Chapter 2

The Broadband Fulcrum

The Broadband Incumbent left the reader with a company of scale — 58.4 million passings, 31.8 million customer relationships, a four-product bundle — and a stock down 71.6% from its three-year high. This act is about the machine underneath that scale: how a dollar of Spectrum revenue is actually made, and how little of it survives the trip to free cash flow.

The short version is that almost the entire model now turns on one line. Residential Internet is where the pricing happens, where the margin pools, and where the whole business is either defended or lost. Everything else — a melting video product, a fast-scaling mobile business, a stable commercial book — either feeds that line or rides alongside it.

Source: FY2025 Annual Report (Form 10-K), Adjusted EBITDA table [1]; margins and FCF margin derived from reported financials, FY2021–FY2025 [2].

A 41.5% Adjusted EBITDA margin and a 23.6% operating margin are both the widest in the peer group — and the 8.1% free-cash-flow margin is the narrowest. The distance between those two numbers is the subject of this chapter.

The line the model turns on

Charter groups its residential revenue into four service lines. In 2025, only one of the two large ones grew. Internet revenue rose 1.7% to $23.8 billion; mobile service jumped 22.0% to $3.8 billion; the two together, which management now brands "Connectivity," rose 4.1% to $27.5 billion. Against that, video fell 9.4% to $13.7 billion and voice fell 6.0% to $1.35 billion [3].

Source: FY2025 Annual Report (Form 10-K), MD&A revenues by service offering [4].

Internet is the line to watch because it carries the profit and because it is being priced against a base that no longer grows. The 10-K decomposes its residential-Internet revenue change into two forces that pull in opposite directions: rate and product mix added $785 million, while a lower average customer count subtracted $380 million, for a net gain of $405 million [5]. Behind the volume figure: residential Internet customers fell by 393,000 during the year, to 27.64 million, and total Internet customers fell to 29.68 million from 30.08 million [6].

Source: FY2025 Annual Report (Form 10-K), residential Internet revenue bridge [7].

This is the mechanism the equity rests on: price beat volume by $405 million, so the line grew — but the margin of victory is thin and it is being won on rate, not units. Rate here means promotional roll-offs, list-price adjustments, and a favorable shift in how bundled revenue is allocated, not a headline price hike; monthly residential revenue per customer was essentially flat at $119.05, versus $118.71 a year earlier [8]. Flat per-customer revenue with a shrinking base is a model that grows only as long as rate keeps out-running attrition by a small margin each year. Whether it can, as fiber and fixed-wireless press on both the rate and the base, is a competitive question the next act takes up.

A growth core inside a shrinking top line

Total revenue fell 0.6% in 2025, to $54.77 billion from $55.09 billion — the first annual decline in the company's modern history, after growth had already decelerated from 7.5% in 2021 to 1.1% and 0.9% in the two prior years [9]. Read in isolation, a top line turning negative looks like the beginning of a contraction. Read against the service lines, it is something more specific: the growth core (connectivity, +4.1%) is now slightly smaller than the melting core (video and voice, together down roughly $1.5 billion), so the two nearly cancel and the sign flips. The decline is a mix crossover, not an operating collapse — the company has passed the point where its shrinking legacy products can be fully papered over by its growing ones.

The video melt is large and orderly. Residential video revenue fell $1.43 billion, of which $813 million was lost customers (255,000 fewer residential video subscribers), $322 million was a reallocation of streaming-app costs netted inside video revenue, and $291 million was rate and package mix as customers migrate to cheaper tiers [10]. Voice lost 804,000 wireline customers and $87 million of revenue [11]. What holds the picture together is the relationship, not the product: residential customer relationships slipped only to 29.61 million from 29.96 million, because the bundle keeps the household even as video and voice fall away underneath it [12].

Why the melt lifts the margin

The counter-intuitive part is that video's decline makes the company more profitable, not less. Video is a low-margin, largely pass-through business: most of its revenue leaves again as programming payments to content owners. As video shrinks, that cost shrinks with it. Programming was the single largest operating expense at roughly $8.8 billion in 2025, down from $9.7 billion, falling from 29% to 27% of total operating costs as the video base and package mix declined [13].

