Fifteen days ago, SpaceX completed the largest initial public offering in financial history, raising $75 billion at a $1.77 trillion valuation. On its first day of trading, the stock blew past $160. Within 48 hours it had nearly reached $225. Then, just as quickly, it fell — shedding more than 30% in a week, settling near $153 where it trades today.

The consensus on Wall Street is split in a way that is genuinely unusual. Morningstar’s discounted cash flow model places fair value at $62. The sell-side consensus average sits at $188. The most bullish estimate on record is $310. That is a $248 spread — wider than many mid-cap companies’ entire stock price — and it exists because analysts are not disagreeing about near-term revenue. They are disagreeing about the physical and economic feasibility of a rocket program and an AI compute venture that has never generated a dollar of profit.

So is $153 justified?

You Are Not Buying One Company

SpaceX’s S-1 filing, released in May 2026, disclosed something investors in the private market never had access to: segment-level financials that make clear this is not a single business with a single risk profile. It is three.

Connectivity (Starlink) generated $11.4 billion in revenue in FY2025, up 49.8% year-over-year. Operating income was $4.4 billion. Adjusted EBITDA was $7.2 billion — a 63% margin that rivals the best hyperscalers in the world. Subscriber growth has been staggering: from 2.3 million in 2023 to 4.4 million in 2024 to 8.9 million in 2025, and above 10 million as of March 2026. This business, if it were a standalone public company, would command a premium valuation in its own right.

Space (Launch) contributed $4.1 billion in revenue in 2025, but posted a $657 million operating loss — the direct result of approximately $3 billion poured into Starship research and development. Strip out Starship spend and the launch business is operationally profitable. Falcon 9 costs $67 million per launch, has achieved a 34-reuse record on a single booster, and commands roughly 80% of global mass-to-orbit market share. The economics of Falcon are extraordinary. The question is how much you pay today for what Starship might eventually enable.

AI (xAI and X) is the most consequential and the most contested. Following SpaceX’s merger with Musk’s xAI in February 2026, the AI segment reported $3.2 billion in revenue against $6.4 billion in operating losses in 2025. Capital expenditure for AI infrastructure — primarily the COLOSSUS and COLOSSUS II data centers housing 220,000 Nvidia GPUs in Memphis — reached $12.7 billion in FY2025 alone. In the first quarter of 2026, AI capex was $7.7 billion in a single quarter. Annualized, that is a $30 billion burn rate.

The consolidated picture: $18.7 billion in revenue, $6.6 billion in adjusted EBITDA, and a GAAP net loss of $4.9 billion. The accumulated deficit since SpaceX’s founding in 2002 is $41.3 billion. Total debt stands at $29.1 billion.

The Bull Case: Optionality at Scale

The case for $153 — or higher — rests on compounding several scenarios that are individually plausible but together form a heroic growth requirement.

Starlink is a genuine monopoly in formation. No competitor has the launch cost structure, the satellite density, or the subscriber base to challenge it meaningfully in the near term. Amazon’s Project Kuiper is scaling, but from a standing start. Starlink at 63% EBITDA margins with 10 million subscribers is already a category-defining infrastructure asset. At 50 million subscribers — a plausible extrapolation over five years — the revenue line alone starts to justify a significant portion of the current valuation.

Starship changes the economics of everything. A fully reusable Starship capable of lifting 100 metric tons to low Earth orbit, with turnaround times approaching commercial aviation, would represent the most significant structural shift in the space industry’s cost structure since the invention of the rocket. SpaceX expects to begin payload deliveries in the second half of 2026. If the program delivers, it unlocks orbital data centers, satellite-to-mobile connectivity, and a dramatic reduction in the cost of Starlink’s own network replenishment.

AI infrastructure has a real revenue story emerging. Google has reportedly agreed to approximately $1 billion per month in compute contracts for COLOSSUS capacity. If that figure is accurate, it implies a $12 billion annual revenue run rate from a single customer — validating that SpaceX’s AI infrastructure competes with hyperscale clouds at market rates. The Cursor option agreement (giving SpaceX the right to acquire the AI coding platform at a $60 billion implied valuation) adds another layer of strategic positioning in the AI application stack.

Index inclusion mechanics will drive forced buying. Nasdaq 100 eligibility is expected around July 2–3. Estimates place the total passive inflow at approximately $400 billion in mandated purchases as funds tracking the index are required to absorb SPCX at its index weight. That is mechanical demand, not judgment-based demand, and it will compress the float further in a stock where insiders already control the overwhelming majority of shares.

