Summary

  • NOS has the assets of a serious converged operator: a large Portuguese fixed base, national mobile scale, high reported 5G coverage, extensive fibre reach, pay-TV strength, business connectivity, RIPE NCC membership and a public internet footprint. Those assets make the company strategically relevant, but they do not automatically make each retained household value-creating.
  • The centre of gravity has moved from growth to conversion. In 2025 NOS reported EUR 1.823 billion of consolidated revenue, EUR 813.5 million of EBITDA, EUR 366.4 million of capex excluding leases and EUR 243.4 million of free cash flow. In the first quarter of 2026 revenue rose, EBITDA expanded and operating cash flow improved, but telecom consumer revenue and ARPU were under pressure.
  • The threat is not only MEO or Vodafone matching a promotion. Digi's entry and the broader low-price response have changed the reference price for mobile and fibre. NOS can defend value only if its bundle has a service, content, coverage or household-benefit reason to stay above the cheapest available substitute.
  • My judgment is conditional but clear: NOS can make convergence pay if capex really steps down, fibre take-up rises without buying customers with permanent discounts, and business/IT growth becomes a margin contributor rather than a distraction. If consumer ARPU keeps falling while lease, content and wholesale costs rise, convergence becomes a retention expense rather than a cash engine.

The Bundle Is A Financing Contract

The right way to read a Portuguese household telecom bill is not as a menu of services. It is a financing contract for infrastructure. One household may think it is buying television, broadband, Wi-Fi, two mobile numbers and a fuel or supermarket discount. NOS has to ask a harder question: which part of that bundle pays for the fibre drop, the router, the set-top equipment, the mobile radio network, the spectrum licence, the tower rent, the billing system, the customer service load and the next retention offer?

That is why fixed-mobile convergence can be both a powerful moat and a dangerous illusion. The moat comes from friction. A family with broadband, pay TV, mobile, security and loyalty benefits in one account has more to unwind than a SIM-only customer. The illusion comes when management counts every extra mobile card or promotional add-on as value creation while the incremental economics are thin. A fourth mobile card that stops churn can be worthwhile. A fourth mobile card that forces the operator to price the whole household at a challenger level is not.

NOS has spent years building the case that its household proposition is more than access. The group points to fibre, mobile network recognition, Wi-Fi quality, television experience, customer benefits through its Combina offers, and a brand attached to entertainment and everyday Portuguese consumption. The strategic logic is coherent. A household that links telecom service with grocery savings, fuel discounts, entertainment and a better in-home experience may become less price-sensitive. But that only works if the discount budget is lower than the churn and reacquisition cost it prevents.

The economic incentive is therefore explicit. NOS wants each additional service to increase household lifetime value. Customers want a lower total bill and fewer service failures. Rivals want to detach the most price-sensitive part of the account, usually the mobile SIM or the fixed broadband line, and then use that wedge to attack the rest of the relationship. The party carrying the downside is the network owner if the market trains customers to expect premium coverage, sports and fibre reliability at low-cost prices.

This article judges NOS on that incentive, not on the marketing language of convergence. The relevant questions are whether reported ARPU and revenue can hold, whether churn is controlled without permanent discounting, whether fibre homes passed become paying customers, whether mobile spectrum earns a return, whether content is a profit lever rather than a defensive tax, whether wholesale access reduces capital intensity without weakening differentiation, and whether tower economics remain manageable after passive infrastructure was monetised.

What NOS Comunicacoes Actually Operates

The public directory name is NOS COMUNICACOES, S.A., but the economic evidence is mostly reported through the listed parent, NOS, SGPS, S.A. That distinction matters. NOS COMUNICACOES is the operating telecom company inside a wider Portuguese communications, technology and entertainment group. The telecom boundary includes fixed broadband, pay TV, voice, mobile, business connectivity, wholesale and associated services. The wider group also includes information technology, cinema and audiovisual activities, and interests or partnerships in content and adjacent businesses.

For this research, the telecom company should be treated as a national converged operator, not merely as a RIPE NCC member and not merely as an article subject in a directory. RIPE membership and public routing evidence show that the business sits in the internet number-resource and interconnection world. They support the conclusion that NOS is part of Portugal's operating communications infrastructure. They do not, by themselves, prove the profitability of retail broadband, mobile, cloud, transit or managed services. The economic case has to come from revenues, customers, network coverage, capital spending and competitive position.

