Ferrotec Corporation (6890.T) - Deep Dive Research Report

Prepared 2026-05-31. Fiscal year historically ended March 31; the company is transitioning to a December 31 year-end. The four most recent earnings calls used throughout: Q1 FY3/26 (14 Aug 2025), H1/Q2 FY3/26 (28 Nov / 8 Dec 2025), Q3 FY3/26 (16 Feb 2026), and Full-Year FY3/26 (15 May 2026).

Per the brief, this report contains no valuation, no absolute revenue or profit figures, and no price targets. Segment scale is expressed only as approximate revenue mix. Dividend and share-count figures appear in Section 10 as required.

Section 1: What the company does

Ferrotec makes the unglamorous physical parts that sit deep inside a semiconductor factory and never appear in a finished chip, but without which no chip gets made. When a wafer is being etched or coated inside a vacuum chamber, something has to spin a shaft or move a robot arm through the wall of that chamber without letting air leak in and ruin the vacuum. Ferrotec's original product - the ferrofluidic vacuum seal - solves exactly that. It uses a magnetic liquid (ferrofluid) held in place by magnets to form a liquid O-ring that seals a rotating shaft. The company has built roughly 60% of the world market for these seals and, by its own account, employs more engineers and manufacturing staff dedicated to ferrofluid sealing than every competitor combined.

From that single component the company expanded outward into a wide catalogue of "consumables and components" that semiconductor equipment makers and chip fabs burn through: fabricated quartz chambers and tubes, advanced ceramics, silicon and silicon-carbide parts, reclaimed wafers, and parts-cleaning services. In parallel it built a second leg in "electronic devices" - thermoelectric (Peltier) modules that heat or cool by running current across a junction, and power-electronic ceramic substrates that carry the heat away from the chips inside EV inverters. The thread tying it all together is materials science applied to thermal and vacuum problems.

The company traces directly to NASA. Ferrofluid was developed under NASA sponsorship in the 1960s to move liquid rocket fuel in zero gravity. Ferrofluidics Corporation commercialised it in the US in 1968. Akira Yamamura set up the Japanese arm, Nippon Ferrofluidics, in 1980, led a management buyout in 1987, and in a striking reversal acquired the original American parent in 1999, renaming the whole group Ferrotec. Yamamura then made the defining strategic bet of the company's life: he moved manufacturing aggressively into mainland China through the 2000s and 2010s, building scale in low-cost, high-volume production of semiconductor materials long before "China semiconductor localisation" became a theme. That bet is the source of both Ferrotec's cost advantage and its biggest risk today. Yamamura died in April 2024; the group is now led by Group CEO He Xian Han, with the founder's relative Takeru Yamamura as Vice President. The holding company simplified its name from Ferrotec Holdings to Ferrotec Corporation in July 2025.

The essence of the business: Ferrotec sells the picks and shovels of chipmaking and power electronics, manufactured at Chinese-scale cost, to a global customer base that increasingly wants those same parts made outside China.

Section 2: Business segments

Ferrotec reports in three segments. A legacy photovoltaic/solar-materials business and miscellaneous industrial equipment now sit inside "other" lines rather than as a headline segment.

Semiconductor and Other Equipment-related Business (the core, roughly 55-60% of revenue)

This is the heart of the company and the part the market cares about. It makes the vacuum seals, fabricated quartz ware, advanced ceramics, silicon parts (sold under the SiFusion and related brands), CVD silicon carbide parts, silicon wafers, reclaimed wafers, and offers parts-cleaning and refurbishment. These are the items inside an etch or deposition tool that wear out and must be replaced, or the precision components a tool builder designs into a new machine.

