Verizon Communications Inc. (VZ)

Deep Dive Research Report

Sector: Communication Services | Listing: NYSE | Report date: 10 September 2026 Latest reported period: Q2 2026 (quarter ended 30 June 2026), released 24 July 2026

1. What The Company Does

Verizon sells connectivity. That is the whole business, stripped of jargon. About 147 million American phones, tablets, watches, cars and home routers connect to the internet through radio equipment and fibre-optic cable that Verizon owns, and roughly every month each of those connections generates a bill. Strip away the branding and Verizon is a landlord: it owns exclusive federal licences to use specific slices of radio spectrum across the United States, it has spent decades burying glass fibre in the ground to carry the traffic those radios collect, and it charges rent for access.

What makes the business interesting rather than boring is the physics and the balance sheet underneath it. Radio spectrum is a finite, government-rationed resource. You cannot manufacture more of it. The company that holds the most usable mid-band spectrum in a given city can serve the most customers at the highest speeds from the fewest towers, and that translates directly into cost per gigabyte. Verizon spent $52.9 billion in the 2021 C-band auction to fix exactly this problem, winning between 140 and 200 MHz of mid-band spectrum and more than doubling its mid-band holdings (Verizon C-band explainer). That single cheque is the largest capital allocation decision in the company's modern history and it still shapes everything about how Verizon competes today.

Where the company came from, and why the history matters

Verizon Communications was created on 30 June 2000 by the merger of Bell Atlantic Corp. and GTE Corp. (Verizon company history). Bell Atlantic was one of the original Regional Bell Operating Companies spun out of the 1983-84 breakup of the old AT&T Bell System. That inheritance is the reason Verizon's fixed-line footprint is concentrated in the US Northeast and Mid-Atlantic while its wireless network is national, an asymmetry that has driven strategy for twenty-five years and drove the Frontier acquisition that closed in January 2026.

Three transactions define the modern company:

Verizon Wireless (2000). On 3 April 2000, Bell Atlantic and Vodafone AirTouch contributed their US mobile businesses into a partnership doing business as Verizon Wireless, with Bell Atlantic holding 55% and Vodafone 45% (Verizon Wireless Form S-1, 2000). For fourteen years, Verizon owned only a majority of its own crown jewel.

The Vodafone buyout (2014). Verizon acquired Vodafone's 45% interest for approximately $130 billion, entering the stock purchase agreement on 2 September 2013 and completing on 21 February 2014 (Jones Day transaction record). Consideration was roughly $58.9 billion cash, about 1.27 billion Verizon shares worth roughly $61.3 billion, and $5.0 billion of senior unsecured notes. This is the single most consequential event in understanding Verizon's balance sheet. The company took on an enormous debt load to buy something it already controlled operationally, and it has been managing that debt ever since. Total debt stood at $158.2 billion at the end of 2025.

Frontier (2026). Closed 20 January 2026, expanding Verizon's fibre footprint to 31 US states plus Washington DC (Verizon 10-Q, quarter ended 30 June 2026). This is the company buying its way out of the geographic asymmetry the Bell Atlantic inheritance left behind.

The value proposition, and why it broke

For most of the last decade, Verizon's proposition was straightforward: we have the most reliable network, so we charge more. That worked until it didn't. Between roughly 2023 and 2025, Verizon leaned on repeated price increases to generate revenue growth while the subscriber base stopped growing and then shrank. In Q1 2025 the company lost 356,000 postpaid phone subscribers, driven by pricing actions taken in December and January (Q1 2025 concall, 22 April 2025). The losses continued through Q2 and Q3.

Dan Schulman, who became CEO in October 2025, diagnosed it in public on his first earnings call:

"For years, our financial growth relied too heavily on price increases, a strategic approach that relies too much on price without subscriber growth is not sustainable."

• Dan Schulman, Q3 2025 concall, 29 October 2025

The rebuilt proposition, launched in Q2 2026, is a different bargain: pay a simple, all-in, transparent price, get the network, and get rewarded for staying. Verizon One offers mobility and broadband together at $70 per month with all taxes and fees included, a router and professional installation bundled, and no activation or upgrade fees. Simplicity is a single wireless plan at $45 per month for existing customers switching to it, or $30 per line with autopay for customers porting in, with no network tiers - everyone gets the same 5G (Verizon plan comparison; Fierce Network, June 2026). Schulman's framing on the Q2 2026 call was blunt:

"One plan, one price, no games."

• Dan Schulman, Q2 2026 concall, 24 July 2026

What it actually looks like for a customer

Walk through a household that signs up for Verizon One. A technician arrives and either connects the home to Verizon's fibre - a glass strand running from a central office to a terminal at the kerb and then into the house - or, where fibre is not built, installs a fixed wireless access receiver that pulls a signal from the same C-band radios on the cell tower down the road that serve mobile phones. Speeds on the fixed wireless product typically exceed 300 Mbps where fibre is unavailable (Q2 2026 concall, 24 July 2026).

The same household's phones attach to the same tower. Both services land on one bill at one price with taxes included. If the family buys a phone, they can finance it over 36 months on a device payment plan, and Verizon books that receivable and, in many cases, securitises it - the company carries $28.8 billion of asset-backed debt secured by device payment plan receivables. That is the mechanical structure: a physical network, a recurring subscription, a consumer credit book attached to the handset, and a loyalty program designed to make leaving feel like a loss.

The strategic bet underneath all of it is convergence. Fibre customers churn roughly 40% less than mobile-only customers (Q3 2025 concall, 29 October 2025). If Verizon can put a wire in the ground at a house and a phone in the hand of the person living there, the relationship becomes very hard to unwind.

2. Business Segments

Verizon reports two segments: Verizon Consumer Group and Verizon Business Group, with a residual of corporate and other items (FY2025 Form 10-K, filed 17 February 2026). In FY2025, Consumer accounted for roughly 77% of consolidated revenue and Business roughly 21%, with the remaining ~2% in corporate and other.

The split is not cosmetic. These are genuinely different businesses with different economics, different competitors, different sales cycles and different growth trajectories, and one of them is a great deal healthier than the other.

2.1 Verizon Consumer Group (~77% of revenue)

What it does. Consumer sells wireless service, home broadband, and video to individuals and households across the United States. At the end of 2025 the segment carried approximately 116 million wireless retail connections - including fixed wireless access - of which 83% were postpaid, plus 11 million total broadband connections and 2 million Fios video connections. Postpaid means the customer is billed monthly under a contractual relationship with a credit check; prepaid means they pay up front, churn more, and pay less.

Consumer covers three distinct customer economies:

• Postpaid wireless. The profit engine. Higher-credit customers on multi-line family accounts, sold through Verizon's own retail stores, its website, and third-party retail. This is where device financing lives and where churn matters most.
• Prepaid, the Verizon Value portfolio. A collection of brands - Straight Talk, Visible, TracFone, Simple Mobile, SafeLink, Walmart Family Mobile and Verizon Prepaid - serving over 22 million customers, largely acquired through the TracFone transaction (Fierce Network op-ed on carrier prepaid portfolios). These brands ride the same network at lower price points and lower service costs, and they let Verizon compete at the value end without cannibalising the premium brand. Total Wireless opened its 2,000th store in 2026 (Verizon news, 2026).
• Home broadband. Two delivery mechanisms, one product. Fios is fibre-to-the-home, historically concentrated in the Northeast. Fixed wireless access (FWA) delivers home internet over the 5G mobile network. Post-Frontier, the fibre base reached 10.913 million subscribers at Q2 2026, and combined fibre plus FWA reached 17.1 million connections.

The core capability. Consumer's real asset is a national radio network with mid-band depth, plus the operational machinery to acquire and bill tens of millions of households. Verizon has deployed C-band on macro towers and roughly 17,000 mmWave cell sites, and by 2024 the company expected more than 250 million people to have access to its 5G Ultra Wideband service on C-band (Verizon 5G FAQs). Building that took a decade of tower leases, fibre backhaul, municipal permits and spectrum clearing that no new entrant can compress.

The second capability, newer and less obvious, is fibre scale. Frontier added over 30 million fibre passings. Verizon has raised its medium-term target to 40-50 million passings, up from a prior 35-40 million, with at least 2 million added organically in 2026 (Q4 2025 concall, 30 January 2026).

