Durability

Durability

Charter clears the mechanical disqualifier — no three-year high-single-digit revenue decline — and its reported free cash flow has never turned negative, with a smooth five-year rolling average. Two conviction sources hold firmly: the sunk-cost plant and the essentialness of broadband. But the year-10 gate carries a genuine doubt on the revenue leg: the broadband subscriber base peaked in 2023 and has fallen two years running against fixed-wireless and fiber entrants that are structurally adding capacity, and Internet revenue is now held up on price alone [1].

The conviction sources, graded for Charter

The year-10 gate draws conviction from five places. Charter holds two firmly, one partially and weakening, one with a caveat, and lacks one outright. Each is graded on this company's own evidence, not on an "infrastructure is forever" prior. The market-structure exposition is built in Business; this tab adds the durability grade and the movement in share over time.

Capital intensity as a moat — applies, strongly

The clearest source. Spectrum's network reaches 58 million homes across 41 states, a hybrid fiber-coaxial last-mile plant carrying gigabit speeds across the entire footprint and symmetrical multi-gigabit in part of it [2] [3]. Management puts the incremental cost of upgrading that plant to symmetrical multi-gig at just over $100 per passing — cheap because the plant already exists [4]. A new wireline rival, by contrast, must sink the whole cost first: Charter itself has spent $7.7 billion since 2022 to activate roughly 1.3 million rural passings [5]. This is the capital-heavy essential the framework means — a business that survives because the replacement cost of its footprint runs to tens of billions.

The qualification is the whole of the durability debate: capital intensity protects against a new wireline rival, not against a wireless one. Fixed wireless rides cell towers that already exist, sidestepping the last mile entirely. The moat is real against the threat it was built for and porous to the threat that is actually growing.

Essentialness — applies

Broadband is a genuine utility, and demand for it is rising rather than flat. Non-video Internet customers used roughly 825 gigabytes per month in early 2025, with over 30% of them above one terabyte, and usage compounds every year [6]. A household without connectivity cannot work, stream, or run its phones. What essentialness looks like in the data is churn: management describes it as "at or below historic lows," meaning customers leave over price and competitive offers, not because they stop needing the product [7].

Market structure — applies partially, and weakening

Broadband is a regional oligopoly, not a monopoly: most of the footprint is a wireline duopoly against a single telco (see Business). The durability complication is share stability, and here the record has turned. Charter's total Internet base rose through the pandemic to a peak of 30.59 million in 2023, then fell two consecutive years to 29.68 million — a loss of roughly 908,000 customers, 508,000 in 2024 and 403,000 in 2025 [8] [9].

Source: Charter FY2021, FY2023 and FY2025 Annual Reports (Form 10-K), customer statistics [10] [11] [12]. The FY2025 filing restates 2024 at 30,083.

The structure is no longer two players contesting a stable pool. A third pipe — fixed wireless from the national carriers — and a growing fourth — fiber overbuilders — now contest the footprint, and Charter faces terrestrial competition from AT&T fiber across roughly 27% of its footprint and Verizon across 16%, both higher than a year earlier [13]. The pace worsened into 2026: total Internet customers fell about 120,000 in the first quarter, versus roughly 59,000 a year earlier, with the Affordable Connectivity Program disruption now fully behind the base [14]. An oligopoly whose incumbent is losing units every year is a weaker conviction source than a stable duopoly, and it is graded that way.

Operating history — applies, with a caveat

Charter was founded in 1993, and the cable systems it runs are decades older [15]. The caveat is that today's Charter — its scale, brand, and levered-equity model — dates to the 2016 Time Warner Cable and Bright House combination, the point from which its buyback program runs [16]. The survival-through-cycles record of the current entity is closer to a decade than to the 30-to-50-year ideal, and it has not yet been tested through a full recession in its present, heavily levered form.

Regulatory entry barriers — does not apply

This is the source Charter lacks, and it matters because it is the one that shields the framework's banks and insurers. Charter states plainly that its cable franchises "are non-exclusive," so franchising authorities "can grant additional franchises to competitors in the same geographic area" [17]. A regulator does not stop T-Mobile or Verizon from selling home Internet, and the carriers already own the spectrum to do it. Federal and state policy is in fact pushing capital toward new broadband supply — Charter notes "billions of dollars in subsidies" for deployment through programs such as RDOF, ARPA, BEAD and the IIJA [18]. There is no regulatory gate keeping entrants out; if anything, the state is funding them in.

The structural threats, hunted

Execution is not a moat, so the question is not whether Charter operates well — it plainly does — but whether the structure of its market lets a well-capitalized attacker take year-10 revenue. A search of Charter's filings and of five peers' filings surfaces named, quantified threats. They are not waved at.