The core capability is process knowledge accumulated over decades: how to machine quartz to the tolerances and purity a plasma chamber demands, how to braze ceramics, how to grow and finish silicon parts, and how to do all of it at volume and at a cost point Western fabricators struggle to match. Quartz is illustrative - Ferrotec does not make the raw quartz; it is one of the largest fabricators certified by raw-material suppliers including Heraeus, Tosoh, Saint-Gobain and Momentive, and the value it adds is the shaping, fusing and qualification. This segment was the standout in FY3/26, with management citing strength in vacuum seals and parts for manufacturing equipment, and it is the engine the entire growth plan is built around.

Electronic Devices Business (roughly 25-30% of revenue)

This segment contains three distinct product families. First, thermoelectric (Peltier) modules - solid-state heat pumps used to cool laser diodes, optical transceivers, scientific instruments, and increasingly the optics in data-centre networking. Management explicitly attributed this segment's FY3/26 strength to "capturing demand for generative AI," which points at thermal management for AI networking and compute. Second, ferrofluid itself, sold as a material and into audio (loudspeaker cooling) and sensing. Third, and strategically the most interesting, power-electronic ceramic substrates - DCB (Direct Copper Bonding) and AMB (Active Metal Brazing) substrates that insulate and conduct heat away from the power chips inside EV inverters and industrial converters. Ferrotec built the first domestic Japanese mass-production line for silicon-nitride AMB substrates in 2019 and its Shanghai and Dongtai plants together rank first in China and second in the world by volume for these substrates.

This segment exists separately because its customers (electronics, automotive, data-centre) and its physics (thermal device engineering) differ from the consumables-and-cleaning logic of the equipment segment, even though both draw on the same materials base.

Automotive-related Business (roughly 10-15% of revenue)

Thermoelectric and thermistor-type components and ceramic parts sold into vehicles, with significant exposure to EV thermal management. This was the weak segment in FY3/26 - revenue and profit both fell as the EV market went through what management called an "adjustment." It is the smallest leg and currently the drag, but it shares manufacturing and materials with electronic devices.

Section 3: Products and business detail

Ferrofluidic vacuum seals (FerroSeals). The founding product. A ferrofluid trapped by a magnetic circuit forms a hermetic, frictionless seal around a rotating shaft, letting motion pass into a vacuum chamber without leaks. Used in etching and deposition tools for chips, plus flat-panel display, LED and solar equipment, and in panel-transfer robots. The standalone ferrofluidic-seal device market is small in absolute terms (around US$244M in 2024, growing roughly 5% a year), but Ferrotec dominates it and sells higher-value sealing sub-assemblies on top.

Fabricated quartz. Quartz tubes, chambers, bell jars and fixtures for plasma and thermal processes. Requires extreme purity and dimensional precision; Ferrotec is a top-tier certified fabricator.

Advanced and machinable ceramics. Insulators, structural and functional ceramic parts for vacuum and high-temperature semiconductor processes; also feed into medical and optoelectronic uses. Ceramics capacity expansion is a named capex priority.

Silicon and silicon-carbide parts. Silicon electrodes, rings and CVD-SiC components consumed inside etch chambers; the SiFusion brand covers silicon products used as consumable parts.

Silicon wafers and reclaimed (reclaim) wafers. Prime and test/reclaim wafers. Note: Ferrotec sold a majority of its mainland China silicon-wafer subsidiary to Chinese government-linked and private funds (cutting its stake below 30%) with the aim of a separate China listing - an early example of its strategy of monetising Chinese subsidiaries while keeping the materials core.

Parts cleaning and refurbishment. A recurring service business - chamber parts are removed, cleaned, recoated and returned. The new Beijing plant under construction is a cleaning facility.

Thermoelectric (Peltier) modules and thermistors. Solid-state heat pumps and temperature sensors for optics, lasers, instruments, AI networking and automotive.

Power-electronic substrates (DCB and AMB). Copper-clad ceramic substrates - alumina, aluminium-nitride and silicon-nitride - that insulate and cool power chips. AMB silicon-nitride is the high-end product for EV/SiC inverters.

