AT&T at $23.40 as 4.78% Dividend Meets a 4.731% Treasury — Buyback Raised to $10 Billion

AT&T traded in a $23.15 to $23.62 band on Friday against a prior close of $23.94, opening at $23.41 and holding roughly 2.2% lower into the final session of July. Market capitalization sits at $159.13 billion across 6.85 billion shares outstanding.

The round trip since the earnings print tells the whole story. AT&T reported second-quarter results on July 22, beat on the bottom line, and ripped 4.8% to close at $24.06 on more than 80 million shares — roughly 57% above the average daily turnover. KeyBanc's Brandon Nispel upgraded the stock to Overweight from Sector Weight with a $30 target, arguing satellite fears were overdone. Then a Semafor report landed describing SpaceX plotting a direct assault on the U.S. wireless oligopoly, and T fell 4% alongside Verizon while T-Mobile slid 2%.

Every dollar of the earnings gain is gone. The stock sits closer to its 52-week low of $19.89, printed July 2, than to the $29.79 high set September 15, 2025 — a 21% drawdown from that peak with the stock roughly 17% above its July trough.

The broader tape offered no help. The S&P 500 traded 7,424, down 0.18%, and the Dow held 52,155 as Treasury yields ripped. The 10-year jumped almost 7 basis points to 4.731% with an intraday print at 4.737%, the highest since January 2025, and the 30-year surged to 5.263%. For a stock with a beta of 0.17 and a 4.78% dividend yield, the long end is the primary valuation input, and it moved directly against the name.

The dividend math is now uncomfortably tight. AT&T pays $0.2775 quarterly, $1.11 annualized, for a 4.78% yield. The 10-year Treasury pays 4.731% with no equity risk, no competitive disruption and no capital intensity. That spread has compressed to five basis points.

Underneath the price action the operating business is performing. Second-quarter adjusted EPS of $0.65 beat the $0.59 consensus by 10.2%. Free cash flow hit $4.7 billion against guidance of $4.0 billion to $4.5 billion. Postpaid phone net additions came in at 432,000 against a 338,500 forecast. Management raised the 2026 buyback target to roughly $10 billion from $8 billion, explicitly citing a disparity between operating fundamentals and the stock's valuation.

That disparity is the trade. The company is executing and the multiple is compressing, and the reason is a competitor that has not yet launched a product.

The Q2 Print: $0.65, $4.7 Billion, and 432,000 Phone Adds

AT&T delivered the cleanest quarter it has produced in several years, and the market gave it three sessions of credit.

Adjusted EPS came in at $0.65 against a $0.59 consensus, a 10.2% beat and 20% year-over-year growth. Revenue of $31.56 billion rose 2.3% but missed the $31.80 billion estimate by 0.6%. Adjusted EBITDA climbed 5.2% to $12.3 billion with the margin expanding 110 basis points to 39.1%. Free cash flow reached $4.7 billion, up roughly $300 million or 6.3% year over year, and cleared the top of management's own $4.0 billion to $4.5 billion guidance range.

The subscriber line is where the competitive question gets answered empirically. AT&T added 432,000 postpaid phone subscribers against a 338,500 consensus — a 28% beat on the metric the entire satellite thesis says should be deteriorating. Total advanced connectivity additions exceeded one million, driven by both fiber and fixed wireless.

Guidance held across every line. Full-year adjusted EPS stays at $2.25 to $2.35 against a $2.31 Street estimate. Free cash flow remains $18 billion or more. Capital investment holds at $23 billion to $24 billion, with the second-half spend described as more ratable than last year's fiber ramp. Service revenue growth stays low single digit and adjusted EBITDA growth stays 3% to 4%. Multi-year targets through 2028 were reaffirmed.

CFO Pascal Desroches guided third-quarter free cash flow to roughly stable year over year with stronger growth in the fourth quarter, and flagged higher cash flow from operations in the back half.

The revenue miss is the one soft spot, and it is the recurring pattern in this business — AT&T grows subscribers and margins faster than it grows the top line, because legacy revenue is declining underneath the advanced connectivity growth. The Legacy segment provides copper-based voice and data services that the company is actively shutting down, and every wire center retired removes revenue alongside cost.