That is why the operating margin has climbed even as growth has faded. Over FY2019–FY2025, operating margin rose from 14.2% to 23.6% while revenue growth fell from 4.9% to below zero.

Source: derived from reported financials, FY2019–FY2025 10-Ks [14]; programming-cost mechanism per FY2025 10-K [15].

The rising line and the falling line are the same story told twice. Margin strength here is not being manufactured by growth; it is being manufactured by the retreat of a low-margin product. That is a genuine and durable form of mix benefit — but it is finite. Programming cost cannot fall forever, and when the video base is small enough, the margin tailwind from its decline runs out.

Mobile crosses over

Mobile is the one line that is both growing fast and beginning to help the bottom line. Service revenue rose 22.0% to $3.76 billion, from $1.70 billion in 2022, as residential mobile lines grew by 1.8 million in the year to 11.37 million [16]. For most of its life this business was a customer-retention subsidy: Charter sold mobile cheaply to lock the broadband relationship, and the direct costs ran ahead of the revenue. In 2025 that inverted. Management states Adjusted EBITDA grew "with mobile revenues growing at a faster rate than mobile direct costs" — the first clean statement that mobile is now additive to profit, not a drag on it [17].

That crossover is why a 41.5% Adjusted EBITDA margin held — indeed edged up 0.6% to $22.71 billion — in a year revenue fell. Two caveats belong on the record and are developed elsewhere. First, the filing does not disclose mobile as a standalone profit line; its direct costs sit inside "other costs of revenue," which rose $353 million on higher mobile volumes, so the crossover is management's characterization rather than a reported segment margin [18]. Second, Charter's mobile network is not its own — it resells a competitor's, an economic dependence the competitive chapter examines.

From 41.5% EBITDA to 8% free cash flow

Here is the gap the chapter opened on. A business generating $22.71 billion of Adjusted EBITDA — 41.5% of revenue — delivered $4.42 billion of free cash flow, an 8.1% margin. Two deductions consume the difference, and both sit below the operating line.

Sources: Adjusted EBITDA and income from operations, FY2025 10-K [19]; capital-expenditure table [20]; interest wedge derived as income from operations less pretax income, FY2025 10-K [21]; FCF per reported financials.

The first deduction is capital expenditure — $11.66 billion, or 21.3% of revenue, the highest intensity in the company's decade and roughly double the 10.4% peer median. Its composition is telling: $7.72 billion of core capital (customer equipment, scalable infrastructure, upgrade/rebuild, support), plus $3.94 billion of line extensions, of which $2.20 billion is subsidized rural construction [22]. The discretionary growth spend — rural expansion plus the DOCSIS 4.0 network evolution introduced in The Broadband Incumbent — is the portion management frames as peaking, and the entire cash-recovery case rests on it rolling off.

The second deduction is interest. Operating income of $12.91 billion falls to $7.46 billion of pretax income — a roughly $5.45 billion annual wedge, the cost of carrying $94.3 billion of net debt [23]. That wedge consumes about 42% of operating income before a dollar of it reaches equity. Leverage is not a footnote to this model; it is the second-largest claim on the cash the business produces, ahead of taxes and behind only capex.

Flat profit, collapsing cash

The clearest way to see that free cash flow here is a capital-and-financing story, not an earnings story, is to watch profit and cash diverge. Net income attributable to shareholders held in a narrow $4.6–5.1 billion band every year from FY2021 to FY2025. Over the same window free cash flow fell from $8.60 billion (2021) to $3.16 billion (2024) — a 63% collapse — before recovering to $4.42 billion in 2025. Nothing in earnings moved; capital intensity did.

Source: derived from reported financials, FY2020–FY2025 10-Ks — FCF margin and capex/revenue series.

The two lines are near-mirror images: as capex intensity climbed from 14.8% of revenue to 21.3%, the free-cash-flow margin fell by roughly the same amount, then ticked up in 2025 as capex growth flattened. That leaves the 2025 recovery to explain — and it did not come from the operating business.

