护城河对比
最多三家已覆盖的公司，逐项对照。每一项判断都是人工编辑的判断，而非公司披露的数字——并且每一项都附有其引用依据。
| Marvell Technology | IBM | Pfizer | |
|---|---|---|---|
| 护城河评级 | 窄 FY2026 10-K (filed 2026-03-11): differentiated platform IP — over 10,000 issued patents and pending applications as of 2026-01-31, plus a proven custom ASIC platform leveraging ultra-high-speed SerDes, silicon photonics, co-packaged optics and custom HBM — but Marvell itself calls its markets 'intensely competitive' with 'pricing pressures', notes customers 'have chosen to develop certain semiconductor products internally', and discloses two >=10% customers with the ten largest at 82% of FY2026 net revenue. Real, defensible IP in a concentrated, contestable customer base = narrow, not wide. https://www.sec.gov/Archives/edgar/data/1835632/000183563226000011/mrvl-20260131.htm | 窄 IBM's FY2025 10-K describes "a highly competitive environment" with "hundreds of competitors" and says IBM is "regularly exposed to new competitors" as it executes its hybrid-cloud/AI strategy, while differentiating through "incumbency with enterprises" and "client relationships and trust" — a real but narrow moat, not a wide one. | 正在侵蚀 The FY2025 10-K publishes the expiry schedule of its own protection. The Patents table dates the basic U.S. product patent on Eliquis ($7,961 million of 2025 revenue) to 2027, the Prevnar family's Prevnar 13 ($6,494 million for the family) to 2026, the Vyndaqel family ($6,380 million) to 2026 with a 2028 extension only pending, Ibrance ($4,122 million) to 2027, Xtandi ($2,194 million) to 2027 and Xeljanz ($1,087 million) to 2026 - against Total revenues of $62,579 million. The filing then states it anticipates 'a significant reduction of revenue from patent-based or regulatory exclusivity expiries in 2026 through 2030 as several of our in-line products experience these expirations, with the rate of the reduction of revenues from patent-based or regulatory exclusivity expiries expected to significantly accelerate over the next few years', and puts the 2026 instalment at $1.5 billion. A protection whose end dates the issuer tabulates product by product, and whose revenue effect it forecasts as accelerating, is being consumed on a clock it has already disclosed. |
| 护城河类型 | 无形资产与知识产权 The moat rests on hard-to-replicate mixed-signal IP: the 10-K describes the custom ASIC platform built on ultra-high-speed SerDes, ARM compute, security, storage, silicon photonics and advanced packaging (die-to-die interconnects, chiplets, CPO, custom HBM), with multiple 5nm designs executed, 3nm in progress and a 2nm platform in development; a secondary switching-cost element comes from multi-year custom design wins co-developed to individual customer specifications. https://www.sec.gov/Archives/edgar/data/1835632/000183563226000011/mrvl-20260131.htm | 转换成本 The 10-K roots IBM's differentiation in "incumbency with enterprises" and "client relationships and trust" — installed-base lock-in across its hybrid-cloud platform (Red Hat) and enterprise software and infrastructure raises switching costs. | 无形资产与知识产权 The filing locates the protection in the patent estate, not anywhere else. The Patents and Other Intellectual Property Rights section tabulates 'the patent rights we consider most significant in relation to our business as a whole, together with the year in which the basic product patent expires', and states that once protection is lost 'we typically lose market exclusivity on these products, and generic and biosimilar pharmaceutical manufacturers generally produce identical or highly similar products and sell them for a lower price.' The Competition section names patents as the thing rivals attack: 'several of our competitors operate without large R&D expenses and make a regular practice of challenging our product patents before their expiration', and 'Generic pharmaceutical manufacturers pose one of the biggest competitive challenges to our branded small molecule products because they can market a competing version of our product after the expiration or loss of our patent protection.' No network, switching cost or cost-curve advantage is claimed anywhere in the section. |
