Within a single day, Oracle founder Larry Ellison announced and then abandoned a plan to sell $7.5 billion of his own company's stock — a reversal that speaks less to indecision than to the weight of a moment. At 82, still holding more than 40 percent of the company he built in 1977, Ellison's choice to hold signals that he sees Oracle's costly transformation into an AI cloud giant as a wager still worth keeping. The episode unfolds against a backdrop of aggressive debt, restructuring, and a market that has not yet decided whether to trust the vision.
Ellison cancels $7.5B Oracle stock sale plan after one day
He's not taking money off the table right now
Why would Ellison announce a stock sale and then cancel it the very next day? That's an unusual move.
It signals confidence, actually. He's saying: I believe in where this company is going. I'm not taking money off the table right now. The timing matters too—Oracle just reported strong first-quarter earnings and a $26 billion jump in its revenue backlog.
But we should be careful about reading too much into his motives. A 10b5-1 plan is a mechanical thing—once you file it, you can cancel it. We don't know if he changed his mind, if there was pressure from the board, or if market conditions shifted. The company's statement doesn't explain why.
Fair point. What we do know is that Oracle is spending heavily—$45 to $50 billion this year on cloud infrastructure—and Ellison still owns more than 40 percent of the company. If he thought the debt load was unsustainable or the strategy was wrong, selling would make sense.
So the cancellation is a vote of confidence in the AI strategy?
It could be. Or it could be that selling 50 million shares would have spooked investors who are already nervous about how much debt Oracle is taking on. The stock fell 23 percent this year. A founder dumping $7.5 billion in stock might have triggered a panic.
The backlog number is real, though—$26 billion in new revenue commitments. That's concrete demand from major customers like Meta, Nvidia, OpenAI. The question is whether Oracle can turn that into cash flow faster than it's burning through money on infrastructure.
And the restructuring costs rising by $700 million—what does that tell us?
It tells us the company is under pressure. Job cuts are part of the plan. That's not a sign of smooth sailing. It's a sign they're trying to optimize costs while they're in the middle of this massive spending cycle.
But Oracle has done this before. They've been aggressive with debt and capital spending. The real test is whether the AI bet pays off and whether customers actually need the capacity they're building.
So Ellison's decision to hold his shares—is that the end of the story?
Not really. We should watch whether he tries to sell again, whether the stock stabilizes, and whether that revenue backlog actually converts to cash. The cancellation is one day's news. The outcome depends on what happens over the next year or two.
Le Pouls
- A $7.5 billion stock sale announced on Friday was cancelled by Saturday — one of the fastest reversals of a formal insider trading plan in recent memory.
- Oracle is carrying enormous financial weight: $45–50 billion in planned 2026 financing, $700 million in restructuring costs, and a stock that had already fallen roughly 23 percent before this episode.
- The company's cloud infrastructure push is real and in demand — customers like OpenAI, Meta, and Nvidia are lined up — but converting that backlog into cash flow remains the unresolved question.
- Markets responded with ambivalence: shares briefly surged 7.8 percent on strong revenue backlog news, then closed 2 percent lower, reflecting investor tension between belief in AI-driven growth and anxiety about how it is being funded.
- Ellison's decision to hold his stake rather than cash out is being read as a confidence signal — but one that the balance sheet has not yet fully earned.
Within a single day, Oracle founder Larry Ellison announced and then abandoned a plan to sell $7.5 billion of his own company's stock — a reversal that speaks less to indecision than to the weight of a moment. At 82, still holding more than 40 percent of the company he built in 1977, Ellison's choice to hold signals that he sees Oracle's costly transformation into an AI cloud giant as a wager still worth keeping. The episode unfolds against a backdrop of aggressive debt, restructuring, and a market that has not yet decided whether to trust the vision.
On Friday, Larry Ellison disclosed a plan to sell up to 50 million Oracle shares — a stake worth $7.5 billion. By Saturday morning, Oracle announced the plan was cancelled. No shares had been sold, and Ellison had no intention of selling any going forward.
