Telecommunications (U.S.) — Subsector Rollup

North American Industry Classification System (NAICS) 2022 code: 517. NAICS is the standard framework U.S. statistical agencies use to sort businesses. This is a NAICS subsector (the three-digit level) — the whole of "Telecommunications" — made of three child industry groups that are wildly different in size but share one boundary: they move communications (voice, video, internet, data, machine signals) as a service, over infrastructure of their own or someone else's. The three children are 5171 — Wired and Wireless Telecommunications (except Satellite), the terrestrial network core; 5174 — Satellite Telecommunications, the space-based layer; and 5178 — All Other Telecommunications, the catch-all residual.[1][2][3][4]

---

1. Overview

Telecommunications is the connectivity layer of the U.S. economy, and at the subsector level it is defined by one overwhelming fact: it is almost entirely one child. The terrestrial carriers of 5171 — the companies that own the fiber, cable, copper, spectrum, and cell sites behind every phone and home-internet connection — are ~95.7% of the subsector's receipts and ~96% of its jobs. The other two children are rounding errors by measured dollars: satellite (5174) is ~1.1% and the "all other" residual (5178) is ~3.2%.[1][2][3][4]

So a naive reading of NAICS 517 says telecom is a mature, cash-rich, consolidating utility dominated by a handful of giant dividend-paying carriers. That reading is correct about the money — but it hides the interesting part. The two tiny children are where the industry's disruption and its worst statistical blind spots live. Satellite (5174) is the fastest-growing telecom business in the country, yet its single dominant asset — SpaceX's Starlink — is private and classified almost entirely outside this subsector. And the "all other" bucket (5178) hides both a commoditizing cloud-phone business and a fast-growing satellite-ground niche, much of whose revenue is coded as software or done in-house by the government. The distinctive job of this rollup is to show that the official $669 billion is a near-perfect measure of the terrestrial core and a poor measure of the edges — and that the edges are where the growth and the undercount are.[2][3][4]

Both public-market and private investors have routes here, but they diverge sharply by child: the terrestrial core offers a deep public menu (large-cap dividend carriers, cable, wholesale fiber, tower real-estate investment trusts) and deep private access (fiber-infrastructure funds, high-yield debt); satellite's best asset is unbuyable (private) with only niche or mid-acquisition public proxies; and the residual is a scatter of small-cap public names plus a mostly private, defense-adjacent ground segment.

---

2. What's inside — the three children and how they differ

The three children sit side by side (unlike a vertical parent-child chain): they are three distinct ways of carrying communications, not three layers of one business. What unites them is the boundary; what separates them is medium (wire/spectrum vs. orbit vs. everything else), scale, ownership, and how much of their real economics the federal code actually captures.

Contrast at a glance

NAICS = North American Industry Classification System. CR4 = combined revenue share of the four largest firms (CR8/CR20/CR50 likewise for the top 8/20/50). HHI (Herfindahl-Hirschman Index) = the standard market-concentration score; regulators treat 1,500–2,500 as "moderately concentrated," above 2,500 as "highly concentrated." REIT = real-estate investment trust. PE = private equity.

How they differ — the three things that matter

1. Size: this subsector is one child wearing a subsector's number. By receipts, jobs, and firm-count, 5171 is telecommunications — it dwarfs the other two by roughly 30-to-1 and 45-to-1 respectively. Satellite and "all other" together are ~4.3% of measured receipts. Anyone modeling the subsector's financials is, to a first approximation, modeling the terrestrial carriers. But size by receipts is the most misleading single lens here, because of point 3.[2][3][4]

2. Ownership inverts across the children — and so does investability. The terrestrial core is the most ownable telecom business in America: public giants hold most of the value, and even the privately held parts (Cox; fiber-infrastructure funds; the sector's large high-yield bond market) are reachable through familiar private channels.[2] Satellite is the opposite — its dominant, fastest-growing asset (Starlink) is private and not for sale to public investors, leaving listed players that are either niche, foreign-listed, or in the middle of being acquired.[3] The residual splits again: its public face is a set of small- and micro-cap cloud-phone names, while the genuinely space-related part is a private-market and defense-contractor story.[4] So as you move from 5171 to 5174 to 5178, the quality of public access falls even as the growth rate can rise — the classic telecom trade-off in miniature.

