Larry Ellison Just Cancelled a 7.5 Billion Dollar Oracle Stock Sale

Larry Ellison was supposed to sell 50 million Oracle shares worth approximately 7.5 billion dollars. He has now cancelled that sale entirely. No detailed explanation has been offered, which has sent the tech finance world into a fairly predictable spiral of speculation.

The planned sale had been disclosed publicly, as required under SEC rules. Executives at public companies in the US must file a 10b5-1 plan when they intend to sell shares, which gives advance notice to the market and is meant to demonstrate that the sale is not based on insider information. Cancelling one of these plans is less common, and cancelling one at this scale is genuinely unusual.

Why This Is a Big Deal

To put 7.5 billion dollars in context: that is larger than the GDP of several countries. It is also a significant chunk of Oracle's market activity. Ellison is Oracle's chairman and chief technology officer, and he owns an enormous stake in the company he co-founded in 1977. His decision to sell, or not sell, sends a signal to the market whether he intends it to or not.

The question everyone is now asking is why. There are a few plausible explanations, none of which are confirmed.

The first is simple confidence in Oracle's near-term performance. Oracle has been one of the unexpected winners of the AI infrastructure boom. The company's cloud business, which provides computing power to AI developers, has been growing at a pace that surprised a lot of analysts. If Ellison believes Oracle's share price is going to climb significantly in the next year, selling now looks like leaving money on the table.

The second explanation involves the Stargate project. Oracle is a key infrastructure partner in Stargate, the joint venture announced earlier in 2026 to build AI data centres across the United States. Stargate involves commitments worth hundreds of billions of dollars over several years. If Oracle is about to announce a major expansion of its role in that project, selling a large block of shares right before that announcement would raise uncomfortable regulatory questions. Cancelling the sale sidesteps that problem cleanly.

A third, more speculative possibility is that Oracle is exploring something that would make a large cash sale less appealing right now. Acquisitions, for instance, or a significant capital raise that changes the calculus around Ellison's personal holdings.

Oracle's Position in the AI Race

What makes this interesting beyond the pure finance story is what it tells us about Oracle's momentum. This is a company that most people thought of as a legacy database and enterprise software business five years ago. The cloud computing and AI wave has repositioned it significantly.

Oracle's infrastructure cloud, known as OCI, has become a genuine alternative to Amazon Web Services and Microsoft Azure for AI workloads, partly because Oracle invested aggressively in the kinds of GPU clusters that AI training requires. Its partnership with Nvidia has been central to that. Several major AI labs now run significant portions of their training infrastructure on OCI rather than the more dominant hyperscalers.

The company reported strong revenue growth in its most recent quarter, with cloud infrastructure revenue growing at a rate that outpaced AWS and Azure growth figures for the same period. Ellison himself has been unusually energetic in public appearances, clearly relishing Oracle's moment in the spotlight after years of being written off as a dinosaur of enterprise tech.

What Happens Next

Ellison is not obligated to explain the cancellation beyond the basic regulatory filing. He may refile a new 10b5-1 plan in future, at different terms or a different scale. Or he may not.

For investors watching Oracle stock, the cancellation is generally interpreted as a bullish signal. If the founder and CTO does not want to sell at the current price, the inference is that he thinks the current price is too low. Markets tend to respond accordingly.

What this week's news reinforces is that Oracle is not the boring legacy tech play it appeared to be three years ago. The AI infrastructure buildout has created winners in unexpected places, and Ellison's decision to hold onto 7.5 billion dollars worth of stock suggests he thinks Oracle's role in that buildout is far from finished.