Oracle Shares Whipsawed After OpenAI Revenue Scare and Morgan Stanley Debt Warning

Oracle shares fell roughly 5.5% on Thursday after a report that OpenAI revenue came in below investor expectations rattled the entire AI infrastructure trade, then partially rebounded Friday as analysts argued the selloff was overdone. Because Oracle's cloud backlog is heavily concentrated in contracts tied to OpenAI and other cash-burning AI startups, any sign of demand softness hit the stock disproportionately hard.

Morgan Stanley issued a pointed warning that delays in Oracle's data center buildout could pressure the company's already stretched balance sheet, framing the risk as "delayed revenue, delayed cash flow" that could weigh on its ability to service debt. The concern is not trivial: S&P has cut Oracle's credit rating to BBB-, one notch above junk, and the cost of insuring Oracle's debt recently hit a record high. The stock is down roughly 59% from its peak. Oracle has been borrowing, selling stock, and spending more cash than it earns as it races to build capacity for contracts that are not expected to generate meaningful revenue until around 2027.

The infrastructure strain is visible in unusual operational details. Oracle has resorted to trucking natural gas to some of its AI data centers because pipeline capacity has not kept pace with the rapid expansion, a workaround that illustrates both the urgency and the execution risk embedded in the buildout.

Analyst Paul Kedrosky identified four distinct bubble risks facing Oracle, while a separate Yahoo Finance piece noted that a recent share surge had narrowed the margin of safety enough to justify exiting the position. On the other side, a Seeking Alpha analysis argued that doomsday fears are far overblown and that the selloff created an opportunity. A Form 4 filed October 9 disclosed insider stock activity at the company. The debate between the two camps is essentially a disagreement about whether Oracle's 2027 revenue ramp will materialize as contracted or whether counterparty risk from money-losing AI startups makes that assumption fragile.