Questions on the call from Morgan Stanley, UBS and BNP Paribas pushed management on pricing, fiber monetization, and the trade-off between broadband and wireless growth. The answers were consistent: converged accounts carry higher lifetime value and lower churn, and the company would rather optimize for that than for headline revenue.

Desroches retires at the end of the year, with Jennifer Byry returning to succeed him. Continuity at the CFO seat through a spectrum integration and a copper decommissioning is not trivial.

A $10 Billion Buyback Is Management Calling Its Own Stock Cheap

The single most informative disclosure on the call was not an operating metric. AT&T raised its 2026 share repurchase plan to approximately $10 billion from $8 billion — a 25% increase — and John Stankey stated the reason directly: to capture what management sees as a disparity between the company's operating fundamentals and the valuation of its stock.

That is a board authorizing capital deployment against its own share price on the explicit judgment that the market is wrong. The buyback represents a pull-forward of repurchases previously planned through 2028 rather than incremental capacity, which matters for how much runway remains, but the timing decision is the signal.

The combined return math is aggressive. Planned repurchases plus expected dividends total approximately $18 billion in 2026 — essentially 100% of the free cash flow outlook. AT&T repurchased close to $1 billion of stock in July alone. Total shareholder returns from 2026 through 2028 are guided at $45 billion or more.

At a $159.13 billion market capitalization, $10 billion of buybacks retires roughly 6.3% of shares outstanding in a single year. Layer the 4.78% dividend on top and the total shareholder yield runs above 11% against a business guiding to 3% to 4% EBITDA growth. Those two numbers do not normally coexist.

The constraint is the balance sheet, and management addressed it. Net debt to adjusted EBITDA closed the second quarter at 2.68 times, essentially flat with the first quarter, and the company reiterated its commitment to reduce leverage following the EchoStar spectrum acquisition. Total debt stands at $157.31 billion against $12.25 billion of cash. AT&T secured $29.5 billion in credit agreements to support the transaction.

Running buybacks at 100% of free cash flow while carrying 2.68 times leverage and integrating a $23 billion spectrum purchase is a stretched position if anything goes wrong operationally. It works only if the free cash flow guide holds and the copper cost removal delivers.

The market's response was to sell the stock 4% on a competitor headline within a week. Management's read that the disparity between fundamentals and valuation is wide has been confirmed, though not in the direction they intended.

The $23 Billion EchoStar Close Adds 50 MHz

AT&T completed its acquisition of wireless spectrum licenses from EchoStar for approximately $23 billion, closing on schedule by the end of July under a deal first announced in August 2025.

The assets add roughly 50 MHz of low-band and mid-band spectrum covering virtually every market in the country. Low-band carries propagation — it travels far and penetrates buildings — which is the specific characteristic that determines coverage quality in rural areas and inside structures. Mid-band carries capacity. Together they address the two constraints that limit 5G performance.

The strategic logic runs through traffic growth. Management argued that AI-driven connectivity will substantially increase network traffic, reinforcing both the fiber and 5G investment cases. Whether that materializes at the scale implied is the open question, but the spectrum is a defensive necessity regardless: capacity purchased today is capacity a competitor cannot buy.

The cost history is the uncomfortable part. AT&T has spent more than $60 billion on spectrum licenses since 2020 with limited prospects for incremental revenue growth attached. Maintaining position in wireless through spectrum control has been expensive, and this transaction adds $23 billion to that total. Legacy carriers collectively spent $110 billion over the past decade acquiring low-band spectrum to defend their territory.

The timing is what makes it awkward. SpaceX purchased $2.6 billion of airwaves from the same EchoStar to extend Starlink's reach, and is reportedly hunting urban-grade spectrum either by acquiring rivals or bidding in next year's federal auction. AT&T is buying terrestrial spectrum at $23 billion while a competitor assembles a hybrid orbit-to-ground network that may not require the same asset base.

Funding is covered. The company stated it is well positioned to complete the transaction while sustaining the dividend, accelerating buybacks and maintaining its deleveraging commitment. The $29.5 billion credit facility provides the liquidity bridge.