The 10-K walks the $747 million year-over-year increase in free cash flow to its drivers, and the largest by far is a $669 million reduction in cash paid for taxes, after the One Big Beautiful Bill Act restored 100% bonus depreciation in July 2025. Lower cash interest added $347 million and favorable mobile-device working capital added $398 million. Higher Adjusted EBITDA contributed just $139 million; capital expenditure actually rose, subtracting $390 million, and other working capital subtracted $455 million [24].

Source: FY2025 Annual Report (Form 10-K), free-cash-flow drivers and new tax legislation [25].

Separating a durable inflection from a timing benefit matters because the peak-capex-then-recovery story is the one the equity is priced on — the said-versus-did record on that promise, and the price it produced, are the ground of the closing act.

Cash quality: clean above the line, capex-governed below

None of this is an earnings-integrity problem. Above the free-cash-flow line, the accounting is clean and the cash conversion is high. Operating cash flow was $16.08 billion against consolidated net income of $5.77 billion, a 2.79x conversion, bridged by roughly $8.71 billion of depreciation and amortization and a recurring deferred-tax build (the deferred-tax liability rose to $19.84 billion from $18.85 billion) — standard non-cash items for a capital-heavy operator, not accrual engineering [26]. KPMG issued an unqualified opinion, internal control over financial reporting was assessed effective with no material weakness, and there was no restatement [27]. The largest judgmental estimate on the balance sheet is the $67.47 billion of indefinite-lived franchise intangibles — 44% of the $154.2 billion asset base — carried without impairment [28].

Two things below the headline deserve a reader's attention. Charter steers on Adjusted EBITDA, a figure 76% above income from operations ($22.71 billion versus $12.91 billion), and reports its own free-cash-flow definition of $5.00 billion, about $0.6 billion above the raw operating-cash-flow-minus-capex figure of $4.42 billion, the gap being an add-back of capex-related accruals [29]. Both measures are reconciled, but both flatter the statutory numbers, and leverage is governed against the higher one.

The second is a growing consumer-credit exposure hiding inside a connectivity balance sheet. Accounts receivable grew 42.6% from FY2021 to FY2025 — to $3.68 billion — while revenue grew just 6.0%; in FY2025 alone receivables rose $583 million as revenue fell [30]. The driver is device financing: as Charter sells phones on equipment-installment plans, it books a receivable for the unpaid handset. Equipment-installment receivables inside accounts receivable reached $1.28 billion (from $967 million), with a further $1.11 billion classified as noncurrent [31].

Sources: consolidated balance sheet, FY2025 10-K [32]; EIP receivables note [33]; EIP Financing Facility note [34].

Charter funds this book the way a lender would: it sells the receivables into a bankruptcy-remote special-purpose vehicle and borrows against them under an EIP Financing Facility, whose carrying value grew to $1.45 billion from $1.07 billion; the SPV holds $2.2 billion of pledged receivables (up from $1.6 billion) plus $121 million of restricted cash [35]. It is on the balance sheet and modest against a $94 billion debt stack, but it is a securitization structure scaling with mobile growth — a financing of working capital, not an operating cash source — and the allowance for doubtful accounts ($219 million, up from $204 million) is growing more slowly than the receivable it covers. The same mobile crossover that lifts EBITDA quietly embeds a consumer-credit line inside the model.

Where this leaves the reader

The Spectrum machine is a rate-versus-volume engine bolted to a mix shift and a heavy fixed-cost tail. Price still beats volume on the one line that matters, by a narrowing margin; the decline of video manufactures operating margin even as it shrinks the top line; mobile has crossed into profit and holds EBITDA flat. But the cash that reaches an owner is set below the operating line — by a record capex bill management says is about to fall, and by a $5.4 billion interest load that will not. FY2025's cash uptick was mostly a tax-timing gift, not proof the build is over.

That hands the next question to the field. The whole model depends on Internet rate continuing to out-run subscriber losses — and both the rate and the base are set by competition. The next act examines the moat that is supposed to protect them, and why it is leaking.