Fiber Optic Cable Manufacturing (United States)

NAICS 2022 code 335921 — an industry primer for public-market and private investors

1. Overview

Fiber optic cable is the physical backbone of the modern internet: hair-thin strands of ultra-pure glass, bundled and sheathed into rugged cables that carry data as pulses of light. This industry covers the U.S. factories that take purchased glass fiber and turn it into finished, jacketed cable — the product that gets buried under streets, strung on utility poles, run under oceans, and threaded through data-center racks.

For an investor, the appeal is that fiber sits at the center of two of the largest capital-spending waves of the decade: the build-out of artificial-intelligence (AI) data centers, and the federally funded push to wire rural America for broadband. Demand has run so far ahead of supply that lead times for some cable stretched toward a year, and prices have firmed after a long stretch of commodity-style pressure.[14][19]

There is a catch for stock-pickers: almost no large U.S. company is a pure fiber-cable maker. The business is embedded inside diversified industrial firms. Public exposure runs mainly through Corning and, after a landmark 2025–26 deal, Amphenol; the rest of the field is European-listed groups, Asian conglomerates' U.S. subsidiaries, and one tiny domestic pure-play. Private-market investors reach the industry through those subsidiaries, through cable-plant assets, and through the connectivity and installation firms that surround it.

2. What it is, and how it's structured

Official scope. NAICS 335921 comprises establishments that manufacture insulated fiber optic cable from purchased fiber optic strand.[3] The defining word is "purchased." A firm in this code buys glass fiber and does the cabling — the coloring, buffering, stranding, armoring and jacketing that produces a deployable cable.

The manufacturing process. A cable plant purchases coated strand and colors, buffers, or ribbons it; strands the fibers or buffer tubes around strength members; adds water-blocking material and, where required, armor; extrudes protective jackets; tests optical attenuation and mechanical performance; and winds finished cable onto reels. Loose-tube constructions are common outdoors because they isolate the glass from stress and temperature changes; tight-buffer cable is common indoors; and ribbon and rollable-ribbon designs increase density and permit mass splicing.[20][21]

What it excludes (this matters for sizing the industry honestly):

• Making the glass fiber itself — drawing optical fiber from a glass preform — is classified in NAICS 32721, Glass and Glass Product Manufacturing, not here.[3] So the single most valuable step in the fiber value chain sits in a different code.
• Other insulated wire and cable (copper communication and energy wire from purchased wire) is NAICS 335929, Other Communication and Energy Wire Manufacturing.[3]
• Connectors, transceivers, patch panels and passive optical hardware fall under electronic-component and connector codes, not 335921.
• Installing fiber (trenching, splicing, network construction) is telecom construction and services, not manufacturing.

That classification boundary is the industry's biggest analytical trap. Consultant "fiber-optic market" estimates frequently combine glass preforms, drawn fiber, finished cable, connectors, electronics, installation, and sometimes network services. Likewise, Corning's Optical Communications sales or Amphenol's acquired CommScope business cannot be treated as NAICS 335921 market share.

Ownership mix. The domestic footprint is a handful of large plants owned by a few global groups, plus specialty shops. The Census counts 99 firms operating 130 establishments — evidence that many players run more than one plant.[1][2] Ownership is heavily foreign-parented or diversified-parented: Japanese (OFS/Furukawa, AFL/Fujikura, Sumitomo), Korean (Superior Essex/LS Cable & System), Italian (Prysmian, which owns the former General Cable), and U.S.-headquartered diversified industrials (Corning, Amphenol). The lone independent U.S. public pure-play is small.

3. How big it is

Federal statistics for the narrowly defined industry (all figures U.S. Census Bureau unless noted):

This is a small, high-value manufacturing industry — roughly 8,300 workers producing about $3.8 billion of cable — with output per worker far above a typical factory, reflecting automation and the value of the product inside.

