Oracle closed Thursday down sharply after the Financial Times reported OpenAI’s annualized revenue was far below earlier estimates, rattling investors already nervous about Oracle’s infrastructure spending. By Friday afternoon, ORCL had reversed course, rising 4.76% to $141.62 as new AI products, a data center update, and bullish analyst coverage supported the rebound. The question worth asking is not whether the bounce was warranted. It is whether the bull case behind it can actually be executed.
The bull case rests almost entirely on one number. Mizuho elevated Oracle to the top spot on its October Top Picks list, setting a $320 price target roughly 25% above the Bloomberg consensus and projecting 35% annual revenue growth from fiscal 2026 through 2030, underpinned by an estimated $638 billion AI cloud order backlog. That backlog figure represents contracted demand. Converting it into recognized revenue requires data centers to come online on schedule.
Project Jupiter is where the conversion either happens or stalls. Oracle issued a public statement affirming that its Project Jupiter AI data center campus in New Mexico remains on its planned schedule and that the company is “fully committed to New Mexico,” directly countering the prior day’s headlines about pipeline delays and construction setbacks. Investors needed to hear it. Two weeks ago, Oracle had sent a force majeure notice to developer Blue Owl, seeking to preserve its right to defer payments if its 2.45-gigawatt New Mexico AI data center campus misses its 2028 opening. The underlying obstacle is a natural gas pipeline: New Mexico’s State Land Office has twice denied access across state land for 0.6 miles of a proposed 17-mile gas pipeline serving Project Jupiter, citing emissions and pressure on water and natural resources. Meanwhile, Project Jupiter loans have been quoted at 89 to 91 cents on the dollar, reflecting investor concerns over construction delays and Oracle’s credit profile.
Friday also brought a concrete vote of confidence from inside the boardroom. Director Stephen Rusckowski declared the acquisition of 25,000 shares at about $139.35 apiece, a transaction worth around $3.48 million. Insider purchases at that scale are rarely performative. Executives buying near a multi-week low while the stock trades well below a firm’s published price target is a signal worth registering, even if it tells you nothing about the pipeline permit.
The product announcements added operational texture. Oracle announced Oracle Fusion Claw, a governed AI execution runtime that integrates AI reasoning with deterministic enterprise computation, while its financial services division simultaneously launched Oracle Nexus Case Flow and Oracle Nexus Reach, agentic AI tools built to assist banks and institutions with anti-crime investigations and compliance workflows. These are not headline-grabbing consumer products. They are the kind of enterprise tools that expand Oracle’s installed base and deepen switching costs.
The bear case is straightforward. Mizuho flagged risks including a planned $45 billion to $50 billion raise and thinner AI cloud margins. Oracle generated $23.1 billion of operating cash flow in the fiscal first quarter, up 184%, but free cash flow was negative about $5.4 billion as the company invested in cloud infrastructure. A credit profile already sitting one notch above junk leaves little margin for a further setback at Project Jupiter.
The investment thesis is not complicated to state: Oracle has won the contracts. The $638 billion backlog is real. The next test is whether Oracle can monetize its huge AI backlog fast enough to cover the financial costs of building the infrastructure needed. Investors watching ORCL in the weeks ahead should track the New Mexico permitting calendar, not the price target. One permits the other.
