T — AT&T Inc.

As of 2026-09-13, the tape is constructive but not euphoric: RSI(14) at 61.94, price 9.20% above the 50-day SMA, +7.46% over one month and +10.52% over three months, all following the 2026-07-22 Q2 beat (revenue $31.56B, +2.3% YoY; 1M+ advanced connectivity adds) and the announcement of an accelerated ~$10B 2026 buyback. What caps sentiment is that the hard analyst data is only a 'Hold': 30 of 62 ratings are hold with 6 sells, the consensus price target is $26.53 (average recent target $25.25, i.e. below the $26.05 price) against a $30 high and $20 low. Note a real discrepancy with the live research brief, which frames consensus as 'Buy/Outperform' with $28-$30 targets and a $36 high - FMP's panel does not support that; the market appears to view T as roughly fully valued here, not as a re-rating story.

Twelve quarters of data show a stable, low-growth business rather than an inflection. Revenue went from $30.35B (Q3 2023) to $31.56B (Q2 2026), roughly +4% cumulatively; YoY in Q2 2026 was +2.3% ($31.558B vs $30.847B), consistent with the low-single-digit range throughout. The better trend is underneath: Q2 2026 operating income of $7.038B was +8.3% YoY (vs $6.501B), lifting operating margin from 21.1% to 22.3%, and diluted EPS rose from $0.62 to $0.66 (+6.5%), following $0.61 (Q1 2026) and $0.54 (Q1 2026 vs $0.61 in Q1 2025). Two quarters are distorted and should not be read as trend: Q3 2024 printed -$0.03 EPS (a $174M net loss) and Q3 2025 printed $1.29 EPS on $9.314B net income against revenue of only $30.7B - a clear one-off gain, not operating strength. Free cash flow is the soft spot: TTM FCF through 2026-06-30 is $17.654B versus $19.624B a year earlier, down ~10%, driven by a weak Q1 2026 ($2.68B FCF on $7.557B OCF vs $4.877B capex). Q2 2026 FCF rebounded to $5.17B (+6.2% YoY vs $4.866B), so the FCF decline looks like a Q1 capex/working-capital blip rather than a structural break - but it needs one more clean quarter to confirm. Earnings quality is genuinely good: the accruals ratio is -1.46%, meaning cash generation exceeds reported net income.

The headline numbers look cheap but are partly an illusion. As of 2026-09-13, T's TTM P/E of 8.60x (EPS $3.098) is flattered by the Q3 2025 one-off gain; normalize it out and TTM EPS is closer to $3.01 with roughly $0.60-$0.66 of that being the distorted quarter. On the forward number the stock is not cheap: at $26.05 versus next-FY consensus EPS of $2.339, forward P/E is 11.14x - above the stock's own 5-year average P/E of 9.67x. Price/sales tells the opposite story from price/earnings: 1.403x versus a 5-year average of 1.17x is rich. Cash-flow-based valuation is where real value shows: EV/EBITDA of 6.07x (implied EV of ~$295.6B on $48.7B EBITDA) is at the low end for large-cap telecom, and TTM FCF of $17.65B on a $178.5B market cap is a 9.9% equity FCF yield. Net: fair-to-mildly-attractive on cash flow, unremarkable on forward earnings, and the 5.25% dividend yield is the anchor. The deterministic overlay's 'Neutral vs. its own history' label is the right call, and the 0.119 PEG ratio is a data artifact of near-zero growth, not a buy signal.

As of 2026-09-13, the market is paying ~11.1x forward EPS for ~2% revenue growth and mid-single-digit EPS growth - a multiple that already assumes the fiber/5G convergence strategy keeps working and FCF stabilizes. There is no 'growth the market hasn't noticed' argument here: the stock is +7.46% in a month and +10.52% in three, so the Q2 beat has been priced. The sell-side agrees: the $26.53 consensus target sits essentially on the $26.05 price, and the average recent target of $25.25 is below it. Realistic paths to upside are narrow: (1) sustained EPS beats that force estimate revisions above $2.339, (2) the ~$10B buyback shrinking a 6.852B share count fast enough to add 5%+ to EPS on flat revenue, or (3) multiple expansion on lower rates. The main downside path is equally simple - if the 2026-10-21 Q3 report shows slowing postpaid/fiber adds or FCF that fails to hold the ~$5B quarterly run-rate, there is no valuation cushion at 11x forward with $134.6B of long-term debt behind it.

As of 2026-09-13, RSI(14) stood at 61.94 - firm momentum but not yet overbought (no >70 reading). The stock was 2.02% above its 20-day SMA (implied ~$25.53), 9.20% above its 50-day SMA (implied ~$23.86), and 3.64% above its 200-day SMA (implied ~$25.14), a fully bullish moving-average stack. The catch is the extension: +9.2% above the 50-day is the stretched part of the setup, and it comes after a +10.52% three-month move with a still-negative six-month return (-5.99%), so this is a recovery off the $19.89 low rather than a fresh multi-year uptrend - the 52-week high of $29.79 is ~14% away. Entry-timing implication: chasing at $26.05 with RSI near 62 and price 9% above the 50-day is poor risk/reward; a pullback toward the $25.10-$25.55 zone (200-day/20-day SMA) offers a materially better entry for anyone who wants the 5.25% yield, and the picture only turns genuinely negative on a decisive break below the 200-day.

