Yield

Yield

On the framework's adjusted basis — reported free cash flow, less stock-based compensation, less the five-year average of acquisition spend — Charter generated about $3.7 billion in FY2025: a 21.0% yield on today's $17.8 billion market cap, and roughly four times its own 2019–2024 yield baseline. Against the levered class's 25% reference line, the current year sits about 400 bps short; management's committed capex roll-off and sell-side consensus both carry the adjusted yield above 25% within three years. Acquisitions are zero; SBC is the only adjustment that bites.

The adjustment, line by line

The framework's adjustment removes two things reported FCF flatters: the non-cash cost of paying employees in stock (SBC), and the smoothed cost of buying growth (five-year-average acquisitions). For Charter, only the first applies — the company has made no cash acquisitions since the 2016 Time Warner Cable and Bright House mergers, so every 10-K cash-flow statement from FY2019 on shows no acquisitions line in investing activities [1][2]. The five-year-average acquisition term is therefore zero, matching the feature file's implicit-zero treatment.

SBC is modest for a business this size — $315 million in FY2019 rising to $673 million in FY2025 [3][4] — so the adjustment shaves 1.2% to 1.5% of revenue off FCF, not the double-digit haircut that knocks serial issuers like Accenture out of the framework. Charter's share count is falling, not rising: the SBC dilution is more than offset by buybacks.

Adjusted FCF = reported FCF − SBC − 5-yr avg acquisitions; derived from company filings. Reported FCF = operating cash flow − capex, per the numeric feed (this is stricter than Charter's own FCF definition, which adds back the change in accrued capital expenditures — e.g. $5,004M vs $4,418M in FY2025). SBC and the absence of acquisitions anchor to the filed cash-flow statements: FY2023–25 [5], FY2021–22 [6], FY2019–20 [7].

Source: derived from the filed cash-flow statements [8][9]. The gap between the two bars is SBC alone.

The shape of the adjusted line is the whole story of this tab: adjusted FCF peaked near $8.2 billion in FY2021, fell to $2.5 billion by FY2024, and recovered to $3.7 billion in FY2025. That collapse-and-partial-recovery is capex, not the business — the section below on conversion shows why.

The yield, three ways

Adjusted FCF ÷ market cap. Market cap $17.8B = $129.22 close (2026-07-22) × 137.74M Class A shares, per the feature file's market-cap derivation. Current = FY2025 adjusted FCF $3,745M; 3-yr avg = mean of FY2023–25 adjusted FCF ($2,960M); baseline = median of the FY2019–2024 year-end yields shown below.

Current adjusted FCF of $3,745M on a $17,799M market cap is a 21.0% yield. The three-year average of adjusted FCF ($2,960M) on today's market cap is 16.6%. Both are extreme numbers by any absolute standard — and both are the mechanical result of a 70% price collapse (covered in Dislocation) sitting on top of a temporarily depressed FCF denominator.

The baseline is where the fortress-versus-levered distinction shows up. Reconstructing the same adjusted-FCF yield at each fiscal year-end market cap, Charter's yield ran between 3.9% and 9.1% across FY2019–FY2024, with a median near 5.0%. Today's 21.0% is roughly four times that baseline — the feature file's jump test (current ≥ 2× a positive baseline) is met several times over.

Source: adjusted FCF (derived, above) ÷ year-end market cap (year-end close × year-end shares, from the price and share-count feeds); "Now" uses the 2026-07-22 close. A rising yield here is a falling price, not rising cash: adjusted FCF in FY2025 ($3.7B) is below its FY2020–21 level.

Unlike the framework's fortress signature — a stable ~3.5–4% name (Microsoft, Meta) jolted to 8–9% on a scare — Charter's baseline was never a clean, low fortress yield. It sat in the 4–9% band because the equity was always priced against a large, growing FCF stream carrying substantial leverage. The jump is real, but it is a levered name re-rating from a mid-single-digit equity yield to a 21% one, which is why the 25% reference line — not the 8–9% one — is the right ruler here.