The asset quality argument holds. Following this period of investment AT&T owns a nationwide 5G network with deep spectrum behind it and a fiber network capable of reaching roughly half the U.S. population. No competitor can match that combination of assets today. The question the market is asking is whether that combination still constitutes a moat in 2030.

SpaceX Removed $46 Billion From the Sector and It Hasn't Shipped a Product

The dominant force on AT&T's multiple in 2026 has been a company that does not yet sell a phone plan.

The sequence started June 26, when reports surfaced that SpaceX President Gwynne Shotwell told IPO roadshow investors the company intends to launch a Starlink-branded retail mobile service for U.S. consumers. Verizon fell 7%, AT&T slid 5.8% to $21.41 and touched its 52-week low of $21.29 intraday, and T-Mobile dropped 6% to $171.78. Barron's calculated $15.74 billion wiped from AT&T's market capitalization alone, part of a $46 billion combined slide across the three carriers.

It got worse. Bloomberg reported that Charter Communications held executive-level talks with SpaceX about a consumer mobile offering, potentially routing traffic through Charter's own network. Verizon dropped 6.5%, AT&T fell 4.9% and T-Mobile lost 4.6% in a single session, making them among the S&P 500's worst performers. Across four consecutive sessions AT&T and Verizon each lost more than 11% of their value, with AT&T shedding $17.89 billion and Verizon $21.44 billion.

The most recent leg came this week from a Semafor report describing SpaceX hunting urban-grade spectrum to build dense coverage in major cities, alongside prototype handset demonstrations and plans for ground-based tower networks. AT&T and Verizon both fell 4%.

The numbers SpaceX has put in front of investors explain the reaction. Its IPO pitch deck pegs the global mobile addressable market at $740 billion, with Starlink's mobile division alone projected at $15 billion of annual revenue this year. Starlink already serves more than 12 million subscribers across over 160 countries. Shotwell has stated she expects Starlink Mobile users to far exceed Starlink broadband users. Oppenheimer's Tim Horan put it bluntly: SpaceX will disrupt the $1.6 trillion communications industry.

The pushback is that nothing has launched. SpaceX has not confirmed a timeline or pricing for consumer mobile, which means the threat is currently priced entirely on investor communications rather than a product. KeyBanc upgraded AT&T to Overweight with a $30 target on that basis, and separate analyst work argues AT&T will not need to worry about SpaceX for years.

Bernstein cut targets across Verizon, Charter, AT&T, Comcast and T-Mobile on satellite-driven valuation risk. Oppenheimer downgraded AT&T from Outperform to Perform. Goldman removed it from the US Conviction List.

The carriers are hedging. AT&T and Verizon both hold agreements with AST SpaceMobile, and the three majors are in discussions to form a satellite backup joint venture.

Internet Air at 2 Million and the Convergence Case

The operating response to satellite risk is convergence, and AT&T produced a hard number on it this week.

AT&T Internet Air, the company's 5G fixed wireless home internet product, reached 2 million subscribers in roughly a year — a milestone the company disclosed July 29. More than half of those customers pair the service with AT&T wireless.

That attachment rate is the entire strategic argument compressed into one statistic. A customer holding both wireless and home internet on a single account churns materially less than one holding either alone, generates higher lifetime value, and costs less to retain. Stankey has repeatedly framed convergence between fiber and wireless as the retention mechanism, and the Internet Air data supports it.

Fixed wireless also solves an economic problem fiber cannot. Fiber reaches roughly half the U.S. population under AT&T's build plan, and extending it further carries costs that do not clear the return hurdle in low-density markets. Internet Air uses existing 5G spectrum capacity — including the newly acquired EchoStar licenses — to serve households where trenching fiber makes no sense. It converts spectrum from a defensive asset into a revenue-generating one.

Total advanced connectivity additions exceeded one million in the second quarter across fiber and fixed wireless combined, which is the composite metric management now steers by. The Advanced Connectivity segment covers domestic 5G and fiber-based wireless, internet and related services for consumer and business customers, and it is the growth engine offsetting Legacy decline.

