SpaceX’s Boost Mobile shortcut still runs through AT&T and T-Mobile

SpaceX’s fourth-carrier plan needs more than 7.2 million subscribers

In sum – what to know:

• The consent problem – LightShed Partners and a person familiar with the contracts say SpaceX cannot buy Boost Mobile and inherit its MVNO agreements with AT&T and T-Mobile.
• Rivals have said no – T-Mobile CEO Srini Gopalan has rejected a satellite MVNO, and Deutsche Telekom’s Tim Höttges told carriers on an investor webcast that courting SpaceX that way is not a good idea.
• Two very different price tags – Verizon’s Dan Schulman puts a competing national network at $300 billion to $500 billion over four to six years, while LightShed estimates an initial build at $10 billion to $15 billion.

SpaceX has held talks with EchoStar about buying Boost Mobile or gaining access to the network Boost uses, according to a Washington Post report. Either arrangement could give Elon Musk a faster route into a U.S. wireless market that, according to The Post, MoffettNathanson estimates at roughly $235 billion. No deal has been announced or priced, and none of the companies commented.

Musk met EchoStar chairman Charlie Ergen in October, the Post reported, and has also approached Verizon and other carriers about network access. They rebuffed him. Boost offers an alternative, but not an uncomplicated one. EchoStar reported nearly 7.2 million wireless subscribers as of Sept. 30, about 4% fewer than a year earlier, after losing 182,000 customers in the third quarter alone, most of them prepaid customers on Boost and Gen Mobile.

The complication is that Boost leases capacity from AT&T and T-Mobile, and an acquisition would not automatically carry those arrangements with it. LightShed Partners analysts Walt Piecyk and Joe Galone wrote that SpaceX cannot “simply acquire Boost Mobile and inherit its MVNO agreements with AT&T and T-Mobile, as [the] Washington Post story suggests.” DISH’s July 2021 8-K on the AT&T network-services agreement describes restrictions on reselling that service and defines a “restricted party” as one that generally includes “certain U.S. wireless providers, certain U.S. cable companies and certain large technology companies.” If a restricted party comes to own more than half of the voting power or economic value of DISH or any parent of its wireless business, or buys substantially all of that business’s assets, AT&T may terminate. DISH could then keep using AT&T’s network for its MVNO customers for up to two years afterward. On EchoStar’s August earnings call, Raymond James’ Brent Penter asked how restrictive the “NSA and MNSA contracts with T-Mobile and AT&T” are, “and changing control provisions there.” Ergen said, “I can’t get into the actual contracts,” offering only that EchoStar retains “a fair amount of flexibility contractually” to pursue “M&A or partnering with people.” A person familiar with the contracts was less equivocal, telling Light Reading that “every MVNO contract has a change-of-control provision” and that clearing it would still mean renegotiating AT&T’s pricing: “That doesn’t give Elon what he wants.”

SpaceX’s route into wireless

The broader ambition is a fourth nationwide U.S. carrier under the Starlink Mobile brand, planned as a hybrid in which satellites reach places towers cannot and ground sites supply dense-urban capacity and indoor coverage. Boost would supply what spectrum and satellites cannot: roughly 7.2 million customers, more than 3,000 Boost-branded stores, and billing, support and retail operations SpaceX would otherwise build itself. None of that is a nationwide tower network, which is why the AT&T and T-Mobile arrangements stay central to any valuation.

On Wednesday, Qualcomm announced that SpaceX has selected its Dragonwing Cellular Infrastructure platforms for Starlink Mobile’s U.S. nationwide terrestrial 5G network. SpaceX plans for the rollout to begin in 2027, with tens of thousands of sites over the multi-year agreement. Starlink senior vice president Michael Nicolls said the company is working with Qualcomm to “accelerate our timeline to develop our base station end-to-end architecture.” It is a supplier selection, not a transaction, and says nothing about Boost — but its timing suggests the buildout is moving in parallel rather than waiting on a deal.

Bloomberg reported in late October that SpaceX wants at least 80,000 U.S. cell sites and has talked with Crown Castle, American Tower and SBA Communications, near T-Mobile’s roughly 82,000 cell sites. Crown Castle said it has about 40,000 cell sites and, asked about satellite operators’ lease talks, said “you can assume that we’re talking to all parties.” MoffettNathanson’s Nick Del Deo read the 80,000 as counting rooftop and small-cell attachments alongside macro towers, or as a long-term rather than near-term target.

