People

Management & Governance

The verdict in one line: Oracle is a founder-controlled company where the economics are superbly aligned — Larry Ellison owns 40.6% — but minority shareholders inherit two concentrated risks they cannot vote away: roughly 346 million of Ellison's shares pledged against personal loans, and an unproven dual co-CEO leadership handed $350M of new option grants weeks into the job.

Source: DEF 14A filed September 26, 2025 (beneficial ownership table; board roster). Grade is our assessment.

This is a control story. One man's 40.6% economic stake makes every other governance feature — independent committees, a lead independent director, improving say-on-pay — true but secondary. The right question for an outside investor is not "is the board independent?" (mostly yes) but "what happens to me if Ellison's financing, health, or judgment turns?"

The control: one owner, one big lien

Source: DEF 14A September 26, 2025, beneficial ownership table (as of record date Sept 19, 2025).

Ellison holds 1,158,232,353 shares — 40.6% of the company, with no dual-class structure inflating it. Every vote he casts is bought with his own capital, and Oracle's relentless buybacks quietly concentrate his stake further each year. That is the bull case for alignment: the largest decision-maker is also, by a vast margin, the largest owner. The entire rest of the board and officer group holds under 1% combined; the only other 5% holder is Vanguard at 5.3%.

The catch sits in footnote (2) of that same table. Of Ellison's 1.16 billion shares, 346 million are pledged as collateral to secure personal lines of credit. Oracle maintains a strict anti-pledging and anti-hedging policy for every employee and director — except Ellison, who is carved out by name. The Governance Committee reviews the arrangement quarterly, hires outside advisors, and concludes Ellison "has the financial capacity to repay his personal term loans without resorting to the sale or transfer of pledged shares," noting none of the shares back margin accounts. That oversight is real and better-than-average disclosure. But the structural exposure remains: a pledged block of this size is a latent forced-seller in any severe drawdown, and the policy's own existence — applied to everyone but the one person it matters for — tells you the board knows it.

The leadership handoff: from a 20-year CEO to two untested co-CEOs

On September 22, 2025, Safra Catz stepped down after 11 years as CEO (and two decades running Oracle's operations and finance) to become Executive Vice Chair. The board did not name one successor — it named two co-CEOs, splitting the company along its two growth engines.

Source: DEF 14A September 26, 2025 (director bios; CEO equity grants).

Clay Magouyrk (39) is the OCI builder — he joined Oracle in 2014 from Amazon Web Services and scaled Oracle's cloud to 100+ public regions. He is the more strategically central of the two, given that OCI and the AI-infrastructure buildout are the entire investment thesis. Mike Sicilia (54) came in through the 2009 Primavera acquisition and runs Oracle's vertical/industry applications and applied AI. Both are deep insiders and credible operators — but neither has run a public company, and a co-CEO structure layered beneath an 81-year-old Executive Chair who still controls the company and "continues to lead product engineering" is a genuinely untested arrangement. In practice, Ellison remains the decider; the co-CEOs are, for now, powerful lieutenants.

The board paid up for them immediately. In September 2025 the Compensation Committee granted Magouyrk options to buy $250 million of stock and Sicilia $100 million — each 80% time-vested over four years and 20% performance-vested over three years against revenue targets.

Source: DEF 14A September 26, 2025, Compensation Committee disclosure.

These are large, retention-driven packages for executives with thin track records at the top — the kind of grant that aligns them to the share price but offers little downside discipline if the AI-capex bet disappoints. Note the 80/20 split favors time-based vesting: most of the value vests simply for staying, not for performing. That is generous to the executives and a point proxy advisors will scrutinize at the 2026 meeting.

Pay: lean cash, equity-heavy, and a board that actually used the brakes

The most encouraging governance signal this year sits in the compensation table. Catz and Ellison drew almost no incentive pay, and the Compensation Committee voluntarily zeroed out $5.2 million bonuses for several executives "to preserve capital in furtherance of the company's strategic priorities" — a discretionary cut against actual achievement. A committee willing to claw back earned bonuses is not a rubber stamp.

Source: DEF 14A September 26, 2025, Summary Compensation Table. Ellison salary is $1 (voluntary); his $5.6M is almost entirely security and aircraft-related "other."

Two things stand out. First, Ellison takes a $1 salary and zero equity — he doesn't need a paycheck, and his $5.6M "other" is overwhelmingly personal-security and aviation cost, not incentive pay. His incentive is his 40.6% stake. Second, for the salaried NEOs (Henley, Levey, Smith, Screven), pay is heavily equity-weighted, tying realized value to the share price. This is a defensible, alignment-oriented structure — Oracle's historical sin was the size of Ellison's and Catz's old mega-grants, and that eight-year performance-option program (FY2018–FY2025) has now fully run off.