| 领先地位 | 并列领先者 Leader in its optics niche, #2 in custom silicon: ~60% of high-end PAM4 DSP share (36kr, 2026-06-27, https://eu.36kr.com/en/p/3870758441178373) but an estimated 20-25% of custom AI ASIC design services versus Broadcom's ~70% (hashrateindex, 2026-05-13, https://hashrateindex.com/blog/design-partners-ai-asic-market-part-2/) — net, a co-leader in AI data-center connectivity/custom silicon behind Broadcom overall. | 快速追随者 An enterprise incumbent that differentiates via "incumbency with enterprises" but is "regularly exposed to new competitors" in its hybrid-cloud/AI push (FY2025 10-K) — a follower in the current AI cycle rather than its leader. | 并列领先者 The 2026 Pharma 50 (Drug Discovery & Development, published 31 March 2026, ranking pharmaceutical revenue only) places Pfizer third at $62.58 billion of FY2025 revenue inside a five-company band that runs from Johnson & Johnson's Innovative Medicine at $60.40 billion to Eli Lilly's $65.18 billion, with Merck & Co. at $65.01 billion and AbbVie at $61.16 billion in between - scale-equal to its nearest peers rather than ahead of them. The same piece attributes the top slot to Lilly's tirzepatide franchise, which it puts at $36.5 billion combined and says passed Keytruda as the world's best-selling drug; Pfizer's 10-K names no comparable single franchise, discloses no market-share position for any product, and concedes in Item 1A that 'some of our competitors may have competitive, technical or other advantages over us' and that it faces 'an increasing number of potential competitors worldwide, including from China, that have expanded R&D capabilities.' |
| 定价权 | 中等 The 10-K characterizes Marvell's markets as having 'pricing pressures' and intensifying competition (https://www.sec.gov/Archives/edgar/data/1835632/000183563226000011/mrvl-20260131.htm), yet the Q1 FY2027 release reports 52.1% GAAP / 58.9% non-GAAP gross margin on record revenue (https://www.sec.gov/Archives/edgar/data/1835632/000183563226000014/q127_8kx522026ex-991.htm) — differentiated-IP margins, tempered by hyperscaler buyer power. | 中等 Gross margin is high and rising (54.9% FY2021 to 58.2% FY2025) on a software-mix shift, but "price" is a principal method of competition per the 10-K and consulting/infrastructure remain price-competitive. | 弱 Price is the lever the filing shows moving against it. CMS selected Eliquis - which the 10-K says 'accounted for 13% of Total revenues in 2025' - for the Medicare Drug Price Negotiation Program, and 'its government-set Maximum Fair Price became effective January 1, 2026', with Ibrance and Xtandi following in 2027 and Xeljanz in 2028. In September 2025 Pfizer 'voluntarily agreed' with the U.S. administration to make certain U.S. prices 'more comparable to those in other developed countries' and to sell through the TrumpRx.gov platform 'at significant discounts to current retail prices'. Gross-to-net product revenue deductions rose to $36,374 million in 2025 from $30,048 million in 2023 while Total revenues fell to $62,579 million, and the IRA Medicare Part D Redesign alone 'negatively impacted our 2025 revenues by approximately $1 billion'. Lower U.S. net price is named as a drag on Vyndaqel, Ibrance, Xeljanz, Nurtec ODT and Lorbrena; every 2025 growth line the filing explains is attributed to demand, patient share or launch uptake, none to price. Cost of sales at 25.7% of revenues shows the manufactured margin is intact - what is not intact is the ability to set the price. |
