ORCL

The company began as Software Development Laboratories, started by Bob Miner, with Larry Ellison and Ed Oates among the early team. The founders had read Edgar Codd's research on the relational model, a mathematical approach to organizing data that was elegant in theory but had no commercial product behind it. They built one. The name Oracle came from an intelligence agency project that was the company's first customer. By shipping a working relational database before IBM commercialized the idea it had pioneered in research, Oracle captured an early lead in a category that would become foundational to all of business computing. Through the 1980s and 1990s the database spread across corporate data centers, and Oracle grew through both rapid sales expansion and a long series of acquisitions. The purchases of PeopleSoft, Siebel, and Sun Microsystems in the 2000s pushed Oracle into enterprise applications, customer relationship management, hardware, and the stewardship of the Java programming language and the open-source MySQL database.

Oracle's business today is organized around three broad areas. The first is cloud services and license support, which includes the database business in all its forms. This covers the flagship Oracle Database, the Autonomous Database that automates tuning and patching, the Exadata engineered systems that run it at high performance, and MySQL HeatWave. The database franchise is the company's profit engine and the source of its durability. The second area is applications, the cloud software suites that companies use to run their operations. Oracle Fusion Cloud Applications cover finance, human resources, and supply chain for large enterprises, while NetSuite serves the same needs for smaller and mid-sized businesses. These are subscription products with high switching costs, because replacing the system that runs a company's payroll, ledger, and procurement is expensive and risky. The third and most closely watched area is Oracle Cloud Infrastructure, known as OCI, the raw computing, storage, and networking layer that competes with the largest cloud platforms and increasingly hosts the artificial intelligence workloads driving Oracle's recent growth.

The economic foundation of the company is the database moat. Enterprise databases are among the stickiest products in all of software. The data inside them is mission-critical, the applications built on top assume their specific behavior, and migration projects are slow, costly, and prone to failure. A large bank or insurer that has run Oracle for decades faces enormous friction in moving to a competitor, which gives Oracle durable pricing power and a recurring stream of license and support revenue. This support revenue carries very high margins and funds the rest of the business. The applications layer reinforces the same dynamic. Once a company runs its core operations on Fusion or NetSuite, the cost of leaving is measured in years of effort and operational risk, not in a simple software comparison. This combination of a sticky database and sticky applications is what allowed Oracle to remain highly profitable even during the years when its cloud infrastructure was widely seen as a distant follower.

That follower position is what changed. For most of the past decade, Oracle was a minor player in cloud infrastructure relative to Amazon Web Services, Microsoft Azure, and Google Cloud, collectively known as the hyperscalers. OCI was a second-generation rebuild of Oracle's cloud, architected by an engineering team that included Clay Magouyrk, who joined from Amazon, and it was designed to run demanding workloads at competitive cost. The pivot that reshaped the company came with the artificial intelligence boom. Training and serving large models requires vast clusters of graphics processing units connected by high-speed networking, and Oracle positioned OCI as a place to rent that capacity at scale. The result has been extraordinary reported growth. By Oracle's fiscal 2026, OCI revenue was growing at rates far above the rest of the business, and total cloud revenue had risen to roughly half of company revenue, a milestone that would have seemed unlikely a few years earlier.

The clearest measure of the shift is the contracted backlog, which Oracle reports as remaining performance obligations, or RPO. This figure represents future revenue that customers have committed to under signed contracts but that has not yet been delivered or recognized. Oracle's RPO expanded dramatically through fiscal 2026, climbing past 400 billion dollars and then toward roughly 500 billion dollars and beyond as the company signed very large multi-year capacity contracts. Much of this backlog is tied to a small number of enormous artificial intelligence customers. The most prominent is the Stargate effort, a multi-hundred-billion-dollar data center buildout led by OpenAI with SoftBank and Oracle as partners, under which Oracle committed to delivering several gigawatts of computing capacity over roughly five years. The first large Stargate facility, in Abilene, Texas, runs on OCI. Oracle has guided to a steep multi-year revenue ramp for OCI on the strength of these commitments, projecting figures that would, if delivered, transform the size and shape of the company.

