The 7.6x P/E is diagnostic of a value trap, not a mispricing. Revenue has been flat for four consecutive years ($121.4B → $121.6B → $123.7B → $123.7B), operating income dropped 11% from 2024 to 2025 ($23.3B → $20.7B), forward EPS growth is only 14.7% above TTM, and the PEG ratio is ~143x — signaling the market prices earnings growth as functionally zero.

• Revenue flat at $121.4B → $121.6B → $123.7B → $123.7B for four consecutive years
• Operating income dropped 11% from 2024 to 2025 ($23.3B → $20.7B)
• Forward EPS $3.58 vs TTM EPS $3.12 (14.7% growth above depressed base)
• PEG ratio of ~143x
• Dividend yield 5.4% matches 10-year Treasury yield, providing no equity risk premium

↩ rebuts: “The 'obscene' valuation being called a bond with equity optionality”

FCF math doesn't support the bull case. TTM FCF of $19.2B was inflated by working capital benefits post-Versant spinoff; normalized run-rate is ~$12.7B. Capital returns (dividends $4.9B + buybacks $7.2B = $12.1B) fully consume FCF, leaving only $600M-$1.6B annually for deleveraging. Additionally, $14-15B in annual CapEx is non-discretionary — the company is spending to defend a depreciating asset base against fiber overbuilders.

• TTM FCF of $19.2B inflated by working capital benefits post-Versant spinoff
• Normalized FCF ~$12.7B (2023 and 2024 both showed this level)
• FCF yield at normalized $12.7B on $84B market cap = ~15.1%
• Dividends $4.9B + buybacks $7.2B = $12.1B annual capital returns
• Net capital return burn leaves only $600M-$1.6B annually for deleveraging
• $14-15B annual CapEx is non-discretionary
• Interest expense $4.4B annually on $90B of debt
• Every 100bps rate increase = ~$900M incremental interest expense or ~$0.25 EPS dilution

↩ rebuts: “FCF/Market Cap of 20.1% being celebrated as 'self-funding through any rate environment'”

Hawkish Fed regime is hostile to leveraged cash-flow stories. Higher-for-longer rates means multiple compression, not expansion. Beta of 0.66 means less upside in recoveries while defensive names don't re-rate in hawkish environments. The 54bps equity risk premium over Treasuries doesn't compensate for operational leverage, execution risk, and secular decline. Recession probability ticked from 7% to 10% in one week, and CMCSA's consumer-adjacent exposure (theme parks, advertising, ARPU-sensitive broadband) makes it vulnerable.

• 10-year Treasury yields ~5% vs CMCSA 5.4% yield = only 54bps equity risk premium
• Beta of 0.66
• 95% probability on Polymarket of no rate cuts
• Recession probability ticked from 7% to 10% in one week
• Interest expense $4.4B annually on $90B debt

↩ rebuts: “Macro 'tailwind' argument for defensive rotation”

Broadband losses are structural, not cyclical. The CFO publicly used the word 'irrational' to describe competitor pricing — the language of a company conceding price competition it cannot win. T-Mobile and Verizon added over 1 million fixed wireless subs in their last reported quarters combined. Frontier, AT&T Fiber, Brightspeed, and regional fiber players are passing 3-4 million new homes per year. The overbuilders are being absorbed by carriers (Verizon buying Frontier, T-Mobile buying Lumos), not Comcast. Comcast is on the wrong side of consolidation. Fixed wireless and fiber players have lower debt-to-EBITDA and better cost structures than Comcast on a per-passing basis.

• CFO used word 'irrational' to describe competitor pricing on public conference call
• T-Mobile and Verizon added over 1 million fixed wireless subs combined in last reported quarters
• Frontier, AT&T Fiber, Brightspeed, and regional fiber players passing 3-4 million new homes per year
• Verizon buying Frontier
• T-Mobile buying Lumos
• Overbuilders have lower debt-to-EBITDA than Comcast on per-passing basis
• Xfinity Mobile runs on Verizon's network — Comcast pays wholesale rates

↩ rebuts: “167K subscriber loss being dismissed as '0.5% of base' and cyclical noise”

The bear's SOTP math shows the split destroys value. Using the bull's own numbers: Connectivity worth $30B EV at 6x EBITDA on $5B EBITDA, plus NBCU/Sky worth $28B EV at 7x EBITDA on $4B EBITDA = $58B combined EV. Add net cash adjustments = $70-80B equity value versus $84B current market cap. That's a 5-17% discount, not a premium. The split destroys value through stranded corporate costs (~$500M-1B), duplicated G&A, tax friction, lost synergies, and multiple compression on underperforming standalone entities.