Why it exists separately. Consumer buying decisions are made in minutes at a retail counter or on a phone; Business decisions take months and involve procurement committees. The distribution channel, the marketing spend, the credit model and the churn dynamics have nothing in common. Verizon runs Consumer as a mass-market retail operation and Business as an enterprise sales organisation because they are, functionally, different companies.

Competitive position. Consumer competes head-on with AT&T Mobility and T-Mobile US in wireless, with Comcast and Charter in broadband and increasingly in mobile via their MVNO products, and now with satellite. Verizon's traditional edge was network reliability. Its traditional weakness was price perception - one analyst framed the problem as Verizon being seen as neither the best network nor the best value (IEEE ComSoc blog on Q3 2025 call, 30 October 2025). The 2026 repositioning is an attempt to fix the second without giving up the first.

How it fits into the group. This is unambiguously the engine. In Q2 2026 the Consumer segment produced an operating margin of 30.6%, up from 28.7% a year earlier, and an EBITDA margin of 45.0% (TelecomLead Q2 2026 breakdown). Every strategic initiative Schulman has announced - Verizon One, Simplicity, the loyalty program, the convergence push - is aimed at this segment.

2.2 Verizon Business Group (~21% of revenue)

What it does. Business sells to enterprises, government agencies, small and mid-sized businesses, and other carriers. The 10-K describes a portfolio spanning mobility, fixed wireless and wireline broadband, IoT connectivity, advanced communications, corporate networking, local and long-distance voice, and security and managed network services. At end-2025 the segment had approximately 31 million wireless retail postpaid connections and 3 million total broadband connections.

The segment is organised by customer type:

• Enterprise and Public Sector. Large corporations and government at federal, state and local level. FY2025 revenue in this sub-group was $13.5 billion. This is the part that has been shrinking.
• Business Markets and Other. Small and mid-market businesses. Growing.
• Wholesale. Selling network capacity to other carriers, MVNOs and cable operators. Structurally declining as legacy wholesale voice and data contracts run off.

The core capability. Two things Consumer cannot do. First, running a network that meets government security and reliability standards - Verizon explicitly states it does not use network equipment from vendors under US national-security restrictions, which matters enormously for federal contracts. Second, the long-haul and metro fibre estate. Verizon owns route miles of fibre laid over decades, much of it inherited from the MCI-era long-haul network and Bell Atlantic's regional plant. That estate is the reason a business built for enterprise voice traffic is suddenly interesting to AI hyperscalers.

Why it exists separately. The Business segment carries a fundamentally different cost and margin structure - Q2 2026 operating margin was 13.9% against Consumer's 30.6%. It also carries regulatory and procurement obligations Consumer does not. And critically, chunks of it are declining assets. Legacy wireline voice, wholesale, and international enterprise networking are melting ice cubes, which is exactly why Verizon has been reshaping the segment rather than growing it.

Competitive position. Against AT&T Business and Lumen in enterprise networking; against T-Mobile for Business in wireless; against cloud providers and specialist managed-service firms in the value-added layers. The 2025 revenue decline came from Enterprise and Public Sector plus Wholesale, partly offset by Business Markets growth. Federal government accounts were a specific drag in 2025, hit by government efficiency measures that disconnected lines (Q1 2025 and Q3 2025 concalls).

How it fits into the group. Business has been treated as a portfolio to be rationalised rather than a growth engine - until AI arrived. Two 2026 decisions reframed it:

The BT joint venture, announced 29 June 2026, combines Verizon's international enterprise wireline arm with BT International into a 50:50 JV serving more than 3,000 multinational customers across more than 180 countries with roughly $4 billion in combined annual revenue. Verizon pays BT a $625 million equalisation payment. Expected cost synergies are approximately $200 million annually, and completion is expected in 2027 (Verizon press release, 29 June 2026). Verizon classified the contributed business as held-for-sale in Q2 2026. Martijn Blanken has been appointed CEO-designate, joining BT from 1 September 2026; per BT and Verizon's joint announcement, he has spent nearly 30 years in senior telecom, technology and digital infrastructure roles including at Telstra, Openwave Systems, EXA Infrastructure and KPN.

AI Connect, launched January 2025, is the other half. This is Verizon selling fibre capacity and edge compute to AI infrastructure builders. On the Q2 2026 call Verizon disclosed an agreement with Google valued at over $1 billion to use Verizon dark fibre to connect Google's data centres (Fierce Network, July 2026). Schulman's framing captures why a declining enterprise segment suddenly has an option attached to it:

"We built that infrastructure for a different era, but it has turned out to be exactly the right asset for this one."

• Dan Schulman, Q2 2026 concall, 24 July 2026

2.3 Segment comparison

3. Products And Business Detail

3.1 The wireless product catalogue

Simplicity. Launched 2026 as the flagship consumer wireless plan. One tier, no network segmentation - every Simplicity customer gets Verizon's best 5G performance rather than the throttled-then-upsold structure that dominates US wireless. Pricing is $45 per month per line with autopay for existing customers moving onto it, and $30 per line with autopay for customers porting a number in from another carrier (Verizon plan comparison). The strategic point is that it removes the two things that drove Verizon's churn: opaque add-on pricing, and the sensation that new customers get better deals than loyal ones.

Verizon One. The converged offer: mobility plus home broadband from $70 per month with autopay, with all taxes and fees included in the headline number, a router and professional installation included, and no mobile activation or upgrade fees. Aimed at customers new to Verizon who want both services.

The loyalty program. Available to every Verizon customer with no plan change required - a deliberate design choice, since requiring a plan migration is exactly the friction that makes loyalty programs fail. It provides monthly cashback in "Verizon Dollars," daily promotional offers, and eliminates activation and upgrade fees. As of June 2026, Verizon was reported to be the only major US carrier to have removed those fees for opted-in customers (Fierce Network, June 2026).

Legacy myPlan and myHome. The prior generation of tiered plans with paid "Perks" add-ons - streaming subscriptions, cloud storage, international calling - bolted on at $10 increments. Verizon also launched the Verizon Value Guarantee in 2025, offering a three-year price lock on the network portion of myPlan and myHome plus a free-phone guarantee with trade-in (Q1 2025 concall, 22 April 2025).

The Verizon Value prepaid stable. Straight Talk, Visible, TracFone, Simple Mobile, SafeLink, Walmart Family Mobile and Verizon Prepaid, serving over 22 million customers. Visible is the digital-native, app-only brand. Straight Talk and Walmart Family Mobile are distributed through Walmart. SafeLink serves the federal Lifeline subsidy programme. Total Wireless is the growth vehicle in physical retail, at 2,000 stores in 2026. Core prepaid has now posted eight consecutive quarters of positive net additions through Q2 2026.

3.2 Broadband products

Fios. Fibre-to-the-premises, delivering symmetric gigabit-class service. Historically confined to Verizon's legacy Bell Atlantic footprint in the Northeast and Mid-Atlantic. Fios also carries a video product, though the video base has been shrinking for years as households cut the cord - roughly 2 million Fios video connections remained at end-2025.

Fixed Wireless Access (5G Home / Business Internet). A router in the home receives a 5G signal from a nearby cell site. This is a genuinely clever piece of capital arbitrage: Verizon bought C-band spectrum and built C-band radios for mobile phones, then discovered that the same radios had spare capacity that could be resold as home broadband at close to zero incremental network cost. Where fibre is unavailable, FWA typically delivers speeds above 300 Mbps.

The economics only work where there is spare capacity, which is why FWA availability is address-by-address rather than blanket. Verizon has targeted doubling its FWA footprint to 90 million homes and businesses covered by 2028, using C-band plus mmWave (Verizon broadband strategy update). The subscriber target is 8-9 million by 2028, raised after the original 4-5 million goal was hit fifteen months early. At Q2 2026, combined FWA and fibre stood at 17.1 million connections against a fibre base of 10.913 million.

Verizon also acquired Starry to strengthen its multi-dwelling-unit capability, announced during 2025 (Q3 2025 concall, 29 October 2025). Apartment buildings are the hardest broadband target - fibre requires landlord agreements and in-building risers - and mmWave-based MDU service is a workaround.

3.3 Business products

Enterprise networking and managed services. Software-defined wide-area networking, private IP, managed security, unified communications. Sold on multi-year contracts to corporations and government.