Geographies and manufacturing. The manufacturing base is concentrated in China (Hangzhou, Shanghai, Dongtai, and a new Beijing cleaning plant) and Japan, with US and European sales and engineering. The pivotal current build-out is a large-scale mass-production base in Kulim, Malaysia, explicitly created to serve the "ex-China" production needs of European and US customers. Customer qualification and ramp at Kulim were reported as progressing smoothly through FY3/26.

Section 4: Customers

The primary customers are the semiconductor capital-equipment makers (the "SPE" makers - the likes of Applied Materials, Tokyo Electron and Lam Research are the archetypes for this kind of consumable and component supply) and, downstream, the chip fabs that buy replacement consumables directly. A second customer pool is automotive and industrial power-electronics makers buying substrates and thermoelectric parts, and a third is the optics/AI-networking and instrument makers buying thermoelectric cooling.

The buying decision inside an equipment maker or fab is made by sourcing and process-engineering teams, and the binding criterion is qualification. A quartz chamber, a silicon electrode or a vacuum seal has to be qualified on a specific tool running a specific process; once a part is qualified and running in production, swapping suppliers means re-testing and risking yield, which fabs are extremely reluctant to do. That is the switching cost, and it is real: it is process-qualification lock-in rather than a contract. Sales cycles for new qualifications run many months to years.

Why customers buy Ferrotec specifically: for vacuum seals, there is effectively no one else at its scale or depth, so it is the default. For quartz, ceramics and silicon parts, the pull is the combination of qualification track record and a China-based cost structure that undercuts Western fabricators. The newest and most powerful reason is geopolitical - US and European customers want a supplier who can give them the same parts made outside China, which is precisely what Kulim is being built to do.

Concentration is more about end-market than single accounts: the equipment segment rises and falls with the wafer-fab-equipment spending cycle and, recently, with AI-driven capex. Revenue predictability is mixed - the consumables and cleaning streams recur as long as fabs run, while the component-into-new-tools business swings with the equipment cycle.

Section 5: Competitive landscape

The competitive picture differs sharply by product, so a single "moat" statement would be misleading.

Vacuum seals. This is Ferrotec's strongest position by far - roughly 60% global share, a fragmented tail of small competitors, and a workforce dedicated to the niche that exceeds all rivals combined. High share in a small, specialised, qualification-bound market is a genuine moat, though the market itself grows only mid-single-digits.

Quartz, ceramics, silicon parts. Here Ferrotec competes as one fabricator among several. In quartz it sits alongside Shin-Etsu Quartz, Heraeus and Tosoh (some of whom are also its raw-material suppliers). In ceramics it faces Kyocera, CoorsTek and others. The edge is cost (Chinese manufacturing) plus qualification breadth; the exposure is that these are more contestable, more commoditised markets where a customer's ex-China demand could equally be served by a Western fabricator if Ferrotec's Malaysia ramp stumbles.

Power-electronic substrates. Competitors include Rogers/Curamik, Heraeus, Kyocera and Toshiba Materials (a leader in silicon-nitride). Ferrotec is the volume leader in China and #2 worldwide in DCB/AMB combined, but high-end silicon-nitride AMB is a technology race tied to the SiC/EV ramp, and demand here softened with the EV adjustment.

Silicon wafers. A market dominated by Shin-Etsu, SUMCO and GlobalWafers; Ferrotec deliberately stepped back by selling down its China wafer subsidiary rather than fighting that battle directly.

Barriers to entry are highest in vacuum seals (decades of ferrofluid know-how, tiny market not worth a giant's effort) and in qualification-locked consumables; they are lower in standardised substrates and wafers. The defining structural shift is supply-chain regionalisation - the "China + 1" / ex-China push - which is simultaneously Ferrotec's largest opportunity (it can build the alternative base) and its largest threat (its existing Chinese cost base is the thing customers are diversifying away from).