The business side is expanding through partnership rather than capital. AT&T Business and Palo Alto Networks announced an expanded collaboration integrating security directly into distributed enterprise environments. AT&T Express Waves launched, which Raymond James flagged as negative for Cogent — a signal the company is competing more aggressively in wavelength services.

The convergence thesis has a specific vulnerability. If SpaceX partners with Charter, the resulting product bundles satellite mobile with cable broadband, which is a converged offering assembled without owning towers or trenching fiber. That combination attacks the exact retention mechanism AT&T is building.

Two million Internet Air subscribers in a year with 50%-plus wireless attachment is real evidence the strategy works today. The market is pricing whether it works in 2029.

The Copper Shutdown Is the Margin Story

The least discussed and most quantifiable margin lever in this business is AT&T retiring its copper network, and regulatory approvals accelerated sharply this quarter.

Stankey disclosed that AT&T now has approval to discontinue legacy services in more than 30% of its wire centers, effective by late 2026. Separately, the FCC granted permission to discontinue legacy copper voice service at approximately 60% of wire centers in California, allowing those customers to be migrated to AT&T Phone Advanced, fiber or wireless. By year-end, the company expects a couple hundred wire centers to have zero customers remaining.

The economics are straightforward. Copper plant requires maintenance, powering, technician dispatch and real estate regardless of how few customers it serves. A wire center with zero customers still carries fixed cost until it is formally decommissioned, and decommissioning requires regulatory approval that AT&T has spent years pursuing. Approval to shut 30% of wire centers converts a stranded cost base into removable expense.

The second-order benefit is real estate. Retired central offices unlock property that can be sold or repurposed, which is why management framed it as a path to unlock access to descaling the network footprint.

The margin math already shows it. Adjusted EBITDA margin expanded 110 basis points year over year to 39.1% while revenue grew only 2.3%. Margin expanding five times faster than revenue in a mature telecom business is cost removal, not pricing power.

Management credited FCC Chairman Carr and the Commission directly for recognizing the urgency of modernizing communications infrastructure. Regulatory posture is a genuine variable here — the prior approval pace made copper retirement a decade-long project, and the current pace makes it a multi-year one.

The Legacy segment provides domestic legacy voice and data services over the copper network, and every wire center retired shrinks that revenue line. That is the mechanical explanation for why AT&T keeps missing revenue estimates while beating on EPS and free cash flow. Legacy revenue declining faster than Advanced Connectivity revenue grows produces a soft top line and an expanding margin simultaneously.

Investors valuing this business on revenue growth will keep getting the wrong answer. The correct frame is free cash flow per share against a shrinking share count, and on that measure $18 billion of free cash flow against $159.13 billion of market capitalization produces an 11.3% free cash flow yield.

2.68x Leverage, $157 Billion of Debt, and a 5.26% Long End

The balance sheet is where rising Treasury yields stop being an abstraction.

AT&T closed the second quarter with net debt to adjusted EBITDA of 2.68 times, essentially unchanged from the first quarter. Total debt stands at $157.31 billion against $12.25 billion of cash and equivalents. Debt to equity runs 130.34%. The company secured $29.5 billion in credit agreements to fund the EchoStar transaction.

Against that structure, the 30-year Treasury at 5.263% — a 19-year high — and the 10-year at 4.731% set the refinancing environment. Every maturity that rolls in the current market reprices upward relative to the coupons issued during the easing cycle. A business carrying $157 billion of gross debt sees material cash interest expense increases from a 100 basis point move in the curve, and the curve has moved considerably more than that from the 2024 lows.

The Fed held at 3.50% to 3.75% on July 29 in a 9-3 vote with Hammack, Kashkari and Logan dissenting in favor of a hike — the most hawkish dissent since September 2016. September hike odds sit near 63%. Core PCE runs 3.3% year over year with headline at 3.7%. Nothing in that configuration points to cheaper refinancing.

Management committed to reducing leverage following the spectrum close while simultaneously running $18 billion of shareholder returns against $18 billion of free cash flow. Those two commitments are in tension. Deleveraging from 2.68 times requires either EBITDA growth of 3% to 4% doing the work through the denominator, or debt paydown from cash that is currently allocated to buybacks and dividends.