A separate USGS estimate puts U.S. shipments in two optical-cable product categories (communications and other applications, produced principally in 335921) at $3.132 billion in 2021, with U.S. optical-fiber-cable production at 48 million fiber-kilometers — 9.7% of world output — in 2022. "Fiber-kilometers" multiply route length by the number of strands and should not be read as kilometers of installed route.[22]

The undercount caveat is large here, and it runs both ways. The $3.76 billion figure captures only cabling from purchased strand. It leaves out the glass fiber itself (counted in glass manufacturing), the connectivity hardware, and installation — so the economic footprint of "fiber" is several times this number. Private third-party estimates that bundle those pieces put the U.S. fiber-optic-cable market around $5–6 billion and the global market near $16 billion in 2024, growing double-digits.[18] A telling illustration: Corning's Optical Communications segment alone reported $4.66 billion in 2024 sales[5] — larger than the entire federal receipts figure for 335921 — precisely because that segment folds in fiber (glass), cable, and hardware that the narrow NAICS code splits apart. Read the $3.76 billion as the cabling slice, not the whole fiber economy.

4. The investable universe

There is no large, U.S.-listed pure-play in fiber optic cable. Exposure is a slice of bigger companies. Tickers and scale below are for orientation, not recommendations.

Major private / foreign-subsidiary owners (not directly investable on U.S. exchanges, but they define the competitive field):

• OFS — Norcross, Georgia; subsidiary of Furukawa Electric (Tokyo-listed). A vertically integrated fiber-and-cable maker.
• AFL — Duncan, South Carolina; subsidiary of Fujikura (Tokyo-listed). AFL announced more than $50 million of South Carolina cable-manufacturing investment in 2024.[25]
• Superior Essex Communications — Georgia-based; owned by LS Cable & System (South Korea).
• Sumitomo Electric (Tokyo-listed) and Sterlite Technologies (India-listed) also supply the U.S. market.

Bottom line for stock-pickers: to "own the industry" in size you effectively buy a diversified company (Corning, Amphenol, Prysmian) and accept that fiber cable is one revenue stream among several.

5. How the money works

This is a capital-intensive, cyclical manufacturing business, and owners make money the way factory owners do — by filling expensive plants with high-margin volume. The metrics that matter:

• Capacity utilization and operating leverage. Cabling lines are expensive fixed assets. When plants run full, incremental volume drops largely to profit; when carriers pause spending, fixed costs crush margins. Corning's optical sales fell 17% year-over-year in early 2024 during a customer inventory correction, then swung to record growth as AI demand returned — the same plants, very different economics.[19][5] Utilization is the single biggest profit lever.
• Volume × average selling price (ASP). Output is measured in fiber-kilometers. Commodity single-mode cable competes largely on price and carries thin margins; specialty and high-fiber-count cable (ribbon, bend-insensitive, high-density data-center cable, armored/harsh-environment) commands premiums. Mix shift toward specialty is how a maker lifts margins without more volume.
• Pricing is cyclical. The BLS producer-price index for cable made from purchased fiber fell from 99.7 in April 2023 to 85.9 in December 2024 — a 13.9% decline — before recovering to 87.0 in June 2025. The series was subsequently discontinued.[26]
• Input costs. Key inputs are the purchased glass fiber, polymers for jacketing/buffering, and steel or aramid for armor. CommScope's CCS financial statements identify aluminum, copper, steel, bimetals, optical fiber, plastics, and other polymers as principal inputs subject to commodity-price movements.[24] Because the industry buys strand rather than draws it, vertically integrated players (Corning, OFS, Sumitomo — who make their own fiber) enjoy a structural cost and supply-security edge over pure cablers when glass is scarce.
• Operating leverage in practice. Optical Cable Corporation's fiscal 2025 sales rose 9.5% to $73.0 million while gross profit rose 24.1% to $22.6 million — gross margin expanding from 27.3% to 30.9%. Yet the company still recorded a $0.5 million operating loss because gross profit did not cover corporate and selling expenses. OCC attributes margin movement to product mix, manufacturing volume, efficiency, material costs, and the spreading of fixed manufacturing costs across output.[10] Broader integrated businesses demonstrate upside from scale and richer connectivity content — CommScope CCS recorded 36.5% gross margin and 19.7% operating margin in 2025 — but those figures include copper cable, connectivity, software, and enclosures, not just NAICS 335921 output.[24]
• Backlog and long-term supply agreements. Because capacity takes 12–24 months to add,[14] buyers increasingly lock in multi-year contracts. Corning signed a >$1 billion multi-year fiber deal with AT&T and a ~$6 billion arrangement anchored by Meta for its North Carolina cable capacity — de-risking new plants before they're built.[6][7] For investors, backlog and anchor customers are the clearest read on forward revenue.
• Capital intensity and cyclicality. Returns swing with carrier and hyperscaler capex cycles. The business rewards owners who add capacity counter-cyclically and punishes those who expand into a downturn.

