCI revenue and believes Oracle may eventually sell AI server racks into this customer base. Legacy hosting, by contrast, declined to about $307M from about $322M in FY25 and has become a very small part of infrastructure services. Oracle's AI infrastructure business barely existed two years earlier but has become a major OCI contributor. Bernstein estimates FY26 OCI AI revenue at about $6.6B. Oracle signed AI data-center contracts beginning in June 2025 with OpenAI, Meta, NVIDIA, AMD, ByteDance and others, helping produce more than $300B of AI RPO. The contracts generally ramp across multiple data centers coming online between FY26 and FY30, are non-cancelable, and initially run for 5–6 years, likely with renewal provisions. In many cases, one client takes an entire data center from its first day of operation. Oracle has signed leases for the required capacity, plus some additional capacity, but does not begin paying until facilities are delivered. It has servers and power and cooling systems under construction while retaining flexibility over hardware-component purchasing because it lacks meaningful long-term component commitments. The report believes most current OCI AI revenue comes from model training rather than inference, with GPU-centered demand concentrated among AI labs, digitally native companies and semiconductor companies. It notes, however, that the mix could evolve as customers gain more flexibility to run inference across platforms. Oracle itself does not disclose the precise scope of OCI AI revenue, so the $6.6B figure is Bernstein's estimate rather than a company-reported amount. Database migration creates another growth and margin bridge. Oracle's multicloud arrangements place Oracle Database at Azure, Google Cloud and AWS. The business remained relatively small but grew 531% year over year in Q3 FY26. Bernstein estimates that moving a database customer to multicloud can produce a 3x-plus revenue lift relative to database support, reduce the likelihood that the customer switches away from Oracle Database and place much of the incremental revenue in OCI. The report believes this revenue could carry SaaS-like gross margins and potentially reach about 80%, depending on revenue recognition between Oracle and its partner, partly offsetting the lower margins of AI training infrastructure. Bring Your Own License also helps explain why support revenue remains flat even as ERP and database workloads move to the cloud. Core SaaS remains a steady second pillar. The detailed FY26 table reports total SaaS revenue of $15.889B, up 11%; a later narrative sentence states 15%, while the product detail assigns 15% growth to Strategic Back Office. Strategic Back Office—ERP plus Fusion HCM—reached an estimated $9.919B from $8.625B in FY25, growing 15%. ERP revenue was about $8.55B, up 14%, including Fusion ERP at about $4.325B with 15% growth and NetSuite at about $4.225B with 13% growth. Fusion HCM generated about $1.368B and grew about 21%. Growth has moderated in core ERP, but these products still outgrow Oracle overall and represent an increasing portion of SaaS revenue. Bernstein's historical bucket analysis shows the growing businesses taking a larger share. The narrative estimates Growth, Stable and Declining businesses at 38%, 40% and 22% of software revenue, respectively, compared with 33%, 44% and 24% in FY24. The Exhibit 5 caption separately displays Growth, Stable and Declining at 48%, 34% and 18%, indicating a different displayed aggregation from the software-only narrative. In both presentations, the direction is the same: the growth bucket is expanding, while stable and declining activities are becoming less important. Organic software-license revenue fell 9% as workloads migrated to cloud, while support stayed broadly flat. Outside software, hardware grew 5% year over year in FY26, but Bernstein expects it to resume a gradual decline as customers move from Oracle hardware to SaaS and OCI Gen 2. Services grew about 10%, yet the report regards this activity as immaterial to the investment thesis because of its low contribution margin. Bernstein expects Oracle to reduce professional services, particularly Cerner-related services; a lower services mix could improve consolidated gross margin and offset some of the cloud mix pressure. The forecast translates these operating drivers into substantial growth. Bernstein estimates revenue of $67.358B in FY26, $90.424B in FY27 and $136.842B in FY28, a 42.5% CAGR across the displayed forecast period. Operating earnings are projected at $28.925B, $37.935B and $55.881B, while adjusted operating margin moves from 42.9% to 42.0% and 40.8%. Adjusted EPS is forecast at $7.63, $8.57 and $12.86, representing growth of 26.5%, 12.3% and 50.1%, respectively. The model also assumes capital expenditure of $55.663B in FY26, $92.813B in FY27 and $97.603B in FY28, illustrating the scale of the infrastructure cycle even as Bernstein expects revenue, earnings and eventually free cash flow to accelerate. For valuation, Bernstein applies a 24.5x adjusted forward P/E multiple to estimated 12-month adjusted EPS of $15.22 one year ahead, then incorporates net cash per share of negative $47.6 to derive a USD 325 target. Against the USD 141.32 closing price on 1 September 2026, the reported implied upside is 130%. The institution reiterates Outperform and argues that Oracle is in the early stages of a cycle that could establish it as the world's third-largest hyperscaler.

Analysis framework

Bernstein first reconstructs Oracle's FY20–FY26 product-level revenue history, then classifies activities into growing, stable and declining buckets to identify the changing mix. It separately decomposes OCI into conventional infrastructure, cloud database, sovereign cloud, legacy hosting and AI data centers, and breaks SaaS into Fusion ERP, NetSuite, Fusion HCM and other products. The report then links revenue mix to gross margin, operating margin, capital expenditure, cash requirements, earnings and free cash flow. Finally, it builds FY27–FY28 financial forecasts and applies a forward adjusted earnings multiple to derive the price target.

Methodology notes

• Other

  Growth, Stable and Declining revenue-bucket analysis

  Bernstein classifies Oracle's product-level revenue streams by growth trajectory and tracks their shares over time. The method is used to show that faster-growing cloud businesses are becoming larger while mature and declining products become less important.