NOS reports a broad operating base. Its 2025 investor materials show total RGUs of about 10.935 million, mobile service RGUs of about 5.7 million, 1.6 million unique fixed accesses, 6.1 million households passed with fixed next-generation network coverage, 89.5 percent FTTH coverage and 99.6 percent 5G coverage. The same materials present market shares sourced to the Portuguese regulator for retail revenue, pay-TV subscribers, fixed broadband, used mobile access and multiple-play services.

Those numbers put NOS in the top tier of the Portuguese market, even if MEO remains the reference incumbent and Vodafone remains an aggressive converged competitor.

The listed group's 2025 revenue was EUR 1.823 billion, but the telecom segment remained the core, with EUR 1.592 billion of revenue and EUR 745.5 million of EBITDA in the group's institutional presentation. Cinema and audiovisuals are strategically useful because they strengthen content and entertainment positioning, yet they are not the main infrastructure funding source. IT became more important after the Claranet Portugal acquisition, but it also has different margins, sales cycles and integration risk.

That operating boundary is the first discipline for valuation. NOS is not a pure mobile operator, not a pure cable company, and not a software services group. It is a capital-intensive Portuguese converged operator trying to add higher-growth enterprise and IT exposure without letting consumer price pressure erode the cash produced by fixed and mobile scale. The economic promise is that these pieces reinforce each other. The risk is that each piece asks for capital, management attention or discount support at the same time.

Revenue Growth Is Not The Same As Value Creation

The top line gives NOS a respectable first answer. In 2025, consolidated revenue rose 1.6 percent to EUR 1.823 billion, EBITDA rose 4.3 percent to EUR 813.5 million and the EBITDA margin expanded to 44.6 percent. The group also reported free cash flow before dividends, financial investments and treasury shares of EUR 243.4 million, with EBITDA after leases less capex excluding leases rising to EUR 314.1 million. Net financial debt was about EUR 1.022 billion, with net debt to EBITDA after leases at 1.50 times.

Those figures are not weak. They show that NOS entered 2026 with a balance sheet that management and rating agencies could describe as conservative by sector standards, and with a free-cash-flow story that had moved beyond the heaviest phase of mobile and fibre expansion. The first quarter of 2026 reinforced that story at group level: consolidated revenue rose 1.9 percent to EUR 460.2 million, EBITDA rose 3.1 percent to EUR 203.3 million, the margin reached 44.2 percent, capex excluding leases fell 5.1 percent and operating cash flow rose 21.6 percent to EUR 85.9 million.

The value-creation question starts when the segment detail is separated. In the first quarter of 2026, telecom revenue slipped 0.2 percent to EUR 389.8 million. Consumer revenue fell 0.7 percent to EUR 285.8 million, while enterprise revenue rose 5.5 percent to EUR 81.3 million. Wholesale and other revenue fell 11.2 percent to EUR 22.7 million. NOS said consumer ARPU declined 0.8 percent, reflecting the new competitive context and storm-related effects. It also reported positive net additions in fixed services and postpaid mobile, suggesting that the customer base was not collapsing even as pricing was pressured.

That is the central tension. Revenue and EBITDA can improve while the consumer unit is losing pricing power. A converged operator can add RGUs and still dilute value if the incremental lines come at a lower unit price, higher equipment cost or higher retention discount. Conversely, a small decline in consumer revenue can be acceptable if it protects a large, profitable base and lets enterprise and IT growth carry the group. The difference depends on lifetime value, cost to serve and capital intensity, none of which is fully visible from public tables.

Management's claim is that the heavy investment cycle is ending and efficiency work is converting more EBITDA into cash. The numbers support that direction for now. The uncertainty is whether the market gives NOS enough time to harvest that investment before low-price competition forces another round of promotional spend. Free cash flow is the metric that should decide the argument. If reported EBITDA improves but cash is absorbed by leases, content, discounts, installation costs and maintenance capex, the shareholder does not receive the benefit of convergence.

Fixed Network Economics Decide The Household

Fixed access is the anchor of convergence because the home broadband line carries the most household friction. A mobile customer can port a number quickly. A home broadband customer has to deal with installation dates, router placement, Wi-Fi coverage, television equipment, family complaints and a higher perceived risk of disruption. That friction gives the fixed operator a chance to hold price. It also forces the operator to keep investing because the household notices poor Wi-Fi, slow installation and video failures more than it notices a corporate coverage claim.