3. The federal code captures the core almost perfectly and the edges badly. This is the rollup's key insight. The terrestrial carriers are a small number of giant, incorporated firms that file complete data, so 5171's ~$640.8B is close to reality.[2] But both small children are dramatically undercounted, and — critically — the undercount is concentrated in exactly the fast-growing, disruptive parts of telecom. Starlink's connectivity revenue (~$11.4B in 2025 — already larger than the entire measured satellite industry) is folded into SpaceX, classified as a space-vehicle manufacturer outside code 517 altogether; the biggest cloud-communications firms are coded as software (NAICS 513210) or data processing (518210); and much satellite tracking, telemetry, and defense communications is done in-house by government agencies that the business census excludes.[3][4][5][6] The result: the true telecom economy is materially larger than $669 billion, and its growth edge is largely booked outside the subsector that bears its name.

---

3. How big it is

Federal figures for the subsector follow our ground-truth statistics for NAICS 517, from the 2022 Economic Census (the U.S. Census Bureau's five-year business census) for receipts, firms, and concentration, and 2023 County Business Patterns (CBP — the Census Bureau's annual establishment/employment series) for headcount and payroll.[1]

The numbers reconcile cleanly with the children. Receipts: $640.8B (wired/wireless) + $7.25B (satellite) + $21.26B (all other) = $669.3B — an exact match to our subsector ground truth.[1][2][3][4] Firms: 9,628 + 332 + 1,673 = 11,633, just above the subsector's 11,600 — the ~33-firm gap is businesses operating in more than one child, counted once at this level.[1] Employment reconciles by subtraction: the level's 912,645 CBP jobs minus satellite's 9,919 and "all other's" 26,776 leaves ~876,000 in the terrestrial carriers — directional (it is a residual, not a directly reported 5171 figure), but consistent with the child primer's ~850,000–900,000 estimate.[1][3][4]

Concentration is "moderate" at the subsector, and lower than the core. The blended CR4 of 64.7% and HHI of 1,248.4 sit below the terrestrial core's own CR4 (67.6%) and HHI (1,360).[1][2] The reason is arithmetic: the subsector's top four firms are the big terrestrial carriers, and adding satellite's and the residual's combined ~$28.5B of receipts — earned by different companies — to the denominator dilutes the leaders' share without adding to their numerator. So the subsector looks slightly less concentrated than the industry that actually drives it. As always, a national NAICS-classification HHI is a concentration signal, not a finding about any specific local market or a formal antitrust market definition.[1]

Revenue is far more concentrated by dollars than by headcount is spread by firms. The subsector earns ~$733,000 of receipts per employee (2022 receipts over 2023 headcount — mixing years, so directional), a very high figure that reflects capital-heavy, low-labor network economics; payroll is only ~13% of receipts.[1] Value concentrates in the network owners; the long tail of ~11,600 firms is mostly small resellers, agents, niche satellite and ground operators, and cloud-phone shops.

Undercount / scope caveat — the core is solid, the edges are understated (and it matters most where growth is).

• The terrestrial core is unusually well captured. A small number of giant, incorporated carriers file complete data, so ~96% of the subsector's measured dollars are reliable. The one gap: the business census excludes government-owned networks and most no-payroll operators, so municipal broadband, member-owned rural telephone and electric cooperatives, tribal networks, and very small owner-operated wireless internet service providers (WISPs) are undercounted — this barely dents the concentration picture but understates the true number of small carriers.[2][6]
• Satellite is understated by an order of magnitude, and in the wrong direction for investors. The dominant, fastest-growing operator (Starlink, ~$11.4B connectivity revenue in 2025) is classified inside SpaceX, outside this code; direct-to-home satellite TV and military SATCOM (satellite communications) also sit elsewhere. The measured $7.25B is a floor for a slice, not a measure of the whole space-connectivity economy.[3][5]
• The residual leaks by design. Its fastest-growing cloud-communications dollars are coded as software or data processing, and its government tracking/telemetry/radar work is largely done in-house at the Department of Defense (DoD), NASA, and NOAA (National Oceanic and Atmospheric Administration). Treat the measured $21.26B as an establishment-based floor.[4][6]
• It is a 2022/2023 snapshot. The mix has shifted since — broadband, wireless data, fixed wireless, and low-orbit satellite up; cable TV and legacy voice down.