Private 5G and IoT. Dedicated cellular networks on customer premises - ports, factories, stadiums - plus embedded connectivity for connected vehicles, industrial sensors and fleet telematics.

Wholesale. Selling capacity to MVNOs and other carriers. Structurally shrinking.

AI Connect. The newest line, launched January 2025. Verizon sells dark fibre - unlit glass strands the customer lights with its own optical equipment - along with data-centre interconnect and edge inference capacity. Verizon has described retrofitting central offices as edge inference nodes, which turns hundreds of legacy telephone exchanges from stranded real estate into distributed compute locations. The Q2 2025 call disclosed the AI Connect sales funnel had nearly doubled to $2 billion since launch (Q2 2025 concall, 21 July 2025).

3.4 How the network is actually built

The delivery chain runs roughly like this:

1. Spectrum acquisition. FCC auctions grant exclusive licences to transmit on specified frequencies in specified geographic areas. Verizon's holdings span low-band (wide coverage, low capacity), mid-band C-band (the workhorse), and high-band mmWave (enormous capacity, blocked by walls and leaves).

2. Radio access network. Antennas and radios mounted on macro towers - typically leased from tower REITs - and on small cells attached to street furniture in dense areas. Verizon had deployed roughly 17,000 mmWave cell sites as of 2021 and has been layering C-band equipment onto macro towers since.

3. Fibre backhaul. Every cell site needs a fibre connection back to the core. This is the hidden reason mobile and fixed businesses converge: the fibre Verizon builds to a neighbourhood to serve homes can also backhaul the tower serving that neighbourhood, and vice versa. One trench, two revenue streams.

4. Core network and IT. Routing, authentication, billing, provisioning. Increasingly automated - by Q1 2026 Verizon reported that AI systems autonomously resolved 85% of network issues and had generated over $200 million in energy savings, with its proprietary AI tech stack described as substantially complete by July 2026 (Q1 2026 concall, 27 April 2026).

5. Distribution. Company-owned stores, authorised retailers, big-box channels for the value brands, digital direct, and enterprise field sales.

6. Device financing. Handsets sold on 36-month interest-free instalment plans, creating a receivable that Verizon securitises - $28.8 billion of asset-backed debt at 30 June 2026 was secured by device payment plan receivables.

3.5 The fibre supply chain, and a notable September 2026 commitment

On 8 September 2026, Verizon and Corning announced a multi-year, multi-billion-dollar supply agreement covering 80+ million miles of high-density optical fibre and connectivity solutions from 2027 to 2032 (GlobeNewswire, 8 September 2026). Corning will supply its Contour Flow Cable, a flexible ribbon design that fits substantially more fibre strands into existing conduit - which matters because the expensive part of building fibre is digging, not the glass.

The stated dual purpose is the tell: the same fibre programme serves both the 40-50 million broadband passings target and the long-haul backbone for AI hyperscalers. Kyle Malady, Executive Vice President and CEO of Verizon Business Group, framed it as building "the network of the future at an unprecedented scale." Corning's Mike O'Day, Senior Vice President and General Manager of Corning Optical Communications, described generative AI and broadband expansion as "part of the same transformative opportunity." The partnership spans three decades.

3.6 Geography

Verizon is a domestic US business with an international enterprise tail. The wireless network is national. The fibre footprint, post-Frontier, covers 31 states plus Washington DC - a step change from the legacy Northeast concentration. The international enterprise wireline business, which served multinationals across more than 180 countries, is being contributed to the BT joint venture, after which Verizon's direct international presence largely disappears while it retains the US connectivity relationship with those same multinational customers.

3.7 Milestones that changed the business

4. Customers

4.1 Who actually buys

Households. Roughly 116 million Consumer wireless retail connections at end-2025, of which 83% postpaid, plus 11 million broadband connections. The typical postpaid account is a family with three to five lines, an average tenure measured in years, and a switching decision that is made perhaps once every three to five years - usually triggered by a phone upgrade, a bill shock, or a move to a new home.

Prepaid customers. Over 22 million across the Value brands. Different behaviour entirely: monthly decision-making, high price sensitivity, churn several multiples of postpaid, and distribution through Walmart and independent dealers rather than Verizon stores.

Small and mid-sized businesses. Bought largely on the same wireless economics as consumers but with volume pricing and light managed services. This is the growing part of the Business segment.

Large enterprises and the public sector. FY2025 Enterprise and Public Sector revenue was $13.5 billion. Federal, state and local government agencies, plus Fortune-scale corporations. These accounts buy networking, security and managed services on multi-year contracts.

Other carriers and MVNOs. Wholesale customers buying network capacity.

AI infrastructure builders. The newest category, and the one with the largest single contract disclosed to date - Google, over $1 billion for dark fibre.

4.2 Who makes the decision, and on what basis

Consumer. The decision-maker is whoever pays the family bill. The criteria, in rough order: does the phone work where I live and work; what does the bill actually come to after taxes and fees; how painful is switching; what do I get for the phone I'm trading in. Sales cycle is minutes to days. This is exactly why the Simplicity design - one price, taxes included - is a competitive act rather than a marketing one: it attacks criterion two directly, and criterion two is where Verizon was losing.

SMB. An owner or office manager. Criteria are coverage, price per line, and whether the account gets a human being on the phone when something breaks. Cycle: days to weeks.

Enterprise. A procurement committee involving the CIO or CISO, network architects, and procurement. Criteria are security posture, service-level agreements, geographic coverage, integration with existing cloud and SD-WAN estates, and vendor financial stability. Sales cycles run six to eighteen months.

Public sector. Formal RFP processes, often with security clearances, domestic-supply requirements, and multi-year framework contracts. Verizon's explicit statement that it uses no network equipment from restricted vendors is a procurement qualification, not a marketing line.

AI infrastructure. A hyperscaler's network infrastructure team. The criteria are route diversity, available conduit and fibre count on specific corridors, latency, and delivery timelines. There is no brand loyalty here; there is only whether you own glass on the route between two data centres. Verizon's advantage is that it already does.

4.3 Why they choose Verizon

For consumers, historically: network reliability. Verizon's brand equity was built on coverage, and that equity is real but was being monetised too hard. The 2026 answer adds price transparency and a loyalty program with no migration friction.

For enterprises: the combination of national wireless, owned fibre, and a security-vetted supply chain. Very few vendors can deliver all three under one contract in the US.

For AI builders: physical fibre on the specific routes they need, available now, without a permitting cycle.

4.4 Switching costs

Consumer switching costs are moderate and have been falling for a decade. Number portability is a legal right. Devices are increasingly unlocked. What remains:

• Device payment balances. A customer mid-way through a 36-month instalment plan owes the remaining balance on switching, unless a competitor buys them out - which competitors routinely do.
• Family plan gravity. Multi-line accounts require the whole household to agree to move.
• Convergence lock-in. This is the one Verizon is deliberately building. Fibre customers churn approximately 40% less than mobile-only customers, and convergence penetration exceeded 18% of the consumer base as of Q3 2025. A converged household has to schedule a technician, change a router, and re-run every smart device in the house to switch. That is real friction, and it is the strategic point of Verizon One.
• Loyalty accrual. Verizon Dollars accumulate and are forfeited on departure.

Enterprise switching costs are high: multi-year contracts, integration work, security recertification, and the operational risk of migrating a live network. Public-sector switching costs are higher still because of procurement rules.

4.5 Concentration

Consumer revenue is atomised across roughly 116 million retail connections - about as unconcentrated as a customer base can be. There is no single-customer risk in Consumer.

The Business segment carries genuine concentration risk in a different form: the federal government is a large aggregate customer, and government efficiency measures produced public-sector disconnects that management flagged as a specific drag in both the Q1 2025 and Q3 2025 calls. That is concentration by counterparty type rather than by named account.

AI Connect is currently the opposite extreme - one disclosed contract of over $1 billion with a single customer. Management has said more agreements are expected, but as of the Q2 2026 call this is a business with a very short customer list.

4.6 Contract structure and revenue predictability

Consumer postpaid is month-to-month service with a separate 36-month device instalment obligation. There is no service contract lock-in in the traditional sense, which is precisely why churn is the metric management obsesses over: revenue predictability comes from behaviour, not from paper.

Prepaid is genuinely month-to-month with no obligation.

Enterprise is multi-year contracted with committed spend and SLAs, giving high visibility but slow growth and long re-bid cycles.