NOS reports 6.1 million homes passed with fixed next-generation network coverage and FTTH coverage of 89.5 percent in 2025. The improvement from earlier years is material: the institutional presentation shows FTTH coverage rising from 54.0 percent in 2021 to 89.5 percent in 2025, with homes passed rising from about 5.1 million to about 6.1 million. That shift matters because a cable-heavy legacy can be profitable for television and broadband but can become a defensive liability when rivals sell fibre symmetry, Wi-Fi 7, 10 Gbps headline speeds or a simpler all-fibre story.

The economics are not solved by homes passed. Passing a home creates an option. The option becomes valuable only when take-up is high enough, ARPU is high enough and churn is low enough. If the operator builds or rents coverage but households choose a cheaper rival, the fixed network becomes a cost base waiting for revenue. If households subscribe only after a discount, reported penetration can hide weak returns. If customers demand repeated installation visits or equipment upgrades, service quality can protect the brand but still consume cash.

NOS has tried to make fixed access more than a commodity through television, Wi-Fi, customer experience and household benefits. That is rational. In a country with high fixed broadband availability, raw fibre coverage is not enough. The defensible unit is the household relationship, not the cable in the street. A sticky fixed account can carry mobile lines, security, streaming and loyalty benefits; a weak fixed account merely gives a rival a target.

The first quarter of 2026 showed both promise and risk. Fixed services added 23.6 thousand net RGUs, and management said telecom customer numbers improved relative to a difficult prior period affected by a new entrant. But consumer revenue still fell and ARPU declined. That implies that NOS was defending or expanding the base in a market where the price per household was under pressure. The fixed network bill is paid by the customer who stays without being over-subsidised. If the next year shows more fibre customers, stable ARPU and lower installation or maintenance cost, the fixed strategy will look sound.

If it shows more customers at lower unit economics, the network is being filled at the expense of return.

Mobile Spectrum Gives Capacity, Not Automatic Pricing Power

Mobile is the second half of the household lock-in. It brings recurring subscription revenue, makes the bundle harder to leave and gives the operator a way to sell family accounts. NOS's 2025 materials show about 5.7 million mobile service RGUs, 99.6 percent reported 5G coverage and 4,886 5G sites or towers in the company's presentation. The group also presents itself as having a strong 5G position and independent recognition for mobile network speed and experience.

The mistake would be to treat spectrum and coverage as pricing power by themselves. Spectrum is a right to use scarce radio frequencies; it is also a prepaid claim on future revenue. The operator has to turn that right into capacity, coverage, lower unit cost, enterprise use cases or superior customer experience. In Portugal, every national operator can advertise modern mobile service. MEO has incumbent scale, Vodafone has a sophisticated mobile brand and Digi has the option to use low prices to win attention where its own economics permit.

Mobile spectrum does help NOS in three ways. First, it protects service quality in dense areas where data consumption keeps rising. Second, it lets NOS add family SIMs to converged bundles without fully depending on another network. Third, it gives the business segment a platform for private networks, 5G-enabled industrial projects, IoT and low-latency use cases. These are real advantages, but only the first two are visible in mass-market cash flow today.

The consumer problem is that mobile has become the easiest part of the bundle to reprice. A low-cost mobile offer can force an incumbent to respond even if the challenger cannot yet match every fixed or content feature. The household may keep the fixed line but move secondary SIMs. Or it may use the cheaper SIM as negotiation leverage. NOS can answer with network quality, family integration, customer service, entertainment benefits and loyalty discounts, but each answer has a cost.

The enterprise opportunity is more attractive but slower. Business customers may value reliability, security, managed connectivity and integration with IT services more than a household values the difference between two mobile apps. The Claranet acquisition gives NOS more routes into cloud, cybersecurity and managed services conversations. Still, mobile spectrum only creates enterprise value when it is attached to contracted use cases with acceptable margins. A pilot project, a stadium demonstration or an innovation claim does not pay for spectrum on its own.

The mobile judgment is therefore pragmatic. NOS has enough spectrum and coverage evidence to remain credible. The question is whether that infrastructure lets the company preserve ARPU and reduce churn, or whether it simply raises the minimum investment required to stay in the game.

