SpaceX Stock Has Fallen Back to Earth. The Revenue Bridge Hasn't Broken.

SpaceX went public at $135 on June 12 and briefly hit $226 eight days later. Today, the stock is at $133 — essentially back where it started, with over $1 trillion of peak market value evaporated. The headlines say a bubble burst. The operating numbers say something messier: the business is accelerating even as the cash-flow bridge to a $2 trillion enterprise value remains under construction.

The old story hasn't caught up to the new numbers

The market has been oscillating between two extremes. For five days in mid-June, SpaceX was treated as the next Apple — a $2.9 trillion company running on IPO mania and Musk gravity. Since then, investors have swung the other way, fixated on a -$33.4 billion free-cash-flow deficit over the trailing twelve months and a staggering $43.3 billion in capital expenditures. That second frame is more honest. But it's also incomplete.

Revenue jumped 92% year-over-year to $7.8 billion in Q2. The net loss narrowed to $541 million, down from $1 billion a year earlier. Starlink — the connectivity segment that actually prints operating profit — brought in $4.29 billion on 12 million subscribers. AI infrastructure revenue more than tripled to $2.56 billion, with $14.1 billion in contracted cloud services sitting in the pipeline. Management's stated target is a $100 billion annual revenue run rate by year-end.

The operating trajectory is moving in the right direction. The question isn't whether the business is growing. It's whether the spending required to get there leaves enough cash for the stock to earn its $1.75 trillion market cap.

The capex hole is the real question

Here's where the thesis gets tested. SpaceX spent $18.4 billion on capital expenditures in Q2 alone, of which $15.8 billion went to AI infrastructure. Annualized, that's a $73 billion/year buildout pace — though the TTM run rate of $43.3 billion suggests the Q2 spike may be a one-quarter surge as Project Colossus data centers come online. Operating cash flow over the trailing twelve months came in at $9.9 billion. Free cash flow, the number that matters, was -$33.4 billion.

That deficit is the single strongest argument against the stock right now. A company burning $33 billion in free cash flow while trading at 76 times trailing sales and 320 times trailing EV/EBITDA is asking investors to accept a very long bridge between current spending and future returns. The balance sheet can absorb it for now — $93.5 billion in cash against $65.6 billion in debt leaves a net cash cushion of $60.7 billion — but the clock starts ticking once that buffer thins.

The proof point that matters most: AI compute revenue needs to grow fast enough to offset the infrastructure spend before the cash pile runs low. SpaceX's contracted pipeline of $14.1 billion in cloud services gives some forward visibility, but those deals with partners like Anthropic and Google are cancellable within 90 days, which is a far cry from the multi-year durability of traditional hyperscaler contracts. The company is also spending $60 billion on an all-stock Cursor acquisition, a move that bolsters its AI portfolio but adds to the capital intensity picture.

What the market is actually pricing wrong

Dan Ives of Wedbush initiated coverage in July with a $190 price target and a sum-of-the-parts model projecting $2.5 trillion in enterprise value by fiscal 2028. That would imply roughly 44% upside from today's $133 close. Ives' framework treats Starlink as the profit engine, the launch business as a cost-efficient platform, and AI compute as the segment that deserves the richest multiple. The logic isn't wrong — it's the timing that's contested.

Even Ives acknowledged in late July that "valuation support is not there" as the stock slipped below its IPO price. He's since left Wedbush to launch an independent venture, which means his $190 target no longer carries the weight of a major-bank research desk. But the underlying framework — Starlink matures into a cash generator while AI infrastructure scales — is the only path that connects today's price to a higher one.

The crowd is currently anchored to two things that are distracting from the operating signal. First is the lockup expiration schedule. Starting August 6, approximately 911 million shares — worth roughly $98 billion at current prices became eligible for sale. The release is staggered through 16 tranches through December, with Elon Musk's stake locked until June 2027. The float is going from roughly 5% today to about 66% by year-end. That's real supply pressure, and it's been the dominant narrative for two weeks. But supply pressure is a tape problem, not a business problem. The stock actually rose 6% on the August 6 unlock date, and has since bounced back to $133 from lows near $105.

Second is the obsession with GAAP losses. Yes, the company lost $541 million in Q2. But adjusted EBITDA nearly tripled to $3.5 billion, and Starlink is the only segment producing an operating profit. The losses are front-loaded investment, not operational decay. The distinction matters because it determines whether this spending is a bridge to profitability or a treadmill.

The financial bridge

Here's what needs to happen over the next 12 months for the setup to work. Starlink needs to keep scaling subscriber growth while ARPU stabilizes or improves — it's currently at $66, down from $85 a year ago due to lower-priced international expansion. AI infrastructure revenue needs to accelerate from the current ~$2.5 billion quarterly pace toward a level that meaningfully offsets the $15+ billion in quarterly capex. And the capex run rate itself needs to show deceleration once the initial data center buildout completes.

Management's $100 billion annual revenue target for 2026 would require roughly $18 billion per quarter — more than double Q2's $7.8 billion. That's an aggressive ask, though the Q2 revenue beat ($7.8 billion versus a $6.9 billion consensus) and 66% sequential growth leave some room for momentum. A $100 billion run rate on today's capital structure would start making the valuation look less absurd, even before free cash flow turns positive.

If the bridge holds, a simple framing works. Assume SpaceX reaches that $100 billion annual revenue run rate and Starlink continues to mature as a profit center. Applying a conservative 4x revenue multiple — still below the current 76x but more grounded than the IPO frenzy implied — yields a $400 billion market cap floor. That doesn't get you to Ives' $2.5 trillion. But it shows the stock at $133 is pricing in the old, pre-IPO revenue base rather than the trajectory.

For a target, the Ives $190 figure has a cleaner financial justification than most analyst calls on this stock. It would represent a market cap roughly 40% above today's level, still well below the $2.9 trillion peak and below the company's own $1 trillion revenue-by-2030 ambition. That implies a roughly 12-to-18 month window, giving time for the lockup overhang to clear, the Starship program to generate revenue, and the AI buildout to show returns.

The timeframe: mid-2027. That aligns with the final Musk lockup expiration and the point where SpaceX should have two full quarters of public earnings history to anchor expectations.

What breaks it

The tripwire is the free-cash-flow trajectory. If AI capex stays at or above the Q2 $15.8 billion quarterly run rate while revenue growth decelerates below 50% year-over-year, the spending isn't paying off fast enough. A sustained breach below $100 — the level where the stock fell 25% below the IPO price — would signal the market has decided the capex hole is structural rather than transitional.

More specifically: if Q3 results show revenue growth decelerating, AI contracted revenue shrinking, or the cash pile dropping below $70 billion, the financial bridge to Ives' thesis starts fraying. At that point, the story shifts from "investing in future returns" to "questioning whether those returns ever arrive."

Discipline over ego. The operating trajectory is improving. But free cash flow is the hard proof, and it's not there yet.

The setup: SPCX at $133, below its $135 IPO price. Target of $190 over a 12-to-18 month window (mid-2027), contingent on the $100 billion annual revenue run rate materializing and capex decelerating post-buildout. Tripwire: free-cash-flow trajectory worsening alongside revenue deceleration, or a sustained move below $100. Sit through the lockup noise. Watch the Q3 report for the first real proof point on whether AI spending is converting to revenue or just burning the cash pile.