Three Moderate Scores, One 62% Upside

The Matchup

American telecom has carried the same story for a decade: three carriers, a saturated market, no growth left, capital intensity that eats every dollar of cash flow. Buy them for the dividend, expect nothing else.

Three weeks in July put that story under pressure. AT&T reported on July 22 and posted its best margin since it refocused around fiber and wireless. Verizon reported on July 24 and posted the highest adjusted EBITDA margin in its corporate history. T-Mobile, the company that spent ten years taking share from both, reported on July 23, beat on earnings, and lost more than a tenth of its market value in a single session.

This matchup was flagged in an earlier scan because the platform showed Verizon as the only Bullish outlook in an eight-ticker telecom screen. That flag is gone. As of the August 8 and August 10 calculations, all three carriers score Moderate on Company Strength, within four points of each other, with a Neutral outlook. The signal that made this battle look easy decayed before the battle was written, which leaves a cleaner question: if quality cannot separate them, the verdict rests on what the market charges for each and which direction the numbers travel. On those two measures the three could hardly be further apart.

The Tale of the Tape

Prices and performance as of the August 7 close, from the platform's daily price history.

Read the year-to-date column twice, because it inverts the narrative. T-Mobile, the share taker with the best network scores in the industry, is the worst performer and sits nearly a third below where it traded a year ago. Three companies in the same business serving the same saturated market, separated by 28 points of performance. Something split them, and it was not the quality of the quarter.

The Case for AT&T

AT&T is the cheapest large-cap telecom in the United States, and it is cheap for a recognizable reason: the company spent a decade destroying capital in media and is still paying the trust penalty. The July 22 print showed that the underlying business stopped being the problem some time ago.

Consolidated revenue came in at $31.6 billion, up 2.3%, a small miss against the company's own bar and the only line that failed to clear it. Everything below revenue was better. Adjusted EBITDA rose 5.2% and the margin expanded 110 basis points to 39.1%, the highest since AT&T refocused around advanced connectivity. Adjusted earnings per share reached $0.65 against $0.54, up roughly 20%. Free cash flow of $4.7 billion beat the top of the company's $4.0 billion to $4.5 billion range while capital investment climbed to $6.1 billion from $5.1 billion, funding the fastest fiber construction in company history.

The mix shift is the real story. AT&T added more than one million fiber locations in the quarter, posted record second-quarter fiber net additions, and added 432,000 postpaid phone subscribers. Advanced Connectivity, now more than 90% of service revenue and nearly all adjusted EBITDA, grew service revenue 5.1% and EBITDA 8%, while the legacy segment shrank 26%. The declining business is now too small to determine the consolidated result. Advanced home internet service revenue grew more than 27%, with 42.5% of those customers attaching a wireless line. Management raised the 2026 buyback target to roughly $10 billion from $8 billion, with CEO John Stankey calling the stock's valuation suppressed and undervalued.

The complication arrived at the end of July, when AT&T closed its roughly $23 billion purchase of about 50 MHz of EchoStar spectrum, lifting net leverage to around 3.2 times from the mid-2s with a three-year path back to target. The spectrum is valuable and the low-band coverage improves the network materially. It also means the cheapest stock in the group is adding balance sheet risk at exactly the moment its cash flow story was becoming clean.

The Case for Verizon

Verizon spent years as the group's designated share loser, leaning on price increases that drove churn while T-Mobile took the growth narrative. The July 24 print is the first evidence the correction is structural rather than cosmetic.

Adjusted EBITDA of $13.7 billion rose 7.2% for a margin of 40.1%, the highest Verizon has ever reported. Free cash flow of $6.4 billion rose 24.4%, taking first-half free cash flow to $10.2 billion. Adjusted earnings per share of $1.30 rose 6.6%. The company raised full-year guidance on mobility and broadband service revenue, adjusted EPS, and free cash flow, the second consecutive quarter of increases.

The subscriber numbers make the margin credible. Verizon delivered 184,000 postpaid phone net additions, its best consumer second quarter in five years, with gross adds the best in eight. First-half net adds improved by 537,000 against the prior year, and consumer postpaid phone churn fell six basis points to 84. Broadband added 348,000 connections, taking the total past 17.1 million.