Content Keeps The Bundle Sticky And Expensive

Television and content are where convergence becomes emotional. Broadband can be compared by speed and price. Mobile can be compared by data allowance and coverage. Television, sports, cinema, children's channels and household entertainment create habits. For NOS, that history is part of the business: the group descends from cable and media assets as well as mobile, and it still owns or participates in cinema, audiovisual distribution and content-related businesses.

This is a genuine advantage in Portugal, where pay-TV remains an important part of the multi-play bundle. NOS reports a strong pay-TV share in its 2025 materials, and the brand has long been associated with television experience, cinema and entertainment. Those assets help explain why a household might pay more than the cheapest fixed broadband offer. If the television interface works, the content package is familiar and the family uses the operator's entertainment benefits, churn can be lower than a spreadsheet model would predict.

The cost side is just as real. Content rights and distribution economics can absorb cash without creating durable differentiation. Sports rights are especially dangerous because they are valuable precisely when rivals also need them. If an operator cannot afford to lose a must-have channel or sports package, the supplier has leverage. If the operator wins exclusivity or a better position, regulators and rivals may respond. If it spreads the content across wholesale arrangements, the differentiation may fade.

The correct test is not whether content attracts customers. It clearly can. The test is whether the gross margin saved by lower churn and higher bundle ARPU exceeds the rights cost, platform cost and customer support burden. A sports-heavy household may be profitable if it buys broadband, TV, mobile and premium add-ons at a stable price. It may be unattractive if the operator must subsidise the package to stop that household from moving to a cheaper fibre provider and an over-the-top subscription mix.

Cinema and audiovisuals add a related but separate benefit. They give NOS a media identity and a source of non-telecom revenue, but their performance depends on box-office cycles and content slate quality. In 2025, group commentary noted that cinema and audiovisuals were affected by fewer box-office successes, while the first quarter of 2026 benefited from stronger ticket volumes. That volatility makes the segment useful to brand and content positioning, but not a substitute for telecom cash generation.

Content should therefore be treated as retention infrastructure. It is not fibre, spectrum or towers, but it performs a similar economic role: it raises the cost of customer loss. The danger is that retention infrastructure can become an entitlement. Once households expect a rich TV experience at a bundled discount, withdrawing value is hard. NOS must keep enough content to protect the household relationship while refusing to buy every piece of programming at any price.

Wholesale Access Changes The Build-Or-Buy Calculation

Wholesale access is the least visible but one of the most important parts of the NOS story. In the fourth quarter of 2025, management linked lower capex partly to the maturity of mobile infrastructure and progress in fixed expansion through rental of third-party wholesale networks. That is a major strategic choice. Instead of owning every incremental metre of fixed network, NOS can use rented access to extend reach, reduce upfront capital and focus owned construction where returns are better.

The benefit is obvious. A national operator does not have to win by digging everywhere. If a third-party network already passes a home at an acceptable access price, renting can protect coverage, speed up availability and lower construction risk. It can also help NOS respond to rivals without committing scarce capital to areas where expected take-up is uncertain. In a mature market, capital discipline is as important as ambition.

The downside is equally important. Wholesale access turns capex into recurring cost and can reduce differentiation. If several operators can use the same physical access network, the battle shifts to price, installation, router quality, television, customer service and bundle economics. The operator may have less control over repair times or upgrade timing. A rented network can improve free cash flow in the investment phase while creating a longer-term margin ceiling if access prices are high or if retail competition is intense.

The first quarter of 2026 gives a mixed signal. Wholesale and other telecom revenue fell 11.2 percent to EUR 22.7 million, affected in part by changes in wholesale models that stopped recognising some revenue and cost. That accounting and business-model shift makes headline revenue less informative. What matters is contribution: whether the wholesale arrangements improve returns by removing low-margin gross flows, or whether they reveal weaker economics in the less visible parts of the telecom segment.

For fixed-mobile convergence, wholesale access should be judged by three tests. Does it let NOS enter or defend households that would otherwise be unreachable at acceptable return? Does it lower the group's capital intensity without weakening the customer experience? Does it leave enough margin to absorb discounts, content and service costs? A yes on all three makes rented access a rational capital tool. A no on any of the three makes it a hidden margin leak.