6. What drives demand

Three overlapping waves, all currently pulling the same direction:

1. AI data centers (the dominant new driver). AI compute clusters need enormous amounts of fiber inside and between buildings — AI-optimized facilities use roughly 5–10× more fiber than conventional cloud data centers, and dense GPU racks can require many times the fiber of a standard server rack.[14] Data-center fiber is projected to jump from under 5% of global fiber demand in 2024 toward ~30% by 2027.[15] This demand favors ribbon, smaller-diameter cable, pre-terminated assemblies, and connectivity systems rather than undifferentiated long-haul cable. Corning's enterprise-network sales rose from $1.98 billion in 2024 to $3.20 billion in 2025, driven substantially by generative-AI products and data-center interconnect.[23]
2. Rural broadband / fiber-to-the-home. The Fiber Broadband Association reported 76.5 million unique U.S. homes passed by fiber at year-end 2024, up 13% during that year, with average subscriber adoption of 45.2%.[27] FCC data separately showed fiber service of at least 100/20 Mbps available to 63.5 million of 115.8 million broadband-serviceable U.S. locations — approximately 55% — as of December 31, 2024.[28] The federal Broadband Equity, Access, and Deployment (BEAD) program — $42.45 billion — is moving from planning into construction, with most states choosing fiber to reach the large majority of eligible locations.[16] Separately, USDA's ReConnect program awarded 24 fiber-to-the-premises projects totaling $313 million in Round 5 (fiscal 2025).[29]
3. 5G backhaul and network upgrades. Wireless carriers need fiber to connect cell sites; densification keeps a steady baseline of demand.

A caveat on BEAD. The $42.45 billion is program authorization, not a cable-purchasing budget. Funds also pay for construction, labor, electronics, planning, and other eligible costs. Moreover, the program was revised in 2025 to require technology-neutral selection, allowing fixed wireless and satellite to compete with fiber in some areas.[30] Cable demand from BEAD will be substantial but should not be equated with the headline appropriation.

Because capacity is slow to add, this demand has produced a genuine supply shortage — demand running well above supply, extended lead times, and firming prices into 2026.[14][19]

7. Regulation

Fiber cable makers aren't rate-regulated like utilities, but policy shapes the market powerfully:

• Build America, Buy America (BABA). Federal broadband money (BEAD and other Infrastructure Investment and Jobs Act funds) generally requires domestically manufactured cable. NTIA's waiver framework requires specified optical-fiber and fiber-cable manufacturing processes to occur in the United States and previously estimated that close to 90% of BEAD equipment spending would go to U.S.-manufactured equipment.[31] This is a direct tailwind for U.S.-based production and a reason global players expand American plants.[16]
• Trade and tariffs. The sector is a live trade battleground. U.S. antidumping/countervailing duties and Section 301 tariffs have long applied to categories of Chinese optical products, limiting cheap imports; conversely, in September 2025 China imposed a ~37.9% antidumping duty on U.S.-made single-mode fiber, including Corning's, curbing U.S. exporters' access to that market.[17] Trade policy can swing both input costs and export demand.
• Telecom and safety codes. Indirect but real: FCC broadband rules, pole-attachment and permitting regimes, and building/fire codes for cable jacketing all influence what gets built and which cable qualifies.