The mechanism matters more than the count. The Simplicity plan, unlimited at $45, is producing gross adds roughly 16% ahead of the company's own forecast and net new accounts 31% ahead, and every account on it is subsidy-free. That is a structural change to unit economics, not a promotional quarter.

Two items shape the forward case. Net unsecured debt to adjusted EBITDA improved to 2.5 times, moving opposite AT&T, with total unsecured debt down to $136.5 billion. And Verizon disclosed a dark fiber and AI infrastructure agreement with Google worth more than $1 billion, with revenue beginning in 2027, alongside a target of at least $9 billion in combined operating and capital expenditure savings. The dividend, 70.75 cents quarterly and roughly 6.0% at the current price, carries two decades of consecutive annual increases.

The blemish is in the headline. Total revenue fell 0.7% to $34.3 billion and GAAP net income dropped 22.9% on $1.8 billion of pre-tax special items. Verizon is generating record cash from a shrinking top line, a real tension addressed below.

The Case for T-Mobile

By the platform's own measures, T-Mobile is the best company in this matchup: the highest Company Strength at 51.0, the highest Financial Health at 55 out of 100, the widest Moat at 8 out of 15, and a record wireless Net Promoter Score of 46. None of that stopped the stock from falling 10.75% on July 23, from $190.94 to $170.42.

The quarter was strong on almost every profitability line. Service revenue rose 9% to $19.0 billion, postpaid service revenue rose 13%, and core adjusted EBITDA rose 12%. Free cash flow margin reached 25%, and management raised full-year adjusted free cash flow guidance to $18.4 billion to $18.8 billion. Postpaid phone churn improved to 0.85%, and more than 60% of new account customers selected premium plans.

Two numbers undid all of it. Postpaid net account additions came in at 277,000, a 13% decline year over year, and management guided third-quarter net account additions down to approximately 250,000, attributing the step-down to a temporary spike in churn from rate plan modernization. Revenue of $22.79 billion also came in modestly below consensus.

The substance is this: T-Mobile is converting from a volume growth story into a pricing and mix story. Raising prices on existing customers produces revenue and margin, and it produces churn. That is a defensible way to run a mature wireless business. It is not the model the stock was priced for, and the market repriced immediately.

The forward case deserves stating at full strength. T-Mobile trades 31.6% below its 52-week high with the best network quality scores, the best customer satisfaction, the widest moat score here, and roughly $18.6 billion of guided free cash flow against a market capitalization near $190 billion. If the deceleration is as temporary as management says, this is the cheapest entry the market has offered on the sector's best-run operator in over a year.

The Starlink Question

On August 4, SpaceX (SPCX) reported its first quarter as a public company, and President and Chief Operating Officer Gwynne Shotwell used the call to state that Starlink Mobile intends to compete with the three national carriers for their own customers. She noted the three generate roughly $600 billion in combined annual revenue and said she expects to take a meaningful number of them.

The substance is more than rhetoric. SpaceX holds roughly 65 MHz of spectrum from its EchoStar transaction, more than 650 direct-to-cell capable satellites in orbit, and 12 million Starlink subscribers. Shotwell described a next-generation satellite launching in 2027 delivering 5G and LTE speeds up to 150 Mbps to unmodified handsets, plus low-cost terrestrial base stations so the network is not satellite-only. This platform covered the SpaceX capital story in The IPO Is the Exit and tested the Starlink revenue projections against the filings in the Ron Baron valuation piece.

The market response is the data point that matters here. All three carriers fell between 2% and 4% on August 5. By the August 6 close they had recovered nearly all of it: T-Mobile up 3.75%, AT&T up 2.82%, Verizon up 1.12%, with all three finishing the week higher than they started. A 48-hour round trip is not a sector absorbing a structural threat. It is a sector treating 2027 as far away.