AI Connect dark fibre agreements are long-dated capacity leases - the closest thing in the portfolio to genuinely contracted, multi-year, high-visibility revenue, though management has guided that meaningful revenue contribution begins in 2027 rather than 2026.

The overall shape: a very large recurring base with modest but real monthly attrition, a shrinking transactional equipment layer that management is deliberately compressing, and a small but growing contracted infrastructure business.

5. Competitive Landscape

US wireless is a three-player oligopoly with two adjacent invasions in progress - cable operators coming in from below via MVNO economics, and satellite coming in from above. Understanding Verizon requires taking both seriously.

5.1 The core three

T-Mobile US. The share leader by most counts - roughly 35% of the US mobile market against Verizon's ~34% and AT&T's ~27% as of December 2024 (mobile market share compilation). T-Mobile's structural advantage came from the Sprint merger, which handed it a deep 2.5 GHz mid-band position years before Verizon's C-band was cleared. It has spent the intervening years converting that head start into subscriber share and into the largest FWA base of the three. Verizon competes against T-Mobile on network quality perception and, since 2026, on price transparency. It loses to T-Mobile on brand momentum with younger and lower-ARPU customers.

AT&T. The closest structural analogue to Verizon: similar Bell heritage, similar debt-heavy history, similar convergence strategy built on a large fibre build. AT&T added 421,000 net postpaid wireless subscribers in a recent 2026 quarter, described as broadly in line with expectations. Verizon and AT&T compete for the same premium household and the same enterprise and federal accounts. Neither has a decisive edge; both are executing the same fibre-plus-wireless convergence playbook.

Cable: Comcast and Charter. These are not really wireless companies - they are broadband companies that resell wireless as a retention tool, running as MVNOs on other carriers' networks and offloading most traffic to their own WiFi. Their strategic function is to make broadband stickier by bundling mobile at aggressive prices. In 2026 both are in visible distress on their core product: Comcast lost 65,000 broadband subscribers in Q1 2026 and Charter lost 120,000 (Fierce Network, 2026). Of broadband switchers, roughly 43% move to fibre and roughly 25% to FWA - meaning Verizon, AT&T and T-Mobile collectively added more than 1.4 million broadband subscribers in Q1 2026 while cable shed customers. Comcast's CFO Jason Armstrong publicly described fibre broadband pricing as having grown "irrational" in September 2026, which is what a competitor says when someone else is winning on price.

Satellite: SpaceX / Starlink. This is the newest and least settled threat, and it repriced the entire sector in 2026. SpaceX listed on 12 June 2026 at $135 per share, closing its first day at $161 for a market capitalisation of roughly $2.1 trillion (CNBC IPO coverage, 12 June 2026). Its connectivity unit, primarily Starlink, generated $11.39 billion in 2025 - 61% of SpaceX's total sales - and Starlink reported 12 million subscribers, double the prior year (CNBC, 4 August 2026). On its first post-IPO earnings call SpaceX confirmed plans to build a consumer wireless carrier under the Starlink Mobile brand. Verizon shares fell 7% on 29 June 2026 on the news and T-Mobile hit a 52-week low.

Schulman's rebuttal on the Q2 2026 call was a physics argument rather than a marketing one: terrestrial networks are "100 times to 1,000 times more efficient" in suburban and urban markets, because a satellite beam covers an enormous area and must share its capacity across everyone under it, while a cell tower serves a few hundred metres. That is directionally correct for dense areas. It is much weaker for rural coverage, which is precisely where Starlink is strongest and where the incumbents' cost per subscriber is worst.

The competitive response is telling: on 14 May 2026, AT&T, T-Mobile and Verizon announced a joint venture to pool spectrum and create a unified direct-to-device satellite platform - the first time all three have formally aligned (AT&T announcement, 14 May 2026). Each keeps its existing satellite partnerships: T-Mobile with SpaceX, AT&T with AST SpaceMobile, and Verizon with both Skylo and AST SpaceMobile.

5.2 Peer comparison

5.3 Where Verizon wins, and where it is exposed

Wins on: network quality in dense markets, where mid-band depth and small-cell density give a real throughput advantage; converged household economics, where owning both the wire and the phone produces both lower churn and lower cost to serve; federal and security-sensitive enterprise, where the supply-chain position is a qualification barrier; and, newly, on owning fibre on routes AI builders need, where being incumbent beats being cheap.

Exposed on: rural coverage economics, where satellite structurally undercuts a tower-based cost curve; brand price perception, which was the diagnosed cause of three years of subscriber losses and which the 2026 repositioning is only one quarter into fixing; balance sheet flexibility, since a large debt load constrains how aggressively Verizon can respond to a price war; and equipment revenue, which fell roughly 20% in Q2 2026 as a deliberate consequence of pulling handset subsidies - a choice that improves margin but shrinks the top line and can only be sustained if churn stays low.

5.4 Barriers to entry

Genuinely high for terrestrial wireless, and it is worth being precise about why:

• Spectrum. Federally auctioned, finite, and expensive. Verizon's 2021 C-band cheque was $52.9 billion for one auction. There is no way to buy your way in without matching that order of magnitude.
• Site acquisition. Tens of thousands of tower leases, municipal permits, and rooftop agreements accumulated over decades. This is a slow, litigious, unglamorous barrier and probably the most durable one.
• Fibre in the ground. The Corning agreement covers 80+ million miles of fibre over 2027-2032. Digging trenches under existing streets is the binding constraint, not glass supply.
• Regulatory qualification. Federal and public-sector contracts require security postures new entrants cannot assemble quickly.

The barrier that is genuinely falling is the one satellite attacks: for a customer in a low-density area, the capital cost of terrestrial coverage is enormous and the satellite alternative bypasses all four barriers above. That is a real structural shift, not a cyclical one.

5.5 Structural shifts in progress

Three at once. Cable's broadband base is eroding to fibre and FWA, which is a tailwind for Verizon. Satellite is entering from the coverage edge, which is a headwind and has already repriced sector equities - T-Mobile's market capitalisation fell roughly 31% year on year and AT&T's roughly 15% as of mid-2026. And AI infrastructure demand has turned dark fibre from a stranded legacy asset into a contested resource, which is a tailwind for whoever owns route miles.

The honest framing is that these three shifts do not net out cleanly. Verizon is winning a share war against cable while simultaneously facing a technology entrant it cannot outspend, while holding an asset (fibre) whose value is being repriced upward by a customer set (hyperscalers) it has served for less than two years.

6. Industry

6.1 What drives demand

Data consumption. The volume of data per connection grows relentlessly with video resolution, cloud dependence and, increasingly, on-device AI. Verizon's own filings note that continued demand for spectrum is driven by growth in customer connections and increased usage of wireless broadband. This is the industry's structural tailwind: usage grows, but pricing per gigabyte falls faster, which is why volume growth does not automatically produce revenue growth.

Household formation and mobility. New households need connections. Moves trigger switching decisions.

Broadband substitution. Households are migrating away from cable, with roughly 43% of switchers going to fibre and 25% to FWA. This is a share reallocation, not market growth.

Enterprise digitisation. Cloud migration, remote work, IoT deployment and network security spend.

AI infrastructure buildout. The newest and most violent demand driver. Schulman described it on the Q2 2026 call as "one of the largest capital cycles of our lifetime." AI data centre designs use roughly 36 times more fibre-optic cable than traditional designs, and the connectivity between data centres - metro and long-haul - runs over exactly the telecom fibre plant Verizon and its peers already own.

6.2 Size and growth

The US wireless telecommunications carriers industry was sized at approximately $336.1 billion in 2026 by IBISWorld (IBISWorld market size). The narrower US telecom services market was valued at approximately $218.9 billion in 2026 and projected to reach $244.7 billion by 2033, a CAGR of about 1.6% (Persistence Market Research). Fixed internet and broadband services account for roughly 30% of that market in 2026.

The number that matters most for understanding Verizon: this is a low-single-digit-growth industry. There is no organic market expansion to ride. Every point of revenue growth is either a price increase, a share gain, or a new adjacent business. Verizon spent 2023-2025 taking the first option and lost subscribers; it is now attempting the second and third.

