It is a rare sight for a company with a market capitalization of up to $400 billion to see its stock price shed more than half of its value while its sales and profits continue to climb. Yet Oracle (ORCL) is exactly such a case. The stock currently trades at around $144, down 58% from its all-time high of $345.72 set in September last year, even as the business just delivered its best annual performance in history: in fiscal 2026, revenue grew 17% to approximately $67.4 billion, and net income rose 37% to roughly $17 billion.
There is no sign of operational difficulty in Oracle’s latest financial results. In the fourth quarter ended May 31, 2026, revenue increased 21% year over year to $19.2 billion. Total cloud revenue rose 47% to $9.9 billion, with Oracle Cloud Infrastructure—which provides computing power rentals for AI workloads and is the most closely watched segment by investors—posting a 93% jump to $5.8 billion, accelerating from what was already rapid growth at the start of the year.
Even more striking is the scale of its backlog. As of year-end, Oracle’s remaining performance obligations—that is, contract amounts not yet recognized as revenue—stood at $638 billion, an astonishing 363% year-over-year increase, with a single-quarter rise of $85 billion, reflecting a series of large-scale AI computing deals signed with multiple hyperscale customers.
However, such a massive backlog also entails customer concentration risk. The recent surge has come primarily from a handful of giant AI transactions, in which clients either prepay Oracle to procure chips or supply their own chips. This concentration cuts both ways—on one hand, it generates spectacular growth numbers; on the other, it ties Oracle’s fate to the spending appetite of a few large customers.
Demand is beyond question; the issue lies in the cost of meeting that demand.
The core issue that the market repeatedly focuses on is cost. Building data center capacity is enormously expensive, and Oracle is raising funds through debt. Capital expenditures in fiscal 2026 amounted to roughly $55.7 billion, with total new debt added during the year at about $43 billion and further financing plans in place. This former cash-rich company is now spending far more than it earns, as management races against time to build infrastructure and install chips to meet demand, causing free cash flow to turn deeply negative.
This is precisely why the stock price continues to be under pressure.
The realization of the $638 billion backlog depends on Oracle’s ability to secure financing, complete construction, populate data centers, and convert contracts into profitable revenue on schedule—and the majority of that backlog will unfold over several years rather than over the coming quarters. If demand for AI computing slows, or if the returns on massive investments fall short of expectations, the debt burden will remain while earnings shrink and are further delayed.
After a 58% decline, a considerable portion of the risk may have already been priced into the current stock price. Oracle now trades at a price-to-earnings ratio of about 25, with a forward P/E of roughly 17—not a high valuation for a company that still posts double-digit revenue growth and even faster cloud business expansion.
Yet a cheap stock does not necessarily mean it should be bought.
The bullish case is straightforward: if Oracle can translate that massive backlog into actual revenue, the current price will look highly attractive. But the bearish case is equally clear: the high-debt, high-capital-consumption expansion model leaves management with very little room for error, and the sharp stock price swings over the past month have fully demonstrated how quickly market sentiment can turn.
Between these two scenarios lies an Oracle that is vastly different from the company of just a few years ago—its fortunes now depend far more on financing and building data centers than on software sales, and its risk profile has changed significantly.
Overall, Oracle’s current stock price finally reflects serious skepticism about its prospects. For investors who believe that the AI cloud computing boom will continue, establishing an initial position may have some justification. However, given the company’s high leverage and aggressive spending pace, a more prudent approach would be to wait until Oracle converts more of its backlog into actual cash inflows before adding to positions. Investors might consider a small trial stake, allowing actual performance—rather than a discounted price—to validate the investment thesis.