That read has support. All three carriers have publicly refused SpaceX an MVNO agreement, so Starlink must build retail distribution, billing, device subsidies, and customer service from nothing. Craig Moffett of MoffettNathanson has called competitive direct-to-consumer service within five years extraordinarily challenging without one, and T-Mobile chief executive Srini Gopalan calls the threat exaggerated, noting satellite traffic remains a negligible share of network usage. The counter, from David Barden of New Street Research, is that the threat is structurally significant regardless of whether a retail product ever launches, because its existence changes every negotiation the carriers enter. The field is filling in either way: AST SpaceMobile won approval in April 2026 for a 248-satellite constellation, and Amazon acquired Globalstar for $11 billion the same month.

For this matchup, the exposure is not symmetrical. T-Mobile is the carrier whose satellite partner just announced it intends to compete for T-Mobile's customers; its T-Satellite service runs on the Starlink network. AT&T and Verizon have no such entanglement. The company with the widest moat score in this battle is the one whose supplier became a declared competitor.

What the Wealth Engine Scores Say

Before the editorial verdict, here is what the Wealth Engine Pro platform's systematic scoring shows for all three stocks right now.

One number is deliberately missing. The platform does generate a fair value estimate for AT&T, and it implies upside above the reliability threshold this publication applies to every article. Wealth Engine Pro excludes calculated upside beyond roughly 80% on data integrity grounds, because the blended model becomes unstable at that range. AT&T is the case the rule exists for: roughly $174 billion of total debt, where small changes in discount rate or terminal assumption swing the equity value enormously, and where a $23 billion spectrum acquisition just moved leverage from the mid-2s to roughly 3.2 times. The directional read is that AT&T screens cheap. The specific figure is withheld rather than dressed up.

The remaining scores show genuine convergence. T-Mobile leads every quality measure: strength, financial health, and moat. AT&T leads growth. Verizon leads nothing outright and finishes second on almost everything. All three carry a Neutral outlook, dominant across the last eight calculations. The trend signals diverge: AT&T and Verizon register an emerging tailwind, T-Mobile a headwind with severe compression.

These scores are systematic. They evaluate companies based on reported financials, balance sheet quality, moat characteristics, and valuation models (DCF, peer comparison, earnings power). They measure what a company is today, not what it might become. That is by design: the scoring system is built to keep emotion and forward speculation out of the numbers.

This article is doing something different. It argues about which of three statistically similar companies has the most favorable combination of price, direction, and balance sheet trajectory heading into 2027, when the AI infrastructure contracts, the spectrum integration, and the satellite competition all begin to show up in reported results. The scoring system cannot price those in until they appear in filings.

Both perspectives are real data. The platform says these three companies are of similar and unremarkable quality. The article argues that when quality is a tie, the valuation gap and the direction of travel decide. Transparent investors use both.

The Valuation Verdict

The valuation spread is the widest this series has scanned where the businesses are this similar. Verizon carries a fair value estimate of $76.68 against a $47.06 share price: 62.9% of upside as the platform expresses it, or roughly 38.6% below the model's estimate stated the other way. T-Mobile's $193.73 against $177.19 leaves 9.3%, a rounding error and a Fair Value label. AT&T screens cheapest of all and cannot be quoted.

Income tells the same story from another angle. Verizon yields roughly 6.0% on a payout backed by two decades of consecutive increases. AT&T yields roughly 4.7% on a dividend held flat at $1.11 and cut within recent memory, with capital returned through buybacks instead. T-Mobile yields roughly 2.3%, a growth company's payout on a stock that just lost its growth multiple.

The balance sheets are where the two cheap stocks part company. Verizon's net unsecured debt to adjusted EBITDA improved to 2.5 times, with management committed to at least $9 billion of operating and capital expenditure savings plus more than $1 billion in Frontier synergies by 2028. AT&T's leverage moved to roughly 3.2 times at the end of July and will take approximately three years to return to target. Both are cheap. One is deleveraging into its cheapness and one is levering up. That distinction turns a two-way tie into a verdict: AT&T has the larger discount and the larger execution burden, while Verizon has a discount the model can stand behind and is closing it from improving credit rather than deteriorating credit.

