CMCSA — Comcast Corporation

As of 2026-09-13 the tape is the message: $25.20 is 16.45% below where it stood six months earlier, -4.87% in the last week alone, and 6.63% under the 200-day SMA, with the stock sitting in the lower third of its $21.28-$32.86 52-week range. The Q2 2026 print (filed 2026-07-23) showed revenue down 1.2% YoY to $29.94B, confirming that traditional broadband/video is still shrinking, and the research brief notes recent downward revisions to price targets as analysts mark down persistent top-line headwinds. Sector sentiment also took a hit from SpaceX's Starlink outlining terrestrial mobile ambitions, which pressures the cable-bundling economics that Comcast's wireless push depends on. The offsetting positives - a $1.04 adjusted EPS beat vs. $0.97 consensus, Peacock's first-ever profitable quarter, a record 10.2M wireless lines - have not been enough to change the direction of the tape, and the stock is trading below the $28.33 average price target by a wide margin.

Revenue is decelerating into outright decline. TTM revenue through 2026-06-30 was $124.9B versus $124.2B a year earlier (+0.6%), and the two most recent clean comps were 2026 Q2 at -1.2% YoY ($29.94B vs. $30.31B) and 2025 Q3 at -2.7% ($31.20B vs. $32.07B); the +5.3% print in 2026 Q1 ($31.46B vs. $29.89B) reads as an anomaly, since Q2 immediately reverted to negative. The margin trend is clearer and worse: summing each year's four quarters, operating margin fell from ~18.8% in 2024 to ~16.7% in 2025 to 14.7% TTM (operating income $18.31B on $124.9B revenue). Quarterly detail confirms it - 2026 Q2 operating margin of 17.2% vs. 19.8% in 2025 Q2 (-2.5 pts), and 2026 Q1 at 13.1% vs. 18.9% (-5.8 pts) - so this is roughly 200bps of compression per year for two straight years, not noise. Reported EPS is distorted by an outsized 2025 Q2 (net income $11.1B, $2.97 diluted EPS), which makes any trailing comparison look like a collapse; the clean recent run-rate is $0.60-$0.99 per quarter, with Q2 2026 adjusted EPS of $1.04 beating the $0.97 consensus. Cash generation is the genuine bright spot and the one line inflecting positively: quarterly free cash flow has been $4.5B-$5.6B for six consecutive quarters (TTM $20.4B), versus a lumpy 2024 that ranged from $1.3B to $4.5B. Net: a structurally challenged top line with no margin floor yet visible, offset by meaningfully improved cash conversion.

Cheap, but cheap for identifiable reasons. As of 2026-09-13 the stock trades at 8.18x TTM EPS ($3.137) and 0.716x TTM sales, against five-year averages of 14.18x and 1.35x - roughly a 42% and 47% discount to its own history. On next-FY consensus EPS of $3.508 the forward P/E is only ~7.2x. EV/EBITDA is 5.06x and trailing FCF yield is ~22.9% ($20.44B FCF on an $89.4B market cap), with the 5.56% dividend covered at roughly a quarter of free cash flow. PEG of -0.166 is negative and therefore meaningless here - the formula needs a positive growth rate, which this company does not currently have on the top line. What the multiple is discounting is real: negative revenue growth, ~200bps/yr margin erosion, $84.3B of long-term debt, a 0.796 current ratio, and secular broadband share loss to fiber overbuilders and fixed wireless. I would call it cheap-versus-history with a legitimate value-trap risk, not an obvious mispricing.

The market and the sell-side disagree, and that gap is the whole setup. Consensus next-FY EPS of $3.508 implies roughly 12% growth off the current run-rate, and the 11-analyst price target average is $28.33 (low $23, high $36) - yet the stock trades at ~7.2x forward earnings, a multiple that prices in permanent EPS erosion rather than growth. Either the Street's estimate is stale and gets cut toward the low-target case, or the market is over-discounting broadband risk and 9x forward EPS alone gets you to $31.6. Two datapoints feed the bull case: the Q2 2026 adjusted beat ($1.04 vs. $0.97) and Peacock's first profitable quarter. The counterargument is the tape itself - a stock that has de-rated from 14x to 8x earnings is usually right about something. Trade the Oct 29 print (consensus ~$1.00 EPS, ~$29.5B revenue) and the broadband net-add line inside it, not the narrative on either side.