Section 6: Industry

Demand for Ferrotec's core products is driven by the semiconductor capital-equipment cycle and by chip-fab utilisation. When equipment makers build more etch and deposition tools, they buy more seals, quartz and silicon parts; when fabs run hot, they consume more replacement consumables and cleaning. Layered on top of the normal cycle is the AI capex super-cycle - the build-out of AI compute and data centres is pulling both leading-edge wafer-fab-equipment spending and, through thermal management, Ferrotec's thermoelectric and substrate products.

The wafer-fab-equipment market is large (tens of billions of dollars annually) and notoriously cyclical, swinging with memory pricing and capex appetite; Ferrotec's consumables exposure dampens but does not eliminate that cyclicality. The thermoelectric module and power-substrate markets ride the separate cycles of optics/AI networking and of EV/industrial electrification.

Ferrotec sits in the middle of the chip supply chain - between raw-material producers (it buys quartz feedstock, for example) and the equipment makers and fabs. The single most important industry dynamic shaping its future is supply-chain regionalisation. Export controls and customer risk-management are pushing US and European chip and equipment makers to source critical parts outside China. Ferrotec, being a China-heavy manufacturer, is responding by building ex-China capacity (Malaysia) to defend and grow that demand. Regulation here is the demand driver, not just a constraint: the tighter the controls, the stronger the pull toward qualified non-China suppliers - provided Ferrotec is on the right side of the line.

Cyclically, the equipment and consumables business is volatile; the cleaning/refurbishment and recurring-consumable streams are steadier; automotive substrates are tied to the EV cycle, which is currently in a downswing.

Section 7: Growth triggers

All points below are drawn from the four FY3/26 concalls/disclosures.

• Kulim, Malaysia ex-China production base ramping. A large-scale mass-production base built to serve European and US customers' ex-China needs; customer qualification and ramp reported as progressing smoothly. Repeated across multiple FY3/26 disclosures. (Full-Year FY3/26 disclosure, 15 May 2026; reiterated from earlier quarters.)

  "The customer qualification and ramp-up at the Kulim, Malaysia factory are progressing smoothly." (Mid-Term Plan / Full-Year disclosure, 15 May 2026)

• New Beijing plant (parts cleaning) within the raised capex plan. Named as a specific use of expansion capital. (Full-Year FY3/26 / Mid-Term Rolling Plan, 15 May 2026.)

• Ceramics capacity expansion. Additional ceramics production capacity called out as a dedicated investment line. (Mid-Term Rolling Plan, 15 May 2026.)

• Cumulative capex plan raised to ¥174.6 billion (March 2026 through December 2027). Management lifted the multi-year investment envelope in direct response to "robust semiconductor demand." (Mid-Term Rolling Plan, 15 May 2026.)

• Ex-China qualification wins with US SPE customers. Management cited "progress on the ex-China production needs of US SPE customers" with "strong inquiries and capacity expansion requests." (Full-Year FY3/26, 15 May 2026.)

  "Strong inquiries and capacity expansion requests from customers." (Full-Year FY3/26, 15 May 2026)

• Generative-AI demand in Electronic Devices. Management attributed the segment's growth directly to capturing generative-AI demand (thermal management for AI), framed as a continuing tailwind. (Full-Year FY3/26, 15 May 2026; theme present since Q1.)

• Upgraded forward plan for the new December fiscal periods. Management raised its medium-term targets and guided the December 2026 period to materially higher sales and profit than the year just ended, citing AI and data-centre demand. (Mid-Term Rolling Plan, 15 May 2026.)

Section 8: Key risks

China concentration - the defining risk. The bulk of manufacturing sits in mainland China (Hangzhou, Shanghai, Dongtai, Beijing). The mechanism cuts two ways. First, tightening US/European export controls or customer risk-aversion could strand or shrink demand for China-made parts faster than Ferrotec can shift it to Malaysia. Second, the very strategy of building ex-China capacity is an admission that the existing base is a liability for a chunk of the customer book. If the Malaysia ramp slips while customers de-risk, Ferrotec loses on both ends. This is a high-probability, high-impact structural risk, and management's entire capex program is a response to it.