One honest denominator caveat. The 21.0% divides total-company adjusted FCF by the Class A market cap only. Advance/Newhouse holds roughly a 10.9% common-equity interest through exchangeable Charter Holdings units that sit outside the 137.7M Class A count [10]. On a fully-exchanged basis (about 154.6M units, ~$20.0B), the same adjusted FCF yields roughly 18.7%. The framework and the feature file both use the Class A market cap, so 21.0% is the headline figure; 18.7% is the economically-consistent floor. Either way the name sits below its 25% bar today.

Which bar applies — the balance-sheet class

Charter is unambiguously in the levered class, so the 25% reference line governs. The computation, from the filed statements:

Source: FY2025 consolidated balance sheet — long-term debt $94,006M, current portion $750M, cash $477M [11]; operating income and D&A from the FY2025 statements of operations [12]. Net debt $94,279M matches the feature file.

Net debt of $94.3 billion against roughly $21.6 billion of EBITDA is 4.4× net debt / EBITDA (4.15× on Charter's own $22.7 billion Adjusted EBITDA basis [13]). Management reports the ratio at 4.54× on its last-twelve-months leverage definition and targets a range it is now moving down to 3.5–3.75× after the pending transactions [14]. Any of these clears the levered threshold (≥3.0×) with room to spare. The feature file leaves the class as "unknown" only because it lacked an EBITDA figure; the primary statements resolve it cleanly to levered.

State the position plainly: 21.0% on FY2025 adjusted FCF against the 25% levered bar — about 400 bps short on the current year; 16.6% on the three-year average — about 840 bps short. The name does not clear its bar today. Whether it does within Ruchir's window turns entirely on the capex normalization below.

Normalized mid-cycle yield — the capex roll-off

Charter's current-year FCF is depressed by a capital-spending peak, and normalizing for it is where judgment sits on top of arithmetic. The depression is not cyclical demand — it is two self-imposed, finite investment programs: the 1.2 GHz / DOCSIS 4.0 network evolution and the subsidized rural construction build ($7.7 billion spent since 2022) [15]. Capex rose from $7.6 billion in FY2021 to $11.7 billion in FY2025 while Adjusted EBITDA still grew — so FCF fell even as the business expanded.

Management states the timing directly: 2025 was the peak year of capital expenditure, and capital expenditures after this year will decline significantly; free cash flow will increase from an already significant amount, with capital intensity returning to 13–14% of revenue by 2028 and run-rate capex below $8 billion once the evolution and expansion programs conclude [16].

The normalization, with every assumption stated so a skeptic can recompute under an adjacent window:

• Operating cash flow: held flat at the FY2025 level of $16,077M. This is conservative — it assumes zero EBITDA growth and zero cash-tax relief over the normalization horizon.
• Normalized capex: $8,000M, the top of management's stated post-program run-rate (they say "below $8 billion"; using the ceiling is the conservative choice).
• SBC: held at ~$700M, roughly the FY2024–25 level.
• Acquisitions: zero, consistent with history.

Normalized adjusted FCF ≈ $16,077M − $8,000M − $700M ≈ $7,377M, a 41.4% yield on today's $17.8 billion market cap — the "~40% after the further fall" the framework associates with this exact name. Even a stressed alternate (operating cash flow falling to $14,500M, capex a higher $8,500M, SBC $700M) yields adjusted FCF of ~$5,300M, or 29.8% — still above the 25% bar. To fail the bar on normalized capex, one has to assume operating cash flow falls roughly 20% while capex stays near its peak — the opposite of what management has committed to and consensus models.

Source: analyst normalization on FY2025 operating cash flow [17] and management's stated post-program capex run-rate below $8B [18]. The 25% levered reference line sits between the stress case and the actual.

The reduction from $11.7 billion of capex in 2025 to below $8 billion by 2028 is, on management's own arithmetic, worth over $28 of free cash flow per share on today's share count [19] — against a $129 share price. This is also the framework's absurdity check: today's $17.8 billion market cap divided by normalized adjusted FCF of ~$7.4 billion retires the entire Class A float in roughly 2.4 years. A price making that claim is under strain.