The trading vehicle, a 10b5-1 plan adopted in June, is a standard mechanism allowing corporate insiders to sell shares on a predetermined schedule without the appearance of acting on inside information. This one was switched on and switched off within 24 hours.
The speed of the reversal draws attention to what Oracle is navigating. Ellison, who founded the company in 1977 and still controls more than 40 percent of it, is presiding over an expensive reinvention. Oracle has committed to raising between $45 and $50 billion in 2026 to build out cloud infrastructure — already completing a $20 billion stock offering in the first quarter alone. The goal is to serve an AI-hungry roster of clients including OpenAI, Meta, Nvidia, and xAI. But the spending has strained the balance sheet, and restructuring costs tied to job cuts are expected to climb by $700 million.
The market's reaction captured the underlying tension. Oracle shares briefly jumped nearly 8 percent after the company reported a $26 billion increase in its revenue backlog — a sign of genuine customer demand. But the stock closed the day about 2 percent lower, as analysts pointed out that strong bookings have not yet translated into strong cash flow.
Ellison's choice to hold rather than sell is a statement of conviction. Whether that conviction is vindicated depends on Oracle's ability to turn a growing pipeline of AI infrastructure contracts into the kind of financial returns that can justify the debt it has taken on to pursue them.
Larry Ellison announced on Friday that he would sell up to 50 million shares of Oracle stock—a transaction worth $7.5 billion at current prices. By Saturday, Oracle said the plan was off. No shares had been sold, the company stated flatly, and Ellison had no intention to sell any Oracle stock going forward.
The reversal came within 24 hours of Ellison's initial disclosure. The trading plan, formally known as a 10b5-1 Plan, had been adopted on June 22 and was scheduled to run through October 24. Such plans allow corporate insiders to sell shares on a predetermined schedule, a mechanism designed to avoid accusations of trading on privileged information. This time, the mechanism was activated and then immediately deactivated.
Ellison, 82, remains Oracle's largest shareholder. He founded the company in 1977 and continues to control more than 40 percent of it. His decision to hold onto his stake matters because Oracle is in the middle of an expensive transformation. The company has taken on substantial debt and is spending aggressively to build cloud infrastructure capacity. In February, Oracle said it expected to raise between $45 billion and $50 billion during 2026 through a combination of debt and equity offerings. The company has already completed a $20 billion stock sale in the first quarter and plans to raise $40 billion total through debt and equity financing in the current fiscal year.
The cloud infrastructure buildout is meant to serve a roster of major customers—AMD, Meta, Nvidia, OpenAI, TikTok, and xAI among them—all hungry for computing power to support artificial intelligence systems. But the spending spree has weighed on Oracle's balance sheet and on investor sentiment. The stock dropped roughly 23 percent over the year before Ellison's announcement. Restructuring costs, which include planned job cuts, are expected to rise by about $700 million as the company navigates this volatile period.
The market's reaction to Ellison's reversal was mixed. On Friday, Oracle's shares initially climbed as much as 7.8 percent after the company disclosed a $26 billion increase in its revenue backlog—a sign that customer demand for its services remains strong. But the stock later retreated, closing about 2 percent lower as analysts noted that a meaningful recovery in cash flow remains distant. The backlog increase and strong first-quarter earnings did help the stock recover from earlier underperformance, yet the underlying tension persists: investors are divided between confidence in Oracle's artificial intelligence-driven growth and unease about how the company is financing that growth.
Ellison's decision to cancel the stock sale, just one day after announcing it, suggests he believes Oracle's trajectory justifies holding his stake rather than cashing out. Whether that confidence is warranted depends on whether the company can convert its growing revenue backlog into actual cash flow while managing the debt burden it has accumulated in pursuit of cloud dominance.
Citations marquantes
No Oracle stock was sold under that plan, and he has no other plans to sell any of his Oracle stock— Oracle company statement