Semiconductor cyclicality. A large part of the equipment-related segment swings with wafer-fab-equipment spending. A capex downcycle - if the AI build-out cools or memory rolls over - would hit the component-into-new-tools business hard, only partly cushioned by recurring consumables and cleaning. High-probability over a multi-year horizon, moderate-to-severe impact.

Execution risk on a large, front-loaded capex program. The ¥174.6B plan, multiple new plants and the Malaysia ramp all carry start-up costs. Management already flagged that FY3/26 profit growth was constrained by inventory write-downs at new factories - the risk that new capacity comes on faster than qualified demand fills it is live and self-disclosed.

FY3/26 profit growth was "constrained due to increased inventory write-downs at new factories." (Full-Year FY3/26, 15 May 2026)

EV/automotive downturn. The automotive segment already fell in FY3/26 on an EV-market "adjustment," and AMB substrate demand is tied to the SiC/EV ramp. A prolonged EV slowdown keeps this leg a drag and slows the high-end substrate growth story. Moderate probability, moderate impact (small segment).

Commoditisation in quartz/ceramics/wafers. Outside vacuum seals, Ferrotec competes on cost in contestable markets. If Chinese competitors (some of them its own former subsidiaries or JV partners) scale up, or if Western fabricators win the ex-China business instead, pricing power erodes. Moderate probability, moderate impact.

Governance and complexity. A China-heavy structure with partially divested subsidiaries (the silicon-wafer sale, China listings) creates a web of minority stakes and related-party complexity that is hard for outside investors to fully see through. The founder's death in April 2024 also marks a leadership transition to a non-founder CEO. Low-probability but worth watching.

Section 9: Walk the talk

Four reference points: Q1 FY3/26 (14 Aug 2025), H1 FY3/26 (28 Nov / 8 Dec 2025), Q3 FY3/26 (16 Feb 2026), and Full-Year FY3/26 with the Mid-Term Rolling Plan (15 May 2026). The most recent is well within 90 days of today.

Through the year the consistent message was a recovering semiconductor environment led by AI, with the equipment-related and electronic-devices segments carrying the load and automotive lagging. That narrative held from Q1 to year-end without reversal - management did not flip-flop on direction, which is the first marker of credibility. The equipment segment they pointed to early as the strength did finish the year as the strongest performer (driven by vacuum seals and parts for manufacturing equipment), and the AI-pull they flagged in Electronic Devices showed up in that segment's full-year strength. On the things they could control operationally, the story they told in the autumn matched the result they printed in May.

Where management showed honesty rather than spin was on profit quality. They did not bury the fact that bottom-line profit attributable to owners actually declined for the year even as sales and operating profit rose, and they attributed it plainly to inventory write-downs at the new factories. That is the kind of self-disclosure that builds trust - they named the cost of their own expansion rather than letting it surprise the market.

Profit growth was "constrained due to increased inventory write-downs at new factories." (Full-Year FY3/26, 15 May 2026)

The forward-looking test is the bigger one. Alongside the May results, management raised its medium-term plan and its capex envelope (to ¥174.6B through December 2027) and guided the new December 2026 period to materially higher sales and profit, explicitly on AI and ex-China demand. This is management leaning further in, not trimming. It is consistent with the conviction they expressed all year, but it has not yet been delivered - the Malaysia ramp, the Beijing plant and the ceramics expansion are promises, not outcomes. The pattern that can be judged - a year of accurate directional calls and candour about a profit shortfall they caused themselves - reads as a management team that does roughly what it says and does not hide the warts. The open question is execution on an ambitious, front-loaded capex plan; the record on big China-scale build-outs is the founder's legacy, but this is the first such cycle being run by the post-founder leadership.

Plain assessment: directionally accurate and refreshingly candid about self-inflicted costs; the heavy forward guidance is credible but unproven, and the right thing to track over the next two December periods is whether new-plant ramps convert into the profit the raised plan implies.

