[Industry Report Vol.1 (Revised & Expanded Ver.2)] Diversification in the Semiconductor & Electronic Components Industry — Why 56 Companies Succeeded or Failed in Next-Gen Power Investments, and What Truly Separated Them Was Not 'Technical Capability'

Archetype Inc. (Headquarters: Minato-ku, Tokyo; CEO: Tatsuhiko Kanno; hereinafter 'Archetype'), a company specializing in business development support, has released an industry report titled 'Industry Report 2026 Vol.1 — Semiconductor & Electronic Components Industry: Realities of New Business and Diversification (Revised & Expanded Ver.2)' (49 pages total) on July 14, 2026 (Tuesday). The report is available for free download.

▼ Download the report for free

This report uses a broad selection of 65 major industry players as its population, with 56 companies whose consolidated operating profit margins could be verified through securities reports (EDINET) as the analysis target, and drills down to the segment level for 18 core companies. All figures are mechanically extracted from each company's securities reports (FY2025, March 2026 fiscal year, etc.) via EDINET, while the four overseas reference companies (NVIDIA, TSMC, ASML, ON Semiconductor) are sourced directly from SEC EDGAR. Rather than vague estimates like 'around X%', the report precisely identifies which segment of which company achieved what operating profit margin, with full source attribution. While the initial version released in April 2026 provided a bird's-eye view of industry trends based on public and financial data, this Ver.2 is an expanded and revised edition that verifies each company's data through individual securities reports and analyzes down to the segment level.

Key Findings

Finding 1: Specialization was a necessary condition for successful diversification — Profits concentrate at the 'ends where specialization works'

When overlaying each company's segment operating profit margins onto the value chain, a 'smile curve' emerges, with high margins at both ends and low in the middle. Upstream materials (Shin-Etsu Chemical's electronics materials at 33.7%) and equipment (Advantest's test systems at 50.9%), and downstream AI high-performance components and proprietary technologies (Murata's components at 27.2%, Renesas' automotive at 30.7%, Hamamatsu Photonics' imaging and measurement at 29.6%) show high profitability, while midstream general-purpose modules and assembly show low margins. Among the 18 core companies, consolidated operating profit margins range from a high of 51.0% at Keyence to a low of -5.9% at Sanken Electric (among 56 companies, 10 exceed 20%, 32 are in single digits, 2 are in deficit, median 7.1%).

Profit Map (Smile Curve)

Ranking of consolidated operating profit margins for 18 core companies (51.0% to -5.9%, median line)

Finding 2: But even with specialization, some companies collapse — What separates success from failure is not technical capability, but 'business design'

Specialization is necessary but not sufficient. A symbolic example is next-generation power semiconductors (SiC). Despite investing in the same technology, ROHM (consolidated -8.9% → +2.3%) and Kyocera (1.4% → 5.7%) overcame the initial deficit valley, while Sanken Electric (-3.1% → -5.9%) sank. What separated them was not technical capability, but the presence or absence of a 'diversification buffer' to absorb deficits during the valley period, and the design of exit strategies. This report categorizes these failure patterns into three vulnerabilities: narrow demand base, lack of technology ownership, and lack of capital buffer.

SiC 'Valley' Outcomes (ROHM and Kyocera succeeded, Sanken failed) / 5 Types of Specialization × 3 Vulnerability Patterns

Finding 3: Many large deficits are not due to 'core business failure' — Distinguishing impairment losses

Many notable deficits in FY2025 were due to one-time accounting factors such as goodwill impairment and amortization from acquisitions (e.g., Murata Manufacturing's deficit was primarily due to a full 43.8 billion yen impairment related to Resonant Inc.). Distinguishing between 'impairment-driven deficits' and 'structural core business deficits' using primary data significantly changes business diagnosis. Misjudging this could lead to dismissing promising diversification efforts as 'failures'.

Internal Polarization Dumbbell (High-profit segments × Loss-making segments)

Finding 4: Companies that designed not just 'where to profit' but 'how to build' the business are the ones that cross the valley

Beyond financial analysis, the latter half of this report dissects 'how the business was created — through what business model and organizational structure'. It extracts, with real examples, four types of business model transformation, three types of organizations that generate new ventures, and a large-enterprise version of the 'valley of death'. It connects financial analysis with the analysis of how new ventures are built, all within one report.

Census distribution of 56 companies (10 above 20%, 32 in single digits, 2 in deficit, median 7.1%)

The Report's Unique Framework

To help companies with world-class technology but struggling to convert it into stable profits or new business exits determine 'where and how we should compete', this report analyzes the industry through three layers: 5 Types of Specialization × 3 Vulnerability Patterns × How New Ventures Are Built.

5 Types of Specialization: Categorizes 'places where specialization pays off' to determine where a company's core business stands

3 Vulnerability Patterns: Organizes the conditions under which specialization turns into weakness — narrow demand base, lack of technology ownership, lack of capital buffer

How New Ventures Are Built: Connects financial outcomes with implementation through four types of business model transformation, three types of organizations that generate new ventures, and the large-enterprise 'valley of death'

Overseas benchmarks (NVIDIA 60.4%, TSMC 50.8%, ASML 34.6%) serve only as reference points for 'profit ceiling', while organizational design is discussed using Japanese company examples. These connect to Archetype's free organizational diagnostic tool, 'Organization Score Diagnostic for New Ventures', and its 6-axis scoring system.

Executive Comment

Tatsuhiko Kanno (CEO, Archetype Inc.)

'What we face daily is the challenge of Japanese mid-sized and quasi-large companies that possess globally competitive technology but fail to convert that advantage into stable profits or viable new business exits. After verifying the securities reports of 56 major companies with primary data, we were once again confronted with the clear fact that even with identical technology investments, some companies clearly succeed in crossing the initial deficit valley while others sink. And what separates them is not technical capability itself, but the design of diversification strategies that absorb deficits and the decisions on exit strategies. What matters is not to treat this as a matter of individual company superiority or inferiority, but to reframe it as a design problem of creating a 'winning structure' that any company can replicate.'

About the Industry Report 2026 Series

Archetype is sequentially releasing the 'Industry Report 2026' series, which organizes industry-specific new business and diversification strategies from primary sources.

Vol.1 Semiconductor & Electronic Components Diversification Strategy (Revised & Expanded Ver.2) (released today, free download)

Vol.2 Realities of Agricultural Entry and Diversification Strategy (published, free download)

Vol.3 New Business in Media & Entertainment Industry (published, free download)

Future reports on automotive, materials & chemicals, machinery, food, and other industries will be sequentially released.

Survey Overview

Survey Name: Semiconductor & Electronic Components Industry: Realities of New Business and Diversification Report (Revised & Expanded Ver.2)

Publisher: Archetype Inc.

Population: 65 major companies / Analysis Target: 56 companies with verifiable consolidated operating profit margins via securities reports / Segment Analysis: 18 core companies

Primary Sources: Each company's securities reports (EDINET, FY2025 / March 2026 fiscal year)