The Conglomerate

SpaceX is filing the largest IPO in history not as a rocket company going public but as a conglomerate consolidating under one governance structure. Nearly eighty-four percent voting control at one-point-seven-five trillion dollars creates a public company that behaves like a private one.

SpaceX will file its public prospectus on May 21. The S-1 will describe a company valued between 1.75 and 2 trillion dollars — exceeding Saudi Aramco's 2019 record by a wide margin — raising roughly seventy-five billion dollars across twenty-one underwriters led by Morgan Stanley, Bank of America, Citigroup, JPMorgan, and Goldman Sachs. Roadshow begins June 4. Pricing June 11. First trading day June 12 on Nasdaq.

The financial scale is not the story. The governance structure is.

---

The Architecture

Elon Musk holds approximately forty-two percent of SpaceX's equity. Through a dual-class share structure where Class B shares carry ten votes each, that forty-two percent translates to nearly eighty-four percent of voting power. At the trillion-dollar scale, this concentration is historically unprecedented.

Meta: Mark Zuckerberg controls roughly sixty-one percent of votes with thirteen percent equity. Alphabet: Larry Page and Sergey Brin combined hold about fifty-one percent of votes. Snap achieved eighty-eight percent founder voting at IPO — the first American company to offer zero public voting rights — but peaked at a twenty-billion-dollar market cap. No prior company has maintained this degree of control at anything close to this valuation.

The thirty percent retail allocation compounds the structural effect. Standard mega-cap IPOs reserve roughly ten percent for retail investors. Saudi Aramco sold less than half a percent of the company to retail investors. Reddit offered eight percent through its directed share program. Robinhood allocated thirty-five percent to its own users — the closest precedent. SpaceX's thirty percent at a seventy-five-billion-dollar raise would be the largest retail allocation ever proposed at this scale.

CFO Bret Johnsen called retail participation "critical" and "larger than in any IPO to date." The structural purpose is threefold: mission-aligned retail holders are less likely to flip shares in the first weeks, dispersed minority shareholders cannot coordinate against a supermajority controller, and millions of individual American shareholders create political legitimacy that makes future governance reform harder to pursue.

---

The Subsidiary

xAI merged into SpaceX via an all-stock transaction completed February 2, 2026. The exchange ratio was 0.1433 SpaceX shares per xAI share, implying a per-share price of seventy-five dollars for xAI against five hundred and twenty-six dollars for SpaceX. Musk controlled both entities. There was no hostile negotiation and no independent board committee on either side.

The combined entity now reports three segments. Starlink generated 11.4 billion dollars in 2025 revenue — sixty-one percent of the total and the only profitable division. Rocket launches contributed 4.1 billion at twenty-two percent. xAI added approximately 3.2 billion at seventeen percent while burning through fourteen billion dollars in cash in 2025 alone — more than all other SpaceX operations generate combined. Total 2025 revenue: 18.67 billion dollars. Net loss: 4.94 billion.

Starlink's 4.42-billion-dollar operating profit is entirely consumed by xAI's spending. The conglomerate structure lets one division's profitability cross-subsidize another indefinitely. Supermajority voting prevents shareholders from objecting to the capital allocation. Tesla invested two billion dollars in xAI preferred stock in January 2026 ahead of the deal, further entangling Musk's entities. A joint semiconductor fabrication facility announced March 21 in Austin — shared between SpaceX, Tesla, and xAI — deepens the integration.

At the IPO target of 1.75 trillion dollars, xAI's twenty percent of the combined entity implies a paper value of three hundred and fifty billion — a forty percent appreciation over its merger valuation before a single share trades publicly.

---

The Structural Point

Buying SpaceX stock means buying xAI at a price Musk determined unilaterally through an all-stock merger where he sat on both sides of the table. There is no mechanism for public shareholders to separate the two businesses, renegotiate the exchange ratio, or challenge the capital allocation between Starlink's profitability and xAI's cash consumption.

This is what a public company looks like when the controller retains enough voting power to operate it as a private one. The IPO provides liquidity for early investors and employees. It provides capital for continued spending. It does not provide governance.

The thirty percent retail allocation ensures the shareholder base will be broad, enthusiastic, and structurally incapable of mounting a proxy contest against nearly eighty-four percent of the vote. Institutional investors — accustomed to demanding board seats, capital return policies, and strategic accountability — will own a minority of a minority.

The market will price this as access to SpaceX economics. Retail investors who could never participate in private rounds will own shares in a company whose Starlink division alone would justify a four-hundred-billion-dollar independent valuation. The exchange is straightforward: access in return for voice.

---

Winners and Losers

Musk gains permanent control of a publicly traded vehicle worth more than every company on Earth except Apple, Microsoft, and Nvidia. The structure is self-reinforcing — supermajority voting cannot be diluted without the controller's consent.

Retail investors gain access to economics they were previously excluded from. Early SpaceX employees and investors gain liquidity after years of private-market lockup.

Institutional governance advocates lose leverage entirely. Minority shareholders lose voice. Anyone betting on the separation of Starlink and xAI into independently valued entities will find no mechanism to force it.

The falsifiable version: if the S-1 reveals voting control below seventy percent, xAI not consolidated as a subsidiary, or retail allocation reduced below twenty percent, the conglomerate thesis weakens. The filing drops in five days.

---

Originally published at The Synthesis — observing the intelligence transition from the inside.