6.3 Where Verizon sits in the supply chain

Verizon is the retail-facing layer. Upstream it buys spectrum from the FCC, radio equipment from a small set of vendors (explicitly excluding restricted vendors), fibre and cable from Corning and peers, tower space from tower REITs, and handsets from Apple, Samsung and others which it resells largely at or near cost. Downstream it sells directly to households, businesses and government, and wholesale to MVNOs.

The AI Connect business inverts Verizon's usual position: instead of being the retail layer, Verizon becomes the infrastructure supplier to hyperscalers who are themselves the retail layer for compute. That is a lower-margin, longer-contract, higher-visibility business than consumer wireless, and it uses assets that are already built and largely depreciated.

6.4 Regulation

Heavily regulated, in several distinct ways:

• Spectrum. The FCC allocates and auctions licences and enforces build-out obligations. Spectrum policy - how much is released, on what timeline, at what auction terms - is the single biggest regulatory lever on the industry's long-term cost structure.
• Merger review. The Frontier acquisition required FCC, DOJ and state-level approvals, taking roughly a year from announcement to close and requiring approvals across thirteen states. This is a real constraint on consolidation as a strategy.
• Supply chain security. The FCC restricts equipment from certain vendors deemed national-security risks. Verizon states it uses none of that equipment.
• Cybersecurity and privacy. Verizon's 10-K notes that Congress and some states are considering increased regulation of provider cybersecurity practices, and that cyberattacks against companies including Verizon have increased in frequency, scope and potential harm - while stating none to date have been material to its operations or financial condition.
• Universal service and subsidy programmes. Lifeline and related programmes underpin part of the prepaid base (SafeLink). Changes to federal subsidy programmes flow directly into subscriber counts.

6.5 Cyclicality

US telecom is among the least cyclical large industries in the economy. Phone service is close to the last bill a household stops paying. That is why the sector is a defensive holding and why the equity has historically traded on dividend yield rather than growth.

But the industry has two non-obvious cyclical exposures. The first is handset cycles: equipment revenue swings with device launch timing and replacement cycle length, and lengthening replacement cycles compressed Verizon's equipment revenue by roughly 20% in Q2 2026. The second is interest rates: with a very large debt stack, refinancing costs move with the rate cycle, and the equity's yield-substitute character means it trades partly as a bond proxy.

6.6 Tailwinds and headwinds

Tailwinds: AI infrastructure demand repricing fibre assets upward; cable broadband share losses flowing to fibre and FWA; convergence economics rewarding operators who own both wire and spectrum; and the industry-wide shift away from promotional subsidy warfare, which improves everyone's unit economics if it holds.

Headwinds: satellite direct-to-device entry attacking the rural cost base with a fundamentally different capital model; near-zero underlying market growth, so all growth is zero-sum; regulatory scrutiny of consolidation limiting the classic telecom playbook; and lengthening handset replacement cycles reducing the equipment revenue line.

7. Growth Triggers

All items below are drawn from the six concalls listed in Section 9, or from company announcements referenced on those calls.

• Fibre passings expansion to 40-50 million medium-term, with at least 2 million added organically in 2026. Target raised from a prior 35-40 million range. (Q4 2025 concall, 30 January 2026; repeated Q1 2026 concall, 27 April 2026, where management said Verizon was on track to exceed 32 million passings by year-end 2026.)

• Frontier integration synergies of over $1 billion run-rate by 2028 - double the original estimate at announcement. (Q4 2025 concall, 30 January 2026; reaffirmed Q1 2026 concall, 27 April 2026, described as on track.)

• AI Connect revenue contribution beginning meaningfully in 2027, anchored by the disclosed Google dark-fibre agreement valued at over $1 billion, with management indicating further agreements expected before year-end 2026. (Q2 2026 concall, 24 July 2026.)

  "An agreement with Google, valued at over $1 billion, to use Verizon dark fiber to connect their data centers."

  • Dan Schulman, Q2 2026 concall, 24 July 2026

• $5 billion of operating expense reduction targeted for 2026 through workforce reduction, legacy network decommissioning, marketing efficiency, real estate rationalisation and contract renegotiation. (Q4 2025 concall, 30 January 2026; progress reported Q1 2026 concall, 27 April 2026, with cost of acquisition and retention down approximately 35% since Q4.)

• Verizon One, Simplicity and the loyalty program as the new customer acquisition engine, foreshadowed one quarter ahead and launched in Q2 2026. (Foreshadowed Q1 2026 concall, 27 April 2026; launched and reported Q2 2026 concall, 24 July 2026.)

  "We are in the final stages of extensive market research that will inform a new generation of offers built around the principles of transparency, simplicity and genuine value delivery."

  • Dan Schulman, Q1 2026 concall, 27 April 2026

• BT joint venture completion expected in 2027, combining Verizon's international enterprise wireline arm with BT International, targeting approximately $200 million of annual cost synergies. (Q2 2026 concall, 24 July 2026; announced 29 June 2026.)

• Fixed wireless access subscriber base to reach 8-9 million by 2028, with FWA coverage doubling to 90 million homes and businesses. (Reaffirmed Q1 2025 concall, 22 April 2025 and Q2 2025 concall, 21 July 2025; repeated across multiple periods.)

• Second half of 2026 stronger than the first half, and 2027 stronger than 2026, driven by core business momentum plus first contributions from AI infrastructure revenue. (Q2 2026 concall, 24 July 2026.)

  "The build-out of AI infrastructure across the United States is one of the largest capital cycles of our lifetime."

  • Dan Schulman, Q2 2026 concall, 24 July 2026

• Multi-dwelling-unit broadband expansion, using mmWave-based service plus the Starry acquisition, targeting apartment buildings that fibre cannot economically reach. (Q1 2025 concall, 22 April 2025, launching in 50+ markets; Starry acquisition announced and discussed Q3 2025 concall, 29 October 2025.)

• Tillman partnership for capital-light fibre expansion, extending the fibre footprint without full balance-sheet funding. (Q3 2025 concall, 29 October 2025.)

• Deleveraging to a net unsecured debt-to-EBITDA ratio of 2.0-2.25x by 2027, from 2.5x at Q2 2026. (Q4 2025 concall, 30 January 2026; reaffirmed Q1 2026 concall, 27 April 2026.)

• Corning fibre supply agreement covering 80+ million miles from 2027-2032, securing the physical input for both the broadband passings target and the AI long-haul backbone. (Announced 8 September 2026, after the Q2 2026 call.)

Trigger summary

8. Key Risks

8.1 Satellite direct-to-device structurally undercuts the rural cost base

Mechanism. Verizon's cost to serve a rural household is enormous: a tower, backhaul fibre, power, and site lease amortised across a handful of subscribers. Starlink's marginal cost to add that same household is near zero once the constellation is in orbit. If Starlink Mobile launches a credible consumer wireless service, the economics of Verizon's most expensive coverage tier invert, and Verizon cannot respond by cutting price without destroying the returns on assets it has already built.

Calibration. High probability of some impact, uncertain magnitude, and the market has already begun pricing it - Verizon fell 7% on 29 June 2026 when SpaceX confirmed its carrier plans, and T-Mobile hit a 52-week low. Starlink already had 12 million subscribers globally as of Q2 2026, doubled year on year.

What management says. Schulman's Q2 2026 counter-argument is a capacity argument: terrestrial networks are "100 times to 1,000 times more efficient" in suburban and urban markets because satellite beams must share capacity across huge footprints. That is a sound defence of dense markets and a weak defence of rural ones. The defensive move - the three-carrier direct-to-device joint venture announced 14 May 2026 - is an acknowledgment that the threat is real enough to require the first formal alliance in the industry's history.

8.2 The transformation depends on churn staying low, and churn is a behavioural bet

Mechanism. Verizon has deliberately dismantled its promotional machine. Equipment revenue fell roughly 20% in Q2 2026 as a direct consequence of reduced handset subsidies, and cost of acquisition and retention was down approximately 35% from Q4 2025 by Q1 2026. This produces margin expansion, but only works if customers stay without being paid to stay. If churn reverts - because a competitor launches an aggressive subsidy campaign, or because the loyalty program proves less sticky than modelled - Verizon faces a bad choice: re-enter the subsidy war and surrender the margin gains, or hold the line and lose subscribers again.

Calibration. Moderate probability, high impact. Consumer postpaid phone churn improved from 0.91% in Q3 2025 to 84 basis points in Q2 2026 - real progress across three quarters, but three quarters is a short series against a competitor set with strong incentives to disrupt it.