Our statistics file gives no subsector-wide operating metrics (average revenue per user, churn, capital spending, subscribers, fiber passings, spectrum, satellite fill rates), so those must be read company by company (Section 5). Where small and individually owned operators dominate — the municipal/co-op/tribal carrier tail, independent wireless agents, non-employer satellite and ground operators, and the government activity above — read the counts as an employer-business benchmark, not a full census.

---

4. The investable universe — where value concentrates across the children

The investable heart of NAICS 517 is almost entirely the terrestrial core; the two small children add strategic optionality but little of the subsector's dollars. Tickers below size an entry point only — the largest companies are diversified, so company-wide revenue is not the same as any single NAICS line. Reserve valuation detail for Section 10 and the child primers.

Terrestrial core (5171) — where ~95.7% of the value sits

• Diversified & wireless giants (the income core): AT&T (T), Verizon (VZ), and T-Mobile (TMUS) — the biggest dividend/free-cash-flow names in the subsector; T-Mobile is the fastest-growing of the three.[2]
• Cable-broadband leaders: Comcast (CMCSA) and Charter (CHTR), whose mobile brands resell a national carrier's network; plus pure-plays Cable One (CABO) and Optimum (OPTU).[2]
• Enterprise / wholesale & AI-transport fiber: Lumen (LUMN), Cogent (CCOI), Uniti (UNIT) — infrastructure and turnaround stories tied to the data-center buildout.[2]
• Infrastructure "picks and shovels": the cell-tower REITs American Tower (AMT), Crown Castle (CCI), and SBA Communications (SBAC) — long-term lease income without carrier-level price wars.[2]
• Private / strategic owners: Cox (largest privately held U.S. broadband operator, family-owned); PE/infrastructure-fund carriers (Brightspeed, Astound, Zayo); plus hundreds of municipal, cooperative, and tribal systems reachable only through project finance.[2]

Satellite (5174) — the disruptive child, but the best asset is unbuyable

• Private and dominant: SpaceX/Starlink (low-Earth-orbit broadband + direct-to-cell) and Amazon's "Amazon Leo" (built inside Amazon) — the two mega-constellations, neither a clean public play.[3]
• Public operators in transition: EchoStar (SATS), Viasat (VSAT), SES (Luxembourg), Iridium and Globalstar (both mid-acquisition), AST SpaceMobile (ASTS, a richly priced direct-to-device bet), Telesat, Eutelsat/OneWeb — niche, foreign-listed, or in flux.[3]

All other (5178) — small-cap public voice + private/defense ground

• Non-carrier cloud voice/messaging (all small- or micro-cap): Bandwidth (BAND), 8x8 (EGHT), Ooma (OOMA), Crexendo (CXDO), IDT's net2phone (IDT), Spok (SPOK) — theme exposure, not a code index. Larger adjacent proxies: RingCentral, Zoom, Twilio.[4]
• Satellite-ground / teleport / government RF (mostly private): Ground-Station-as-a-Service (GSaaS) and teleport operators (KSAT, ATLAS Space Operations, RBC Signals, Leaf Space); hyperscaler ground networks buried inside mega-caps (AWS Ground Station, Azure Orbital); and defense-IT contractors (Peraton, Leidos, SAIC) for government tracking/telemetry/radar.[4]

The recurring lesson across the universe: public-market ownership is deepest exactly where growth is slowest (the mature terrestrial core), and thinnest exactly where growth is fastest (private Starlink, private ground). To own this subsector as a public investor is mostly to own the network core; to own its growth edges usually means private markets or waiting for an IPO.

---

5. How the money works

The three children run on three different profit engines, but two levers recur across all of them: who owns the scarce asset (spectrum, orbital slots, fiber routes, antenna sites) and how much fixed cost you must keep replacing.

Terrestrial core (5171) — subscribers on expensive fixed plant. The core equation is subscribers × average revenue per user (ARPU, the recurring monthly revenue per customer) = recurring revenue, earned off a network with very high fixed costs and low incremental cost per added customer. Because networks are expensive, carriers carry heavy debt and are sensitive to interest rates; the figures investors watch most are free cash flow (FCF — cash left after capital spending, to fund dividends, buybacks, and debt paydown), EBITDA (earnings before interest, taxes, depreciation, and amortization), and net leverage. Two product shifts define current economics: home broadband is now the profit engine while cable TV video is a low- or negative-margin drag, and fixed wireless access (FWA — home broadband delivered over the cellular network) turns spare 5G capacity into a second broadband revenue stream at very low marginal cost.[2]