Section 10: Shareholder friendliness index

Dividends. Ferrotec pays a semi-annual dividend that has trended upward over the multi-year horizon (its five-year dividend growth rate is roughly +40%). For the year ended March 2025 the annual dividend was about ¥160 per share. For the year ended March 2026 the dividend was around ¥148 per share (payout ratio in the low-40s percent) - a modest step down that tracked the ~5% decline in bottom-line profit caused by new-factory write-downs, not a policy change. For the transitional December 2026 period management has guided a higher ¥200 per share, split ¥150 ordinary plus a ¥50 special, at a ~47% payout ratio. The willingness to add a special dividend while simultaneously raising capex signals a management that wants to share the up-cycle with holders rather than hoard it.

Buybacks and dilution. Buybacks have been small and tactical rather than a core capital-return tool - the most recent authorised program was for up to 200,000 shares (about 0.43% of shares outstanding) for roughly ¥500 million, framed as capital-efficiency housekeeping. Share count has been broadly stable (on the order of ~47 million shares, with only token treasury holdings), so holders have not suffered meaningful option dilution, but neither is the count being materially shrunk. Note that capital is also returned indirectly through monetising Chinese subsidiaries (e.g. the silicon-wafer stake sale) rather than via buybacks.

Verdict: Returns Capital (modestly) - a rising, special-dividend-topped payout funded alongside heavy growth capex, with buybacks immaterial; the dividend, not repurchases, is the real return vehicle.

Section 11: Insider activities

Ferrotec trades on the Tokyo Stock Exchange, where insider and large-shareholder activity is disclosed through EDINET Large Shareholding Reports (大量保有報告書) for 5%+ stakes and TDnet timely disclosures for officer holdings. I attempted to retrieve the underlying EDINET filings directly; the primary EDINET large-shareholder feed for 6890 was not accessible within the search budget (the portal returns auth/script-gated responses to automated retrieval), and no clean primary filing for the trailing 12 months could be confirmed. Granular insider transaction data for this TSE name could not be verified against a primary source within the search budget, and is therefore not reported as fact below. What can be stated from confirmed disclosures:

• Founder transition. Founder and Honorary Chairman Akira Yamamura died in April 2024. Any subsequent movement of his holdings (estate transfer, foundation, or family succession) would appear as EDINET large-shareholder changes; this is the single most likely source of any large insider share movement in the period and should be checked directly on EDINET before relying on it.
• Family still in management. Takeru Yamamura serves as Representative Director and Vice President, indicating continued family involvement and, typically, a continued family shareholding.
• Company buyback. The issuer itself ran a small repurchase authorisation (up to 200,000 shares / ~¥500M, valid to 31 March 2025) - a corporate action rather than an individual insider trade, but the only confirmed "insider-side" buying of stock in the window.
• Disclosed substantial holders seen in secondary databases include institutions and corporates such as GLOBERIDE, The Kita-Nippon Bank, Sawakami Asset Management and Micronics Japan, but specific dated transactions could not be tied to primary filings here.

Net assessment: No verifiable pattern of individual open-market insider buying or selling could be established from primary sources within the search budget, so the honest read is inconclusive / not assessable rather than bullish or bearish. The only confirmed signals are mildly shareholder-friendly at the corporate level (a small buyback and a raised, special-topped dividend). A user who needs a definitive insider read should pull EDINET 6890 large-shareholder reports and TDnet officer-holding disclosures directly, with particular attention to any post-April-2024 reallocation of the founder's estate.

