Compare moats

MACOM’s FY2025 10-K describes advantages that are real but contested. On the durable side: a catalog of “thousands of standard and custom devices” serving “over 6,000 end customers”; its own compound-semiconductor fabs (GaAs, GaN and InP) in Lowell, Research Triangle Park, Ann Arbor and Limeil-Brévannes; a Lowell fab accredited by the U.S. Department of Defense with “Trusted Foundry” status, in markets where “a domestic fabrication facility may be a requirement to be a strategic supplier”; and product life cycles of five to ten years, “with some of our products generating revenue for over 20 years”. On the limiting side, the same filing calls its markets “highly competitive”, names ADI, Broadcom, Credo, Marvell, MaxLinear, Microchip, NXP, Qorvo, Semtech, Skyworks and Sumitomo among its significant competitors, “some of whom have greater financial resources and scale than us”, adds “increased competition from Chinese companies”, sells “primarily on a purchase order basis” with no minimum purchase commitments, and warns that “the ASPs of our products may decrease over time”. Its income statement reports gross profit of $385,797 thousand on revenue of $648,407 thousand in fiscal 2023, $393,773 thousand on $729,578 thousand in fiscal 2024 and $529,002 thousand on $967,258 thousand in fiscal 2025. Process know-how and defense accreditation that larger rivals can contest, product generation by product generation, is a narrow moat rather than a wide one.

source: sec.gov

The advantage is narrow and confined to one small unit. In the FY2025 20-F (filed 2026-03-05), Nokia Technologies turned EUR 1 501m of net sales into EUR 1 059m of operating profit (70.6%) while the group earned EUR 885m on EUR 19 889m (4.4%) — licensing out-earned the whole company on 7.5% of sales. Mobile Networks ran a 2.8% margin, down 270bps from 5.5% in 2024, against risk factors citing "equipment price erosion and aggressive price competition". Nokia Technologies net sales fell 22% from EUR 1 928m and operating profit 30% on 2024 catch-up, yet it is not eroding: over EUR 800m of contracted recurring revenue runs through 2030 and Technology Standards grew 12% cc in H1'26.

source: sec.gov

The FY2025 10-K's own Competition section undercuts any claim of insulation: 'The industries in which we operate are highly competitive, and we compete with thousands of companies that range from large multinational corporations to local manufacturers', competition is 'generally based on breadth of product offering, product innovation, price, quality, delivery, and service', and TE has 'experienced, and expect[s] to continue to experience, downward pressure on prices'. The same filing explicitly disclaims IP as the basis of its position: 'we do not believe that our competitive position or our operations are dependent upon or would be materially impacted by any single patent or group of related patents.' What TE does hold is durable rather than absolute — it calls itself 'one of the leading providers of advanced automobile connectivity solutions', its Industrial segment 'a leading supplier', and it converted that position into rising profitability across the same filing's three income statements (gross margin 31.5% of net sales in FY2023, 34.4% in FY2024, 35.2% in FY2025). A position that earns expanding margins into admitted price pressure is an advantage; one held against thousands of competitors and no decisive patent is not a wide one.

source: sec.gov

The FY2025 10-K places the advantage in process and design know-how: “We continue to invest in proprietary processes, circuit design and packaging technologies”, “we utilize a broad array of internal, proprietary process technologies and commercially available foundry technologies”, and U.S.-based wafer fabrication “enables us to offer proprietary processes” and provides “a domestic source for U.S. I&D customers”. The filing itself says patents matter less than people: MACOM held 729 U.S. and 497 foreign issued patents as of October 3, 2025, but “we believe that our future success will be determined by the innovation, technical expertise and management abilities of our engineers and management more than by patent ownership”. Switching costs are the weaker candidate: long product life cycles help in Industrial & Defense, but sales are made on purchase orders and new business must be won through “a competitive selection process to develop semiconductors for use in our customers' systems, known in the industry as a ‘design win.’”

source: sec.gov

The advantage is standard-essential cellular patents, and it lives in one unit. Nokia Technologies conducts "cellular, multimedia and WiFi research and standardization, protects Nokia's innovation by securing patents" and "monetizes Nokia's innovation through patent licensing", across "more than 26 000 families, including over 8 000 essential to 5G" and "more than 250 licensees" including Apple, Samsung and Mercedes-Benz. It is not network effects or switching costs: customers "may also consolidate their supplier base to our disadvantage - all the way to a one-supplier model". From 1 January 2026 the unit reports as Technology Standards inside Mobile Infrastructure.