8.3 Fixed wireless access growth is decelerating against a public 2028 target

Mechanism. FWA is capacity-constrained by design: it only works where the mobile network has spare capacity at a given address. As mobile data usage grows, that spare capacity shrinks. FWA net additions have been trending down through the six quarters covered here - 261,000 in Q3 2025, 319,000 in Q4 2025, 214,000 in Q1 2026, 193,000 in Q2 2026. The 8-9 million subscriber target for 2028 requires that trend to reverse or the fibre build to carry more of the load than planned.

Calibration. High probability of a miss on the FWA-specific target; moderate impact, because fibre net additions have been growing in partial offset (155,000 in Q2 2026) and total broadband additions rose 12.3% year on year. This is a mix risk more than a total-broadband risk, but the mix matters: FWA is high-margin capacity arbitrage, fibre requires capital.

8.4 Debt load constrains strategic response

Mechanism. Total debt was $158.2 billion at end-2025, and Verizon assumed roughly $12.9 billion of Frontier debt at fair value on closing, of which $12.4 billion had been repaid by 30 June 2026. Net unsecured debt-to-adjusted-EBITDA stood at 2.5x at Q2 2026 against a stated target of 2.0-2.25x by 2027. A company deleveraging toward a target has limited capacity to absorb a price war, fund an unplanned capex acceleration, or bid aggressively in a future spectrum auction. It also has a large refinancing schedule exposed to rate moves. The 10-K explicitly lists the debt load as limiting financial flexibility.

Calibration. High probability of constraint, moderate impact under normal conditions, severe impact if it coincides with a competitive shock. The company is simultaneously deleveraging, funding a 40-50 million-passing fibre build, paying a growing dividend and running a $25 billion buyback authorisation. Something gives if cash flow disappoints.

8.5 Business segment decline is structural, not cyclical

Mechanism. Enterprise and Public Sector and Wholesale revenues both declined in 2025. Wholesale is running off as legacy contracts expire. Public sector was hit by government efficiency measures producing line disconnects, flagged in both the Q1 2025 and Q3 2025 calls. The BT joint venture removes the international enterprise wireline business from consolidated results entirely. Roughly 21% of revenue sits in a segment where most sub-lines are shrinking and operating margin is under half the Consumer level.

Calibration. High probability, moderate impact. The AI Connect opportunity is being asked to offset a structural decline, and as of Q2 2026 it consists of one disclosed contract with revenue contribution guided to begin in 2027.

8.6 Concentration risk in AI Connect

Mechanism. The AI infrastructure story rests on a single disclosed agreement, with Google, worth over $1 billion. Hyperscaler capital plans are volatile and change quickly. If the AI capex cycle slows - and there is active public debate about whether it will - the dark fibre pipeline could compress before it ever becomes material revenue. Verizon has committed to an 80+ million-mile fibre purchase from Corning over 2027-2032 partly on the strength of this demand.

Calibration. Moderate probability, moderate impact. The downside is bounded because most of the fibre serves the broadband passings target regardless, but a hyperscaler pullback would remove the most exciting part of the current equity narrative.

8.7 Execution risk under a new and rapidly changing management team

Mechanism. Schulman became CEO in October 2025. Sowmyanarayan Sampath stepped down as CEO of the Verizon Consumer Group on 4 February 2026 and separated from the company on 27 March 2026; Alfonso Villanueva, EVP and Chief Transformation Officer, took the Consumer role on an interim basis (Verizon executive leadership transition). As of the search window, the largest segment in the company - roughly 77% of revenue - is being led on an interim basis during the most significant strategic repositioning in a decade. Thirteen thousand positions were eliminated in Q4 2025.

Calibration. Moderate probability of execution slippage, moderate impact. The counter-evidence is that results have improved in each of the three quarters since the transition. But large cost reductions, a leadership vacancy in the core segment, and a full repricing of the consumer proposition happening simultaneously is a lot of concurrent change.

What management acknowledged. Schulman was explicit on his first call that this is a multi-year rebuild, not a quick fix:

"Cost reductions will be a way of life for us."

• Dan Schulman, Q3 2025 concall, 29 October 2025

8.8 Cybersecurity

Mechanism. Verizon holds location data, call records, payment information and network access for well over 100 million connections, plus federal government network contracts. The 10-K states that cyberattacks have increased in frequency, scope and potential harm, while noting that none to date have been material to operations or financial condition. A material breach would carry regulatory, litigation, remediation and reputational costs, and reputational damage in a business where the product is trust in the network is qualitatively worse than in most industries.

Calibration. Low-to-moderate probability of a material event in any given year; potentially severe impact. This is a tail risk, not a base-case drag.

9. Walk The Talk

Concalls used (six most recent):

1. Q1 2025 - 22 April 2025 (Hans Vestberg, CEO)
2. Q2 2025 - 21 July 2025 (Hans Vestberg, CEO)
3. Q3 2025 - 29 October 2025 (Dan Schulman, first call as CEO)
4. Q4 2025 - 30 January 2026 (Dan Schulman)
5. Q1 2026 - 27 April 2026 (Dan Schulman)
6. Q2 2026 - 24 July 2026 (Dan Schulman)

This is an unusual credibility assessment because it spans a CEO change. The first two calls and the last four have to be judged separately, and the contrast is the most informative thing in this section.

The Vestberg period: a diagnosis that did not survive contact with the data

The Q1 2025 call, on 22 April 2025, opened with a bad number: 356,000 postpaid phone net losses, driven by pricing actions taken in December and January. Management's framing was that this was temporary. Sam Path, then Consumer Group CEO, described the elevated churn as "transitory" and told the market to expect normalisation in the second half of 2025. He also delivered a line that reads uncomfortably in hindsight:

"We are back to leading the market, not reacting to it."

• Sam Path, Consumer Group CEO, Q1 2025 concall, 22 April 2025

By Q2 2025, on 21 July 2025, consumer postpaid phone losses had narrowed to 51,000 from 109,000 a year earlier - genuine improvement - and the company raised full-year guidance for adjusted EBITDA, adjusted EPS and free cash flow. Vestberg's read on it:

"Our performance in the first half of the year demonstrates that our strategy is working."

• Hans Vestberg, Q2 2025 concall, 21 July 2025

But churn had not moved. Consumer postpaid phone churn stayed at 0.90%, identical to Q1, which management attributed to "residual effects of first-quarter pricing actions" and "elevated competitor promotional activity." The "transitory" diagnosis had been given a quarter and had not been validated. Meanwhile the Business segment deteriorated sharply: phone net additions fell to 42,000 from 135,000 a year earlier, mostly on public sector weakness.

Then came Q3 2025, on 29 October 2025. Postpaid phone net additions were a loss of 7,000, and churn ticked up to 0.91%. The second-half normalisation promised in April did not arrive. Three consecutive quarters of postpaid phone losses is not a transitory phenomenon; it is a business model producing the outcome it is designed to produce.

Assessment of the Vestberg-era calls: consistently optimistic, and wrong on the specific, testable claim. The company hit its financial guidance - free cash flow guidance of $19.5-20.5 billion was reaffirmed and ultimately delivered at $20.1 billion - but the operational diagnosis was wrong for three straight quarters, and management kept describing a strategy as working while its central metric moved the wrong way.

The Schulman period: three quarters, three deliveries

Schulman's first call, Q3 2025, is notable because he refused the prior narrative entirely:

"For years, our financial growth relied too heavily on price increases, a strategic approach that relies too much on price without subscriber growth is not sustainable."

• Dan Schulman, Q3 2025 concall, 29 October 2025

He made three specific, checkable commitments on that call. First, that Verizon would "aggressively" exit or shut down legacy businesses lacking a clear path to profitability. Second, that the detailed cost and portfolio plan would be presented on the January 2026 earnings call. Third, that "Verizon will no longer be the hunting ground for competitors" - a churn commitment.

He also set the tone that a lot of subsequent capital allocation flowed from:

"We will be a simpler, leaner and scrappier business."

• Dan Schulman, Q3 2025 concall, 29 October 2025

Promise one, the January detail: delivered. The Q4 2025 call on 30 January 2026 laid out the full plan - $5 billion of 2026 operating expense savings, 13,000 positions eliminated in Q4 2025, Frontier synergies doubled to over $1 billion by 2028, capex cut to $16-16.5 billion (down $4 billion from combined Verizon and Frontier 2025 spending), a $25 billion buyback authorisation, and a fibre passings target raised from 35-40 million to 40-50 million. That is a specific, dated commitment kept precisely.