• Bull's SOTP: Connectivity $30B EV at 6x EBITDA on $5B EBITDA
• Bull's SOTP: NBCU/Sky $28B EV at 7x EBITDA on $4B EBITDA
• Combined EV $58B + net cash adjustments = $70-80B equity value
• Current market cap $84B = 5-17% premium to SOTP value
• Stranded corporate costs ~$500M-1B
• CFO cited 'diverging competitive conditions, investment needs and strategic priorities' as split driver

↩ rebuts: “SOTP unlocking $70-80B in equity value versus $84B today”

Technical analysis shows a confirmed downtrend, not a basing pattern. Lower highs sequence: April 23 high $31.65, August 26 high $27.20 (–14%), September 2 high $26.81 (–1.4%). Price below lower Bollinger Band with expanding bands indicates trending breakdown, not exhaustion. MACD flipped negative with widening histogram — bearish momentum confirmation. RSI at 34.8 with no bullish divergence (RSI making new lows with price). September 9 sell day had 47M shares — over 2x average — indicating institutional distribution, not accumulation.

• April 23 high: $31.65
• August 26 high: $27.20 (–14%)
• September 2 high: $26.81 (–1.4%)
• June low: $21.83
• Recent low: $23.73
• Below lower Bollinger Band with expanding bands
• MACD negative with widening histogram
• RSI at 34.8
• September 9 sell day: 47M shares (over 2x average)

↩ rebuts: “Technical basing pattern argument”

AI edge and Equinix partnerships are rounding errors, not re-rating catalysts. Fastly (Comcast's partner) has market cap under $1B and has been struggling to grow. Comcast is entering the edge market as a reseller and last-mile provider — they don't have the software stack, developer ecosystem, or AI/ML inference tooling. Cloudflare ($50B+), Akamai (~$11B), and hyperscaler CDNs dominate. The Equinix integration is a carrier partnership where Comcast provides last mile, receiving only a sliver of revenue against a $12-13B FCF base.

• Fastly market cap less than $1B
• Cloudflare market cap $50B+
• Akamai market cap ~$11B
• Hyperscaler CDNs (AWS, Azure, Google) dominate enterprise edge spend
• Comcast providing only last-mile connectivity in Equinix partnership
• ~$1-2B revenue opportunity at best against $12-13B FCF base

No institutional accumulation signal. Zero insider purchases in past 6 months — insiders are selling, not buying. Tape shows steady grinding lower on elevated volume, not the block trade signature of stealth accumulation. Analysts have been steadily cutting price targets; consensus 12-month target is below where stock traded 6 months ago.

• Zero insider purchases in past 6 months
• Steady grinding lower on elevated volume
• No heavy block trades at premium prices
• Analysts steadily cutting price targets

↩ rebuts: “Quiet retail accumulation argument”

Probability distribution shows negative expected return. Bear case (45%): broadband losses accelerate, technical breakdown to $20-21 = –10-15%. Base case (40%): sideways drift $22-25 with dividend maintenance = –5% to +5%. Bull case (15%): split catalyst + Fed pivot + broadband stabilization = +20-30%. Joint probability of all three bull case requirements is under 3%. Expected value: roughly –3% to –7% over 12 months, plus 5.4% dividend = negative to flat total expected return.

• Bear case probability 45% → $20-21 (–10-15%)
• Base case probability 40% → $22-25 (–5% to +5%)
• Bull case probability 15% → +20-30%
• Expected value: –3% to –7% over 12 months
• Total expected return including dividend: negative to flat

↩ rebuts: “Bull's 50% probability assigned to $30-32 base case”