The consensus check

CapIQ consensus does most of the work of confirming the normalization is not a private theory. The closest vendor proxy for adjusted FCF is consensus free cash flow (mean estimate), which tracks Charter's own levered-equity FCF definition — it does not subtract SBC, so it runs about $0.7 billion rich of the framework's adjusted basis. Naming that gap, the picture is:

Source: consensus FCF mean from the CapIQ estimates feed (data/sp/estimates.json), divided by today's $17.8B market cap; the feature file's consensus_forward_yield carries the gross column. Adjusted-equivalent column deducts ~$700M SBC from each year. Estimates are Charter standalone — management's guidance and the sell-side series exclude the pending Cox acquisition.

On the gross basis the sell side already has Charter clearing 25% every forward year — 27.6% in FY2026 rising to 47.4% by FY2029. On the SBC-adjusted basis, consensus clears the 10% moderate bar immediately, and clears the 25% levered bar by FY2027 (~30.7%), with FY2026 at ~23.7% just under it. The direction is unambiguous, and the mechanism is the same capex roll-off management has committed to — the sell side is not underwriting an operating turnaround, only the completion of a finite build.

So the framework's favorable reading applies: consensus forward FCF clears the bar, which means the setup is fear rather than a market that has concluded the fundamentals are broken. The 70% drawdown and the 3.9× volume spike (see Dislocation) are the fear; the cash-flow trajectory the sell side already models is the fundamentals.

The forward-path underwrite, stated as a probability with its path. The yield clears 25% on an adjusted basis within one-to-three years with roughly 75–80% probability, on this mechanism: capex falls from $11.7B toward a sub-$8B run-rate as network evolution and rural construction conclude by 2028 — a spending decision within management's control and already scheduled — while operating cash flow need only hold roughly flat. What consensus would have to concede for this to fail is an EBITDA decline large enough to eat the entire ~$3.7B capex tailwind. That is the real risk and it is not trivial: broadband subscribers are declining (119,000–120,000 net residential Internet losses per quarter recently) against fixed-wireless and fiber competition, and FY2025 was the first year of revenue decline. If EBITDA erodes faster than capex falls, the normalized yield compresses toward — or through — the 25% line. The 20–25% probability weight sits there. The two pending transactions (Cox, Liberty Broadband) further cloud the standalone forward math: at close the standalone share count rises to roughly 179 million on an as-exchanged basis [20], so per-share and per-market-cap figures will be redrawn even as Cox's cash flow is added — a point developed in Self-Help and Clock.

FCF-to-revenue conversion — the trend that tests the exception

The levered exception the framework grants Charter is conditional on FCF/revenue not deteriorating structurally. The record needs care, because two things move in opposite directions.

Source: adjusted FCF (derived) and revenue from the income statement feed; Adjusted EBITDA from the FY2025 and prior 10-K reconciliations [21]. Both expressed as a share of revenue.

The EBITDA margin is stable-to-rising — 39.9% in FY2021 to 41.5% in FY2025 — so the operating business is not decaying at the margin line. Adjusted FCF/revenue, by contrast, fell from ~15.8% (FY2021) to ~4.6% (FY2024) and recovered to 6.8% (FY2025). That decline is the capex peak passing through, not a margin problem — which is precisely what makes the normalization defensible rather than wishful. Had the EBITDA margin fallen alongside FCF conversion, the levered exception would not hold.

The genuine counter-fact the reader should carry: revenue itself has plateaued and turned down — $55.1 billion in FY2024 to $54.8 billion in FY2025, the first annual decline in the series, with the framework's own three-consecutive-year high-single-digit-decline disqualifier not yet triggered but the trajectory worth watching. Rising EBITDA margins on flat-to-declining revenue is cost discipline, not growth; the flywheel that a high adjusted yield is supposed to power (Self-Help) needs the revenue line to stop falling for the re-rating to be more than a capex-timing artifact. That tension — a plainly cheap adjusted and normalized yield against a top line that has stopped growing — is the honest shape of the yield case here.