Promise two, churn: delivered so far. Churn went 0.91% in Q3 2025, 0.90% in Q1 2026 with an exit rate below 0.85% in March, and 84 basis points in Q2 2026. Schulman's Q2 2026 framing:

"Consumer postpaid phone churn of 84 basis points is a step change for Verizon."

• Dan Schulman, Q2 2026 concall, 24 July 2026

Promise three, exiting legacy businesses: delivered in substance. The BT joint venture, announced 29 June 2026, removes the international enterprise wireline business. That is the "aggressive exit" made concrete.

A fourth, added mid-stream and delivered on schedule. On the Q1 2026 call, 27 April 2026, Schulman said the company was "in the final stages of extensive market research that will inform a new generation of offers." One quarter later, Verizon One, Simplicity and the loyalty program launched. Saying what is coming next quarter and then delivering it next quarter is the cheapest and most reliable credibility test there is, and Verizon passed it.

Guidance behaviour has been conservative, then raised twice. Initial 2026 adjusted EPS guidance was set at 4-5% growth on the Q4 2025 call. It was raised to 5-6% on the Q1 2026 call, and to 6-7% on the Q2 2026 call. Free cash flow guidance moved from roughly 7% growth to 9-10%. The buyback commitment moved from a $3 billion minimum to up to $4.5 billion. Postpaid phone net additions guidance of 750,000 to 1 million was narrowed to the upper half of that range. Two consecutive raises against initial guidance set only six months earlier is the pattern of management that under-promised.

What has been quietly dropped or is quietly slipping

Two items deserve flagging.

**The FWA 8-9 million by 2028 target ** was reaffirmed in the Q1 2025 and Q2 2025 calls and has not been formally withdrawn, but FWA net additions have declined in each of the last three reported quarters - 319,000 in Q4 2025, 214,000 in Q1 2026, 193,000 in Q2 2026. Management has reframed the discussion around total broadband additions rather than FWA specifically, and on the Q2 2026 call the answer to FWA questions was "broadband is broadband." That is a legitimate strategic point and also a change of subject. The target has not been repeated with the emphasis it once had.

The Perks subscription target of 15 million by year-end 2025, stated on the Q1 2025 call, was not revisited in any of the subsequent five calls reviewed here. The entire Perks construct has been superseded by the loyalty program, so the metric has been rendered obsolete rather than missed, but it was never reconciled.

Revenue is being sacrificed for margin, and management has been transparent about it. Total Q2 2026 revenue declined 0.7% year on year and came in below consensus, because equipment revenue fell roughly 20% as subsidies were pulled. Schulman flagged the trade explicitly a quarter earlier:

"We are purposely shifting our mix towards durable recurring service revenues and away from low-margin, highly promotional activity."

• Dan Schulman, Q1 2026 concall, 27 April 2026

That is a defensible strategy stated in advance rather than an outcome explained after the fact.

Promise versus outcome

Verdict

Two different managements in one dataset. The Vestberg-era calls hit financial guidance but repeatedly mis-diagnosed the operating problem, describing a strategy as working across three quarters in which its central metric deteriorated. That is not dishonesty; it is a management defending a model it had committed to.

The Schulman-era record is short - three calls - but every specific, dated commitment made in that window has been met or beaten, initial guidance was set low and raised twice within six months, and the one genuinely hard number (churn) has moved in the promised direction for three consecutive quarters. Management has also been explicit in advance about the costs of its own strategy rather than explaining them afterwards.

Assessment: on the evidence of the last four quarters, this is management that does what it says. The caveats are that the sample is three quarters long, that the easiest wins in a cost programme come first, and that the FWA 2028 target inherited from the previous regime is drifting without being formally retired. A management team earns its reputation by what it does when the trend turns against it, and that test has not yet arrived.

10. Shareholder Friendliness Index

Dividends. Verizon has raised its dividend every year for over two decades. Dividend per share was $2.56 in 2022 (four quarterly payments of $0.640), $2.61 in 2023 ($0.6525 quarterly), $2.66 in 2024 ($0.665 quarterly), $2.71 in 2025 ($0.6775 quarterly), and $2.83 annualised in 2026, with the quarterly rate lifted to $0.7075 (stockanalysis.com dividend history). The Q4 2025 call on 30 January 2026 announced a $0.07 annualised increase, a 2.5% raise, marking the 20th consecutive annual increase by the company's own count at that time. On 9 September 2026 the board declared a quarterly dividend of 70.75 cents payable 2 November 2026 to holders of record on 9 October 2026 (GlobeNewswire, 9 September 2026). The pattern is unmistakable: a growing dividend, but growing at roughly 2-3% a year, which is a maintenance raise rather than a signal of confidence. The payout ratio of approximately 74% is worth noting only because it explains why the raises are small - the dividend already absorbs the large majority of earnings, leaving the growth in it capped by the growth in the business.

Buybacks and dilution. This is where the record changes sharply, and the two windows must be read separately. For the three years 2022, 2023 and 2024, Verizon repurchased no shares at all under its then-authorised programme, a February 2020 authorisation covering up to 100 million shares, per the company's own 10-K disclosures. That inactivity extended through 2025. Then on 30 January 2026 the board authorised a new programme of up to $25 billion over three years, with a stated minimum of $3 billion in 2026 and a broader commitment to return approximately $55 billion to shareholders through the end of 2028 (Form 8-K, 30 January 2026). Execution has run ahead of the plan: accelerated share repurchase agreements entered in February 2026 delivered 50,758,023 shares in total at an average repurchase price of $49.25, completed in March, with a second ASR completed in June; $3.5 billion was repurchased in the first half of 2026, and the full-year target was raised on the Q2 2026 call to up to $4.5 billion. Shares outstanding stood at 4,154,775,202 at 30 June 2026. The three-year direction of travel on share count was therefore flat-to-slightly-up through 2025 on routine equity compensation, and only began shrinking in 2026.

Verdict: Returns Capital - a reliable if slow-growing dividend for over two decades, now joined for the first time in years by a genuine buyback that has already been executed ahead of its own minimum and raised mid-year.

11. Insider Activities

Verizon is US-listed on the NYSE, so the primary source is SEC Form 4 filings via EDGAR. The transactions below are drawn from Form 4 filings on EDGAR and cross-checked against Form 4 aggregators. No MoatMap database block was supplied for this company, so EDGAR is the sole spine here.

Recent transactions

Buys - reading the signal

There were no open-market purchases by any Verizon insider in the twelve months to September 2026. Not by the CEO, not by the CFO, not by any director.

This deserves to be stated plainly rather than softened, because it is the single most conspicuous fact in this section. Dan Schulman took over in October 2025 promising the largest strategic reset in the company's modern history, delivered three consecutive quarters of improving results, raised guidance twice, and has not bought a share of Verizon on the open market. The recurring Schulman and Skiadas Form 4s that appear frequently in EDGAR are not purchases: they are biweekly accruals of unitised phantom stock under the company's Deferred Compensation Plan, economically tied to the share price but cash-settled, not shares, and the "price" shown on those filings (in the $11.99-$14.11 range) is a unit accounting price, not a share price. Anyone scanning an aggregator quickly could easily mistake them for buying. They are not.

The Villanueva grant of 47,754 shares on 7 July 2026 and the April 2026 director retainer awards of 5,062 shares each are likewise grants, not conviction expressions. An insider who receives shares as pay has revealed nothing about what they think the shares are worth.

The absence of buying is not evidence of a negative view. Verizon executives already hold large equity positions through annual grants, and executives at very large-cap companies buy on the open market far less often than executives at small caps. But if the thesis is that this is a transformation being underestimated by the market, the people running it have not put personal capital behind that view.

Sells - working out the why

Kyle Malady has been selling 1,100 shares roughly weekly since August 2026, at values of $53,000-$56,000 per transaction. Every filing discloses a Rule 10b5-1 trading plan adopted on 18 May 2026 - a pre-scheduled, pre-committed plan established more than two months before the trades began and before the Q2 2026 results. Trades executed under such a plan carry no information about the executive's current view, because the decision was made months earlier and the executive cannot alter it. His holding has declined from 109,866 to 106,566 across the sequence, a reduction of roughly 3% of his direct position. Reason: disclosed 10b5-1 plan; this is routine diversification, not a signal.