Satellite (5174) — two engines under one code. The traditional geostationary-orbit (GEO) model leases capacity wholesale: a satellite costs $150–400 million and earns for 15+ years, so once in orbit each additional unit sold drops almost straight to margin — but wholesale prices have been falling. The low-Earth-orbit (LEO) model is a subscriber business (subscribers × ARPU minus network cost) that demands a brutal capital treadmill because LEO satellites live only ~5 years and must be constantly relaunched. Across both, launch cost (SpaceX's structural edge) and spectrum/orbital slots (scarce, licensed, sometimes worth more than the satellites) decide winners.[3]

All other (5178) — an asset-light/asset-heavy barbell. Cloud voice/messaging looks like software: recurring revenue, ~60–75% gross margins, "land and expand," with churn and customer-acquisition cost as the health metrics — dented by telecom taxes and termination fees. Teleport/GSaaS/government-RF is the opposite: fixed-cost antenna networks where utilization is everything and a globally dispersed footprint is pricing power, plus sticky, lower-margin government backlog. The subsector's high revenue-per-head, low-payroll signature is this barbell writ small.[4]

Our federal file provides no subsector-wide utilization, churn, ARPU, capex, or margin series; all of it must be underwritten company by company.

---

6. What drives demand

Demand is durable and rising across the subsector, but it accrues very differently by child.

• Connectivity is now essential. Households and businesses treat broadband and mobile as utilities; terrestrial penetration is near saturation, so growth comes from taking share and selling more per customer, not a growing pool.[2]
• Data consumption keeps climbing. U.S. wireless users generated 132 trillion megabytes in 2024 (up 32 trillion in a year) across ~579 million connections — supporting ARPU even when customer counts are flat.[7]
• Fixed wireless and fiber for AI. 5G home internet reached ~12 million U.S. subscribers, and the artificial-intelligence/data-center buildout is driving demand for high-capacity fiber routes between data centers — the clearest terrestrial-growth pockets.[2][7]
• Where terrestrial can't reach, satellite wins. Rural/remote broadband, mobility (aviation, maritime, land), government/defense resilience, and especially direct-to-device (D2D — connecting ordinary, unmodified phones from space) are the satellite child's demand pools; demand is strongest where terrestrial alternatives are unavailable or uneconomic.[3]
• The cloud and the constellation feed the residual. Business phone systems keep migrating from on-premises hardware to cloud calling (with AI features lifting revenue per seat), while the LEO satellite boom means thousands of fast-moving spacecraft each needing frequent ground contacts — driving the satellite-ground niche.[4]
• Public subsidy. Federal programs fund rural construction across media (Section 7).[10]

Editor's judgment: demand for connectivity should stay structurally positive, but rising traffic will not automatically produce attractive returns. In the saturated terrestrial core, price competition can offset volume; in satellite, capacity oversupply from overlapping constellations is the swing risk; in the residual, big-tech bundling caps the voice side.

---

7. Regulation

All three children are overseen primarily by the Federal Communications Commission (FCC) under the Communications Act of 1934 and the Telecommunications Act of 1996, with the National Telecommunications and Information Administration (NTIA) managing federal spectrum and, for satellite, the International Telecommunication Union (ITU) coordinating orbital slots globally. Regulation lands differently by child, but several threads cut across the whole subsector.[2][3][4]

• Universal Service Fund (USF) upheld. In June 2025 the Supreme Court upheld the USF, a roughly $9-billion-a-year program funded by fees on carriers (and passed through by resellers and interconnected-voice providers) that subsidizes rural, low-income, and institutional service — a cross-cutting cost whose quarterly contribution factor ran in the high-30% range in 2025.[8]
• Net neutrality is off the table federally (for now). In January 2025 the Sixth Circuit struck down the FCC's attempt to regulate broadband as a common-carrier service, relying on the Supreme Court's 2024 Loper Bright decision that ended automatic judicial deference to agencies; some states keep their own rules.[9]
• Rural buildout and spectrum (terrestrial-led). The $42.45 billion Broadband Equity, Access, and Deployment (BEAD) program was rewritten in mid-2025 to a technology-neutral, lowest-cost-per-location approach — adding substantial satellite and fixed wireless; and the FCC's lapsed spectrum-auction authority was restored in July 2025 with a directed pipeline.[10]
• Scarce-resource licensing (satellite-heavy). Satellite depends on two licensed resources — radio spectrum and orbital positions; the FCC and ITU control both, and the Supplemental Coverage from Space (SCS) framework enables satellite-to-phone service. Orbital-debris, launch (FAA), export-control (ITAR/EAR), and foreign-investment (CFIUS) rules all apply, and a single ruling can shift a company's addressable market overnight.[3]
• Interconnected voice obligations (residual-heavy). Cloud-voice providers carry Enhanced 911, USF contributions, STIR/SHAKEN caller-ID authentication, and lawful-intercept duties — a real drag on otherwise software-like margins; ground stations require FCC earth-station licensing plus export/CFIUS review.[4]
• Merger review is tightening. The FCC and Department of Justice jointly review deals; the DOJ has publicly called wireless an "oligopoly."[2]