This is where MEO and Vodafone comparisons matter. MEO's incumbent footprint and Vodafone's fibre expansion create pressure for near-national availability. NOS cannot tell households in attractive regions that the bundle is unavailable. But matching coverage by any means is not the same as earning a return. The build-or-buy decision has to be ruthless because every rented line still needs a customer who pays enough to justify the monthly access cost.

Towers Turned Into Rent And Discipline

Tower economics show how telecom operators can improve near-term financial flexibility while accepting future fixed obligations. NOS agreed in 2020 to sell a large tower portfolio to Cellnex in a transaction reported at about EUR 375 million for roughly 2,000 Portuguese telecom sites. That kind of deal releases capital and transfers passive infrastructure ownership to a specialist. It can be sensible if the operator receives an attractive valuation and keeps secure long-term access to the sites it needs.

The trade-off is that tower ownership becomes tower rent. The operator may reduce capital employed and improve headline leverage at the time of sale, but lease payments then sit inside the recurring cost base. In 2025, NOS reported lease costs of EUR 133.0 million at consolidated level, up from EUR 126.1 million in the comparable re-expressed 2024 figure. In the first quarter of 2026, lease costs were EUR 33.5 million, up 3.6 percent year over year. These are not only tower costs, but they show why after-lease metrics matter.

For a converged operator, after-lease EBITDA and free cash flow are better guides than pre-lease operating profit. A household bundle that looks profitable before tower, equipment, access and lease obligations may be less attractive once the recurring infrastructure rent is included. That is especially true when mobile data keeps growing and the company has to keep densifying or upgrading the network even after the initial 5G rollout.

The tower sale also changes strategic flexibility. Owning passive sites gives an operator one kind of control; leasing gives another. A sale-and-leaseback can lock in site access, but future amendments, colocations, energy costs and upgrade terms still matter. If demand rises and the operator needs more capacity, the economics depend on contract terms not fully visible to outside readers. If demand disappoints, the lease burden remains.

The discipline comes from recognising that the tower monetisation is not free money. It was a capital allocation choice. The cash proceeds helped the group, but the network still needs those sites. Investors should therefore watch whether NOS's after-lease margins and cash flow continue to improve as capex declines. If they do, tower monetisation can be part of a rational asset-lighting strategy. If lease growth offsets capex relief, the company has merely moved the cost from one line to another.

In this sense, tower economics mirror wholesale access. Both can improve capital intensity. Both can also create recurring claims on the bundle. The customer does not care whether the network cost arrives as depreciation, lease expense or wholesale fee. The shareholder should.

MEO, Vodafone And Digi Define The Price Ceiling

NOS does not price in a vacuum. MEO, Vodafone and Digi define the range of acceptable offers for Portuguese households. MEO carries incumbent weight, broad coverage and a strong bundled position. Vodafone has a long-standing mobile brand, fixed expansion and a willingness to compete on service quality and converged offers. Digi, even at a smaller base, changes the negotiation because it gives customers a visible low-price reference.

The most important change is psychological. Before a low-cost challenger enters a market, incumbents can argue mainly about quality, content, coverage and loyalty benefits. After entry, every retention conversation has a sharper anchor. The customer may not move to the lowest-price operator, but the customer can use that price to demand concessions. That affects ARPU before it affects subscriber counts. NOS's first-quarter 2026 disclosure fits that pattern: consumer ARPU declined even as the company reported positive customer additions and a better RGU trend than the prior year.

MEO is the hardest benchmark because it can make the case for national reliability and incumbent depth. In many households, the choice between MEO and NOS is not low-cost versus premium; it is two full-service operators arguing over coverage, TV, price and service history. Vodafone is similar in quality terms but often sharper as a challenger in fixed and mobile. Digi is different. Its threat is not that it immediately matches every feature; it is that it compresses the price a household is willing to pay for the features it does not actively value.

NOS's own response includes service quality, Combina household savings, fixed and mobile network claims, enterprise diversification and content. That is a better response than simple price matching, because matching the lowest price with the highest cost base destroys value. Yet the response must be measurable. If Combina claims annual household savings, the economic question is who funds those savings and whether they reduce churn enough. If the mobile network wins awards, the question is whether customers pay a premium for that difference. If pay TV is stronger, the question is whether content costs leave margin.