source: sec.gov

The FY2025 10-K locates the advantage at the customer, not in patents. It describes 'close working relationships with many of our customers' whose 'relationships with them typically date back many years', and co-development as the mechanism: 'By working with our customers in developing new products and technologies, we believe we can identify and act on trends and leverage knowledge about next-generation technology across our products.' Roughly 75% of fiscal 2025 net sales were direct to manufacturers rather than through distribution, and the Transportation segment's products 'must withstand harsh conditions' — parts engineered in with the customer and built to survive the application, so the position sits inside the customer's design rather than in a purchase order. The filing rules the alternative out in its own words on intellectual property, so intangibles_ip is not the source; the 'thousands of companies' in Competition rules out efficient_scale.

source: sec.gov

The only independent ranking found is for the optical analog line. Deep Fundamental’s “Deep Dive: Optical Module Market” of September 27, 2024 (https://deepfundamental.substack.com/p/deep-dive-optical-module-market) states “In the driver/TIA market, Marvell ($MRVL) and Macom ($MTSI) are the dominant players, especially in the high-end 400G+ segment” and “in the LPO market, Macom holds a significant advantage, largely because Marvell, strong in DSP, tends to promote solutions that include DSP”. That is a shared front rank, from an equity-research newsletter rather than a market tracker, and now two years old. MACOM’s own FY2025 10-K claims no rank - it says “We believe that we compete favorably” - and names significant competitors from ADI and Broadcom to Marvell and Sumitomo, “some of whom have greater financial resources and scale than us”. No third-party ranking was found for the Industrial & Defense or Telecom lines. A two-year-old newsletter covering one of three segments cannot carry a company-wide co-leader call, so the band is fast follower.

source: sec.gov

Nokia is top-three almost everywhere and decisive only in slices. The 20-F reports that "Dell'Oro and Omdia ranked Nokia third in global Mobile RAN market share for the first three quarters", and Radio Networks is the largest single business unit (EUR 1 765m of EUR 4 815m group, Q2'26). It also holds #2 in global optical, #1 in IP edge routing and #1 in xPON OLT for a sixth year (Dell'Oro/Omdia Q3'25), while in data centers it competes "against large incumbent players" and calls expansion "challenging". 2025 segment sales were Network Infrastructure EUR 7 986m vs Mobile Networks EUR 7 806m; the Q2'26 6-K shows Optical +20%, IP +16% cc, AI & Cloud EUR 446m vs 220m, Radio +7%.

source: sec.gov

The FY2025 10-K claims leadership in qualified form and then names the peers who contest it. Transportation Solutions 'is a leader in connectivity and sensor technologies' and TE is 'one of the leading providers of advanced automobile connectivity solutions' — one of, not the — while the segment's 'major competitors include Yazaki, Aptiv, Sumitomo, Sensata, Honeywell, Molex, and Amphenol'. Industrial Solutions 'is a leading supplier of products that connect and distribute power, data, and signals' and 'competes primarily against Amphenol, Hubbell, Carlisle Companies, Integer Holdings, Molex, Omron, JST, and Korea Electric Terminal (KET)'. Amphenol and Molex appear on both lists, so TE shares the top of the interconnect market rather than owning it.

source: sec.gov

The FY2025 10-K says “the ASPs of our products may decrease over time, and we must introduce new products that can be manufactured at lower costs or that command higher prices based on superior performance to offset price erosion”. On inflation it says MACOM has “generally been able to offset increases in these costs through various productivity and cost reduction initiatives, as well as adjusting our selling prices to pass through some of these higher costs to our customers; however, our ability to raise or maintain our selling prices depends on market conditions and competitive dynamics.” Margins have risen with the Data Center mix: the fiscal Q3 2026 earnings release (https://www.sec.gov/Archives/edgar/data/0001493594/000149359426000036/ex99_1earningsreleaseq3fy26.htm) reports GAAP gross margin of 58.3%, compared to 55.3% a year earlier, and guides fiscal Q4 adjusted gross margin to between 60.0% and 61.0%. The release does not say how much of that is price rather than mix or fab utilisation.