Regulation is a two-way force across the subsector: it raises entry barriers and protects incumbent asset values, but it also subsidizes rural networks and makes spectrum and orbital access available.

---

8. Consolidation

The subsector-wide theme is consolidation toward scale and scarce assets, playing out on three separate boards.

Terrestrial core — rolling up fiber, shrinking the wireless field. A wave of deals is consolidating fiber (Verizon–Frontier, closed January 2026; Charter–Cox, ~$34.5B, pending; AT&T–Lumen consumer fiber; Uniti–Windstream, completed 2025), while wireless has collapsed to three nationals (2020 T-Mobile/Sprint) plus T-Mobile's ~$4.4B absorption of UScellular (2025); the would-be fourth network, EchoStar/Dish, is selling core spectrum to AT&T and SpaceX. Regional carriers are nearly extinct.[2][12]

Satellite — scarce spectrum and orbits command premiums. The field has split into LEO mega-constellations chasing mass-market broadband (Starlink dominant; Amazon Leo the first credible challenger) and legacy GEO/medium-orbit operators retreating to defensible niches — driving a consolidation wave: Viasat–Inmarsat (2023), SES–Intelsat (July 2025), Eutelsat–OneWeb (2023), Rocket Lab's ~$8.0B agreement to acquire Iridium (~2027), and Amazon's ~$11.6B agreement to acquire Globalstar (~2027).[3][13]

All other — bundling squeezes voice, scale consolidates ground. Standalone cloud-phone providers are squeezed between Microsoft Teams Phone and Zoom bundling calling into products customers already own (Ericsson's $6.2B Vonage purchase in 2022; Ooma's 2025 tuck-ins; 26North's 2026 Intermedia buy), while the ground segment consolidates around global antenna footprints and hyperscaler entry.[4][14]

Judgment: across all three children, buyers are paying up for the scarce, hard-to-replicate asset — dense fiber and spectrum, orbital slots, or global ground footprints — while outsourcing or exiting the commoditized parts. Concentration is moderate at the subsector (CR4 64.7%, HHI 1,248.4) but rising in the terrestrial core and in satellite, where the true (Starlink-inclusive) concentration is far higher than the measured figures suggest.[1][2][3]

---

9. Risks

• The subsector is a bet on one child. ~96% of measured revenue and jobs sit in the terrestrial core, so the terrestrial risks — capital intensity, heavy debt, interest-rate sensitivity, near-complete saturation, and share-shifting that erodes ARPU even as traffic rises — are effectively the subsector's risks.[2]
• The best growth is unbuyable or hidden. Satellite's dominant asset (Starlink) is private; the residual's fastest-growing dollars are coded outside 517. Public investors face the structural problem that where the code grows fastest, they can own the least.[3][4]
• Cross-medium competition is permanent. Fixed wireless, fiber overbuilders, and low-orbit satellite (including direct-to-cell) are loosening old local monopolies from three directions at once — good for consumers, corrosive to incumbent pricing power.[2][3]
• Capacity oversupply and the LEO treadmill. Overlapping satellite constellations risk oversupply, and short LEO satellite lifespans demand relentless recapitalization.[3]
• Concentration and single-customer risk. Wholesale/enterprise carriers, tower REITs, and satellite operators can depend on a few large tenants or customers (e.g., government reliance; Globalstar's Apple dependence).[2][3]
• Regulatory and subsidy swings. Net-neutrality standards, USF financing reform, BEAD execution, spectrum and orbital policy, and merger review are all moving parts; interconnected-voice tax/compliance burdens weigh on the residual.[8][9][10][4]
• Physical, cyber, and geopolitical hazards. Fiber cuts, storms, outages, launch failure, orbital collision/debris, jamming, ransomware, and export-control/sovereignty exposure create liability across all three children.
• Analytical opacity — worst at the edges. Private ownership (Starlink, PE roll-ups, defense contractors) and classification leakage make the two small children genuinely hard to size and benchmark; federal statistics never fully capture them.[3][4][6]

---

10. How to invest and the outlook

NAICS 517 is really three investment propositions of very unequal size: a deep, ownable, mature core; a fast-growing satellite child whose crown jewel is private; and a two-speed residual that is mostly small-cap or private.

