Compare moats
Up to three covered companies, band by band. Every call is a curated editorial judgment, never a disclosed figure — and every band carries its cited basis.
| MaxLinear | JCET Group | GitLab | |
|---|---|---|---|
| Moat rating | none MaxLinear’s FY2025 10-K does not show a durable advantage. Its income statement reports revenue of $693,263 thousand in 2023, $360,528 thousand in 2024 and $467,641 thousand in 2025, with operating losses of $38,221 thousand, $223,352 thousand and $126,890 thousand. The risk factors say “Increased competition has resulted in price pressure, decreased demand, reduced revenue and profitability, and loss of market share”; the business section says competitors include “companies with much longer operating histories, greater name recognition, and substantially greater financial, technical and operational resources”; and because its products “often are building block semiconductors” it also faces integrated-circuit makers, “some of which may be existing customers or platform partners”. Two customers were 28% of 2025 net revenue and the ten largest 65%, and “substantially all of our sales to date have been made on a purchase order basis”. Gross profit held up - $385,663 thousand, $194,782 thousand and $265,814 thousand for 2023 to 2025 - and the AI optical ramp has lifted 2026 results, but a revenue base that nearly halved in one year and three straight years of operating losses do not evidence a moat. | none JCET has scale but not excess returns. TrendForce's 2024 OSAT ranking (2025-05-13) puts it third with '$5B in revenue (+19.3% YoY)', behind ASE at $18.54B ('nearly a 45% share among the top ten') and Amkor at $6.32B. JCET's 2025 annual report, citing ChipInsights, says it stayed third globally and first in mainland China, with the top three OSATs holding more than 52% combined. That scale has not lifted returns. The same report gives weighted average ROE of 5.81% in 2023, 6.00% in 2024 and 5.56% in 2025, or 4.86% in 2025 excluding non-recurring items. Main-business gross margin was 13.95% in 2025, up 1.07 percentage points. Net profit attributable to shareholders fell 2.75% to RMB 1.565 billion on revenue of RMB 38.87 billion. ROE for the first half of 2026 was 2.92%. | narrow The FY2026 10-K (year ended 2026-01-31) shows real expansion inside the installed base: 118% Dollar-Based Net Retention, Base Customers up to 10,682 from 9,893, customers above $100,000 of ARR up to 1,456 from 1,229, and $1.0 million ARR customers "to 155 from 123, an increase of 26%". But the filing concedes "limited barriers to entry" and names "Microsoft Corporation, which owns GitHub" as principal competitor. Retention decelerates (130% FY2024, 123% FY2025, 118% FY2026), and since DBNR is reported "on a threshold basis of 130%," the FY2024 figure is a cap and the slide is steeper than it looks. Narrow, not wide: expansion must be re-won against a bundled hyperscaler. |
| Moat type | none No single source of advantage in the 10-K is strong enough to name. MaxLinear has “over one thousand issued patents” and says consideration under intellectual property sale agreements “has previously been and is currently expected in the future be material”, but the same filing describes its products as building blocks that larger vendors can integrate, says some optical-interconnect customers are “module makers who are vertically integrated, where we compete with internally supplied components”, and claims only that “We believe that we compete favorably” on factors from product performance to price. Its RF-CMOS integration know-how is real, but nothing in the filing shows customers locked in or rivals unable to match it. | none None of the usual moat sources holds up in the cited evidence. Cost scale: JCET ranks third, but its ROE has stayed between 5.56% and 6.00% for three years, and TrendForce's 2024 ranking puts ASE at $18.54B against JCET's $5B. IP: the annual report lists 3,123 patents, 2,601 of them invention patents. The same report's 2026 plan says that in high-end advanced packaging the company will 'accelerate catching up' in customer binding, technology R&D and capacity. Its high-end advanced packaging plant, JCET Microelectronics, posted a net loss in 2025 while ramping. Switching costs: the top five customers took 48.80% of 2025 sales. Amkor's FY2025 10-K names JCET Group with ASE Technology and Powertech as established Asian competitors and says IDM customers weigh outsourcing against 'their own in-house capabilities'. JCET's 1H26 report adds that foundries and IDMs are speeding up their own advanced packaging. Policy also limits the high end. TrendForce (2025-02-14, citing Commercial Times) reports that with Chinese OSAT firms excluded from the U.S. PW list, 'high-end chip packaging is blocked in China', pushing those orders to Taiwanese OSATs. | switching costs Stickiness comes from platform embedding, not IP. The 10-K rests differentiation on "our single platform with a unified data model," on "Consolidation of multiple tools into a single platform," and on deployment inside the customer's own perimeter for regulated buyers. Replacement means re-integrating a toolchain and re-certifying compliance. IP is the wrong label: only "16 issued patents and 25 pending patent applications," plus open-source licenses granting "broad permissions to use, copy, modify, and redistribute." The no-lock-in claim is the filing's own characterisation, footnoted to a Forrester study "commissioned by GitLab." |
