SpaceX Earnings: Is a 50% Post-IPO Selloff Overdone, or Are Investors Right to Be Nervous?

SpaceX went public on June 12 in the largest IPO in U.S. history, priced its shares at $135, and briefly soared to an intraday high of $225.64 as investor enthusiasm peaked. Two months later, the stock sits around $114 - below its IPO price and roughly 50% off that all-time high. Tuesday afternoon, it releases its first earnings report as a public company.

The market has arguably baked in a doomsday narrative. But the question is whether the selloff is justified by the numbers - or whether the post-IPO whiplash has created a mispriced setup worth paying attention to.

Starlink: The Cash Engine Investors Are Ignoring

Underneath the AI hype and the rocket launch headlines, SpaceX is a three-segment business, and one of those segments is quietly profitable at scale. Starlink - the satellite broadband network - generated $11.39 billion in revenue in 2025, up roughly 50% year over year, and delivered $4.42 billion in operating profit. Operating income for the connectivity segment rose 120% between 2024 and 2025.

As of the end of March, Starlink had 10.3 million subscribers across 164 countries, roughly double the level a year earlier. Analysts expect second-quarter revenue from connectivity to hit $3.82 billion with $1.42 billion in operating profit, up from $1.19 billion in the first quarter. The growth story here is real.

The complication is ARPU (average revenue per user), which fell 18% to approximately $81 per month between 2023 and 2025 as SpaceX expanded into lower-income markets and introduced cheaper pricing tiers. Volume growth is outpacing the erosion for now, but the margin trajectory of the connectivity segment is a variable to monitor. Still, Starlink remains the financial engine that bankrolls everything else.

The AI Spending Problem

This is where investors' concerns are legitimate. SpaceX's AI segment - built around its February acquisition of xAI, the developer of the Grok chatbot - posted a $6.35 billion operating loss in 2025 and generated only $818 million in revenue in the first quarter of 2026. Capital expenditures in the AI division alone were $7.7 billion in Q1, accounting for roughly three-quarters of the company's total capital spending.

Analysts expect the AI segment's capital spending to surge more than sixfold year over year in Q2, to around $10.2 billion. Full-year capex for the entire company is projected to climb from $48.7 billion in 2026 to $118.4 billion in 2028, with overall debt expected to grow over fivefold to more than $218 billion. One asset manager told Reuters bluntly that "Starlink is executing beautifully, but it cannot single-handedly fund a $30 billion annualized AI capex program."

But there's a counterweight: committed compute contracts. SpaceX has signed AI infrastructure deals with Anthropic, Alphabet's Google, and Reflection AI. Those agreements, once fully ramped, could exceed $25 billion in annualized revenue. Alphabet alone is paying approximately $920 million per month from October 2026 through June 2029 for access to roughly 110,000 Nvidia GPUs. Analysts expect AI revenue to nearly triple in Q2 to $2.33 billion, up from the first quarter.

The AI business is not going to be a free cash flow story - not now, not soon, and maybe not ever, in the near term. But it's not an empty promise either. Those contracts represent real revenue with named hyperscaler customers, not vaporware.

The Valuation Disconnect

Here's the arithmetic. At current levels, SpaceX's market cap sits around $1.5 to $1.6 trillion, down from nearly $3 trillion at its post-IPO peak. For a company that generated $18.7 billion in revenue last year and is expected to report roughly $6.8 to $6.9 billion in Q2, the stock still trades at a very high multiple. That's an extreme multiple by any measure.

Even the bull case has its limits. Morgan Stanley - which is bullish on the stock - doesn't expect SpaceX to turn free cash flow positive until 2035. The company is going to have to raise enormous amounts of capital, through equity or debt offerings, to fund what is an extraordinarily capital-intensive business.

The valuation question is whether the sum of Starlink's cash generation, the contracted AI compute revenue, and the strategic value of Starship (which underpins NASA contracts, lunar missions by 2028, and future satellite deployment) justifies the multiple. Morningstar put a fair value of $62 on the stock. Raymond James has an $800 target. The FactSet consensus sits at $228.

That range is so wide it reveals what's really happening: there is no agreed-upon framework for valuing SpaceX. It's part satellite operator, part AI infrastructure company, part launch monopoly, part Musk vision play. The market doesn't know what multiple to apply, so it's oscillating wildly.

The Lockup Overhang

Then there's the mechanical problem. On Thursday, August 6, the first insider lockup expires, freeing approximately 911.5 million SpaceX shares - roughly 7% of the company, worth about $100 billion at current prices. That more than triples the current tradable float.

More supply without commensurate demand is a textbook formula for additional selling pressure. But lockup expirations can also work the other way: some of those pre-IPO holders become buyers, providing support. The IPO expert at the University of Florida told CBS News that short sellers, who have already made more than $8 billion betting against the stock, could shift into buying mode, soaking up shares and reducing the drop.

Either way, the stock is going to be volatile around this event. The earnings report on Tuesday could either blunt the lockup concern or amplify it.

What I'm Watching

The Q2 revenue estimate - roughly $6.8 to $6.9 billion, up from $4.7 billion in Q1 - is not going to be the make-or-break number. As one Oppenheimer analyst put it, "results are secondary to reading management's body language and outlook". Musk's tone on the call, his comments on AI self-funding timelines, and any update on the rumored Tesla merger will matter more than the quarter itself.

What would change my view? If Starlink growth materially slows - subscriber deceleration beyond the already expected moderation, or a steeper ARPU drop that starts eating into operating profit - the cash engine story cracks. If the AI capex guidance overshoots the $48.7 billion estimate significantly, and the company signals even later paths to monetization, the valuation becomes indefensible regardless of Musk's narrative.

On the flip side, if Starlink continues to generate strong operating cash flow and the AI compute contracts stay on track for their $25 billion-plus annualized run rate, the current selloff looks increasingly like a fear-driven overreaction to a capital-intensive but defensible long-term strategy.

The Verdict

The market has punished SpaceX for two things: the sheer scale of its AI spending, and the upcoming lockup expiration. The spending concern is real - this is a company burning billions on infrastructure that won't be cash-flow-positive for a decade. But Starlink's profitability and the contracted AI revenue provide a floor that the doomsday narrative doesn't acknowledge.

The risk/reward at current levels is arguably more attractive than it was at $200-plus, even if the valuation is still stretched. I'm not suggesting investors chase this blindly into Thursday's lockup event. The better setup is patience: wait to see how the stock reacts to the earnings call and the share unlock, then assess whether the technical action confirms selling exhaustion or just more gravity.

If you believe in the long-term thesis - Starlink's global expansion, contracted AI compute revenue, and Starship's strategic optionality - the current price offers a more reasonable entry than the post-IPO euphoria did. But don't mistake patience for conviction. I would reassess if Starlink's operating margins start deteriorating or if Musk's AI timeline slides significantly beyond what the market expects. Until then, the disconnect between the numbers and the price is worth watching closely.