Own the core (5171 — where the money and the public access are).

• Income / diversified: Verizon (VZ), AT&T (T), Comcast (CMCSA) — high-dividend, debt-heavy, held for yield and free cash flow. Growth / buyback: T-Mobile (TMUS), Charter (CHTR).[2]
• Wholesale & AI-transport fiber: Lumen (LUMN), Cogent (CCOI), Uniti (UNIT). Picks-and-shovels: tower REITs American Tower (AMT), Crown Castle (CCI), SBA Communications (SBAC).[2]
• Private routes are unusually deep: fiber-infrastructure and PE funds (Brightspeed, Astound, Zayo), a large and liquid high-yield telecom bond market, and digital-infrastructure funds targeting fiber, towers, and data-center connectivity.[2]

Reach the satellite child (5174 — the growth is real, the access is not). Public routes are event-driven or in-transition: Iridium and Globalstar (both mid-acquisition), Viasat and EchoStar (diversified operators in transition), AST SpaceMobile (a richly priced D2D bet), foreign-listed SES/Eutelsat/Telesat, and thematic space ETFs — none a clean pure-play. The Starlink gap is the defining feature: the dominant asset is private, reachable only via pre-IPO SpaceX secondaries and funds holding SpaceX shares; a future Starlink IPO would be a landmark that could reprice the whole child.[3]

Approach the residual (5178 — theme exposure, not a code index). Small-cap cloud-voice names (BAND, EGHT, OOMA, CXDO, IDT, SPOK) plus larger adjacent proxies (RingCentral, Zoom, Twilio); the satellite-ground story is largely private (GSaaS/teleport ventures) and defense-contract-based (DoD/NASA/NOAA vehicles). Start from segment exposure, not the NAICS label.[4]

When analyzing any single name, separate recurring service revenue from one-off equipment sales; watch ARPU, churn, net adds, and FWA/broadband adds on the terrestrial side, fill rate/backlog/fleet age and launch cost on the satellite side, and net revenue retention/utilization on the residual side; and check capex, spectrum and orbital commitments, debt maturities, EBITDA-to-FCF conversion, and dividend coverage.

Near-term outlook (forward-looking). Expect the subsector's reported results to keep looking like the mature terrestrial core — near-flat customer totals with winners taking share, strong cash generation despite stalled unit growth, and steady consolidation — because that core is ~96% of the dollars. The clearest terrestrial bright spots are enterprise/wholesale fiber tied to the AI/data-center buildout and fixed wireless expanding into home broadband; the clearest tailwind is BEAD-funded rural construction finally deploying. Meanwhile, the two small children carry most of the change: satellite is in a once-in-a-generation reordering toward LEO broadband and direct-to-device, led by a private company with a structural cost advantage, while the residual's ground niche grows fast off a tiny base.

Bottom line: NAICS 517 is telecommunications in name and, by the numbers, almost entirely the terrestrial carrier core — ~$669.3 billion in classified receipts, of which ~95.7% is wired and wireless (except satellite), ~3.2% is the "all other" residual, and ~1.1% is satellite. That measured picture is accurate about where the money is today and misleading about where the growth and disruption are, because the subsector's most dynamic economics — Starlink's satellite broadband and the largest cloud-communications firms — are booked outside code 517 entirely. For an investor, the practical translation is stark: the deepest, most ownable telecom assets are the mature core's networks (public equities, high-yield debt, private fiber and towers), while the fastest-growing edges are largely private or hidden — so the central choice is between owning a cash-rich, consolidating utility you can buy today and waiting for the disruptive edges to become buyable at all.