8. Competitive dynamics and consolidation

The federal data describe a moderately concentrated industry: the top four firms hold 64.6% of receipts, the top eight 82%, and the top twenty 91.9%, with a Herfindahl-Hirschman Index of 1,364.7 (moderate concentration).[2] In practice a few global groups dominate, trailed by specialty producers.

Competition turns on three things: vertical integration (owning glass fiber supply), manufacturing scale/cost, and specialty capability for premium cable. The commodity end is price-competitive and historically pressured by Asian oversupply; the specialty and data-center end is where margins and differentiation live.

Consolidation is accelerating. The signature event is Amphenol's $10.5 billion purchase of CommScope's Connectivity & Cable Solutions business, announced August 2025 and closed January 2026 — reshuffling one of the largest U.S. fiber-cabling franchises into a diversified interconnect giant.[8] Prysmian has grown through acquisition (it absorbed General Cable), and Corning is expanding capacity via anchor-customer deals rather than M&A.[6][7][11] Expect the field to keep concentrating around vertically integrated scale players.

9. Risks

• Cyclicality and inventory whiplash. Demand tracks carrier and hyperscaler capex, which is lumpy. Corning's Optical Communications sales fell from $5.02 billion in 2022 to $4.01 billion in 2023 as carrier customers deferred projects and reduced inventory, then swung back to $4.66 billion in 2024 and $6.27 billion in 2025 as AI demand returned — a reminder that today's shortage can flip to glut.[32][5][6]
• Customer concentration. A handful of hyperscalers and large telcos drive orders; losing or a pause from one moves the needle.
• Input/glass supply. Pure cablers depend on scarce purchased fiber; a glass shortage squeezes the non-integrated players hardest.[14]
• Trade exposure. Tariffs and foreign antidumping actions (e.g., China's 2025 duty) can raise costs or close export markets abruptly.[17]
• Capacity overshoot. Everyone is adding capacity into the AI boom; if demand normalizes, the industry could face oversupply and margin compression — the classic manufacturing trap.
• Policy/timing risk on BEAD. Federal broadband spending has repeatedly shifted timelines; delays defer a chunk of demand. The 2025 technology-neutral reform may also shift some BEAD-funded projects away from fiber toward fixed wireless or satellite.[16][30]
• Downstream bottlenecks. Labor shortages and permitting delays can slow field deployment more than cable production itself. The Fiber Broadband Association found that smaller providers believed they could build 7% more network absent labor constraints, with permitting and pole access even larger obstacles — slower field construction defers cable orders.[27]
• Technology substitution (long-tail). Advances such as hollow-core fiber, higher-density optics, DOCSIS upgrades over coax, fixed wireless, or satellite could alter cable volumes over time — a slow-moving rather than imminent risk. Wavelength-division multiplexing and faster transceivers mean data traffic can rise without a corresponding increase in route-kilometers.

10. How to invest, and the outlook

Public-market routes. There is no clean pure-play at scale, so the practical choices are:

• Diversified leaders with the biggest fiber exposure — Corning (GLW) for integrated glass-and-cable, and Amphenol (APH) now that it owns the former CommScope fiber franchise. In both, fiber is a large and growing segment, not the whole company, which dilutes both the upside and the risk.
• European cable groups — Prysmian (PRY, Milan) and Nexans (NEX, Paris) — for global cable exposure with a North American slice, via foreign listings or ADRs.
• Pure-play, micro-cap — Optical Cable Corp. (OCC) is the only listed U.S. company that is essentially this industry, but it is tiny and thinly traded, with the volatility that implies.
• Investors wanting the demand theme without factory risk sometimes prefer the fiber deployers (telecom/broadband carriers) or the data-center owners buying the cable.