Low-price competition also exposes segmentation risk. Premium households may stay for service and content. Younger users may split mobile lines. Small businesses may mix fixed access from one provider with cloud, security or mobile from another. Families under inflation pressure may prefer a cheaper bundle even if customer support is weaker. In that environment, a national operator cannot defend every customer at any price. It has to know which customers are worth saving.

The price ceiling is therefore set by rivals, but the value floor is set by NOS's own cost base. The company wins if it uses convergence to keep profitable customers and lets uneconomic discount seekers go. It loses if it treats every gross addition as proof of success.

Business And IT Help, But Do Not Erase Consumer Pressure

The business segment is the best argument that NOS can grow beyond a mature household market. In 2025, telecom business revenue rose 5.8 percent, compared with essentially flat consumer revenue. In the first quarter of 2026, enterprise telecom revenue rose 5.5 percent while consumer revenue fell. The group also added IT scale through the acquisition of Claranet Portugal, a deal presented as a way to expand in cloud, cybersecurity, managed services, data and digital workplace.

This makes strategic sense. Companies buy connectivity differently from households. They may care more about uptime, service-level commitments, cybersecurity, managed equipment, multi-site networking and integration with cloud or workplace systems. If NOS can combine access, mobile, security and IT services, it may earn stickier revenue at better lifetime value. The acquisition of Claranet Portugal for EUR 152 million, at a reported 9.9 times EV/EBITDA multiple in NOS materials, was designed to accelerate that shift.

The caution is that IT services are not automatically higher quality earnings. Resale of equipment and licences can carry lower margins and volatile timing. Managed services need specialised staff and delivery discipline. Cybersecurity and cloud partnerships can deepen customer relationships, but they also depend on suppliers, certification, recruitment and project execution. In the first quarter of 2026, NOS's IT revenue rose 16.0 percent, helped by equipment and licences, while services revenue rose 4.8 percent.

EBITDA in the IT segment rose 5.7 percent, but margin contracted because the mix included more lower-margin equipment and licences.

That matters for the convergence thesis. Enterprise and IT growth can offset household pressure only if the revenue has durable margin and low churn. A burst of resale revenue does not solve consumer ARPU decline. A managed security contract with recurring margin might. A cloud migration project is useful, but the shareholder needs evidence of repeatable cash returns, not only strategic adjacency.

The business segment also changes supplier dependence. NOS becomes more exposed to cloud platforms, software vendors, cybersecurity talent and integration partners. That can be positive if it gives NOS a broader role in Portuguese companies' technology spend. It can be negative if the company becomes a lower-margin reseller standing between global suppliers and local customers. The difference depends on whether NOS owns the customer relationship and adds enough technical value to command margin.

Business and IT therefore improve the story, but they should not distract from the household bundle. The telecom segment is still the cash centre. The enterprise opportunity raises the ceiling; it does not remove the need to defend consumer returns.

Network-Resource Evidence Is Real, But Narrow

BTW tracks NOS COMUNICACOES, S.A. partly because of RIPE NCC membership and number-resource context. That evidence is useful, but it has a narrow meaning. A RIPE NCC member record shows that the organisation participates in the regional internet registry system for number resources. Public routing and interconnection records can show an autonomous-system footprint, peers, upstream relationships or routed prefixes. They do not say that every service sold under the brand is profitable, that the company sells transit to all customer classes, or that an internet number is itself a company.

For NOS, the network-resource evidence fits the operating reality. A national telecom operator needs internet numbering, routing, peering, upstream connectivity, DNS, security operations and resilient transport. Public interconnection sources such as PeeringDB and BGP observation tools point to a visible network footprint associated with NOS. That supports the article's treatment of NOS as an infrastructure operator, not merely a media brand.

The economic interpretation is about control surfaces. The more traffic NOS carries on-net, the more it can manage quality, cost and customer experience. The more it depends on upstream transit, third-party wholesale access or rented infrastructure, the more external cost and coordination enter the bundle. A residential customer sees Wi-Fi speed and video quality. The operator sees routing, peering, backhaul, international capacity, access network cost and customer equipment.

Peering and transit are especially important because data consumption rises faster than household willingness to pay. If traffic grows and unit transport cost falls, the operator can preserve margin. If traffic grows while content delivery, international capacity, security and equipment costs rise, flat-rate broadband becomes harder to monetise. Public network evidence cannot resolve that unit-cost question, but it identifies the technical layer where the economics are fought.