---

Sources

1. U.S. Census Bureau. "2022 Economic Census — Concentration of Largest Firms (receipts, firms, CR4/CR8/CR20/CR50, HHI) and 2023 County Business Patterns (employment, establishments, payroll), NAICS 517 (Telecommunications)": receipts $669,317,530 thousand; 11,600 firms; CR4 64.7%, CR8 79.0%, CR20 85.6%, CR50 90.1%; HHI 1,248.4; 912,645 employees; 58,992 establishments; annual payroll $83,948,432 thousand; Q1 payroll $25,728,666 thousand. (Histometrics ground-truth statistics file for NAICS 517.) 2022/2023. https://data.census.gov/table/EC2200SIZECONCEN?n=517
2. Histometrics. "Wired and Wireless Telecommunications (except Satellite) — NAICS 5171" (child primer and its underlying company filings and federal sources: receipts $640.8B; 9,628 firms; CR4 67.6%; HHI 1,360.1). 2026.
3. Histometrics. "Satellite Telecommunications — NAICS 5174" (child primer and underlying sources: receipts $7.25B; 332 firms; CR4 49.4%; HHI suppressed; CBP 2023: 429 establishments, 9,919 employees, $1.52B payroll; Starlink ~$11.4B 2025 connectivity revenue, private). 2026.
4. Histometrics. "All Other Telecommunications — NAICS 5178" (child primer and underlying sources: receipts $21.26B; 1,673 firms; CR4 59.5%; HHI 1,232; CBP 2023 for 517810: 1,932 establishments, 26,776 employees, $3.34B payroll). 2026.
5. U.S. Census Bureau / NAICS Association. "2022 NAICS Manual and Definitions — code 517 and children 5171 / 5174 / 5178, and adjacent codes 336414 (space-vehicle mfg.), 513210 (software), 518210 (data processing), 517111 (satellite TV)." 2022. https://www.census.gov/naics/reference_files_tools/2022_NAICS_Manual.pdf
6. U.S. Census Bureau. "County Business Patterns & 2022 Economic Census methodology; Nonemployer Statistics" (coverage of employer vs. nonemployer and government activity). https://www.census.gov/programs-surveys/economic-census/year/2022/technical-documentation/methodology.html
7. CTIA. "2025 Annual Wireless Industry Survey Highlights" (132 trillion MB in 2024; ~579M connections; active 5G devices; fixed wireless). 2025. https://www.ctia.org/news/2025-annual-survey-highlights
8. U.S. Supreme Court / Congressional Research Service. "FCC v. Consumers' Research: High Court Rejects Challenge to Universal Service Fund" (June 27, 2025); FCC USF quarterly contribution-factor filings. 2025. https://www.congress.gov/crs-product/LSB11301
9. Congressional Research Service. "No More Deference: Sixth Circuit Relies on Loper Bright to Strike Down Net Neutrality Rules." 2025. https://www.congress.gov/crs-product/LSB11264
10. Congressional Research Service. "The Broadband Equity, Access, and Deployment (BEAD) Program" and FCC spectrum-auction authority restoration (2025). https://www.congress.gov/crs-product/R48666
11. FCC / ITU / NTIA. "Satellite and spectrum regulation: FCC earth-station and NGSO licensing, ITU orbital coordination, Supplemental Coverage from Space (SCS); FCC/NTIA federal vs. nonfederal spectrum MOU." 2022–2025. https://www.ntia.gov/other-publication/2022/memorandum-understanding-between-fcc-and-ntia
12. Quiver Quantitative (Verizon–Frontier); FCC (Charter–Cox approval); T-Mobile (UScellular close, Aug 2025); Light Reading (EchoStar spectrum sale to AT&T/SpaceX). 2025–2026.
13. Rocket Lab (Iridium, ~$8.0B, ~2027); Amazon/Globalstar (~$11.6B, ~2027); SES (Intelsat, July 2025); Viasat (Inmarsat, 2023); Eutelsat/OneWeb. 2023–2026.
14. Ericsson–Vonage ($6.2B, 2022); Ooma (FluentStream, Phone.com, 2025); 26North–Intermedia (2026); MarketsandMarkets, "Satellite Ground Station Market" (~$41B 2025 → ~$83B 2030). 2022–2026.
15. Company financial filings — AT&T, Verizon, T-Mobile, Comcast, Charter, Lumen, Cogent, Uniti (terrestrial); EchoStar, Viasat, SES, Iridium, Globalstar, AST SpaceMobile (satellite); Bandwidth, 8x8, Ooma, Crexendo, IDT, Spok (residual). 2025–2026 Annual Reports / Form 10-K.