The EBITDA path is credible. Adjusted EBITDA grew 5.2% in the quarter with the margin up 110 basis points, and the copper decommissioning provides a visible cost lever into 2027 and 2028. Growing into the leverage rather than paying it down is a workable plan when the operating trajectory holds.

The vulnerability is any interruption to free cash flow. Capital investment guides to $23 billion to $24 billion, which is not discretionary in a business defending against a well-capitalized new entrant. If competitive intensity forces higher capex or pricing concessions, the $18 billion free cash flow figure that funds everything comes under pressure first, and the buyback is the flexible line.

A 4.78% Dividend Against a 4.73% Treasury

The single cleanest way to frame what has happened to AT&T's valuation is the spread between its dividend and the risk-free rate.

AT&T pays $0.2775 quarterly, $1.11 annualized, producing a 4.78% yield at the current price with an ex-dividend date of July 10 most recently passed. The 10-year Treasury closed the week at 4.731% after an intraday print at 4.737%. The spread has compressed to roughly five basis points.

That comparison is the reason this stock trades where it does. For most of the past decade, AT&T's yield carried a 200 to 300 basis point premium over the 10-year, which is what income buyers were paid to accept equity risk, competitive risk and capital intensity. At five basis points, the compensation for those risks has effectively disappeared.

The 30-year at 5.263% makes it worse. An investor can now buy a longer-duration government bond yielding 49 basis points more than AT&T's dividend, with no exposure to a satellite entrant, no exposure to spectrum auction costs, and no exposure to $157 billion of corporate leverage.

That mechanism explains why AT&T fell alongside every other bond proxy this week. Beta of 0.17 means the stock barely responds to equity index moves — it responds to the discount rate. Friday's 7 basis point jump in the 10-year did more to the price than the S&P 500's 0.18% decline.

The counterargument runs through the total return, not the dividend alone. Adding roughly $10 billion of buybacks against a $159.13 billion market capitalization produces about 6.3% of share retirement, which stacks on the 4.78% dividend for total shareholder yield above 11%. A Treasury does not compound. AT&T's payout is covered by $18 billion of free cash flow against roughly $7.6 billion of annual dividend obligation, giving a coverage ratio above 2.3 times.

The dividend is not at risk on the numbers. Management explicitly stated the cash flow and liquidity position sustains the dividend while accelerating buybacks and deleveraging.

What is at risk is the multiple. Income investors mark this stock against the curve, and the curve has moved against it for eight consecutive months while the operating business improved.

AT&T Trades at a Discount to Verizon and a Deep Discount to T-Mobile

The relative valuation is where the disparity management referenced becomes measurable.

AT&T trades at approximately 6.7 times forward EV/EBITDA against Verizon at 7.3 times and T-Mobile at 8.8 times. It sits below its own five-year historical range of 7.5 times and above. On earnings, the stock carries a trailing P/E of 7.67 on TTM EPS of $3.03, a forward P/E of 9.62 on next-twelve-month estimates, and a normalized P/E of 10.50. Price to sales runs 1.33 against TTM revenue of $127.24 billion.

Return metrics do not justify the discount. Return on equity runs 19.92%. Gross margin sits at 59.72% and net margin at 18.15%. TTM EBITDA is $47.30 billion. A business generating those returns at 6.7 times EV/EBITDA is being priced for structural decline rather than 3% to 4% EBITDA growth.

The peer comparison carries its own complication. T-Mobile has been the worst performer among the three on a one-year basis at down 28%, and hit a fresh 52-week low of $165.66 on June 30. It also carries the most direct satellite exposure, since it already partners with SpaceX for direct-to-cell service — a relationship analysts describe as making T-Mobile the clear acquisition or partnership target if SpaceX wants to own a wireless business outright. That T-Mobile still commands 8.8 times EV/EBITDA against AT&T's 6.7 times reflects growth expectations rather than competitive insulation.

Verizon at 7.3 times sits between them with a similar dividend profile and similar leverage. The two have traded as a pair through every SpaceX headline, falling within a percentage point of each other on each of the four major selloff sessions.

