Who can afford to play

Everything below comes from filed accounts. Nothing here is investment advice, and nothing here values anyone’s equity.

The incumbents are cash-committed rather than cash-poor

What decides an incumbent’s room for manoeuvre is not how much cash it generates, because all five of the companies below generate a great deal of it. It is how much of that cash is still free once the network and the shareholders have been paid, because that residue is the money available for anything new: a spectrum position, a satellite partnership, a second network. Everything in the table below is read from each filer’s tagged XBRL lines rather than from adjusted measures, taking FY2025 Forms 10-K for the year and Q2 2026 Forms 10-Q for the most recent balance sheet.

¹ T-Mobile’s long-term debt is as reported at 31 Dec 2025; its Q2 2026 filing does not tag the same line. The same applies to Iridium’s long-term debt in the table below.

The same shape repeats down the columns. Operating cash flow is very large, capital expenditure takes a substantial share of it, and dividends take much of what is left, so the line that describes strategic room is not free cash flow but free cash flow after dividends. Charter, which pays no dividend at all, is the exception that makes the rule legible: its two free-cash-flow lines are the same number. For the big three, operating cash flow splits into capex, dividends and what remains.

The big three together generated $105.4B of operating cash flow in FY2025, spent $47.8B on capex and paid out $23.8B in dividends. What remains, $33.8B, is the discretionary money, and it sits against roughly $337B of combined long-term debt less cash. (We use long-term debt less cash rather than a full net-debt figure because T-Mobile’s current-debt line is not tagged in a recent filing; including a stale figure would flatter or distort the total.) The carrier sector is not short of cash. It is short of uncommitted cash, which is a different condition and a slower one to change.

Calculated AT&T, Verizon and T-Mobile FY2025 cash-flow statements, summed

The constraint is not revenue, since a carrier’s against a satellite operator’s is no contest. It is what an incumbent can still choose to spend, set against what a spectrum position now costs. These companies are not poor; the money they have is already doing something, and every new commitment has to displace an existing one.

SpaceX’s entire 65 MHz position cost $19.6B. The three carriers could cover that between them with about seven months of post-dividend free cash flow. They could not do it and also fund a network build, and they could not do it without the debt markets noticing.

The dividend is the real constraint, and Auction 113 shows it working. Verizon spent $3.16B on AWS-3 licences in 2026, its gross winning bids in that auction, which is 37% of its entire FY2025 free cash flow after dividends of $8,645M. T-Mobile, which pays a smaller dividend and carries less net debt, bought more licences for a tenth of the money. Two companies bidding in the same auction, two different degrees of freedom, and what separates them is the prior claim on the cash rather than anything either of them has said about strategy.

The challengers: freedom without cash

The satellite side is the mirror image of that position. There is no dividend to protect and no legacy network to maintain, so in principle every dollar is available for the build, and there are very few dollars. The four operators below are read on the same basis as the carriers, from the filed tags rather than from adjusted measures, and they do not form a single category: the distance between the most profitable of them and the least is wider than anything in the carrier table.

¹ As with T-Mobile above, Iridium’s long-term debt is as reported at 31 Dec 2025; its Q2 2026 filing does not tag the same line. ² Globalstar tags only PaymentsToAcquireOtherPropertyPlantAndEquipment, which was $5.5M in FY2025; it does not tag constellation spending as capital expenditure. Its FY2025 net cash used in investing activities was $550.4M, which is the better proxy for what it is spending, and is not comparable with the capex line above.

AST’s revenue is real now, at $70.9M in FY2025 against $4.4M the year before. It is also spending $859M in six months while running a $145M operating cash outflow, and against $2,288M of cash that is roughly fourteen months at the current rate before new financing. AST has raised capital repeatedly and may well do so again; the clock is short and visible in the filings, and it runs alongside the May 2028 lease expiry from part 2.

Filed FY2025 Forms 10-K and Q2 2026 Forms 10-Q for EchoStar, AST SpaceMobile, Globalstar and Iridium

EchoStar is the cautionary exhibit, and its FY2025 Form 10-K is where the caution is recorded. It carries its FCC authorisations on the balance sheet at $34,548,952,000 (31 December 2025), and in FY2025 it recorded pre-tax impairments of $5,784,779,000 against indefinite-lived intangibles, $6.5B against long-lived assets and $4.3B against operating leases, in a year it reported a $14.5B net loss. (The charges are pre-tax and do not sum to the loss; they are the scale of the write-down, not an arithmetic explanation of it.)

It has since agreed to sell roughly 115 MHz for about $39.9B before the escrow described in part 4: the AT&T leg closed on 28 July 2026, the SpaceX leg is approved but not yet closed. Separately, EchoStar disclosed that its DISH DBS subsidiary has restructuring proceedings pending before the US Bankruptcy Court for the Southern District of Texas, and that the court authorised DISH DBS’s repayment of $2.0B of senior notes on 28 July 2026 (EchoStar Form 8-K, filed 3 Aug 2026). Owning spectrum is not the same as being able to use it, and a balance sheet full of licences is not a business.

The finding that reframes the whole market

Part 1 identified Globalstar and Iridium as the operators who never appear in the lease records because they own their spectrum outright. Both are now being acquired, and not by telecom companies. The terms below come from each company’s Form 8-K rather than from the announcements that accompanied them.

• Globalstar is being acquired by Amazon under an Agreement and Plan of Merger dated 13 April 2026, at $90.00 per share in cash or, at the holder’s election, stock at an exchange ratio of 0.3210 Amazon shares, with an adjustment tied to an amount payable to Apple.
• Iridium is being acquired by Rocket Lab under an Agreement and Plan of Merger dated 28 June 2026, at $27.00 per share in cash plus Rocket Lab stock, at an exchange ratio between 0.2400 and 0.4000 depending on Rocket Lab’s twenty-day price. Iridium’s lenders consented to the transaction on 15 September 2026.

Iridium was the one unambiguously profitable business among the four satellite operators above, with $872M of revenue and $114M of net income in FY2025, and it is being bought. Globalstar, the other operator that never had to rent the spectrum it uses, is going the same way. Together the two agreements move both owned-spectrum incumbents into other hands.

The pattern is not distress. It is that owned mobile-satellite spectrum has become an input somebody else wants, and the somebodies are Amazon, Rocket Lab and SpaceX. None of them is a telecom incumbent.

So: does anyone have the cash to play in a mobile world?

Three positions emerge from the filings. Cash flow and freedom of action have ended up in different hands, and neither side of this market holds both at once.

• The carriers have the cash flow but not the freedom. Their operating cash flow dwarfs anything on the satellite side, but it is pre-committed to capex, dividends and debt service; they can win an auction, and they cannot fund a second, parallel network.
• The satellite operators have the freedom but not the cash. AST has about fourteen months of visible runway, Globalstar and Iridium have agreed to be acquired, and EchoStar has agreed to sell the asset and has already closed half of it.
• So the capital is arriving from outside the industry. Amazon, Rocket Lab and a private SpaceX are funding this from balance sheets that no US carrier can match and no satellite operator ever had.

SpaceX files nothing, so the most important balance sheet in this market is invisible. Every comparison above is drawn between companies obliged to disclose, and one of the three buyers named above is under no such obligation, a limit on any analysis of this market.