SpaceX's Spectrum Grab Rattles US Carriers, But the Real Bill Is Still Coming

SpaceX's spectrum deal with Grain Management pressures AT&T, Verizon and T-Mobile as analysts flag up to $130B network build costs.

SpaceX has agreed to buy a nationwide portfolio of paired 800-megahertz spectrum from Grain Management, a move that pushes the rocket-and-satellite company squarely into the business of running a conventional mobile network — and immediately knocked the shares of America's incumbent carriers.

AT&T, Verizon and T-Mobile all came under selling pressure following the announcement, as investors weighed what a satellite operator with deep pockets could do to the cozy US wireless oligopoly. SpaceX's own stock, by contrast, has been rewarded: the shares were up 3.9% at EUR 148.86 in today's session.

The deal, struck on Thursday, hands SpaceX as much as 14 megahertz of paired low-band spectrum. The Wall Street Journal puts the cash consideration at roughly USD 8 billion, though neither party has confirmed the precise financial terms.

Why low-band matters

Low-band airwaves are prized for a simple physical reason: they travel well through thick walls and into buildings. Starlink Mobile has until now leaned primarily on frequencies around 2 gigahertz, which deliver respectable bandwidth but struggle when faced with structural obstacles. Marrying terrestrial low-band capacity with a low-earth-orbit satellite constellation is what could turn Starlink Mobile from a niche satellite service into a genuine rival to the established networks — one that reaches ordinary handsets without any extra hardware.

The FCC has separately cleared the deployment of 15,000 second-generation Starlink Mobile satellites, though that approval does not guarantee the specific radio licenses tied to this transaction will follow. Regulatory sign-off on the Grain Management purchase is still pending.

The price tag nobody is talking about

Owning spectrum is not the same as operating a network. Bernstein analysts estimate that building a competitive nationwide terrestrial mobile infrastructure would cost between USD 50 billion and USD 130 billion. Tens of thousands of cell sites would need to be leased or constructed to blanket densely populated areas — a multi-year undertaking that would absorb enormous amounts of liquidity. Ground stations and radio towers cannot simply be replaced from orbit.

That spending lands on top of an already crowded capital agenda. SpaceX is reportedly negotiating a financing package of about USD 40 billion to fund data centers equipped with Nvidia chips for AI infrastructure, with Reuters and the Financial Times naming Apollo as a lead participant and noting the structure would include substantial corporate bond tranches. Back in June, the company had already raised USD 25 billion through bond placements.

The strain is visible in the numbers. Free cash flow stood at minus USD 16.82 billion in the second quarter of 2026, alongside a net loss of USD 541 million.

Wall Street is split on the destination

Analyst price targets suggest the market is willing to look past the near-term cash burn. Goldman Sachs recently lifted its target to USD 230, while Morgan Stanley sees USD 300 as achievable. Even so, the stock remains roughly 25% below its 52-week high of EUR 194.46, a reminder that enthusiasm has limits.

What SpaceX is attempting is a dual-front expansion into two of the most capital-hungry industries imaginable: AI computing on one side, terrestrial mobile infrastructure on the other. Should the FCC-approved buildout proceed and Starlink Mobile genuinely capture US telecom market share, the revenue base would widen dramatically.

But the financial burden is real, and the operational questions are unresolved. Anyone betting on this vision needs the stomach to sit through heavy losses and repeated waves of investment. The hard part — proving the model works on the ground — has barely begun.