• Interest rates: T carries $134.6B of long-term debt against $17.6B of cash (net debt ~$117B), so refinancing costs flow straight to FCF; with a 5.25% dividend yield the stock also competes directly with bond yields for income capital, making it highly sensitive to the direction of the 10-year Treasury.
• Bond-proxy sector rotation: a beta of 0.417 makes AT&T one of the least economically sensitive names in the S&P 500 - it tends to be bought in risk-off/rate-cutting rotations and sold when cyclicals and growth lead, which can override company-specific results for quarters at a time.
• LEO satellite and direct-to-device competition: Starlink-style mobile and direct-to-device offerings remain the market's main long-term structural worry for terrestrial wireless and rural broadband margins, and it is a headline risk that periodically compresses telecom multiples regardless of current subscriber trends.
• Capital intensity and spectrum spend: the EchoStar spectrum license acquisition closed 2026-07-28, and quarterly capex ran $5.70B in Q2 2026 ($4.877B in Q1 2026) - a permanent cash drain that caps how much FCF can be converted into dividends and buybacks.
• Regulation and competitive intensity: a mature three-player-plus-cable wireless market and price competition in fixed wireless/fiber limit pricing power, while any change in broadband subsidy or spectrum policy is a sector-wide swing factor.

• Q3 2026 earnings on 2026-10-21 - the next hard checkpoint for postpaid phone net adds, fiber expansion milestones, and whether FCF holds the ~$5B+ quarterly run-rate after the weak $2.68B Q1.
• Execution of the accelerated ~$10B 2026 share repurchase program - with only ~2% revenue growth, share-count reduction is the single largest per-share EPS lever management controls.
• Continued fiber/5G convergence bundling momentum - Q2 2026 delivered 1M+ advanced connectivity additions (fiber, fixed wireless, postpaid phone), which supports churn and pricing power if the pace continues.
• Deleveraging and refinancing at lower rates, or the possibility of a dividend increase - either would be read by the market as a signal that the balance sheet and FCF are on durable footing.
• Analyst estimate revisions above the current $2.339 next-FY EPS consensus, which at 11.1x forward is the most direct route to any real upside from $26.05.

• Leverage and returns on capital: $134.6B of long-term debt versus $110.4B of stockholders' equity (D/E 1.475), with ROIC of 5.8% - below most reasonable estimates of AT&T's WACC. The company earns less on incremental capital than that capital costs, which means growth alone does not create value without ROIC improvement.
• Free cash flow trend and capital-return sustainability: TTM FCF of $17.65B is down ~10% YoY, and a ~$9.4B annual dividend (implied by the 5.25% yield) plus a ~$10B buyback totals roughly $19B - modestly above TTM FCF, meaning the buyback leans on FCF recovery, working capital, or balance-sheet capacity.
• Valuation is not the cushion it appears: the 8.60x TTM P/E rests on a quarter (Q3 2025) with $9.314B of net income that was not operating earnings; on forward earnings T is at 11.14x versus its own 9.67x 5-year average P/E, while the consensus target of $26.53 implies no upside - a stock already priced for good news.
• Structural competitive threat from LEO satellite direct-to-device and cable fixed wireless, which could pressure the wireless and broadband economics that the fiber convergence strategy depends on.
• Thin liquidity on the balance sheet: current ratio 0.971 and quick ratio 0.927, with $53.4B of current liabilities against $51.8B of current assets as of 2026-06-30 - a business that depends on continuous capital markets access to roll debt.

At $26.05 on 2026-09-13, I would not be a buyer - this is a decent, cash-generative income vehicle that is already priced for good news. The bearish-leaning framework call is not just a formula artifact: revenue grew 2.3%, ROIC of 5.8% sits below the cost of capital, TTM free cash flow fell ~10% YoY, and the eye-catching 8.60x P/E is inflated by a one-off gain in Q3 2025 - strip it out and the stock trades ~11.1x forward earnings, above its own 5-year average of 9.67x, after a 10.5% three-month rally, with a $26.53 consensus target that is essentially today's price. My own blended fair value is about $26.60, so the honest read is 'roughly fair, no margin of safety.' The 5.25% dividend is well-covered by FCF and the ~$10B buyback will quietly compound per-share earnings, so existing holders collecting the yield have no urgent reason to sell - but new money is being asked to pay full price for low-single-digit growth and a highly levered balance sheet, and a pullback toward the $25.10-$23.90 support zone would offer far better risk/reward.

Equal-weighted blend of two approaches using only the 2026-09-13 data: (1) 10.75x forward P/E on next-FY consensus EPS of $2.339 = $25.14/share, justified as a modest premium to AT&T's 9.67x 5-year average trailing P/E because that trailing figure is flattered by the Q3 2025 one-off gain while H1 2026 EPS is already running at $1.20; and (2) 6.35x EV/EBITDA on the ~$48.7B EBITDA implied by the given 6.071x EV/EBITDA and a ~$295.6B enterprise value, giving $309.2B EV less ~$117.1B net debt ($134.6B long-term debt minus $17.6B cash) = $192.1B equity, or $28.03 per share on 6.852B shares. Average of $25.14 and $28.03 is $26.59, rounded to $26.60 - roughly fair value versus the $26.05 price, implying no margin of safety.