Damage Math

Damage Math

Comcast's equity fell 53% — about $93.7B — from its November 2024 peak, while consensus forward free cash flow fell roughly 20% ($16.9B to $13.6B) and FY2026 EPS fell 18.5%. About a third of the drawdown is the cash-flow cut; two-thirds is the market re-rating Comcast from a 9.6% to a 16.5% free-cash-flow yield. That re-rating is warranted only if the decline is durable — the profile's trial puts the probability it is temporary at 0.38.

The near-term hit — the numerator

The trigger is not a single-day guidance cut. Comcast peaked at $45.14 on 6 November 2024 and ground lower over 624 days as a self-imposed broadband repricing collided with fiber, fixed-wireless and satellite competition. FY2025 itself was a strong year on the face of it — revenue $123.7B, diluted GAAP EPS $5.39, adjusted EBITDA $37.4B [1] — but the damage lives in the forward numbers, where consensus marked down the earning power the market capitalizes.

Source: FY2025 reported figures, FY2025 Annual Report (Form 10-K) [2]; forward consensus per CapIQ estimate feed (data/sp/estimates.json, vintage 2026-07-25), derived from fit_features.consensus_forward_yield.

Two lines carry the cut. Normalized EPS troughs at $3.51 in FY2026, down 18.5% from the $4.31 FY2025 actual, then edges back to $3.64 in FY2027 and $3.96 in FY2028 — a shallow V, not a slide, on the current sheet. The forward curve itself was marked down over the last six months: FY2027 EPS fell from $4.10 to $3.64 (−11.2%) and FY2028 from $4.53 to $3.96 (−12.7%); FY2027 revenue slipped 2.0% and FY2028 revenue 3.9%.

Source: CapIQ estimate momentum, 180-day vs current (data/sp/estimates.json, momentum), vintage 2026-07-25 — no backing PDF page.

Free cash flow is the murkier line, and honesty requires flagging why. Management reported FY2025 free cash flow of $19.2B, "the highest year on record," but said plainly it was inflated by roughly $2B of one-time cash-tax benefit that will not recur, favorable working capital, and lower capex — and that the Versant spin-off "removes a significant pool of cash flow" going forward [3]. The vendor feed carries FY2025 FCF at $21.9B and consensus at $16.9B; forward consensus sits at $13.6B (FY2026) and $13.3B (FY2027). So part of the ~$3.3B step down from the $16.9B normalized base is portfolio and timing — one-time tax and the Versant pool — not broadband earning power. The clean numerator is therefore the EPS cut (−18.5%) and the ~20% forward-FCF reset, with the caveat that not all of the FCF drop is the operating problem.

Source: CapIQ consensus FCF (data/sp/estimates.json), via fit_features.consensus_forward_yield; FY2025 shown at the $16.9B consensus/normalized level, not the $21.9B one-time-boosted reported figure.

The price and enterprise-value move, side by side

Consensus FY2026 EPS fell 18.5% and forward FCF about 20%; the market cap fell 53%. That asymmetry is the whole tab. On 3,908M shares at the $45.14 peak, Comcast was worth about $176.4B; at $22.295 on 3,709M shares it is worth $82.7B — a $93.7B loss of equity value, roughly three times the proportional cut to the cash-flow line.

Enterprise value fell less. With about $90B of net debt sitting ahead of the equity — total debt was $90.4B at 30 June 2026 against $98.9B at year-end 2025, and management put net leverage at 2.3× [4] [5] — enterprise value fell from roughly $268B to roughly $165B, about −38%. The debt is a fixed claim, so a 38% fall in the whole firm's value became a 53% fall in the equity beneath it. Leverage amplifies both the damage and, if the cash flows hold, the recovery.

The decomposition matters more than either number alone. At the peak the market capitalized normalized FCF (~$16.9B) at about 10.4× (a 9.6% yield). Today it capitalizes forward FCF (~$13.6B) at about 6.1× (a 16.5% yield). Split the $93.7B drawdown:

Source: derived — peak $176.4B (3,908M shares × $45.14, 2024-11-06) and current $82.7B (fit_features.market_cap); multiple 10.4× normalized FCF at peak, 6.1× forward FCF now (data/sp/estimates.json).

Holding the peak 10.4× multiple and applying only the cash-flow cut takes the equity from $176.4B to about $141.4B — so roughly $35B (37%) of the drawdown is the earnings cut. The remaining $58.7B (63%) is pure re-rating: the market roughly doubled the free-cash-flow yield it demands, from 9.6% to 16.5%. The temporary-versus-permanent question is really a question about that 63%.

How much NPV the problem plausibly destroyed

The re-rating is justified — or not — by what happens to the FCF level after the transition laps. A transparent free-cash-flow-to-equity perpetuity makes the arithmetic visible. Assumptions, stated: base forward FCFE of $13.6B (the FY2026 consensus, already levered — Comcast's free cash flow is struck after cash interest); cost of equity of 9%; terminal growth the variable under test. Because FCFE is post-interest, it is discounted directly to equity value with no further net-debt subtraction.

Source: derived — FCFE perpetuity, base $13.6B × (1+g)/(r−g); base from CapIQ consensus (data/sp/estimates.json). Reproducible from the numbers shown.