Carrier consent and pushback

However much spectrum SpaceX accumulates, Boost’s immediate value still runs through AT&T and T-Mobile. New Street Research’s David Barden called the suggestion that SpaceX may need carrier consent “a gross understatement.” LightShed’s Piecyk and Galone saw “no reason those restrictions would disappear in a change of control or why AT&T and T-Mobile would suddenly want to enable SpaceX as a competitor,” and Verizon CEO Dan Schulman has said there are “no back doors” to any of its MVNO agreements. MoffettNathanson’s Craig Moffett wrote, “We suspect that AT&T would have mechanisms to introduce sand in the gears of any such arrangement, even if legally permissible.” He cited what he described as CEO John Stankey’s repeated emphasis that Starlink Mobile would not fit the bill for a strategic MVNO partner. BNP Paribas analyst Sam McHugh reads the contracts as leaving EchoStar either unable to white-label the MVNO or facing change-of-control clauses with AT&T: “This means there’s no back door to an MVNO this way.” McHugh estimates the Boost MVNO is only about 1% of AT&T’s mobile service revenue, so preserving it may matter less than blocking a well-funded competitor.

T-Mobile CEO Srini Gopalan has already rejected a SpaceX MVNO arrangement. On the Oct. 28 earnings call he called satellite “a complementary product,” said “we don’t really see the room for an MVNO for satellites or SpaceX in that context,” and said that if SpaceX is a customer of T-Mobile or T-Mobile a customer of SpaceX, “that’s a purely commercial agreement.” Barden argues that a wholesale structure is the better path anyway: “The fastest, cheapest, most profitable and globally scalable business approach for SpaceX’s terrestrial ambition is a wholesale deal.” Moffett also noted that Boost’s T-Mobile MVNO expires July 1, 2027, which “makes it effectively irrelevant.” T-Mobile remains SpaceX’s partner on its first-generation satellite service, so its opposition is specific to a terrestrial deal. It runs above Gopalan, too: Deutsche Telekom CEO Tim Höttges, whose company controls T-Mobile, told mobile carriers on an Oct. 12 investor webcast, in remarks translated from German, “I do not think it would be a good idea to court SpaceX through an MVNO arrangement now.”

Cost is the other objection. Schulman estimates a competing national network would take at least four to six years to build, at a cost of $300 billion to $500 billion spent over seven or eight years, even assuming flawless execution. Musk disputed the timeline on X, putting it at “2 or 3 years.” LightShed’s Piecyk and Galone say Schulman’s comments “wildly overstate” what SpaceX needs to spend, putting an initial build at $10 billion to $15 billion of capex excluding spectrum. Piecyk and Galone also argue that Dish radios already mounted on towers “could give SpaceX a head start” and “throw a big bucket of cold water on Verizon’s four-to-six-year timeline and $300-$500 billion estimate.”

The shortcut has technical limits. Recon Analytics’ Roger Entner calls Dish’s network more or less “scrap metal”: “Charlie could have sold Elon the network, the antennas and everything a long time ago. What is Charlie going to do with it? It is equipment that will be sold as scrap or somewhere else… Probably it costs you more to take this down than what it’s worth.” Ownership has shifted, though: American Tower CEO Steven Vondran said on the company’s Q3 call that it owns Dish’s abandoned radios and antennas on its towers under a bankruptcy abandonment order, and LightShed says that lets SpaceX negotiate directly with tower companies. Ken Schmidt, president of consultancy Steel in the Air, allowed that Dish’s low-band radios might be retuned to SpaceX’s 800MHz holdings, “but that’s a configuration that has never carried commercial traffic on Dish’s network,” and said the idea of a fallow network “just waiting to be switched on” is “as specious as Verizon’s claim that the only path to terrestrial wireless is 4-6 years and $300-500B.” Piecyk put a question to Ergen on the August call about the DBS prepackaged proceedings and what happens to the radios, and got a closer answer: “in our opinion, there’s not a lot of liquidation value there. I think it’s relatively immaterial in the scheme of things.” A hybrid service also depends on modems, antennas and software managing both layers, and no handset compatibility list for the planned Starlink Mobile service has been reported, though T-Mobile publishes a per-model device list for T-Satellite, the first-generation service that runs on SpaceX’s satellites. Starlink’s own base does not settle it: the service costs $55 to $130 a month, and TMF Associates president Tim Farrar estimates roughly 5 million U.S. subscribers, only about 2 million of whom have a realistic wired or cellular home internet alternative (Broadband Breakfast).