Fixed wireless access — the live one, and it is not execution

The national carriers are converting spare mobile-network capacity into home Internet aimed at cable's base. T-Mobile is now, on its own account, the fifth-largest Internet service provider in the country, with 12 million fixed-wireless customers [19]. Its stated strategy names the target directly: "We see our strength as attacking incumbents … a huge opportunity to attack incumbents across fiber and fixed wireless access" [20]. This is capacity added to Charter's markets by a company with a national spectrum position — a structural fact, not an operating stumble Charter can out-hustle.

Charter's read is more benign, and it is a genuinely two-sided point examined below: management says "cell phone Internet growth appears to have plateaued" and argues WiFi, not 5G, carries the household's traffic [21]. That call has been made repeatedly since 2023, and the unit losses have not yet stopped.

Fiber overbuild — a slower, permanent second front

Where fiber is built, cable's speed advantage narrows to nothing, and it does not retreat. AT&T ended 2025 with 32 million fiber passings and plans for 40 million by the end of 2026, then five million a year thereafter [22]. That is a decade of steadily expanding overlap with Charter's footprint, on top of the 27% AT&T already reaches. Charter's counter is that new fiber "builds are destined for poor financial returns," and that its own multi-gig upgrade is far cheaper than a fresh fiber build — also weighed below [23].

Video and voice — secular decline, already in the run-rate

The legacy products shrink structurally, not cyclically. In FY2025, video revenue fell 9.4% to $13,703 million and voice fell 6.0%; advertising fell 17.6% [24]. Streaming substitution is the named cause; Charter acknowledges it "negatively impact[s] the number of customers purchasing our video product" [25]. These lines are low-margin, so the revenue drag overstates the profit drag — but they are a persistent headwind on the top line, which is the leg of the gate under question.

Mobile growth rides on a rival's network

Charter's fastest-growing engine is structurally dependent on a competitor. Spectrum Mobile is an MVNO running over Verizon's network: the 22% mobile-revenue growth and 1.8 million net new lines are real, but the wholesale economics cap the margin and the arrangement is a supplier relationship, not owned infrastructure [26]. Verizon in early 2026 said it had "completed a comprehensive long-term agreement with Comcast and Charter to continue our partnership," which de-risks the dependency without removing it — the company Charter most relies on to grow mobile is the same one overbuilding its broadband [27].

Quantifying the year-10 impact

The threat is not abstract. Total Internet customers have fallen for two straight years, at roughly 400,000–500,000 per year against a 29.7 million base, and the pace roughly doubled year-over-year in early 2026 [28]. Held near that pace, the base would shed on the order of 12–15% by the mid-2030s — a meaningful bite out of the profit engine unless offset. The offset so far is price: FY2025 Internet revenue rose only because rate and mix added $785 million against a $380 million drag from fewer average customers, a net +$405 million [29]. Scale is being added on the other side — the pending Cox combination and the rural build expand passings (see Business) — but that is more customers to defend on the same contested terrain, not a structural barrier against the entrants.

The disqualifier check — revenue trajectory (X3)

The framework's one mechanical disqualifier is revenue declining high-single-digit for three consecutive fiscal years after a long existence. The feature file records the answer directly: revenue_trajectory.consecutive_decline_years = 1 and revenue_trajectory.three_year_hsd_decline = false. The disqualifier is not triggered.

Source: fit_features.revenue_trajectory; figures tie to Charter FY2025 Annual Report (Form 10-K), Revenues by service offering [30].

Revenue stepped up 43% in 2017 on the Time Warner Cable and Bright House merger, compounded at mid-single digits through 2022, then flattened: +1.1% (2023), +0.9% (2024), and −0.6% (2025), the first decline in the series and nowhere near the high-single-digit, three-year threshold. On the letter of the test, the structural-decline flag (X3) is checked and absent. The softer signal — a top line that has plateaued and just turned down, on a subscriber base already in a two-year decline — is not a disqualifier; it feeds the year-10 read below.

The year-10 case, both ways

The gate is binary by construction: year-10 revenue and year-10 adjusted free cash flow must be higher than today's, with very high conviction. Genuine doubt fails it.

The strongest case that both are higher

The revenue that is disappearing is the low-margin legacy; the revenue that is growing is the essential core. Connectivity — Internet plus mobile — was $27,527 million in FY2025, up 4.1%, and is 50.3% of total revenue [31].

Source: Charter FY2025 Annual Report (Form 10-K), Revenues by service offering [32].