As of 2026-09-13, RSI(14) stood at 44.71 - neutral, not oversold, so no washout signal. The $25.20 price sat 4.32% below its 20-day SMA (~$26.34), 0.98% above its 50-day (~$24.96), and 6.63% below its 200-day (~$26.92), inside a 52-week range of $21.28-$32.86. Performance: -4.87% over one week, -1.06% over one month, +2.86% over three months, -16.45% over six months. That is a lower-high downtrend contained within a wide range, not a capitulation. Timing implication: there is no technical urgency to buy here. A flush toward the low-$20s with RSI in the low 30s would be the higher-quality entry; alternatively, a reclaim of the 200-day around $27 would be the first real evidence of a trend change. Buying at $25.20 means buying mid-range with momentum still pointed down.

• Broadband competition is the dominant macro/sector force: fiber overbuilders and fixed-wireless access are compressing ARPU and subscriber growth, and Starlink's stated terrestrial mobile ambitions (via spectrum acquisitions) add a long-tail threat to cable bundling economics.
• Interest rates: moderating rate expectations relieve some refinancing pressure on $84.3B of long-term debt and support capital-intensive infrastructure spend, but with a 1.007 debt-to-equity ratio and a 0.796 current ratio there is very little balance-sheet cushion if rates reverse.
• Consumer spending: persistent household belt-tightening pressures the discretionary parts of the portfolio - theme parks, theatrical box office, and premium multi-service cable bundles - where Comcast has no pricing power to offset volume weakness.
• Sector rotation: capital continues to favor AI-adjacent and high-growth names over defensive high-yield media/telecom, which is consistent with the -16.45% six-month move despite a stable dividend and cheap multiple.
• Regulatory and structural: any separation of NBCUniversal/Sky assets raises tax, regulatory and execution questions, while recurring retransmission/affiliate-fee and content-licensing negotiations remain a swing factor for the Content & Experiences segment.
• Media advertising cyclicality: NBCUniversal's network and streaming ad inventory is tied to broader ad-spend cycles, so an advertising downturn would hit the segment already absorbing margin compression.

• Q3 2026 earnings on 2026-10-29 (consensus ~$1.00 EPS on ~$29.5B revenue) - the swing factor is broadband net additions and ARPU, not the headline EPS.
• Concrete structure, timeline and tax treatment for the separation of the media & technology segments (NBCUniversal/Sky) - a value-unlock event if executed cleanly, a distraction if not.
• Durability of Peacock's first profitable quarter plus continued wireless scaling (record 10.2M lines) as the bundling offset to broadband/video losses.
• Capital return: $20.44B of TTM free cash flow funds buybacks and supports growth off the 5.56% dividend yield; a raise or accelerated repurchase would be a signal.
• Competitive response to fixed-wireless and Starlink - any evidence that broadband share loss is decelerating would trigger a multiple re-rating faster than an earnings beat would.

• Structural (not cyclical) broadband erosion from fiber overbuilders and fixed wireless - this is the core reason the stock trades at 7x forward earnings, and it has no visible fix.
• Margin compression with no floor yet: operating margin of ~18.8% (2024) to ~16.7% (2025) to 14.7% TTM, including -5.8 pts YoY in 2026 Q1 alone.
• Leverage and liquidity: $84.26B long-term debt against $7.66B cash, 1.007x debt-to-equity, 0.796 current ratio, and $33.1B of current liabilities versus $26.3B of current assets.
• Consumer-spending sensitivity across theme parks, theatrical and advertising, which are the higher-multiple parts of the portfolio.
• Value-trap risk: the discount to the 14.18x five-year average P/E can persist indefinitely while consensus estimates of $3.508 are cut toward the $23 low-target case; the brief already notes downward target revisions.

I would not buy Comcast at $25.20 as of 2026-09-13. The valuation is genuinely cheap - 8.2x trailing and ~7.2x forward earnings, a 5.06x EV/EBITDA, a ~22.9% FCF yield and a 5.6% dividend covered at roughly a quarter of free cash flow - and Peacock's first profitable quarter plus a record 10.2M wireless lines show the cash engine still works. But revenue is shrinking, operating margin has fallen roughly 410bps in two years, and the broadband pressure driving it is secular competitive share loss rather than a cyclical trough, so a low multiple is not by itself a reason to step in. Existing holders can collect the dividend; new capital should wait for broadband stabilization or a flush toward the $21.28 52-week low.

Primary method: 8.5x forward EPS - a premium to the current ~7.2x forward multiple but still a ~40% discount to the five-year average P/E of 14.18 - applied to next-FY consensus EPS of $3.508, giving $29.82. Cross-checks: 5.7x EV/EBITDA on ~$32.8B of EBITDA (EV of ~$187B less $76.6B net debt, on ~3.55B shares) yields ~$31.1, and an 18% required FCF yield on $20.44B of TTM free cash flow yields ~$32.0. I weight the earnings method highest because consensus EPS carries modest downside risk given the margin trajectory, landing at $29.80 - roughly 18% above the $25.20 price of 2026-09-13 and marginally above the $28.33 analyst consensus.