What Could Go Wrong

Every company here has a real case against it, including the one this article picks.

Against Verizon, the winner

The most uncomfortable fact is that Verizon generated record cash on a declining top line. Revenue fell 0.7% and GAAP net income fell 22.9% on $1.8 billion of pre-tax special items. Margin produced by cost reduction on a shrinking base is a different animal from margin produced by growth, and cost programs have finite runway. A 40.1% margin cannot be beaten by cutting forever.

The trade is also the most crowded of the three. Verizon is up 15.5% year to date and sits only 8.4% below its 52-week high, so much of the re-rating argued for here has already happened. The payout ratio near 72% is the tightest in the group. And the Company Strength edge over AT&T is 1.2 points, which is statistical noise, not a quality verdict. Ranked on business quality alone, Verizon finishes second.

Against AT&T

AT&T is genuinely the cheapest of the three and its turnaround is real, which makes the risk harder to see. Leverage at roughly 3.2 times following the EchoStar close arrives just as the company spends at the fastest fiber pace in its history, and the path back to target assumes nothing interrupts cash generation. The dividend is frozen and carries the memory of a cut. And the fair value model is effectively saying it cannot price this balance sheet with confidence, which is itself information.

Against T-Mobile

The steelman for T-Mobile is the strongest bull case in this article, and the platform's own scores support it. It is the highest-quality company in the matchup by strength, health, and moat, trading 31.6% below its 52-week high, with roughly $18.6 billion of guided free cash flow against a $190 billion market capitalization. If the third-quarter weakness is the temporary rate-plan artifact management describes, buyers here are acquiring the sector's best operator during a mechanical air pocket. The counter is that severe compression and a headwind reading are the platform's way of saying the repricing is unfinished. But the case is real, and anyone who believes the deceleration is temporary has better risk-reward here than the winner offers.

Against all three

If Starlink Mobile reaches market with competitive pricing sooner than consensus expects, every model here is wrong, because all three are valued on the assumption that a three-player oligopoly persists. A fourth national competitor with no legacy cost base would compress industry margins regardless of which incumbent wins the subscriber. Rate sensitivity is the shared risk: these are bond proxies, and a move higher in long yields hits all three at once.

The Data Picks a Winner

Score it category by category. T-Mobile takes Company Strength, Financial Health, and Moat. AT&T takes Growth. Verizon takes the usable fair value gap, the trend and compression reading, year-to-date performance, dividend yield, margin record, free cash flow trajectory, and the direction of leverage. On a count of measurable categories, the verdict is Verizon, six to three.

The reasoning matters more than the count. This series runs on a simple discipline: when quality is a tie, price decides. Here quality genuinely is a tie, a four-point spread across three Moderate scores with identical Neutral outlooks, so the question becomes which company offers a discount the platform can stand behind, moving in the right direction, on an improving balance sheet. Verizon is the only one satisfying all three conditions. AT&T has the bigger discount, a number this publication will not print, and leverage moving the wrong way. T-Mobile has the best business and no valuation cushion.

To be explicit: this is not a claim that Verizon is the best company in American telecom. By the platform's own scoring it is not, and T-Mobile is. It is a claim that Verizon is the one whose price, cash flow direction, and credit trajectory align at the same moment. Record margin, free cash flow up 24%, guidance raised twice running, the best consumer subscriber quarter in five years, deleveraging to 2.5 times, and a fair value estimate 62.9% above the current price.

The narrative said American telecom was three dying utilities and that the disruptor was the only one worth owning. The data from the last three weeks says two incumbents just posted the best margins in their corporate histories while the disruptor guided subscriber growth down. Battle Stocks goes to Verizon, on the judges' cards, decided on price and direction rather than quality. The referee is the October print, and the thing to watch is whether the margin holds without further special items.

What Battle Do You Want to See Next?

Battle Stocks runs every Monday, and the best matchups come from readers. The payments rematch is already on the calendar for late October, when Visa and Mastercard both report again and the closest verdict in series history gets regraded on fresh numbers. Send the matchup you want refereed by the data, and it goes into the rotation.