| 综述 | Marvell is a fabless data-infrastructure silicon supplier whose center of gravity has shifted decisively to the AI data center: the data center end market was $6,100.3M, 74% of FY2026 revenue, up from 40% two fiscal years earlier (FY2026 10-K). Its strongest position is electro-optics — in high-end PAM4 optical DSPs for 400G+ transceivers it holds roughly 60% share on Inphi-inherited SerDes/FEC IP, with Broadcom above 30%, the two together over 90% (36kr, 2026-06-27). In custom AI silicon it is the structural #2 design partner at an estimated 20-25% of the custom AI ASIC design-services market versus Broadcom's ~70%, anchored by AWS Trainium and Microsoft Maia wins (hashrateindex, 2026-05-13). The Q1 FY2027 release (2026-05-27) shows the flywheel turning — record $2.418B revenue (+28% YoY), Q2 guided to $2.7B mid-point (+35% YoY), management citing 'exceptional AI-related bookings' across 800G/1.6T optics, 51.2T switches, CPO/NPO and custom XPU — and the Celestial AI (Photonic Fabric) and XConn (PCIe/CXL switching) acquisitions closed in February 2026 extend the interconnect moat toward scale-up fabrics. The offsets that keep the moat narrow are in Marvell's own filing: intense competition (AMD, Alchip, Astera, Ayar, Broadcom, Credo, GUC, Lightmatter and others), hyperscaler in-housing risk, and heavy customer concentration. | IBM's moat rests on deep enterprise incumbency — sticky installed-base relationships across hybrid cloud (Red Hat), enterprise software, and mission-critical infrastructure, plus brand and a large patent base. But its own FY2025 10-K frames "a highly competitive environment" with "hundreds of competitors" and says IBM is "regularly exposed to new competitors" as it pursues hybrid cloud and AI, so the moat is durable but narrow and contested in the AI/cloud growth arena. | Two clocks run in opposite directions inside the same company. The revenue base is near-dated: Eliquis, the Prevnar family, the Vyndaqel family, Ibrance, Xtandi and Xeljanz all carry U.S. basic product patents expiring 2026 or 2027 in the 10-K's own table, and the filing warns the resulting revenue loss will 'significantly accelerate over the next few years'. The replacement book is later-dated - Padcev and Lorbrena to 2033, Nurtec ODT and Litfulo to 2034, Prevnar 20 to 2035, Abrysvo, Cibinqo and Elrexfio to 2036, Comirnaty and Paxlovid to 2041 - but in 2025 it was still much smaller, and the two products with the longest patent runway are the two the filing says are shrinking fastest (Comirnaty down 20% and Paxlovid down 59% operationally). The 10-K's answer is purchased pipeline rather than defended position: it cites the Seagen and Metsera acquisitions as 'significant investments in obesity and oncology, respectively, which are extremely competitive therapeutic areas', while R&D expense itself fell 4% to $10,437 million. Excluding Comirnaty and Paxlovid the remaining business grew 6% operationally in 2025, which is the honest measure of what is being rebuilt underneath the expiries. A third pressure sits on top of the patent clock and is independent of it: government price-setting now reaches the largest product in the book. |
| 产业链位置 | Fabless supplier spanning 'data center core to network edge': it sits between hyperscaler AI compute (custom XPU/XPU-attach ASICs) and the optical layer (PAM4/coherent DSPs, CPO/LPO, DCI, AEC, PCIe retimers), outsourcing fabrication to independent CMOS foundries; the Feb-2026 Celestial AI and XConn acquisitions push it further into scale-up photonic fabric and PCIe/CXL/UALink switching (FY2026 10-K, https://www.sec.gov/Archives/edgar/data/1835632/000183563226000011/mrvl-20260131.htm). | A hybrid-cloud platform, enterprise-AI (watsonx), and consulting provider to large enterprises — a software/cloud layer of the AI stack. | Demand side of the AI chain, not a supplier into it. The 10-K lists 'Scale AI across our business' as one of four 2026 key priorities and describes 'expanding automation, data-driven decision making, and enterprise AI solutions', crediting 'enhanced digital enablement, including automation and AI' for part of the savings in its cost-realignment and R&D-simplification programs. It describes no AI model, tool or platform that Pfizer sells. |
| 产品（份额 / 壁垒） |
|
|
|
| 长周期投票 | +0.13 权重 0.20 · 群体判断 bullish 编辑先验，未经回测。 | +0.06 权重 0.20 · 群体判断 bearish 编辑先验，未经回测。 | -0.17 权重 0.20 · 群体判断 bearish 编辑先验，未经回测。 |