This is also where the risk concentrates. Delivering on the backlog requires building physical capacity at a pace that demands extraordinary capital. Oracle's capital expenditures rose sharply to roughly 50 billion dollars in fiscal 2026, and the company funded the buildout through a combination of operating cash flow, large debt issuance, and equity. Total debt rose well past 100 billion dollars over the course of the year, and trailing free cash flow turned deeply negative as spending on data centers and chips outran the cash the business generated. Credit-rating agencies and analysts have flagged the increase in leverage, and Oracle's credit risk reached levels not seen before in its history. The logic of the strategy is straightforward. The contracted backlog gives Oracle confidence that the capacity it is building will be paid for, so it is willing to spend ahead of the revenue. The danger is equally clear. If the artificial intelligence demand that underpins those contracts softens, if a major customer delays or fails to consume what it committed to, or if the cost of capital rises while the buildout is still in progress, Oracle would be left carrying very large fixed obligations against revenue that arrived more slowly than planned.

Competition runs along two distinct fronts. In databases, Oracle still faces pressure from cloud-native alternatives, open-source systems, and the database services offered directly by the hyperscalers. Oracle has responded with a multicloud strategy, placing its database inside Amazon, Microsoft, and Google data centers so that customers who have standardized on those clouds can still run Oracle Database close to their other workloads. This multicloud database revenue has grown quickly and represents a notable shift in posture, with Oracle choosing to cooperate with its infrastructure rivals in order to keep its database franchise central. In infrastructure, Oracle is the smallest of the major players by overall scale, but it has positioned OCI as a specialist for large, performance-sensitive workloads and has won business partly on price and partly on its willingness to dedicate enormous blocks of capacity to single customers. Whether OCI can sustain its growth while the hyperscalers, all of which are spending heavily on their own artificial intelligence capacity, defend their positions is one of the central questions facing the company.

Leadership reflects both continuity and a recent transition. Larry Ellison, who co-founded the company and served as chief executive from 1977 to 2014, remains executive chairman and chief technology officer and continues to be the dominant strategic voice, deeply involved in product and infrastructure direction. Safra Catz, who led the company as chief executive from 2014 and is widely credited with the financial discipline and operational rigor that defined that era, moved in September 2025 into the role of executive vice chair of the board. In her place Oracle named two chief executives, an unusual structure for a company of its size. Clay Magouyrk, who came from the OCI engineering organization, and Mike Sicilia, who led Oracle's industry applications business, took the top operating roles, signaling that the company sees its future as resting on both the infrastructure buildout and the applied use of artificial intelligence inside its software. The concentration of influence in Ellison, who remains one of the company's largest shareholders, is a defining feature of how Oracle is governed.

The strategic bet is unambiguous. Oracle is wagering that it can convert its database-funded balance sheet and its engineering capacity into a durable position as a major supplier of artificial intelligence infrastructure, capturing a large share of a category that the hyperscalers have historically dominated. If the demand holds and the capacity is delivered on schedule, the company would emerge meaningfully larger and more central to the computing economy than its software legacy alone would suggest. The risks are the mirror image of that ambition. The backlog is heavily concentrated in a handful of customers whose own businesses are unproven at the scale of their commitments. The capital required is immense and front-loaded, the debt load has grown faster than at any prior point, and free cash flow is negative while the buildout proceeds. Layered on top is the broader uncertainty about whether the current pace of artificial intelligence investment across the industry will be sustained or will moderate.

For an investor trying to understand Oracle, the company is best seen as two businesses fused together at a specific moment in time. One is a mature, highly profitable enterprise software franchise whose database and applications generate dependable cash and whose customers are extremely difficult to dislodge. The other is a young, fast-growing, capital-hungry infrastructure operator making one of the largest spending bets in the technology sector. The first business is what makes the second one possible, because it provides the cash flow and the credit standing that fund the buildout. The forward question is whether the infrastructure bet pays off before its costs strain the foundation that supports it. The contracted backlog suggests the demand is real and committed, but contracts are promises, and the gap between a signed obligation and delivered, profitable revenue is where the outcome will be decided. How Oracle manages that gap, the balance between the certainty of its legacy and the scale of its ambition, will determine which of its two identities defines the company a decade from now.