COMCAST CORP

COMCAST CORP (CMCSA) — Investment Memo

🐂 The Bull Case (Warren's voice)

• Moat is real and sticky. 32 million broadband subscribers won’t switch because of hassle, not loyalty — and 2.5 million business locations have multi-year contracts with penalties. That’s a toll bridge with zero marginal cost per byte.
• Recurring revenue dominates. ~81% of revenue from connectivity (broadband + business services) — predictable, high-margin, inflation-protected. Monthly tolls compound with price increases (revenue up $6.2B in 2025 even with flat subs).
• Free cash flow machine. FCF of $21.9B in 2025 exceeds net income of $20.0B — no accounting fudge. Capital-light operating model (already have the pipes) generates steady cash to buy back shares and pay a $1.24/share dividend.
• Pricing power is underappreciated. Broadband ARPU rose ~7% in 2025 while subscriber losses were negligible. Inertia is a powerful pricing tool. Even if 10% of subs leave, the remaining 90% will pay more.
• At a low enough price, the moat becomes cheap. If the market panics over cord-cutting and fiber overbuilds, Comcast’s cash flows become a screaming bargain. At $50/share (half the DCF of $99.36), the yield on FCF is ~11% — a classic Buffett entry.

🐻 The Bear Case (Charlie inverts)

• The 2–3 permanent impairment scenarios:

  1. Fiber-to-home overbuild. AT&T or Verizon builds fiber to 60%+ of Comcast’s footprint. Switching costs vanish when a competitor offers symmetrical gigabit for the same price. Comcast loses 20–30% of broadband subs → revenue down $15B+, fixed programming costs of $24.5B don’t shrink → margin collapse.
  2. Fixed wireless becomes “good enough.” T-Mobile/Verizon 5G fixed wireless reaches 300Mbps at $50/month. 32 million residential subs realize they don’t need a truck roll — they just plug in a modem. No installation barrier, no contract. Churn accelerates.
  3. Content cost death spiral. Sports rights keep inflating (NFL, NBA, Olympics). Peacock loses $2.5B+/year (implied from programming costs). If broadband revenue stalls, the $24.5B programming anchor drags free cash flow to zero. Comcast becomes a cash incinerator.

• Most likely scenario over 5–10 years: Fiber overbuilds push broadband price competition. Comcast’s pricing power breaks — they can still raise prices, but they lose 2–3% of subs per year. Revenue flat, margins compress from 16% net to 10–12%. FCF drops from $22B to $15B. The DCF of $99.36 assumes 7% FCF growth — unrealistic given flat revenue and rising competition.

• Management risk is structural. Roberts family controls 33.3% voting power through dual-class stock. They can make bad capital decisions (like the $40B Sky acquisition that likely destroyed value) without accountability. The 2022 earnings crater ($14.2B → $5.4B) was never explained — that’s a trust fail.

💰 Valuation & Margin of Safety

• Intrinsic value estimate: $99.36/share (DCF: 7.1% FCF growth, 10% discount, 3% terminal). But this assumes the moat holds — which is optimistic given structural threats.
• 25% margin of safety entry: $74.52/share — still above current market price (~$40–45 range). At this level, you’re paying 12x FCF. Not enough cushion for a moat that’s leaking.
• 50% margin of safety entry: $49.68/share — this is where the math gets interesting. At that price, the FCF yield is 10%+ and the downside is partially priced in. Buffett would buy here, but only if he believed the terminal moat was intact.
• Current valuation: The DCF implies ~2.2x premium to market price (if CMCSA is ~$45). But the DCF is probably too high — real intrinsic value (adjusting for slower growth and higher risk) is closer to $60–70/share. At $45, it’s not a clear bargain. Fairly valued, not cheap.

Verdict: WATCH — Wait for a deeper discount or clearer evidence that the moat isn’t breaking.

At $45/share, Comcast is neither obviously cheap nor expensive — but the risks (fiber overbuild, fixed wireless, management dual-class) are real and unhedged. If the stock drops to $35–40 (a 20%+ decline), it becomes a compelling buy for the cash flows alone. Until then, the inversion machine says watch and wait. The moat is real, but it’s narrowing — and that’s not a Berkshire purchase.