What was canceled
Oracle said Larry Ellison had canceled the trading plan adopted on June 22, which allowed the sale of up to 50 million company shares. At Friday's closing price, the maximum possible block was worth about $7.5 billion. [1 · Reuters]
No shares were sold under the plan. Oracle also said Ellison had no other active plans to sell the company's securities. [1 · Reuters]
The context of the decision
The plan was due to expire on October 24, but Ellison canceled it early. Oracle did not disclose the reason, so there is no basis for linking it to a specific forecast by the executive about the share price. [1 · Reuters]
Ellison remains Oracle's largest shareholder and, according to LSEG[1], owns more than 38% of the company. Oracle shares had fallen nearly 23% since the start of the year at the time of the announcement. [1 · Reuters]
Sources
- Reuters — Larry Ellison cancels his Oracle share-sale plan — September 12, 2026
- Reuters — Oracle’s quarterly results — September 10, 2026
- Cohen, Malloy and Pomorski — Decoding Inside Information — The Journal of Finance, 2012; a study of completed trades
- SEC — disclosure of executives’ trading plans — Primary source on the aims of disclosure reform, 2022
Expert commentary
I see Ellison's cancellation of the plan as a change in the public signal about an owner's intentions. It brings Oracle no new cash and does not replace financing for its projects. A possible easing of concern about a large sale deserves attention, but the reason for the decision was not disclosed. It would therefore be a mistake to connect it with guaranteed business growth. An assessment of the company should distinguish the shareholder's actions from Oracle's own ability to meet its obligations. [1 · Reuters]
The mechanism for expanding computing infrastructure matters more to the industry. Oracle's latest results show a gap between intensive investment and cash returns, with customer prepayments helping to finance it. In my view, competition here will depend on the ability to turn contracts into working capacity without putting excessive strain on cash flow. The founder retaining his stake could support confidence, but does not establish the technical readiness of sites or the economics of serving orders. [2 · Reuters]
Cohen, Malloy and Pomorski's research shows that insider trades differ in how informative they are: some are driven by recurring reasons and say little about a company's future. An important limitation is that the authors studied completed trades, so their findings cannot be applied directly to a canceled plan. The broader research lesson is useful: before interpreting an executive's action as a signal, alternative explanations need to be considered. The absence of a sale does not let us reconstruct the unknown motive behind the decision. [3 · Cohen, Malloy and Pomorski] [1 · Reuters]
For Oracle's customers, the reliability of available capacity and predictability of contracts matter more. If financial pressure begins to affect deployment schedules, support or commercial terms, the consequences will become operational. This is a conditional risk, not an established result of canceling the sale. I would track completion of launch milestones, actual service availability and the ability to distribute critical workloads among providers. Such observation better reflects customers' interests than interpreting the personal intentions of the largest shareholder. [2 · Reuters]
At the societal level, this story shows the importance of disclosure for trust in markets. The SEC explains trading-plan rule reforms as an effort to make executives' actions clearer to outside participants. My conclusion is that transparency is useful even when a trade does not occur, because it allows facts to be separated from assumptions. But a founder's prominence should not replace analysis of the company. This is particularly important where building digital infrastructure depends on the long-term trust of customers and creditors. [4 · SEC] [2 · Reuters]
I distinguish between two possible scenarios. If project execution and cash flow improve, Ellison's decision may be seen as consistent with business results. If the metrics do not improve, this event will retain little explanatory power on its own. Testable signals include customer receipts, spending on new capacity, commissioning schedules and financing costs. They need to be compared across several reporting dates: a single quarter or a share-price reaction does not establish a causal connection between the plan's cancellation and Oracle's prospects. [2 · Reuters] [3 · Cohen, Malloy and Pomorski]