The Bear Case: The Math Is Brutal

Morningstar analyst Nicolas Owens did not arrive at $63 carelessly. He ran a probability-weighted DCF across three scenarios — a “Moonshot” case where Starship succeeds and orbital data centers scale, a “Minimum Viable Product” case, and a “No Go” case where AI infrastructure fails to commercialize. He assigned the Moonshot scenario only a 7% probability. The resulting blended fair value is $63 per share, implying a company worth approximately $800 billion — still enormous, but roughly $1.2 trillion below where SPCX trades today.

The core arithmetic is unforgiving. Goldman Sachs analysts reportedly calculated that sustaining a $1.8 trillion valuation through 2030 requires annual revenue exceeding $100 billion, implying a compound annual growth rate above 40% for four consecutive years. SpaceX grew 33% in 2025 and 15% year-over-year in Q1 2026. Growth is decelerating, not accelerating.

Consider what the price-to-sales multiple of 108 times trailing revenue actually means. Amazon, which generates $600 billion in annual revenue with multiple profitable segments including AWS, trades at roughly 3 times sales. Nvidia, the defining AI infrastructure company of this cycle, trades at approximately 20 times. SpaceX trades at five times Nvidia’s revenue multiple — for a business that is GAAP-unprofitable and has an accumulated deficit approaching the GDP of Luxembourg.

The AI segment is not a speculative footnote. It is the primary driver of the company’s losses. At $7.7 billion of AI capex in a single quarter, the company is absorbing costs faster than any comparable business has ever done outside of the original hyperscaler buildout. If that capex does not yield sustainable revenue at scale, Starlink’s $4.4 billion in operating income gets consumed entirely before a single shareholder sees a dollar.

Governance deserves scrutiny proportional to its significance. Elon Musk controls more than 82% of SpaceX’s voting power through a dual-class share structure. The company’s S-1 disclosed a $530 million accrual for litigation losses stemming from Grok-related AI safety cases across multiple jurisdictions. The Cursor option, the Tesla merger chatter, and numerous related-party transactions between SpaceX, Tesla, and xAI create a web of potential conflicts that minority shareholders have essentially no power to constrain.

And then there is the lockup. From late July through August 2026, up to 20% of insider shares become eligible for sale. Early employees and venture investors who have held SpaceX shares for a decade or more will be staring at the largest liquidity event of their careers. The mechanics of that moment — combined with August 6 earnings and the end of underwriting bank quiet periods — make the next six to eight weeks a critical and unpredictable stress test for the stock.

What Has to Be True at $153?

The most useful way to evaluate a stock with no earnings is to work backwards from the price and ask what the market is requiring you to believe.

At $153, SpaceX’s market capitalization is approximately $2.02 trillion. Apply a reasonable terminal multiple for a mature, profitable SpaceX — say, 30 times earnings — and you need the company to generate approximately $67 billion in annual profit at some point in the future to justify the current price on a discounted basis. With a conservative 10% discount rate and a 10-year horizon, you need the company to reach that profit level by the mid-2030s.

That is not impossible. A Starlink at 50 million subscribers, a Starship that has dramatically reduced launch costs and enabled orbital data centers, and an AI business that has matured into a cash-flow-positive compute provider could collectively generate those numbers. But each of those conditions requires engineering breakthroughs that have not yet occurred, regulatory environments that may not cooperate, and competitive dynamics that are not static.

Morningstar’s $63 requires you to believe most of that does not materialize. Today’s $153 requires you to believe most of it does. Neither position is irrational. The honest judgment is that the market is pricing significant optionality, which is not the same as pricing certainty.

Space X is Priced for a World That Has Not Happened Yet

SpaceX at $153 is not absurd. It is a bet — a large, structured, multi-decade bet on a specific vision of the future in which low-cost orbital access reshapes telecommunications, computing, and eventually transportation. The Starlink segment alone, at its current growth trajectory, will be worth hundreds of billions of dollars. The optionality embedded in Starship and orbital AI is real, even if it is unpriced in any traditional sense.

But the price at which you buy that bet matters enormously. At 108 times trailing revenue and no path to near-term profitability, $153 leaves almost no margin for error. A Starship delay of two years, meaningful Starlink ARPU compression from lower-income market penetration, or an AI infrastructure market that proves more competitive than expected — any of these alone could see the stock revisit its IPO price. All three together could see it trade well below $100.

The Morningstar framework is not pessimistic. It is probabilistic. And when you run the math, the most likely scenario — across a realistic distribution of outcomes — is a company worth considerably less than its current price.

What $153 is buying is the dream. Whether the dream is worth that price depends on your time horizon, your confidence in the engineering, and your appetite for a governance structure where one person holds all the decisions. For a patient, risk-tolerant investor with a 10-year view and genuine conviction in Musk’s execution, the bet may be defensible. For a fundamental investor requiring a margin of safety, the numbers are not yet there.

The first real test arrives August 6. Watch the AI capex trajectory above everything else. That is the variable that makes or breaks this story.