The $638 Billion Oracle Paradox: A Record Backlog Meets a Crushing Cost of Delivery

Oracle's $638B backlog masks a capital-intensive pivot: $55.7B in capex, credit downgrade risk, and a 52% stock plunge as investors fear debt-fueled AI cloud expansion.

Oracle sits on an unprecedented pile of contracted revenue—$638 billion in remaining performance obligations (RPO)—yet its stock is trading within a hair’s breadth of its 52-week low. At €103.00, the shares have shed 38% of their value since January and a staggering 52.6% over the past twelve months. The disconnect between the company’s order book and its market price has become one of the most striking contradictions in technology investing today.

The RPO figure, disclosed at the close of fiscal 2026, represents a 363% surge from the prior year. To put that in perspective: at Oracle’s current revenue run rate, this backlog equates to nearly a decade of business. Major artificial intelligence players including OpenAI and Meta have effectively reserved Oracle Cloud Infrastructure (OCI) capacity for years to come. Yet instead of celebrating this visibility, investors are treating the backlog as a liability.

The reason lies in what it takes to convert those contracts into cash. Oracle is no longer a capital-light software company with fat margins; it has transformed into a capital-intensive infrastructure operator that must spend billions upfront before a single cloud dollar hits the income statement. Capital expenditures for fiscal 2026 hit approximately $55.7 billion—more than double the prior year’s level. To bridge the funding gap, the company plans to raise between $45 billion and $50 billion this year through a combination of new bonds and equity issuance, including an at-the-market stock offering program.

That aggressive financing strategy has rattled credit markets. While Oracle retains investment-grade status, the margin for error has narrowed dramatically. S&P Global downgraded the company’s credit rating to BBB- on July 9—just one notch above junk territory. Two weeks later, Moody’s followed with a warning about cash-flow risks, maintaining a negative outlook. The cost of insuring Oracle’s debt against default has climbed to 200 basis points, as measured by credit default swaps.

Adding an unusual layer of risk: founder Larry Ellison has personally guaranteed $40.4 billion related to a separate acquisition. That exposure has fueled speculation that he might sell shares from his personal stake to meet obligations—a “founder discount” that further weighs on an already beaten-down stock.

Even positive developments struggle to gain traction. A software contract with the U.S. Department of Defense, announced July 23, barely registered with investors. The market’s attention remains fixed on the balance sheet, not the sales pipeline.

Technically, the stock looks ripe for a bounce. The 14-day relative strength index sits at 31.3—deep in oversold territory that historically precedes at least a temporary recovery. The annualized volatility over the past 30 trading days has approached 47%, reflecting extreme nervousness among holders.

Yet the most jarring number may be the gulf between price and analyst expectations. The consensus price target among Wall Street analysts stands at €217.90, implying upside of more than 111% from current levels. For a company with a market capitalization of €303.65 billion, such a disconnect is almost unheard of. Analysts are betting on the long-term economics of that $638 billion backlog; the market is fixated on the short-term cash burn required to deliver it.

Oracle’s quarterly dividend of $0.50 per share—ex-dividend since July 10—offers a modest incentive for patient holders, but it does little to address the core anxiety. The company is running deeply negative free cash flow as it builds out data centers and secures Nvidia chips to fulfill its multi-year agreement with OpenAI, announced in early June.

The coming quarters will determine whether Oracle’s massive RPO can eventually translate into the kind of cash generation that justifies the analyst consensus. For now, the company is a case study in how expensive the pivot to AI infrastructure can be—and how quickly the market can punish a balance sheet, even when the order book has never looked stronger.