Shareholders have noticed the improvement: the FY2024 say-on-pay vote drew ~78% support, which the company itself frames as "ongoing year-over-year improvements" after years of weaker backing. Respectable, not resounding — and the new $350M co-CEO grants put the 2026 vote back in play.

Source: DEF 14A September 26, 2025 (FY2024 say-on-pay result; FY2025 pay; Sept-2025 grants). Co-CEO figure is combined time-vested portion.

The board: independent on paper, gray and entrenched in practice

Eight of the 13 director nominees are independent (62%), all standing committees — Finance & Audit, Compensation, Governance, and Independence — are composed solely of independent directors, and Bruce Chizen (ex-Adobe CEO) serves as a credible Lead Independent Director. By the letter of NYSE standards, this board checks the boxes.

The substance is more strained on two fronts: age and tenure. The independent directors average roughly 74 years old and 16 years of tenure; three are 80 or older (Conrades 86, Seligman 87, Boskin 80), and several have sat on this board for two to three decades. Long tenure brings institutional memory — but at these lengths and ages, "independent" and "genuinely able to challenge a founder you've served alongside for 25 years" are not the same thing. Five of the 13 seats, meanwhile, are held by insiders (Ellison, Catz, Henley, and the two co-CEOs), an unusually heavy executive presence.

Source: DEF 14A September 26, 2025, director biographies (ages as of Sept 19, 2025). Panetta and Parrett did not stand for re-election; the two longest-tenured independents are reported to be retiring as the board refreshes into FY2026.

The scorecard below maps each director against the three dimensions where this board is weakest. Green is a strength, red a concern. The independence column is mostly green; the freshness and age columns are where the red clusters.

Source: derived from DEF 14A September 26, 2025 director data. Scoring is our assessment: Independence (independent=strength), Tenure freshness (under 10y=strength, 20y+=concern), Age (under 70=strength, 80+=concern).

To the board's credit, refreshment is underway: Leon Panetta and William Parrett rolled off this cycle, and the two oldest independents are reported to be retiring as Oracle resizes the board into fiscal 2026. Newer additions — Fairhead (audit expert), Moorman (Chevron director), Ablo — are younger and bring relevant outside experience. The direction is right; the starting point is old.

Alignment: real skin, real entanglements

Source: DEF 14A September 26, 2025 (ownership and related-party disclosures).

On raw alignment, few large caps match this: the founder's net worth rises and falls with the share price by hundreds of billions, and management cash pay is modest. But two qualifiers matter for minority shareholders, who experience Ellison's upside without controlling his downside.

Related-party entanglement is wide but financially small. Oracle does business with a long list of Ellison-affiliated entities — SailGP/F50 League (~$7.5M paid for sponsorship), Glass Aviation (~$2.5M, now owned by son David Ellison), Wing and a Prayer aircraft (~$0.5M), plus sales to the Ellison Institute, Sensei AG, Lanai Resorts, Skydance, and others. In total these ran ~$10.8M of sales and ~$10.5M of purchases — about 0.02%–0.04% of company totals. The Independence Committee approves each, commissions third-party fairness reviews of the aircraft rates, and Ellison signed a "price protection" agreement to reimburse any overpayment. The dollars are immaterial; the pattern — Oracle's purchasing teams keeping an internal list of Ellison entities, updated quarterly with "Mr. Ellison's advisors" — is the texture of a company organized around one man's wider empire. Add a $1.6B investment in Ampere (chaired by a former Oracle director, now being sold to SoftBank) and two relatives on payroll (Ellison's half-brother, Screven's son), and the conflict surface is broad even if no single item is abusive.

Insider behavior cannot be tracked at transaction level here — the proxy notes the Form 4 feed was unavailable — but the beneficial-ownership snapshot is the more important signal: Ellison's holding remains intact at 40.6%, and reporting indicates his periodic sales are diversification against an unchanged 1.1B+ share base, not a directional exit. No insider has signaled loss of conviction.

The verdict

Source: our assessment, from DEF 14A September 26, 2025 and latest reporting.

Oracle earns a B-. The economics are excellent — a $1-salary founder with 40.6% skin in the game, an equity-heavy pay structure, a Compensation Committee willing to zero out earned bonuses, improving say-on-pay, fully-independent committees, and an honest related-party regime. What holds it back is concentrated, structural risk that no governance process can fully neutralize: Ellison's effective control means minority holders ride along rather than steer; 346M pledged shares create a forced-seller overhang the rest of the company is forbidden from creating; the board, though independent, is old and long-tenured; and the company has just bet its leadership on two untested co-CEOs handed $350M in mostly time-vesting options while the founder still runs product.

What would move the grade up: material reduction or full disclosure-with-collateralization detail of Ellison's pledged block, a clearer single-leader resolution to the co-CEO structure with more performance-weighted vesting, and continued board refreshment toward younger, shorter-tenured independents. What would move it down: any forced sale or margin event on the pledged shares, an expansion of related-party dealing, or a 2026 say-on-pay setback over the new mega-grants.