Mobile grew 22% off low penetration with a long runway; data demand compounds; the multi-gig upgrade is cheap relative to a fiber rebuild; the Cox and rural expansions add passings; and management reads fixed wireless as plateauing and new fiber as return-dilutive [33]. On cash, the case is firmer. Capital spending peaks with the network-evolution and rural programs; Charter now expects to finish the network-evolution build in 2027 — a target that has already slipped from an earlier 2026 plan — and as that capex rolls off on stable-to-growing EBITDA, free cash flow inflects up [34] [35]. Consensus reflects the inflection — forward FCF estimates rise from about $4.99 billion (FY2025) to $8.43 billion (FY2029), well inside the ten-year window (fit_features.consensus_forward_yield). The FCF half of the gate is the more defensible half.

The strongest doubt

The doubt sits on revenue, and on the profit engine specifically. Internet revenue grew only because price outran volume — the FY2025 bridge is +$785 million rate and mix against −$380 million fewer customers [36]. Sustaining that requires pricing power, and pricing power against three balance-sheet-rich entrants that are structurally adding capacity — T-Mobile at 12 million fixed-wireless customers and climbing, AT&T heading to 40 million fiber passings and then five million a year — is precisely the out-execute-the-competition protection the framework says does not count on a ten-year horizon [37] [38]. The regulatory barrier that shields the framework's banks and insurers is absent; capital intensity guards the wireline flank but not the wireless one; video and voice keep shrinking; mobile's margin is capped by an MVNO dependency on a rival; and the top line has already turned down. Charter's broadband margin is exactly the opportunity three carriers are spending to capture — the Amazon test, applied here, cuts against durability.

The read

The FCF half of the gate is reasonably underwritten: as the capex peak passes, free cash flow should rise, and consensus already models it nearly doubling by 2029. The revenue half does not clear the "very high conviction" bar. There is a genuine doubt: whether year-10 revenue will be higher, given a broadband base that peaked in 2023 and is losing 400,000-plus customers a year to fixed-wireless and fiber entrants adding supply structurally, with the Internet line already sustained by price alone. The immediate counter-fact sits beside it — the connectivity core still grew 4.1% in FY2025 and mobile has a long penetration runway [39]. What would resolve the doubt toward the gate holding: broadband units stabilizing as fixed wireless saturates and the rural and Cox passings convert to net customers. What would confirm it: a third and fourth year of Internet-unit losses that ARPU can no longer offset, turning connectivity revenue negative.

FCF consistency (P2)

The consistency test is the rolling five-year average of adjusted free cash flow — volatile year to year is fine, unpredictable is not. A data limitation must be stated plainly first: the feature file returns fcf_stability.rolling_5y_avg as empty and not_computable, because stock-based compensation is missing from the cash-flow feed for every year 2016–2025, so the profile's adjusted-FCF series (adjusted_fcf.latest_adjusted = null) cannot be built. This is a feed gap, not evidence that Charter lacks SBC; it is recorded in the data gaps, and the adjusted-yield computation is owned by Yield.

What can be shown is the reported free-cash-flow series and its rolling five-year average, which speaks to the shape of the question even without the SBC and acquisition adjustments.

Source: reported FCF from fit_features.adjusted_fcf.series (company cash-flow statements); the 5-year rolling average is derived. Adjusted FCF (FCF − SBC − 5-yr avg acquisitions) is not_computable for this run — SBC absent from the feed.

Reported FCF swings widely — from $2.6 billion (2018) to $8.6 billion (2021) and back to $3.2 billion (2024) — but the swings track the capital-spending cycle and working capital, not underwriting losses. There are no negative years in the record, and the rolling five-year average is far smoother than the annual figures, holding in a $4.1–5.8 billion band across 2020–2025. That is the P2 signal: volatile but not unpredictable, and never loss-making. The negative-episode question the framework attaches to insurers and banks — a bad year every five to eight years as a healthy feature of the model — does not apply here; Charter's cash volatility is an investment-cycle phenomenon, not an underwriting-cycle one, with the 2020–2021 peak coinciding with a capex trough and the pandemic broadband surge and the subsequent dip reflecting the network-evolution and rural spend. The consistency pillar is met on the reported series, with the honest caveat that the adjusted series the framework prefers could not be computed.

What would change the durability read

The falsifiers are specific and datable. Toward the gate failing: a third and fourth consecutive year of Internet-unit losses that ARPU cannot offset, turning connectivity revenue negative; or fixed-wireless share continuing to compound rather than saturating. Toward the gate holding: broadband net adds returning to positive as fixed-wireless growth plateaus and the rural and Cox passings convert to customers; and free cash flow inflecting up as the capex peak rolls off. Leverage sits at 4.15 times EBITDA on roughly $94.6 billion of principal, a heavy load that the FCF inflection is meant to service and that Self-Help examines [40]. The revenue line is where the evidence must land; the cash line is the more secure half.