source: sec.gov

Pricing power is weak at the consolidated grain, in the filing's own words. The 20-F risk factors describe a market "characterized by maturing industry technologies... equipment price erosion and aggressive price competition", customers "reverting to vendors to compensate" for their own unit-revenue erosion, and warn Nokia "may increasingly be required to agree to less favorable contractual terms in order to remain competitive". Group operating margin was 4.4% in 2025. This band covers equipment, ~92% of sales, not licensing, whose FRAND-bound royalties ran a 70.6% margin; supply-tight optical is our inference, not filing text.

source: sec.gov

The FY2025 10-K states both halves plainly: TE has 'experienced, and expect[s] to continue to experience, downward pressure on prices. However, as a result of increased costs and tariffs, certain of our businesses implemented price increases in recent years.' Pass-through, not price-setting. The realised result is margin expansion rather than erosion — gross margin of 31.5% of net sales in FY2023, 34.4% in FY2024 and 35.2% in FY2025 per the same filing's income statements, and 36.5% ($5,319M on $14,573M) for the nine months to 26 June 2026 versus 35.3% a year earlier, with Q3 FY2026 GAAP operating margin of 19%, per the 22 July 2026 results release (https://www.sec.gov/Archives/edgar/data/1385157/000110465926085589/tel-20260722xex99d1.htm). Strong would require pricing that leads rather than follows cost; weak is contradicted by three years of expanding gross margin.

source: sec.gov

MACOM sells components that customers build into larger systems - the FY2025 10-K lists wireless basestations, high-capacity optical networks, data center networks, radar, medical systems, satellite networks and test and measurement. In the AI chain it sits upstream of optical-module makers, supplying TIAs, drivers, lasers and photodetectors for 800G and 1.6T transceivers. Sales to distributors were 32.3% of fiscal 2025 revenue, two resellers took 12.4% and 11.2%, and no direct customer reached 10%.

TE is an upstream component supplier to the AI build-out, not a participant in compute. It sells 'connectivity and sensor solutions [that] enable the distribution of power, signal, and data to advance next-generation transportation, energy networks, automated factories, data centers enabling artificial intelligence', about 75% of it direct to manufacturers across roughly 130 countries. AI exposure runs through one end market: digital data networks was 28% of Industrial Solutions, and Industrial Solutions was 46% of fiscal 2025 net sales — so the datacenter line is a minority of a company still 54% transportation. That minority is where the growth now is: digital data networks net sales rose 34.2% in Q3 FY2026 and 48.8% over the nine months, and energy rose 34.4% in the quarter, the CEO tying both to AI momentum 'in both the data center and across the broader energy infrastructure' (https://www.sec.gov/Archives/edgar/data/1385157/000110465926085589/tel-20260722xex99d1.htm).

• Optical TIAs and modulator/laser drivers Top 3 · Moderate source: deepfundamental.substack.com
• Photodetectors and lasers for optical transceivers Unknown · Moderate source: marketbeat.com
• RF, microwave and GaN products for Industrial & Defense Unknown · Moderate source: sec.gov
• Telecom optical and RF components Unknown · Moderate source: sec.gov

• Core Software (mobile core and autonomous networks) Leader · Moderate source: sec.gov
• Fixed Networks Leader · Moderate source: sec.gov
• IP Networks Challenger · Moderate source: sec.gov
• Optical Networks Top 3 · Moderate source: sec.gov
• Radio Networks (Mobile RAN) Top 3 · Moderate source: sec.gov
• Technology Standards (patent licensing, formerly Nokia Technologies) Unknown · Deep source: sec.gov

• Aerospace, defense, and marine interconnect Challenger · Deep source: sec.gov
• Automotive terminals and connector systems Top 3 · Deep source: sec.gov
• Digital data networks connectivity Top 3 · Moderate source: sec.gov
• Energy connection and protection products Challenger · Moderate source: sec.gov
• Sensors Challenger · Moderate source: sec.gov

Editorial prior, not backtested.

see exactly how it voted →

Editorial prior, not backtested.

see exactly how it voted →

Editorial prior, not backtested.

see exactly how it voted →