The current $82.7B equity value sits below every cell in a plausible 8–12% discount / −3% to +2% growth grid. To reconcile the price mechanically requires either a discount rate near 16% at zero growth, or a permanent FCF decline of about −6% a year at 9% — i.e., the market is pricing Comcast as a melting annuity, not a stable essential. Whether that is right is exactly the diagnosis question.

Cast as the framework's two scenarios:

• Temporary. FCF is depressed by ~$3.3B a year for three years (FY2026–28) while the repricing laps, then reverts to the ~$16.9B normalized base. NPV destroyed = the present value of that three-year shortfall only: $3.3B × (0.917 + 0.842 + 0.772) ≈ $8B. Intrinsic value barely moves; the ~$94B price move is a ~12× overshoot.
• Permanent (conservative). FCF resets permanently ~$3.3B lower and grows 2%. NPV destroyed = $3.3B ÷ (0.09 − 0.02) ≈ $47B. Real, large — but still short of the $93.7B the price took out, because a level shift at a 9% discount does not, by itself, justify also doubling the yield.
• Permanent (as priced). The market has done both — cut the level and re-rated the yield to ~16.5% — which is internally consistent only if the FCF base keeps declining (g near −6%). On that reading the ~$94B price move is roughly fair and the gap closes.

Source: derived — perpetuity workings above (r = 9%, g = 2%); price damage $93.7B from fit_features.market_cap and the 2024-11-06 peak.

The trial — temporary or permanent, at comparable depth

The diagnosis was argued by two opposing corpus-cited briefs and ruled on by three blind judges. Both cases are strong; the ruling, not this tab, carries the probability.

The case for temporary. The impairment is a chosen, time-boxed transition, not demand destruction. The CFO framed the pain as investment: "we expect incremental EBITDA pressure over the next couple of quarters until we begin to lap these initial investments in 2026," alongside a 4.5% Connectivity and Platforms EBITDA decline the size of a reinvestment, not a collapse [6]. The leading indicators have inflected: broadband net losses improved 34,000 year-over-year to 167,000 in Q2 FY2026 on the new go-to-market strategy [7], with management guiding to "modest improvements starting in the third quarter" [8]. The offsetting engine is scaling — a record 448,000 wireless net adds [9] against only 7% penetration of the footprint, a long runway [10] — cash generation never broke, with Q2 FY2026 FCF of $4.6B [11], and the media drag turned: Peacock posted its first-ever quarterly profit, $189M [12].

The case for permanent. The impairment is the breakdown of the cable compounder, which for years offset video attrition with broadband rate and buybacks. The new record shows the opposite: domestic broadband revenue fell "due to decreases in average rates and declines" in customers — price and volume together, not a chosen price cut alone [13]. The earnings core is shrinking: Connectivity and Platforms revenue fell 3.0% and adjusted EBITDA 5.7% in Q2 FY2026, with residential EBITDA down 8.0% and its margin down 160 basis points [14]. Broadband penetration of passings fell to 47.9% from 49.7%, video customers fell another 280,000 in the quarter, and the wireless offset is promotional — half of residential connects came from free lines [15]. Linear networks "continue to experience accelerated net customer losses" [16], consensus marked the forward EPS and revenue curves down, and the buyback bridge is weaker: Comcast suspended repurchases at the start of Q3 2026 [17].

The ruling. The judges put the probability the impairment is temporary at 0.38 (mean of the panel 0.44), with a spread of 0.19 across seats (0.37 / 0.38 / 0.56) and the ruling recorded as not contested. Reading order moved the panel about 0.10 — judges who read the permanent brief first landed higher on temporary — but not enough to flip the finding. On the panel's math the diagnosis leans toward a durable reset: roughly 62% weight on permanent, 38% on temporary. That ruling stands as the report's diagnosis; nothing in this tab's own arithmetic overrides it.

The buyback suspension deserves one clarifying line, because it cuts both ways: it was declared "in connection with the proposed NBCUniversal Spin-off," not forced by distress [18]. Whether that is a temporary, technical pause or the removal of the flywheel exactly when a 16.5% yield makes it most valuable is taken up in the Self-Help tab.

Which line broke — and whether it self-corrects

The break is in Residential Connectivity and Platforms, and specifically domestic broadband. Broadband ARPU fell 3.8% in Q2 FY2026 as Comcast withheld its usual rate increase and layered in free wireless lines [19], while the subscriber base fell 167,000 and penetration slipped to 47.9% [20]. Price and volume fell at once, so residential EBITDA fell 8.0% and its margin 160 basis points [21]. Video keeps draining underneath it — another 280,000 customers lost in the quarter [22].

The repricing mechanism that would self-correct is specific and dated: the reinvestment stops recurring once the base rolls onto simplified plans, which management dates to a lap "in 2026" [23]; free wireless lines convert to paying relationships after a year and convergence scales from a 7% penetration base [24]; losses are already narrowing rather than widening [25]. The structural case against is equally specific: broadband is losing customers to fiber, fixed wireless and satellite in an "increasingly competitive environment" that is not a chosen price cut, and video decline is accelerating [26] [27]. The evidence that would decide it is the same evidence the trial named: broadband net adds turning positive without further ARPU decline for several quarters after the 2026 lap, and forward FCF revised back up toward the ~$18–19B run-rate rather than holding at ~$13.5B. The repricing anatomy runs alongside the Dislocation drawdown analysis; the durability of the broadband franchise on the long clock is taken up in Durability.