Inside SpaceX’s Three-Day Analyst Tour Before the Biggest IPO in History

SpaceX opened a three-day closed-door analyst tour Tuesday at Starbase Texas, the first demand test for what would be the largest IPO in history.

SpaceX began a three-day, closed-door analyst tour on Tuesday at its Starbase launch facility in Boca Chica, Texas. The meetings are the first structured pitch to sell-side research analysts ahead of what the company’s underwriters hope will be a $75 billion public offering at a $1.75 trillion valuation, a late-June Nasdaq listing, and the largest IPO ever brought to market. SpaceX submitted a confidential draft registration statement to the SEC on April 1, independently confirmed by Reuters, Bloomberg, CNBC, and The Wall Street Journal. The public S-1 is expected to hit EDGAR in the next four weeks, the statutory minimum of 15 days before a roadshow can begin.

The format of this week’s sessions is revealing. Tuesday’s meeting was for one group of research analysts. A separate group of institutional-investor analysts, representing mutual funds and pension plans, is being briefed at Starbase on Wednesday. On Thursday, the combined group travels to Memphis, Tennessee, to tour the Colossus data center and review the xAI integration work, which SpaceX internally refers to as “Macrohard.” According to CNBC’s reporting, attendees are required to surrender electronic devices before entering each session.

The Record in Sight

The size of the deal is the first thing to calibrate against. Saudi Aramco’s 2019 offering, which raised $29.4 billion, is the current IPO record. A $75 billion SpaceX raise would clear that mark by roughly 2.5 times. The $1.75 trillion valuation, if the book clears at that level, would slot SpaceX between Meta and Berkshire Hathaway in the US market cap rankings on day one.

The valuation is also a sharp step up from where the company was twelve months ago. A February 2026 all-stock transaction merged SpaceX, then valued at about $1 trillion, with xAI, valued at roughly $250 billion, creating a combined entity at $1.25 trillion. The $1.75 trillion IPO number reflects the bankers’ view that the combined company deserves a premium to the private mark, not a discount. That is an unusual starting position for a large listing, where pricing typically lands at or below the last private round.

What the Analysts Are Being Shown

The tour’s choreography signals what SpaceX wants the Street to index on. Starbase is where the revenue story lives. Starlink generated approximately $11.4 billion in 2025 revenue, up 50% year-over-year, with EBITDA (earnings before interest, taxes, depreciation, and amortization, a common measure of cash profit from operations) of roughly $7.2 billion, implying a 63% EBITDA margin. That is the profit engine. Direct-to-Cell, the satellite-to-phone service that now covers more than 650 satellites in orbit, has passed 10 million monthly active users. Starlink’s traditional broadband subscriber base reached more than 9 million by year-end.

Memphis is where the cost story lives. Colossus is the compute footprint that absorbs xAI’s roughly $1 billion per month cash burn and houses the Macrohard program, which integrates xAI’s Grok models into Starlink network management and customer operations. The analyst message is that the merger is not two businesses stapled together. It is a single flywheel where launch economics subsidize AI training and AI models lift the unit economics of the satellite fleet.

That is also where the sharpest analyst skepticism lives. UBS research flagged the shift in a pre-briefing note, arguing that post-merger investors no longer see a pure-play space business with strong cash flow, but a combined entity that has to rebalance that cash flow against ongoing AI investments. The question is whether Starlink’s $7.2 billion EBITDA is enough ballast to fund xAI’s $12 billion annualized burn without forcing dilution later.

The Dual-Class Question

The governance structure is doing work at the valuation. SpaceX will list with a dual-class share structure (two share classes carrying different voting rights per share). Elon Musk is expected to hold approximately 42% of the combined company’s equity and control roughly 79% of the votes through super-voting shares. Ordinary shares sold to the public carry one vote each. The super-voting class, held by Musk and a small group of insiders, carries a multiple of votes per share, creating the gap between economic ownership and voting control.

That is a structure public shareholders cannot out-vote at any purchase size short of buying back the super-voting class. It removes the possibility of an activist campaign, a proxy fight, or a hostile bid without insider consent. For one class of investor, that premium for stability is a feature. For another, it is the reason to price the book below $1.75 trillion. This is the single decision each incoming shareholder has to make before they read anything else in the filing.

Wedbush analyst Dan Ives has framed the deal as additive rather than competitive to the existing Musk-company complex, writing in a recent client note that the listing “expands the ecosystem” for investors who want exposure beyond Tesla. Ives had earlier projected a $1.5 trillion entry valuation. The market has repriced upward since.

The Number That Matters

The single number that most tightly links this deal to reality is the ratio of Starlink’s EBITDA to xAI’s cash burn: $7.2 billion in annual operating cash flow from the satellite business against roughly $12 billion in annualized xAI operating losses. The gap is not catastrophic. It is also not comfortable. A $75 billion raise fills it for five to six years at the current burn rate, assuming xAI does not accelerate data-center capex. If xAI’s monthly burn climbs, as it has done for every frontier lab in every public data point since 2023, the raise compresses faster. That number, not the $1.75 trillion headline, sets the real ceiling on how much equity dilution later shareholders should expect. It is also the number most likely to be sanded down in analyst models coming out of this week’s meetings.

The Banker Lineup

The active bookrunner list is the longest and heaviest this decade. Morgan Stanley, Bank of America, Citigroup, JPMorgan Chase, and Goldman Sachs lead the deal, with 16 additional banks in smaller roles. That is a 21-firm syndicate for a single IPO, a configuration typically reserved for the largest sovereign-scale deals. The breadth is a distribution strategy. A $75 billion book cannot be placed through a single desk or a single client segment. The five-bank lead group lets each book-runner allocate to different investor pools: long-only mutual funds, sovereign wealth, pensions, family offices, crossover funds, and retail through the online brokerages.

The retail allocation is structural, not incidental. The roadshow is targeted for the week of June 8, with a dedicated retail investor event on June 11 for approximately 1,500 attendees. Cathie Wood’s ARK Invest has publicly backed the $1.75 trillion valuation, calling it “grounded in plausible” scenarios for Starlink’s subscriber ramp, Direct-to-Cell monetization, and xAI’s enterprise AI revenue. Wood has flagged a $2.5 trillion scenario by 2030 if Starlink’s growth curve holds. ARK and a handful of retail-heavy crossover funds are expected to anchor the retail tranche.

What to Watch

The public S-1: The confidential draft has been under SEC review since April 1. The public version must sit on EDGAR for at least 15 days before the roadshow begins. Watch the SEC EDGAR docket in the coming weeks for the first public filing. The price range comes in an amended S-1/A filing, typically the earliest durable signal that the deal has cleared review.

The Musk-Tesla angle: Any S-1 language describing the relationship between SpaceX, Tesla, and xAI will be scrutinized for related-party transactions, compute-sharing agreements, and the likelihood of a future Tesla-SpaceX structural combination, which some Street analysts have begun forecasting for 2027.

Book-building signals: Watch for banker-desk chatter on whether the book is clearing at $1.75 trillion or compressing toward the $1.5 trillion mark that pre-merger analyst notes implied. An upsized raise above $75 billion would signal demand is overflowing the planned size. A downsize or price-range cut would signal the opposite.

Macrohard disclosures: The confidential filing is light on the xAI integration. The public S-1 should quantify the compute, capex, and expected revenue contribution from Macrohard specifically. That is the swing factor on the 2027 model.