SpaceX Buys 800 MHz Airwaves to Make Starlink a Carrier

A reported $8 billion deal for Grain Management’s low-band licenses gives Starlink Mobile indoor reach, unsettles U.S. and European carriers, and lifts the tower companies that may build its network.

Key Takeaways

• SpaceX agreed to buy up to 14 MHz of nationwide 800 MHz spectrum from Grain Management, adding the building-penetrating low band that its 65 MHz of EchoStar mid-band spectrum lacks.
• AT&T, Verizon and T-Mobile fell 5% to 7% in after-hours and pre-market trading, while American Tower, Crown Castle and SBA gained 5% to 6% on expectations of new terrestrial demand.
• With about $100 billion in cash and securities, SpaceX can easily fund the deal. FCC approval, 2028 license expiries and a late-2027 satellite timeline will still set the pace of competition.

The Missing Frequency

For two decades, the economics of American wireless have rested on scarcity. Spectrum is finite, towers are costly, and three national carriers control most of both. On Thursday evening, SpaceX moved against that arrangement. The company agreed to acquire Grain Management’s entire nationwide portfolio of 800 MHz licenses, up to 14 megahertz of paired low-band spectrum. The Wall Street Journal and Reuters valued the all-cash transaction at about $8 billion. Neither party disclosed terms, and the deal requires approval from the Federal Communications Commission. SpaceX said the spectrum addresses one of the key remaining technical gaps between Starlink Mobile and the standing of a major U.S. carrier. Elon Musk called it the “last critical piece.”

The language is promotional, but the engineering argument holds. SpaceX’s mobile plans have so far depended on mid-band spectrum: roughly 65 MHz of AWS-4, H-block and unpaired AWS-3 licenses acquired from EchoStar for about $19.6 billion, according to the company’s second-quarter 10-Q. Mid-band frequencies carry a great deal of data, but they lose strength quickly through concrete, glass and steel. Low-band signals travel farther and reach indoors, which is where a satellite-first network struggles most. SpaceX also noted that most phones already in use support the 800 MHz band, so a subscriber could switch without buying new hardware. No number of satellites could have bought that advantage. Two days earlier, on October 6, the FCC authorized a 15,000-satellite constellation dedicated to Starlink Mobile. In the same order it waived the rule requiring a satellite operator to lease spectrum from a terrestrial carrier. The timing of the two events is unlikely to be a coincidence.

An Asset With a Clock Attached

The licenses come with a history. T-Mobile inherited them through its acquisition of Sprint, and the Justice Department required it to divest them. A planned sale to Dish collapsed when the buyer could not pay. Grain ultimately took the portfolio for $2.9 billion in cash plus its own 600 MHz holdings. The FCC approved that exchange on July 1, and T-Mobile completed it on August 11. The Commission attached unusually firm conditions. It shortened Grain’s proposed build-out deadlines from six and twelve years to three and eight. It also tied defined direct-to-device performance terms to a competitive solicitation and to D2D filings before the end of 2026. Regulators made clear that this spectrum was meant to be used, not held.

Those conditions turned a private sale into an auction with a deadline. Bloomberg reported in August that SpaceX and AST SpaceMobile were among the interested parties and that the portfolio was valued at roughly $6 billion. AST had more at stake than most. An AST filing cited in the FCC’s July order stated that more than 80% of its Block 2 satellite production included 800 MHz capability. If the reported price is accurate, SpaceX paid a substantial premium for a band around which its closest direct-to-device rival had designed hardware. That premium both buys coverage and denies a competitor. For Grain, the outcome is a reported $8 billion exit two months after closing, although the value of the 600 MHz licenses it gave up was never disclosed.

A Buyer That Pays in Cash

Earlier challengers to the national carriers failed less because of strategy than because of financing. SpaceX does not share that weakness. In June it completed the largest initial public offering on record, selling 638.9 million shares at $135, including the over-allotment, for net proceeds of $85.7 billion. It closed the second quarter with $93.5 billion in cash and $6.5 billion in marketable securities, so an $8 billion payment uses roughly 8% of its available liquidity. The operating business supports the ambition. Second-quarter revenue reached $7.81 billion, up from $4.07 billion a year earlier. The Connectivity segment, which houses Starlink, contributed $4.29 billion, compared with $2.59 billion a year earlier.

Subscriber figures tell a more nuanced story. Starlink ended June with 12 million subscribers, double the prior year. Over the same period, average revenue per user fell to $66 a month from $85 as the service expanded into lower-priced markets. Giving up price in exchange for scale is consistent with a company preparing a mass-market mobile product. The real constraint is how much the business consumes in capital, not how much cash it holds. SpaceX spent $28.5 billion on property and equipment in the first half of 2026, much of it on AI infrastructure, and reported a net loss of $4.8 billion for the period. At $8 billion, the spectrum is the inexpensive part. Turning licenses into a network that customers trust with their primary phone number will cost considerably more.

How the Market Priced the Threat

Investors responded before the details were settled. AT&T, Verizon and T-Mobile fell between 6% and 7% in after-hours trading on Thursday. Reuters reported declines of 5% to 6% in Friday’s pre-market session, and Forbes put AT&T near $23, Verizon near $43 and T-Mobile near $160 before the open. The reaction reached Europe, where the STOXX Europe 600 Telecommunications index dropped 3.3% to its lowest level since February. AST SpaceMobile lost about 4% after hours. SpaceX shares rose 3.6%.

The more instructive move came from tower operators. American Tower, Crown Castle and SBA Communications gained between 5% and 6%. Investors reasoned that a satellite-native carrier with terrestrial ambitions will still need ground infrastructure. According to reported excerpts of its note, Bernstein estimated that a nationwide build from scratch would cost $50 billion to $130 billion across 30,000 to 120,000 sites. Morgan Stanley described the transaction as “incrementally constructive” for towers. The pattern is now familiar. Verizon fell more than 5% on June 29 after reports that a retail mobile product had been pitched during the IPO roadshow. The carriers declined again in August, when Gwynne Shotwell said SpaceX would “definitely” build out terrestrial infrastructure. Thursday’s agreement turned that stated intent into an asset.

The Limits of Disruption

The threat is real, but it will arrive in stages. Morgan Stanley read the Grain deal as evidence that SpaceX will become a more assertive buyer of spectrum. It also expects any erosion of carrier economics to begin gradually, starting in rural markets. The timetable supports that view. The EchoStar licenses remain in a trust until the acquisition closes, which is expected around November 30, 2027. Next-generation Starlink Mobile satellites are not scheduled to begin launching until late 2027. The 800 MHz transaction needs its own FCC consent, and the band sits next to public-safety operations whose interference protections were expensive to secure. Many of the underlying licenses also expire in 2028.

Scale remains the central question. Analysts estimate that the three national carriers together control about 1,020 MHz of spectrum. Even with Grain’s portfolio added to EchoStar’s 65 MHz, SpaceX holds a small fraction of that. It is betting that satellites, low-cost ground cells and rooftop terminals can make up for that gap, a proposition no one has tested at national scale. The direction, however, is settled. Thirteen months ago SpaceX was effectively a supplier to T-Mobile, filling coverage gaps in remote areas. It is now a spectrum owner with a dedicated constellation license, publicly traded stock to use as currency, and about $100 billion in liquidity. The question for incumbents is no longer whether Starlink Mobile competes, but where it competes first and at what price per gigabyte. For the FCC, which pressed for idle spectrum to be put to work, the result is the competitive entry it sought. It has come from a company that already dominates satellite broadband.