Private-market routes. Most U.S. fiber-cable capacity is already privately or foreign held — OFS, AFL, Superior Essex. Private and strategic capital reaches the industry through those subsidiaries, through greenfield plant investment and long-term offtake agreements (the Meta–Corning and AT&T–Corning deals are the template), and through the surrounding ecosystem of connectivity-hardware makers and network-construction firms that private equity has actively rolled up. A buyer's diligence should prioritize actual cable-versus-connectivity revenue, customer and project concentration, backlog cancellation rights, strand-supply contracts, BABA qualification, inventory ownership, plant utilization, scrap and yield, product certifications, and the capital required to change fiber counts or constructions.

Near-term outlook (forward-looking). The setup entering the mid-2020s is unusually favorable: AI data-center demand, BEAD construction, and 5G backhaul are pulling simultaneously against capacity that is slow to add, producing shortage conditions, firm pricing, and multi-year backlogs.[14][15][16][19] Domestic-content rules and reshoring add a policy tailwind for U.S. plants.[16] The credible bear case is not weak demand but overbuild: the whole industry is expanding into the boom, and manufacturing cycles reliably end in oversupply. The likely winners are the vertically integrated, scale producers who control their own glass and lock in anchor customers before adding lines — and who can keep plants full when the cycle eventually turns.