Charter is the wildcard in the comparison set. As the potential SpaceX partner, it has moved from being a converged competitor to being a possible distribution channel for the disruptor, which changes how the entire cable-versus-telco framework should be modeled.

Book value comparisons favor AT&T on assets. The company owns a nationwide 5G network with deep spectrum, now including 50 MHz more from EchoStar, plus fiber reaching roughly half the U.S. population. Replacement cost for that combination runs well above the current $159.13 billion market capitalization.

Analyst Targets Run From $24 to $33 on the Same Quarter

The sell-side dispersion following the second-quarter report is unusually wide, and the split maps precisely onto how each desk handles the satellite question.

The consensus sits at Moderate Buy with an average price target of $29.19, implying roughly 25% upside from current levels. The distribution behind that average is where the information sits.

On the constructive side, KeyBanc upgraded to Overweight from Sector Weight with a $30 target, arguing SpaceX concerns are overblown. Wolfe Research upgraded after the print. TD Cowen raised its target to $33 from $32 while keeping a Hold, citing good second-quarter results. Exane BNP Paribas moved to Outperform with a $28.50 target. Morgan Stanley raised its target to $27 from $25.

On the cautious side, Argus cut to $30 from $33 while keeping a Buy. Citi cut to $28 from $31.50 while keeping a Buy and describing the quarter as favorable. Barclays reduced its target to $24 from $26 with an Equal Weight rating. Bernstein maintained Outperform at $25. Wells Fargo initiated at Underweight with a Sell rating on July 8. Goldman Sachs removed the stock from its US Conviction List. Oppenheimer downgraded from Outperform to Perform.

The pattern is telling. Firms cutting targets are mostly keeping Buy ratings, which means they see the same fundamentals and are simply marking down the multiple they will apply given satellite uncertainty. Firms raising targets are doing so on the operating beat. Almost nobody is arguing the business deteriorated.

Barclays at $24 and TD Cowen at $33 represent a 37% spread on identical financial statements. That gap is the market's honest uncertainty about a competitive threat with no product, no pricing and no launch date.

The forward EPS anchor is stable. Full-year guidance of $2.25 to $2.35 against a $2.31 Street estimate, with next-year growth modeled in the mid-single digits, puts the stock at 9.62 times forward earnings. Applying the 10.5 times normalized multiple to the midpoint of guidance produces roughly $24.15 — close to where the stock trades. Applying the five-year historical 7.5 times EV/EBITDA floor produces a materially higher number.

The Technical Map: $23 Is the Line

The chart is a repaired downtrend that just failed its first retest, and the levels are tight.

AT&T trades $23.40 in a $23.15 to $23.62 daily band against a $23.94 prior close. The 52-week low of $19.89 was set July 2 at the peak of the satellite panic, and the 52-week high of $29.79 dates to September 15, 2025. The stock sits 21.4% below that high and 17.6% above the low.

Immediate support sits at $23.15, Friday's low. Below it, the $22.26 level that preceded the earnings gap is the first meaningful shelf, and $21.29 marks the intraday low from the June 26 selloff. The $19.89 July low is the structural floor and the level that defines whether this is a base or a continuation.

Resistance is layered close overhead. The $23.94 prior close is the immediate reclaim. Above it, $24.06 was the post-earnings close and $24.67 the high of that move. The $24.20 area has acted as a pivot repeatedly through July. Clearing $24.67 opens the $26 to $27 zone where the stock traded before the SpaceX narrative took hold.

Volume tells the more useful story. The earnings session traded more than 80 million shares against an average near 61 million and a ten-day average of 97.29 million — turnover running well above normal for six consecutive weeks. Elevated volume through a range that has not resolved means genuine two-way disagreement rather than drift.

Beta of 0.17 removes the index from the equation. This stock is trading its own narrative and the Treasury curve, not the S&P 500. That decoupling has held through both the July small-cap weakness and the megacap earnings rally.

The stock is down 1.4% year to date. Against a communication services sector that has felt the same satellite pressure and a broad market up double digits, that is significant relative underperformance for a business that beat, raised its buyback and reaffirmed guidance.