Section 12: Scenarios

Bull case. The AI build-out keeps the wafer-fab-equipment cycle hot for years, and ex-China sourcing pressure intensifies. Ferrotec's Kulim plant qualifies fast and ramps cleanly, so US and European equipment makers route a growing share of their seal, quartz, ceramic and silicon-parts buying to a Ferrotec base they can put in front of their own compliance teams. The ceramics expansion and Beijing cleaning plant fill quickly, the early-stage inventory write-downs roll off, and the new factories swing from cost drag to profit engine. Electronic Devices rides AI thermal management - more thermoelectric cooling for optics and networking - while the EV market eventually stabilises and the high-end silicon-nitride substrate line finally scales into the SiC inverter wave. The December-period plan that management raised in May 2026 proves conservative. In this world Ferrotec stops being seen as "the China-risk semiconductor parts maker" and is re-rated as the qualified ex-China consumables supplier the West needs.

Base case. Demand stays solid but not euphoric. The equipment segment grows on AI and a normal cyclical recovery; vacuum seals remain a steady, dominant cash franchise; Electronic Devices grows on AI thermal pull. Malaysia ramps roughly on schedule but with the usual qualification lumpiness, and new-plant start-up costs keep a lid on margins for a period before easing. Automotive stays soft until the EV cycle turns. Management delivers something close to its raised plan - higher sales and profit in the new December periods - but the path is bumpy and the bottom line is gated by how fast write-downs and start-up costs clear. Dividends keep rising modestly. The China-concentration overhang persists but does not detonate. This is broadly the trajectory the four FY3/26 calls describe.

Bear case. The capex plan front-runs the demand. New plants in Beijing, Malaysia and the ceramics lines come on while qualification wins arrive slower than hoped, so inventory write-downs and start-up losses deepen rather than fade - exactly the dynamic management already flagged in FY3/26, but bigger. Simultaneously, an escalation in export controls or a customer-driven China purge shrinks the existing China-made order book faster than Malaysia can backfill it, and Western fabricators (Shin-Etsu Quartz, Heraeus, Kyocera) win the ex-China business instead of Ferrotec. The EV slump drags on, keeping automotive and AMB substrates weak, and a semiconductor capex downcycle hits the equipment segment at the worst possible moment - high fixed costs from new factories meeting falling utilisation. In this scenario the very expansion meant to de-risk China becomes a stranded-cost problem, and the raised plan is quietly walked back.

A note on the charts: the segment revenue-mix percentages are approximations inferred from segment commentary and growth rates (Ferrotec does not split absolute segment revenue cleanly in the public summaries I could access); the YoY direction and dividend/capex figures are from the official 15 May 2026 disclosures. The "prior plan" capex figure in the last chart is illustrative of the upward revision direction, with only the revised ¥174.6B figure confirmed.

Sources:

• Ferrotec Consolidated Financial Results FY ended March 2026 (TDnet/BigGo)
• Ferrotec Mid-Term Management Plan (Rolling Plan) disclosure (TDnet/BigGo)
• Ferrotec News Release / Results Presentation (irstreet PDF)
• Ferrotec Investor Relations & Earnings (Quartr)
• Ferrotec IR library: financial / quarterly results
• Ferrotec Medium-term Growth Strategy
• Ferrotec H1 FY3/26 Results Presentation (Nov 28)
• Ferrotec Markets / end-markets
• Ferrotec Ferrofluidic Vacuum Seals
• Ferrotec DBC & AMB Power Electronic Substrates
• Ferrotec Quartz Products
• Ferrotec company history
• Ferrotec Directors and Executive Officers
• DCFmodeling: Ferrotec history, ownership, how it makes money
• SCMP: Ferrotec founder Akira Yamamura profile
• Ferrotec announces sale of 60% of mainland silicon-wafer subsidiary
• Ferrotec dividend history (stockinvest.us)
• a2-finance: Ferrotec dividends
• Ferrofluidic Seals market size (Valuates)
• Ferrotec Manufacturing Malaysia (Kulim)

A few honesty caveats worth flagging: I could not parse the official results PDFs directly (they returned as binary), so segment and dividend specifics rely on the TDnet summary and reputable secondary aggregators; absolute segment revenue splits are approximations; and TSE insider/large-shareholder filings (EDINET) could not be verified against primary sources within the budget, which Section 11 discloses explicitly rather than papering over.