---

Sources

1. U.S. Census Bureau. County Business Patterns 2023, NAICS 335921 (establishments, employment, payroll). https://www.census.gov/programs-surveys/cbp.html
2. U.S. Census Bureau. 2022 Economic Census — Comparative Statistics & Concentration, NAICS 335921 (firms, receipts, concentration ratios, HHI). https://www.census.gov/programs-surveys/economic-census.html
3. U.S. Census Bureau / NAICS Association. 2022 NAICS Definition — 335921 Fiber Optic Cable Manufacturing (scope and cross-references to 32721 and 335929), 2022. https://www.naics.com/naics-code-description/?code=335921
4. U.S. Small Business Administration. Table of Small Business Size Standards (NAICS 335921: 1,000 employees), 2023. https://www.sba.gov/document/support-table-size-standards
5. Corning Incorporated / Business Wire. Corning Reports Strong Fourth-Quarter and Full-Year 2024 Results — Optical Communications full-year sales $4.66 billion, 2025. https://www.businesswire.com/news/home/20250128076674/en/
6. Converge Digest. Corning Delivers Record 2025 Results, Raises Springboard Growth Targets — Optical Communications $6.27 billion, 2026. https://convergedigest.com/corning-delivers-record-2025-results-raises-springboard-growth-targets/
7. Manufacturing Dive. Corning, Meta ink $6B deal supporting US manufacturing, 2026. https://www.manufacturingdive.com/news/corning-meta-6-billion-data-center-ai-north-carolina-facebook-q4-2025/810772/
8. SDxCentral / Amphenol Corporation. Amphenol to Acquire CommScope's Connectivity and Cable Solutions Business for $10.5B (announced Aug 2025, closed Jan 2026), 2025–2026. https://www.sdxcentral.com/news/commscope-sells-connectivity-and-cable-business-to-amphenol-corp-for-105b/
9. Lightwave Online. CommScope leans on connectivity and cable solutions for revenue growth (CCS ~$2.8B in 2024), 2025. https://www.lightwaveonline.com/business/article/55244162/
10. Optical Cable Corporation. Fiscal Year 2025 Form 10-K and Shareholder Letter (revenue $73.0M, gross margin 30.9%), 2025. https://www.sec.gov/Archives/edgar/data/1000230/000143774925038228/ex_897382.htm
11. Wikipedia. Prysmian Group (2024 revenue €15.6B; North America ~20%; U.S. plants), 2025. https://en.wikipedia.org/wiki/Prysmian_Group
12. —
13. —
14. Tom's Hardware. AI data centers require far more fiber than standard servers — glass shortages push cable lead times toward a year, 2025. https://www.tomshardware.com/tech-industry/ai-data-centers-are-consuming-fiber-optic-cable-faster-than-suppliers-can-make-it
15. McKinsey & Company. Opportunities in networking optics: Boosting supply for data centers, June 2025. https://www.mckinsey.com/industries/technology-media-and-telecommunications
16. The Pew Charitable Trusts. Demand for Broadband Workforce Expected to Rise to Meet BEAD Requirements ($42.45B program), October 2025. https://www.pew.org/en/research-and-analysis/issue-briefs/2025/10/demand-for-broadband-workforce-expected-to-rise-to-meet-bead-requirements
17. CGTN. China imposes anti-dumping duties (~37.9%) on certain U.S. optical fiber goods, September 2025. https://news.cgtn.com/news/2025-09-04/China-imposes-anti-dumping-duties-on-certain-U-S-optical-fiber-goods-1GodLct1sdy/p.html
18. GMI (Global Market Insights) / Market Research Future. Fiber Optic Cable Market Size (global ~$16B in 2024; U.S. market estimates), 2025. https://www.gminsights.com/industry-analysis/fiber-optic-cable-market
19. IEEE ComSoc Technology Blog. How will fiber and equipment vendors meet increased demand in 2026 due to AI data center buildouts?, December 2025. https://techblog.comsoc.org/2025/12/23/how-will-fiber-and-equipment-vendors-meet-the-increased-demand-for-fiber-in-2026-due-to-ai-data-center-buildouts/
20. Fiber Optic Association. FOA Technical Reference: Fiber Optic Cable. https://www.thefoa.org/tech/ref/basic/cable.html
21. Corning. Outdoor Cable Generic Specification PGS131. https://www.corning.com/catalog/coc/documents/generic-specifications/PGS131.pdf
22. U.S. Geological Survey. Optical Fiber Supply Chain Analysis (U.S. shipments $3.132B in 2021; 48M fiber-km production in 2022), 2024. https://pubs.usgs.gov/publication/ofr20241057/full
23. Corning Incorporated. 2025 Form 10-K, Segment Disclosure (enterprise-network $3.20B, carrier-network $3.08B). https://www.sec.gov/Archives/edgar/data/24741/000002474126000124/R27.htm
24. Amphenol Corporation. CommScope CCS Combined Financial Statements (2025 sales $3.755B, gross margin 36.5%, operating margin 19.7%), January 2026. https://www.sec.gov/Archives/edgar/data/820313/000110465926036173/aph-20260109xex99d1.htm
25. AFL. AFL Announces Multi-Million Dollar Investment and Expansion to Its U.S. Fiber Optic Cable Manufacturing, 2024. https://www.aflglobal.com/en/company/company-profile/news/2024/afl-announces-a-multi-million-dollar-investment-and-expansion-to-its-us-fiber-optic-cable
26. Bureau of Labor Statistics. Producer Price Index — Cable made from purchased fiber (PCU3359213359210), via FRED (discontinued July 2025). https://fred.stlouisfed.org/series/PCU3359213359210
27. Fiber Broadband Association. The State of North American Fiber Deployment (76.5M homes passed, 45.2% adoption), January 2025. https://fiberbroadband.org/resources/the-state-of-the-north-american-fiber-deployment-january-2025/
28. Federal Communications Commission. Broadband Deployment Report (63.5M of 115.8M locations with fiber 100/20 Mbps), 2025. https://docs.fcc.gov/public/attachments/DOC-418459A1.pdf
29. U.S. Department of Agriculture. ReConnect Program Round 5 Awards (24 projects, $313M), December 2024. https://www.rd.usda.gov/media/file/download/fy25-reconnect-round-5-121824.pdf
30. National Telecommunications and Information Administration. Trump Administration Announces BEAD Program Technology-Neutral Reform, 2025. https://broadbandusa.ntia.gov/news/latest-news/trump-administration-announces-benefit-bargain-bead-program-removes-regulatory
31. National Telecommunications and Information Administration. American-Made Internet for All — Build America, Buy America, 2024. https://www.ntia.gov/blog/2024/american-made-internet-all
32. Corning Incorporated. 2022 Form 10-K Segment Filing (Optical Communications $5.023B). https://www.sec.gov/Archives/edgar/data/24741/000143774923003123/R27.htm