NextFin News - Oracle’s stock has just endured its worst weekly slump since the 2001 dot-com bust, a sharp reminder that the AI trade is now being judged on financing as much as on growth. The shares fell 19% over the week, after the company’s latest results and outlook pushed investors to focus on a simple question: can Oracle keep building AI infrastructure fast enough to meet demand without stretching its balance sheet too far? The answer, at least for now, is that the market is no longer willing to give the company the benefit of the doubt.
Oracle said fiscal fourth-quarter revenue rose 21% to $19.2 billion, cloud revenue climbed 47% to $9.9 billion, and remaining performance obligations reached $638 billion from $553 billion in the prior quarter. Those figures are strong by almost any standard and would normally support a much friendlier stock reaction. But the same disclosure also showed a far more capital-intensive reality. Oracle said fiscal 2026 capital spending reached $55.66 billion, above its earlier $50 billion target, while free cash flow for the year was negative $23.7 billion. The company also said it expects to raise nearly $40 billion in fiscal 2027 through debt and equity financing to keep expanding its AI data-center footprint.
That combination has become the market’s central concern. Oracle is winning large cloud and infrastructure commitments, including work tied to AI customers such as OpenAI, but the revenue recognition from those deals comes much later than the capital required to build the capacity. As a result, Oracle’s AI story is increasingly a funding story. Investors are asking not whether demand exists, but who pays for the servers, power, land, networking equipment, and interest expense while the buildout is still running ahead of cash generation.
The selloff was severe because it hit a stock that had already been treated as a major AI beneficiary. Oracle’s market value had approached $900 billion at its peak in September, a level that implied a lot of faith in the company’s cloud and infrastructure strategy. Once a stock reaches that size, the market tends to require more than top-line growth. It wants proof that the growth can be funded internally, or at least funded without creating a balance-sheet overhang. Oracle’s latest figures did not clear that bar.
The company’s own numbers explain why the debate has sharpened. Cloud infrastructure revenue rose 93% in the quarter to $5.8 billion, an extraordinary rate for a company of Oracle’s scale. But the same quarter showed an enterprise leaning hard into capital spending to chase that demand. In fiscal 2027, management said it expects to raise nearly $40 billion and to spend as much as $95 billion on capital expenditures, though it also expects some customer reimbursements to offset part of that total. The details still point to a business model that is becoming more capital intensive just as the AI cycle is becoming more competitive.
Market Reaction
The weekly decline was not a one-off response to an earnings headline. Oracle fell at least 2.6% in each of the five trading sessions, making the damage unusually consistent for a large software name. The stock was also down 8.9% in after-hours trading on the day of the earnings release, before the losses deepened through the rest of the week. That pattern suggests investors were not merely disappointed by one forecast. They were repricing the entire funding equation.
That repricing matters because Oracle had become one of the market’s clearest symbols of the AI infrastructure boom. For months, investors had treated the company’s cloud ambitions and data-center expansion as evidence that Oracle could play a much bigger role in the AI supply chain. But once the spending trajectory became explicit, the stock moved from being a growth story to being a capital-structure story. In other words, the market is still willing to pay for AI exposure, but it wants a cleaner path from contracts to cash.
The scale of the funding requirement is the key issue. Oracle ended May with about $130 billion in debt, according to the company’s own disclosure, and the firm said it had already spent $55.66 billion on capital expenditures in fiscal 2026. That is a lot of borrowing and spending for a company whose free cash flow was negative $23.7 billion over the same period. Even for investors who believe the AI opportunity is real, the question becomes whether Oracle is financing a durable asset base or simply pulling forward years of cash burn.
The market’s verdict so far is cautious. A company can usually get away with one of two things: high spending or a stretched balance sheet. Oracle is asking investors to accept both at once, which is a much harder sell. That is why the stock’s move was so violent and why the comparison with 2001 resonated. It was not nostalgia. It was a signal that the market was suddenly less comfortable extrapolating growth from a business model that depends on very large upfront investment.
“Oracle plans to raise nearly $40 billion in debt and equity in 2027.”
That line captures the heart of the problem. The company is not hiding the financing burden. It is acknowledging it. In a market that had been rewarding AI winners for growth alone, the disclosure made Oracle look less like a simple beneficiary of the theme and more like a test case for how expensive that theme can become.