| Leadership | fast follower In its fastest-growing line MaxLinear is a second source. Deep Fundamental’s September 27, 2024 deep dive (https://deepfundamental.substack.com/p/deep-dive-optical-module-market) says “Marvell ($MRVL) and Broadcom ($AVGO) are two major suppliers of DSPs, with Marvell holding the top position in the market”, that “Coherent also sources heavily from Marvell, with Broadcom/ Maxlinear potentially serving as second supplier with 20-30% share”, and that MaxLinear offers “DSPs at about half the price of Marvell's if it can achieve a meaningful mass production volume of at least 100K units per month”. Keystone has since reached volume - management said on the Q2 2026 call that it “continues to ramp into high volume production at major hyperscale customers across U.S. and Asia” - but the 10-K names Broadcom, Qualcomm, Realtek, Skyworks, Credo, MediaTek, Marvell, MACOM, Texas Instruments, Analog Devices, Renesas, Microchip and Semtech as primary merchant competitors, and no third-party source found ranks MaxLinear first in any of its markets. | fast follower TrendForce's 2024 ranking has JCET third at $5B, against ASE's $18.54B and Amkor's $6.32B. The annual report, citing ChipInsights, keeps it third in 2025. JCET is in volume at the high end but describes itself as catching up there. Its 2025 report says the XDFOI chiplet process has entered mass production and the company made 'substantive progress' in 2.5D volume production, but its 2026 plan also says it will 'accelerate catching up' in high-end advanced packaging. TrendForce (2026-01-30) reports that JCET's XDFOI-based silicon-photonics engine completed customer sample deliveries and passed client-side validation, and that UMC is targeting CPO mass production in 2027. | fast follower The 10-K positions GitLab against a leader rather than as one: "Our principal competitor is Microsoft Corporation, which owns GitHub." Competition reads from the challenger side, differentiating "from GitHub through flexible deployment options that work within enterprise security and compliance requirements, LLM neutrality with self-hosted gateway support" and an open core model. On AI it follows: Duo Agent Platform went "Generally available in January 2026," the last month of the fiscal year. What keeps it a follower rather than behind is cadence, "a new version of our software every month for 172 months in a row," plus a regulated-deployment franchise a SaaS-first rival does not cover. |
| Pricing power | moderate Mixed. Gross profit has stayed in proportion through the cycle - the 10-K’s income statement shows $265,814 thousand on revenue of $467,641 thousand in 2025 - and the Q2 2026 release (https://www.sec.gov/Archives/edgar/data/0001288469/000128846926000050/a06302026exhibit991.htm) reports GAAP gross margin of 57.8% against 56.5% a year earlier, guiding Q3 to 57.0%-60.0%. But the 10-K says “From time to time, we have reduced the average unit price of our products due to competitive pricing pressures, new product introductions by us or our competitors, and for other reasons, and we expect that we will have to do so again in the future”, that under some distributor agreements “we provide protection for reductions in selling prices of the distributors' inventory”, and in optical DSPs it entered as the lower-priced challenger. | weak Main-business gross margin was 13.95% in 2025, up 1.07 percentage points, and materials made up 68.65% of packaging and test cost. The 2025 risk section says China's packaging industry 'has attracted many participants', and that rising competition may lower the industry's average unit prices and margins. The 1H26 report goes further: as new capacity comes on line, 'price competition in traditional packaging is especially prominent'. It also says the company improved its price-linkage mechanism and that gross margin 'continued to improve'. | moderate FY2026 gross margin was 87% against 89%, with cost of revenue up $35.6 million "primarily due to an increase of $18.4 million in third party hosting costs for SaaS and cloud usage," and management expects SaaS and Duo mix to bring costs that "may adversely impact our gross margins." Strain is explicit: GitLab "may be required to reduce our prices," competitors "may offer their products and services at a lower price or for free," and it "implemented user limits on our free SaaS product." Power retained: Premium lists at $29 per user/month billed annually, and seats bundle $12 and $24 of GitLab Credits. But "A majority of our subscriptions are on a one-year period." |