The same discipline applies to ASNs, prefixes and route records. They are evidence, not businesses. They help trace operational infrastructure. They should not be inflated into strategic claims. The strategic claim for NOS comes from the combination of those resources with customers, coverage, spectrum, content and cash flow.

This narrow reading is important because telecom analysis often confuses technical presence with economic moat. NOS has the technical presence expected of a national operator. The moat exists only if that presence lowers cost, improves service quality or raises customer willingness to pay. Otherwise it is simply part of the ticket required to compete.

Regulation And Operating Risk Sit Inside The Margin

Telecom regulation is not outside the business model; it sits inside the margin. ANACOM spectrum rules, licence obligations, consumer-protection standards, number portability, wholesale-access policy, quality reporting and market analysis all shape what operators can charge, how fast customers can move and how much capital must be committed before revenue is earned. For NOS, regulation can protect infrastructure investment in some areas and intensify competition in others.

Spectrum is the clearest example. Operators need frequencies to offer mobile service, but spectrum awards and coverage duties create fixed commitments. Once the licence is acquired, the economic question becomes utilisation. Underused spectrum is trapped capital. Congested spectrum is a quality risk. Properly loaded spectrum is a margin tool because it lowers unit capacity cost and supports higher-value services. NOS's 5G position is therefore valuable only to the extent that it produces customer retention, enterprise contracts or lower cost per bit.

Wholesale and access regulation create a second risk. A more open access environment can lower duplication and help competition, but it can also reduce the payoff to owning infrastructure. If NOS relies more on rented networks, wholesale terms matter directly. If rivals gain access to comparable infrastructure, NOS must compete harder on brand, service and content. If regulators push for lower consumer prices or easier switching, the friction value of convergence falls.

Operating risk is also physical. The first quarter of 2026 disclosure notes storm effects, including cases where NOS stopped billing and collecting from customers who lost service. That is a reminder that a telecom network is exposed to weather, energy cost, field-force capacity and local repair constraints. Climate and resilience spending are not optional if the company wants to maintain premium positioning.

Geopolitical and supplier risk enter through equipment, cybersecurity, cloud platforms and content rights. A converged operator depends on network vendors, device supply, software systems, submarine and terrestrial connectivity, data centres, security tooling and entertainment suppliers. Public information does not show a single supplier concentration that dominates the case, but the overall exposure is broad. The more NOS becomes an ICT provider through Claranet and enterprise services, the more this supplier map matters.

Regulatory and operating risks do not make NOS unattractive. They explain why high EBITDA margins must be converted into cash before they can be celebrated. A regulated, capital-intensive operator can look stable until competition, storms, licence duties, content costs or lease escalators reveal where the margin was thin.

Unofficial Signals Are Useful Only At The Edge

Unofficial market signals should be used carefully. Customer forums, social media complaints, speed-test awards, consumer-award pages and review sites can show how a telecom brand is being experienced, but they are biased samples. Angry customers are louder than satisfied ones. Speed tests overrepresent users who choose to test. Awards can be shaped by methodology. Forums can identify pain points before financial statements do, but they cannot replace audited numbers or regulator data.

For NOS, these signals broadly support a mixed view. Independent or semi-independent recognition around mobile speed, 5G experience, Wi-Fi and television quality helps the premium case. It suggests that NOS has service attributes beyond price. At the same time, consumer complaint channels and market commentary around low-price offers indicate that Portuguese customers remain highly sensitive to billing, installation, contract terms, outages and retention practices. That sensitivity is exactly where a challenger can attack.

The useful question is not whether the internet contains complaints about NOS. Every large telecom operator has complaints. The useful question is whether complaints cluster around issues that damage the convergence thesis: poor installation, unreliable fixed service, weak Wi-Fi, confusing bills, hard cancellation, price increases after promotions, or customer service friction. Those problems can turn a multi-service bundle from a moat into a grievance. A household that feels trapped is not loyal; it is waiting for a switching trigger.

Positive signals require the same restraint. A "fastest network" recognition can help advertising and justify premium positioning, but customers usually pay for perceived reliability, not laboratory bragging rights. If the network is fast in tests but the household's router placement is poor, the customer may still churn. If television wins satisfaction awards but a sports package becomes expensive, the value perception can still deteriorate.