Why the AI Buildout Is Still Credible
To be clear, Oracle’s selloff does not mean the AI strategy has failed. In fact, the operating data suggest the opposite: demand is real, the backlog is enormous, and the company is landing work in a market where compute capacity remains scarce. Oracle said remaining performance obligations reached $638 billion, a record figure that indicates a huge amount of contracted future revenue. It also said cloud revenue rose 47% in fiscal fourth quarter, with infrastructure revenue up 93%. Those are not the numbers of a business that is losing relevance.
The challenge is that the market now cares as much about the funding mechanism as about the demand signal. Oracle’s business is becoming more like an infrastructure utility and less like a traditional software vendor. That shift has consequences. Utility-style businesses often require heavy upfront capital, long payback periods, and close attention to financing costs. They can be excellent businesses, but they do not deserve the same valuation logic as a low-capex software model that throws off cash quickly.
That distinction is especially important in AI, where the time between winning a contract and generating cash can be long. Oracle may book massive commitments today, but it still has to spend heavily on data-center capacity before it can recognize the full economic value of those contracts. If the cost of debt rises, if construction timelines stretch, or if customer demand slows, the gap between promise and profit widens fast.
Oracle’s management is clearly betting that scale will solve part of this problem. In the company’s view, a larger installed base of AI infrastructure should help spread fixed costs and improve operating leverage over time. That may prove correct. But the market has become more demanding after several quarters of AI enthusiasm, and it is now asking for evidence rather than narrative. A record backlog and a fast-growing cloud business are a strong start. They are not yet a complete answer.
The pressure is not just financial. It is also strategic. Oracle is competing against larger cloud providers with deeper pockets and more diversified cash flows. Those rivals can often fund data-center expansion with less stress on their balance sheets. Oracle’s smaller scale means each additional dollar of investment carries more risk, because the company has less room to absorb a miss in execution or timing. In a capital-heavy race, balance-sheet flexibility becomes part of the product.
That is why the company’s guidance is so important. Oracle said it expects to keep spending heavily in fiscal 2027 and still reach rapid delivery milestones as it expands its infrastructure. If those milestones hold and the revenue converts as expected, the current selloff may later look like a temporary valuation reset. If they slip, investors will probably conclude that the market was right to demand a higher risk premium.
What Changed This Time
The key change is not that Oracle is spending. It is that the spending has become large enough to overwhelm the usual “growth will fix it later” argument. In earlier AI rallies, investors could focus on the size of the opportunity and ignore the funding mix. Oracle’s latest quarter made that harder. The company is now producing both strong growth and a visibly heavier financing requirement, and those two facts are colliding in the stock price.
This is also why the historical comparison matters. The 2001 reference does not mean Oracle is in the same fundamental position it was during the dot-com era. Today’s Oracle has much more recurring revenue, a broader enterprise footprint, and a far stronger cloud franchise. But the market reaction still reflects an old truth: when financing needs rise faster than confidence in future cash flow, even a good business can trade like a risky one.
Another reason the selloff has bite is that the AI theme has become crowded. A crowded trade can tolerate a little bad news, but it becomes brittle when one of the marquee beneficiaries starts to look overextended. Oracle’s move therefore has implications beyond the stock itself. It forces investors to reexamine the financing assumptions behind the wider AI infrastructure buildout, from cloud capacity to power contracts to debt issuance.
In that sense, Oracle’s week was a repricing of the whole theme. The market is not saying AI spending is worthless. It is saying the financing path now matters enough to change valuations. A record backlog is impressive. A record backlog financed by rising debt and negative cash flow is less comforting.
What Investors Will Watch Next
The next clues will come from how quickly Oracle converts contracted demand into cash flow and how much of the planned financing it actually needs to raise. Investors will also watch whether capital expenditures begin to moderate as new capacity comes online, because that would be the first sign that the spending cycle is moving from expansion to harvest. Any update on debt issuance, equity sales, or customer reimbursements will matter as well.
For the broader market, Oracle is now a case study in how the AI trade matures. Early enthusiasm rewarded companies for simply exposing themselves to the theme. Later-stage enthusiasm will reward only those that can finance it efficiently. That shift is healthy, but it is also punishing. It means the market is starting to differentiate between AI demand and AI economics.
The sharpest lesson from Oracle’s worst week since 2001 is that the AI story has not ended. It has just become more expensive to own. Investors still want the growth. They are now insisting on a more convincing bill of materials.
Explore more exclusive insights at nextfin.ai.