Regulation adds a layer. Peter Adderton, the MobileX chief executive who founded Boost, called for FCC and DOJ scrutiny on LinkedIn, arguing that MNOs owning the networks is not itself the objection — but “when network owners use that control to limit consumer choice and protect their own businesses from competition, that’s a different story.” Any change of control would also draw FCC and antitrust review; carrier consent, if required, is only one approval.

EchoStar’s position and next moves

For EchoStar, the talks arrive alongside results that reshaped its reported financials. It filed third-quarter numbers on Nov. 3, without a press release or conference call, the day before Dish Wireless was due in bankruptcy court to update a judge on two months of mediation with creditors. Revenue fell to $1.13 billion from $3.61 billion, a comparison driven by deconsolidating Pay-TV, Broadband and Satellite into bankruptcy estates rather than a shrinking business, and net income attributable to the company reached $2.78 billion on a $3.1 billion gain tied to the AT&T transaction. Wireless held up better: service revenue rose 2.3% to $855 million, ARPU hit $39.11 against consensus of roughly $38.40. EchoStar’s 10-Q targets November 30, 2027, for closing its spectrum sale to SpaceX.

Ergen described the shortfall himself: “[A]s management, we haven’t cracked the code on how to be successful to the level we’d like to in the wireless business… In general, we’ve treaded water for four years now.” Entner is skeptical Ergen hands any of it over: “Do I see Charlie working for Elon? No. Do I see Charlie retiring? No.”

Boost and Gen Mobile operate outside the bankruptcy, but Dish Wireless remains in Chapter 11 and its tower disputes complicate any read on what is left. The tower companies are pressing roughly $7.6 billion in disputed lease claims across Crown Castle, American Tower and SBA Communications, and Dish Wireless is asking a judge to erase that figure. McHugh sees one way through: “The cleanest way to bridge the gap between what Dish has offered and what the tower companies want would be for someone else (SpaceX) to take on the tower commitments.” What that leaves on the table splits the analysts. LightShed’s view is blunt: “The allure left at EchoStar is its spectrum and Verizon should buy it.” New Street’s Barden asks whether SpaceX wants “the residue of a failed business with novelty radios and broken vendor contracts.”

Farrar has predicted for months that SpaceX eventually acquires the rest of EchoStar, arguing that the pieces Musk said SpaceX is missing, which he alluded to on the third-quarter call, are “widely recognized as SpaceX’s plan to acquire the rest of EchoStar in a share-based transaction” — attractive to Ergen because a share deal can be structured tax-free where the cash spectrum sales were not. He questioned the Post’s premise on X: “I’m not sure why this is pitched as SpaceX buying Boost, which the MNOs would try to veto. Clearly SpaceX wants the underlying Dish Wireless network… My assumption is that Boost is spun off to CONX, unclear if that has change of control implications?” CONX is Ergen’s SPAC, separately taking MobileX public. The tax pressure is concrete: Ergen puts EchoStar’s combined network-termination and tax exposure at “$5 billion-$7 billion.” Barden frames it the other way: “Why are we talking about SpaceX buying EchoStar in pieces? Why not just buy the whole thing outright?”

Citi’s Michael Rollins has called an outright telecom acquisition a tall order that could cut Starlink Mobile’s time to market by a few years. A BNP Paribas research note expects the story to be taken negatively “in the current ‘glass half-empty’ sentiment on wireless.”

The immediate question is narrower than Musk’s national-carrier ambition: SpaceX still has to prove it can secure usable terrestrial access, handset support and regulatory clearance at a workable cost. Whether Boost is a route around AT&T and T-Mobile, or another chance for those carriers to shape Starlink Mobile’s expansion, remains unresolved.