| Summary | MaxLinear is a fabless designer of RF, analog and mixed-signal communications SoCs whose core skill, per its FY2025 10-K, is combining broadband RF and analog front ends with digital signal processing in standard CMOS. It sells into broadband access (cable, fiber PON and DSL gateways - approximately 44% of 2025 net revenue), home connectivity (Wi-Fi, MoCA, G.hn and Ethernet), wired and wireless infrastructure including optical data-center DSPs, and industrial and multi-market interface and power products. After a downturn that took revenue from $693,263 thousand in 2023 to $360,528 thousand in 2024, the AI optical ramp is turning it around: the Q2 2026 release reports revenue of $168,847 thousand, up 55% year over year, with the infrastructure business up 145% on the Keystone PAM4 DSP ramp for 800G, and management raised its 2026 optical data-center revenue outlook to $210 million-$230 million on the call. The moat question is whether that growth rests on anything durable. In optical DSPs MaxLinear entered as a lower-priced second source to Marvell; elsewhere it competes with Broadcom, Qualcomm, Realtek and MediaTek, which can integrate the functions it sells; customers are concentrated; and the 10-K still carries the Silicon Motion arbitration over its terminated merger, whose outcome it says it cannot predict. On this record MaxLinear is a technically capable challenger without a moat. | JCET is mainland China's largest OSAT and the world's third-largest. It runs eight production bases in China, South Korea and Singapore and offers wafer-level, 2.5D/3D, system-in-package, flip-chip and wire-bond packaging, plus wafer probe and final test. Revenue in 2025 was RMB 38.87 billion, up 8.09%. By application, communications was 36.4%, consumer 23.6%, computing 21.3%, automotive 9.6% and industrial and medical 9.1%. Computing revenue grew 42.6% in 2025, then grew 40.4% in 1H26 to reach 30.1% of revenue. Volume is still mostly mainstream packaging: 38,880.05 million traditional units against 18,275.89 million advanced units in 2025. Most revenue comes from abroad. Overseas sales were RMB 30.44 billion of RMB 38.71 billion in main-business revenue, at a 12.20% gross margin against 20.40% on domestic sales. JCET is building in several directions: the XDFOI chiplet platform, a silicon-photonics engine for co-packaged optics that TrendForce (2026-01-30) says passed client-side validation, the JSAC automotive plant in Lingang, and the SDSS flash packaging plant, in which it bought 80% from Sandisk in September 2024. The 1H26 report shows revenue up 5.0% to RMB 19.53 billion, with net profit attributable up 79.4% to RMB 844.6 million. Even with that half-year gain, ROE was 2.92% for the half. The verdict is no moat: JCET is the scale leader among Chinese packagers, and its returns have stayed near 5% to 6% for three years while it catches up at the high end. | GitLab sells a single-platform DevSecOps toolchain defended by workflow lock-in rather than intellectual property: a unified data model plus deployment flexibility, including single-tenant "Dedicated for Government with FedRAMP compliance," makes it hard to remove where data residency is contractual. The installed base expands: 118% Dollar-Based Net Retention, 10,682 Base Customers, 1,456 above $100,000 of ARR, 155 above $1.0 million (up 26%), over 70% of ARR from public sector and enterprise. The filing supplies the counterweight: "limited barriers to entry," Microsoft/GitHub as principal competitor, 16 issued patents. Retention (130 to 123 to 118, the 130 a cap), gross margin (89% to 87%) and a January 2026 AI launch point to compression at the commodity end. No source measures market share, so every share band is "unknown". |
| Chain position | MaxLinear sells chips, much of it through Asian distributors, ODMs and module makers - products shipped to Asia were 82% of 2025 net revenue, including 49% to Hong Kong - that build cable modems, PON terminals, Wi-Fi gateways, base-station radios and optical transceivers for operators and hyperscale data centers. In the AI chain it sits beside the optical-module makers as a DSP supplier, competing with Marvell and Broadcom and, at vertically integrated module makers, with internally supplied components. | Back-end assembly and test contractor for chip designers, IDMs and foundries. In 2025, overseas sales were RMB 30.44 billion of RMB 38.71 billion in main-business revenue, and the five largest customers took 48.80% of sales. | GitLab sits at the software-tooling layer, not compute or models. It consumes third-party LLM capacity — marketing "LLM neutrality and support for self-hosted AI gateways, including air-gapped environments" — so it captures no model-layer economics and bears inference as COGS, visible in the $18.4 million rise in hosting costs. Its distinctive position is the regulated perimeter. |
| Products (share / barrier) |
|
|
|
| Long-horizon vote | +0.01 at weight 0.20 · swarm neutral Editorial prior, not backtested. | -0.06 at weight 0.20 · swarm neutral Editorial prior, not backtested. | +0.06 at weight 0.20 · swarm neutral Editorial prior, not backtested. |