Hans Vestberg sold approximately 225,000 shares on 24 February 2026 at roughly $49.61, representing about 32% of his direct holding at the time, and exercised options for 101,000 shares on 27 February 2026, selling at an average of $50.14. Context matters enormously here: Vestberg ceased to be CEO in October 2025 and remains on the board. A departing chief executive monetising a portion of an equity position accumulated over seven years as CEO is the most ordinary thing in this dataset. The specific reason is not disclosed in the filings, but the timing - four months after handing over the role, and in the window after the Q4 2025 results - is consistent with post-transition diversification rather than a view on the business. Reason: not disclosed in the filing; the post-CEO-transition context is the most plausible explanation and no other is inferable.

Sowmyanarayan Sampath, who stepped down as Consumer Group CEO on 4 February 2026 and separated from the company on 27 March 2026, made no insider transactions in Verizon stock over the preceding eighteen months. That absence is itself mildly interesting: a departing executive who does not sell is not signalling distress.

Routine option exercises with same-day tax-cover sales have been excluded as immaterial housekeeping, and no pattern in them was notable.

Net assessment

Insiders were net sellers over the twelve months to September 2026, with aggregate net selling of roughly $27 million across options and open-market transactions on one aggregator's count. But the composition matters more than the total. Essentially all of it is explained by two people and two ordinary situations: a former CEO monetising part of a long-accumulated position after handing over the role, and one divisional CEO running a mechanical pre-scheduled sales plan. There is no cluster selling, no acceleration into good news, and no selling by the CFO or by any director outside routine compensation mechanics.

Set against that, there is a complete absence of open-market buying during a period in which management has publicly described the company as being at "a structural and meaningful inflection." That combination - no distress signals and no conviction signals - is genuinely uninformative in both directions.

Plain-language read: neutral. Nothing here should worry a shareholder, and nothing here should encourage one. The insider record is silent, and silence is the correct interpretation.

12. Scenarios

Bull case

The convergence thesis works, and it works because the physics of household inertia are stronger than the physics of satellite beams. By 2028 Verizon has pushed past 40 million fibre passings on the way to 50 million, and a large share of those homes buy Verizon One - one bill, one price, one relationship covering the phones and the wire. Churn stays below 85 basis points not because the loyalty program is clever but because switching means scheduling a technician and re-provisioning every device in the house. The subsidy war that competitors keep threatening to reignite never fully returns, because the whole industry discovers what Verizon discovered: paying customers to switch and then paying them again to stay is a machine that destroys its own returns. Equipment revenue stops being a drag once the replacement cycle normalises, and service revenue growth accelerates through the mid-single digits as the mix shifts.

Meanwhile the fibre estate turns out to be the most undervalued asset in American infrastructure. The Google dark fibre agreement was the first of several, not the only one. AI data centre construction keeps demanding metro and long-haul interconnect on exactly the corridors Verizon already owns, the Corning agreement locks in supply for six years at scale when everyone else is scrambling for fibre allocation, and AI Connect becomes a genuinely large contracted business with multi-year visibility - the opposite of consumer wireless in every respect that matters to a capital allocator. Central offices that were stranded real estate become edge inference nodes. The Business segment, which had been a melting ice cube, stops melting.

Satellite turns out to be complementary rather than substitutive, exactly as the three-carrier joint venture is designed to make it. Starlink Mobile takes rural share nobody was making money on anyway, and the direct-to-device platform lets Verizon sell satellite connectivity as a premium feature on its own plans rather than losing customers to it. The cost programme runs deeper than the initial $5 billion, leverage falls into the 2.0-2.25x range on schedule, and with the balance sheet fixed the company runs a genuinely large buyback against a shrinking share count while continuing to raise the dividend. The equity, which spent 2025 and 2026 being priced as a melting telco, gets re-rated as an infrastructure asset with an AI option attached.

Base case

Schulman does roughly what he has said he will do, which on the evidence of three quarters is a reasonable expectation. Postpaid phone net additions land in the upper half of the 750,000 to 1 million range for 2026, churn holds somewhere in the low-to-mid 80s basis points, and total broadband additions keep growing even as the mix shifts from fixed wireless toward fibre. The FWA subscriber base grows but misses the 8-9 million by 2028 target, because capacity constraints bind and the target was set in a different competitive environment; management continues reframing the conversation around total broadband, and nobody much cares because total broadband keeps growing.

The cost programme delivers most of its $5 billion, with the easy wins landing first and the later tranches getting harder. Frontier synergies build toward the $1 billion run rate. The BT joint venture closes in 2027 and removes a declining business from the accounts without materially changing the earnings picture. AI Connect becomes a real but modest revenue line from 2027 - enough to offset structural decline elsewhere in the Business segment, not enough on its own to change the shape of the company.

Satellite remains a competitive irritant rather than an existential event over this horizon, because building a full consumer carrier is a multi-year exercise and the three-carrier joint venture gives incumbents a defensible interconnect standard. Verizon keeps raising the dividend at 2-3% a year, executes several billion dollars of buybacks annually against the $25 billion authorisation, and grinds leverage down toward target. Revenue growth stays in the low single digits. This is a well-run, cash-generative, slow business that has stopped shrinking its customer base - which, after three years of losing subscribers, is itself the achievement.

Bear case

The churn improvement turns out to have been a promotional artefact rather than a structural change. Verizon One and Simplicity attracted customers with genuinely aggressive introductory pricing - $30 a line for porting customers, $70 all-in for converged households - and when those cohorts hit their first repricing event, they behave the way telecom customers always behave. Churn drifts back toward 90 basis points. Verizon now faces the choice it was trying to avoid: re-enter the subsidy war and surrender the margin expansion that has been the whole story, or hold the line and watch the subscriber base shrink again. Equipment revenue, already down roughly 20%, does not recover, so the revenue line keeps declining while the market discovers that a margin story without a revenue story has a finite life.

Simultaneously, satellite stops being theoretical. Starlink Mobile launches with credible pricing and a network already serving 12 million subscribers, and it does not need to win dense markets to hurt Verizon - it only needs to take the rural and semi-rural customers whose cost to serve is highest, leaving Verizon with the same tower and backhaul cost base spread across fewer accounts. The three-carrier joint venture, which took the first formal alliance in the industry's history to assemble, proves to be a standards body rather than a competitive weapon. The sector repricing that already took roughly 31% off T-Mobile's market capitalisation and 15% off AT&T's in a year turns out to have been the market being early rather than wrong.

The AI story disappoints in the most ordinary way: the hyperscaler capex cycle cools, the pipeline behind the Google contract does not convert, and Verizon is left holding an 80-million-mile fibre purchase commitment running to 2032 against demand that was underwritten in a hotter market. Meanwhile the balance sheet, carrying a debt load built to buy out Vodafone twelve years ago and enlarged by Frontier, has no room to absorb a price war, an unplanned capex acceleration and a spectrum auction at the same time. Something has to give, and the things that give in that order are the buyback first and dividend growth second. The Consumer segment is still being run on an interim basis a year into the most important repositioning in a decade, and a new permanent leader arrives with their own view of what needs changing. The transformation becomes a sequence of transformations, which is how large telcos usually spend the decade after a promising first year.

13. Further Reading

• The OpenAI Hype Cycle, Microsoft's Game Pass Failure, Verizon's Satellites - Stratechery, 2025-10-09 [paid]

Sources: Verizon FY2025 Form 10-K (filed 17 February 2026); Verizon Form 10-Q for the quarter ended 30 June 2026; Verizon Form 8-K filings and Form 4 filings via SEC EDGAR; Verizon Q1 2025 (22 April 2025), Q2 2025 (21 July 2025), Q3 2025 (29 October 2025), Q4 2025 (30 January 2026), Q1 2026 (27 April 2026) and Q2 2026 (24 July 2026) earnings call transcripts; Verizon investor relations and press releases; BT Group newsroom; Corning/Verizon joint announcement (8 September 2026); AT&T newsroom (14 May 2026); IBISWorld; Persistence Market Research; CNBC; Fierce Network; Light Reading; Broadband Breakfast; stockanalysis.com.