• Industry AnalysisVolume-price decomposition

  Product and segment revenue decomposition

  The report separates OCI, database cloud, SaaS products, licenses, support, hardware and services to identify which components drive consolidated growth and how their differing margins affect profitability.

• Corporate Fundamentals and FinanceThree-statement linkage

  Integrated income-statement, balance-sheet and cash-flow forecasts

  Bernstein projects revenue, expenses, earnings, capital expenditure, debt, cash and operating cash flow together to assess whether Oracle can finance its data-center expansion while preserving earnings growth.

• Valuation methodsP/E and PEG Valuation

  Forward adjusted P/E valuation

  Bernstein applies a 24.5x adjusted forward earnings multiple to estimated 12-month adjusted EPS of $15.22 one year ahead and incorporates negative $47.6 of net cash per share to obtain its USD 325 target.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Oracle Corp (ORCL.US)

  Primary covered company; Bernstein expects cloud infrastructure, AI capacity, sovereign cloud, multicloud database and core SaaS to drive accelerating growth.

  Strengths

  Rapid OCI growth, more than $300B of AI RPO, non-cancelable contracts, differentiated sovereign-cloud capabilities, a large installed database base, high-margin support and SaaS revenue, and strong Strategic Back Office growth.

  Weaknesses

  AI infrastructure is capital intensive and carries lower gross margin; legacy licenses, hosting and some on-premise activities are declining, while customer concentration is material.

  Comparison

  OCI is growing faster than hyperscaler peers from a smaller base; the report says it has surpassed Alibaba in size but remains well below Microsoft, AWS and Google Cloud.

  Risks

  Customer concentration, an AI-winter scenario, database displacement, slower SaaS ERP adoption, cybersecurity concerns, weaker Autonomous or MultiCloud Database demand, and OCI execution problems—especially in sovereign cloud.

Key data

• FY26 OCI IaaS/PaaS revenue$18.1BUp 77% year over year, versus approximately $10.2B and 50% growth in FY25.
• OCI share of total revenue27%Up from 18% in FY25 and 13% in FY24.
• Estimated FY26 OCI AI revenueApproximately $6.6BBernstein estimate because Oracle does not disclose the amount separately.
• AI remaining performance obligationsMore than $300BSupported by AI data-center contracts that generally ramp between FY26 and FY30.
• Cloud share of software revenueApproximately 58%Up from 50% in the prior year; cloud grew approximately 39% year over year.
• Latest constant-currency organic growth21%Reached this level in the latest two quarters after low-to-mid-single-digit growth through FY21.
• FY26 total SaaS revenue$15.889BThe detailed table shows 11% growth; a later narrative sentence states 15%, while Strategic Back Office grew 15%.
• FY26 Strategic Back Office revenueApproximately $9.919BUp from $8.625B in FY25 and growing 15%.
• Q3 FY26 MultiCloud Database growth531% year over yearBernstein estimates a 3x-plus revenue lift relative to database support.
• FY26 revenue bucketsGrowth 38%, Stable 40%, Declining 22%Software-revenue narrative estimate, versus 33%, 44% and 24% in FY24; Exhibit 5 separately displays 48%, 34% and 18%.
• Revenue forecastsFY26 $67.358B; FY27 $90.424B; FY28 $136.842BThe displayed forecast gives a 42.5% CAGR.
• Adjusted EPS forecastsFY26 $7.63; FY27 $8.57; FY28 $12.86Corresponding growth rates are 26.5%, 12.3% and 50.1%.
• Adjusted operating marginFY26 42.9%; FY27 42.0%; FY28 40.8%OCI's lower gross margin weighs on mix, but the report expects operating margins to remain substantial.
• Capital expenditure forecastsFY26 $55.663B; FY27 $92.813B; FY28 $97.603BReflects the scale of the data-center capacity build-out.
• Valuation inputs$15.22 adjusted EPS, 24.5x P/E, negative $47.6 net cash per shareInputs used to derive the USD 325 price target.
• Price target and upsideUSD 325.00 and 130%Compared with the USD 141.32 closing price on 1 Sep 2026.

Impact & implications

Bernstein's analysis implies that Oracle's growth is broadening beyond AI training capacity. OCI scale, sovereign cloud, cloud database and Strategic Back Office should accelerate revenue, while recurring support, SaaS and potentially high-margin multicloud revenue help absorb the lower gross margin and heavy capital expenditure associated with AI data centers. If the contracted capacity ramps as expected, the institution believes Oracle can produce faster revenue, earnings and free-cash-flow growth and potentially become the third-largest hyperscaler.

Risks

• Customer concentration could make contracted AI infrastructure revenue dependent on a limited number of large clients.
• An AI-winter scenario in which generative AI fails globally could reduce demand for Oracle's AI capacity.
• Cloud-native and alternative databases could place greater pressure on Oracle Database sales.
• SaaS ERP adoption could proceed more slowly than Bernstein expects.
• Cybersecurity or related concerns could reduce demand for Oracle Cloud.
• Demand for Autonomous Database or Oracle MultiCloud Database could fall short of expectations.
• OCI growth could disappoint, particularly within sovereign cloud.
• The profitability, stability and cash requirements of the large AI data-center build-out remain central investor concerns.

What to watch

• Track whether OCI growth accelerates again in FY27 as management guides.
• Monitor the timing of contracted data-center capacity coming online between FY26 and FY30.
• Watch the mix between AI training, inference, sovereign-cloud and conventional CPU-based OCI revenue.
• Track whether Oracle's core software businesses generate enough cash to fund the capital program without further material cash needs.
• Monitor gross and operating margins as lower-margin AI infrastructure becomes a larger revenue contributor.
• Watch the scaling and margin profile of Oracle MultiCloud Database following its 531% Q3 FY26 growth.
• Track continued growth in Fusion ERP, NetSuite and Fusion HCM as core ERP growth moderates.