The pattern since July 2 is a series of higher lows — $19.89, then $21.29, then $22.26 — which is the signature of accumulation into bad news. It breaks if $23.15 fails and $22.26 does not hold.

What Actually Determines the Next Move

Three variables decide this stock over the next two quarters, and only one of them is inside management's control.

The first is the Treasury curve. With the dividend yield at 4.78% and the 10-year at 4.731%, the spread is five basis points. Every basis point of further yield increase compresses the relative attractiveness of the payout and pulls income capital out. September hike odds at 63%, core PCE at 3.3%, and three Fed presidents openly voting for tighter policy point toward more pressure rather than less. A 10-year back below 4.40% would restore a meaningful yield premium and re-rate the stock without any operational change.

The second is SpaceX headline flow, which is unpredictable and asymmetric. Each report has cost the stock 4% to 6% within a session, and none has been accompanied by a product, a price or a launch date. The next catalysts are next year's federal spectrum auction, where SpaceX is expected to bid for urban-grade licenses, and any formal Charter partnership announcement. A confirmed Charter-SpaceX consumer mobile deal would be the most damaging single headline available, because it converts a satellite operator into a converged competitor with cable broadband attached.

The third is execution, and it is the one going right. Postpaid phone adds beat by 28%. Free cash flow cleared guidance. Margin expanded 110 basis points. Internet Air hit 2 million subscribers with over half attached to wireless. Copper shutdown approval covers more than 30% of wire centers effective late 2026. Third-quarter results arrive around October 20.

The buyback is the mechanism that converts execution into per-share value while the multiple stays compressed. At roughly $10 billion against $159.13 billion of market capitalization, AT&T retires about 6.3% of its float annually. Sustained over the 2026 to 2028 window with $45 billion or more of total returns, that is a materially smaller share count against the same free cash flow.

The bear case does not require SpaceX to succeed. It only requires the market to keep pricing the possibility while the discount rate stays elevated. That combination has produced a 21% drawdown from the September high on a business that grew EPS 20% year over year.

Forecast: $23.15 Holds or $22.26 Comes Next

The base case into the third-quarter print is continued range trade between $22.26 and $24.67, with direction set by the Treasury curve rather than by anything AT&T reports.

The bear path is straightforward and needs no new information. A firm July CPI or a strong August 7 payrolls number takes September Fed hike odds from 63% toward certainty, pushes the 10-year through 4.80% and the 30-year above 5.35%, and inverts the dividend spread outright. Income capital rotates to Treasuries, and $23.15 fails. Below it, $22.26 is the first shelf and $21.29 the second. A confirmed SpaceX-Charter consumer mobile announcement layered on top would test $19.89. Barclays at $24 already sits below the stock's July high, and Wells Fargo carries an outright Sell.

The bull path requires the yield picture to cooperate and the satellite narrative to stay theoretical. Reclaiming $23.94 and then $24.67 confirms the higher-low sequence built from the $19.89 July bottom and opens the $26 to $27 band. Applying the five-year historical low end of 7.5 times EV/EBITDA against $47.30 billion of TTM EBITDA produces a valuation meaningfully above spot, and the consensus $29.19 target implies 25% upside. KeyBanc at $30 and TD Cowen at $33 anchor that end.

The variable nobody is modeling correctly is the buyback's compounding effect. Retiring 6.3% of shares annually against flat free cash flow raises free cash flow per share by the same amount before any operating growth. Stack the guided 3% to 4% EBITDA growth on top and the per-share cash generation compounds near 10% annually while the stock trades at 6.7 times EV/EBITDA.

Targets: downside $23.15, then $22.26, then $21.29 on a break. Upside $23.94, then $24.67, then $26.50 on a reclaim. The dividend at 4.78% is covered 2.3 times by free cash flow and is not the risk. The multiple is.

AT&T enters August having beaten earnings by 10.2%, raised its buyback 25%, closed a $23 billion spectrum acquisition, and given back every dollar of the resulting rally — priced by a market comparing its payout to a 4.731% Treasury and its future to a competitor that has not launched a phone.

That's TradingNEWS