Unofficial signals therefore sit at the edge of the judgement. They are early warnings and texture. They help explain why ARPU might fall before subscriber counts fall, or why a brand can hold share despite higher prices. They do not prove free cash flow. For NOS, the signals point to a company with credible service quality but little room for complacency. The market is telling every operator the same thing: prove the premium every month, or accept a lower price.

What Would Change The Judgment

The positive case would become much stronger with three more years of evidence that consumer ARPU can stabilise while churn remains controlled and capex declines. The ideal proof would be simple: rising or stable consumer revenue, growing fixed take-up on the expanded fibre footprint, lower installation and maintenance cost per customer, mobile data growth absorbed without a new capex surge, and after-lease free cash flow rising faster than EBITDA. That would show that convergence is paying for the network rather than masking discount pressure.

A second positive signal would be cleaner evidence from enterprise and IT. If Claranet integration produces recurring services growth, expanding IT EBITDA margins and cross-sell into NOS connectivity customers, the acquisition will look like more than a defensive move away from consumer telecom. If enterprise telecom keeps growing without heavy project volatility, it can reduce reliance on household ARPU. The crucial point is margin quality. Revenue growth from resale or one-off projects would not change the judgment as much as recurring managed services, cybersecurity and cloud contracts with strong contribution.

A third positive signal would be better disclosure of wholesale economics. NOS says rented third-party networks help lower capex. Investors need evidence that rented access is contribution-positive after service, installation, router, support and retention costs. If wholesale-enabled expansion raises free cash flow and take-up without eroding customer experience, it is a smart capital-light tool. If it only preserves coverage optics, it is a future margin problem.

The negative case would harden if consumer ARPU keeps declining while the company continues to add low-value RGUs. It would also harden if lease costs, content costs or wholesale access costs rise faster than service revenue; if Digi or another low-cost brand forces permanent repricing across the base; if MEO or Vodafone uses stronger fixed footprints to take premium households; or if enterprise IT growth becomes lower-margin resale rather than service-led recurring revenue.

Regulation could change the case in either direction. More favourable wholesale terms, lower spectrum burden or a rational market structure would help. Rules that accelerate switching, lower wholesale prices for rivals without adequate investment return, or pressure consumer tariffs could hurt. Weather resilience and cybersecurity incidents are additional swing factors because a premium operator cannot afford repeated visible failures.

The exact fact that would most change my view is not another coverage claim. It is sustained free cash flow after leases and capex, accompanied by stable consumer ARPU. That combination would prove that NOS is not merely defending share. It would prove that the bundle is paying the network bill.

The Judgment: Convergence Can Pay, But Only With Ruthless Capital Discipline

NOS is not a weak operator. It has national scale, a credible fixed and mobile asset base, strong pay-TV positioning, meaningful business revenue, improving reported cash conversion and a balance sheet that is not stretched by European telecom standards. Its 2025 and first-quarter 2026 numbers show management doing the right thing in broad outline: push capex down as the network matures, defend the consumer base, grow enterprise and IT, use fibre and 5G quality to support premium positioning, and focus on after-lease cash flow.

But the core economic question remains unresolved because the consumer unit is under price pressure at the exact moment when NOS wants to harvest prior investment. Convergence is supposed to make the household more valuable. In a low-price market, it can also make the household more expensive to defend. A bundle with fixed, mobile, TV, loyalty benefits and content has many hooks, but each hook has a cost. The company has to be willing to lose customers that do not clear the return threshold.

MEO and Vodafone make the premium fight continuous. Digi makes the price reference harsher. Wholesale access and tower monetisation improve capital intensity but add recurring claims on cash. Content keeps customers but can become a defensive tax. Enterprise and IT growth widen the opportunity but introduce their own delivery and supplier risks. None of these factors breaks the case. Together, they make discipline non-negotiable.

My position is that NOS can make convergence pay, but only if it treats convergence as a cash-return system rather than a share-defence slogan. The group must prove that fibre take-up, mobile family accounts, content, business services and rented access combine into higher lifetime value, not merely higher service counts. The evidence to date is good enough to keep the case open and better than a simple price-war narrative would imply. It is not yet strong enough to declare victory.

The next decisive period will be measured in ARPU, churn, capex, lease-adjusted cash flow and wholesale contribution. If those move together, NOS will have turned network investment into durable free cash flow. If they diverge, the household bundle will still look busy, but the network bill will be paid by shareholders rather than by customers.