Revenue mix is opaque because the company doesn’t disclose segment margins cleanly, but the Thin Film segment is the one driving the recent multiple expansion — that is where ruthenium sputter targets sit.
Tagline. A Japanese precious-metal artisan turned semiconductor materials supplier — the company’s iridium crucibles still grow LED sapphire, but the share-price story is now ruthenium sputter targets for advanced-node interconnects.
What it does (plain English). Furuya buys, refines, and fabricates platinum-group metals (PGM) — primarily ruthenium (Ru), iridium (Ir), and platinum (Pt) — into highly specialized industrial products. Two product families matter for the equity story: (1) sputtering targets sold to semiconductor and HDD customers and (2) iridium and platinum crucibles used to grow single-crystal sapphire (LED) and to melt specialty glass. The rest is thermocouples, fine chemicals, and PGM recycling.
Key business lines (FY25, fiscal year ended June 2025; segment splits approximated from company filings):
| Segment | What it makes | ~% of revenue |
|---|---|---|
| Electronics | Iridium / platinum crucibles for sapphire & glass; precision tooling | ~30–35% |
| Thin Film | Ru / Ir / Pt sputtering targets and evaporation materials (HDD, semis, FPD, MRAM) | ~30–35% |
| Thermal | Thermocouples, temperature sensors using Pt-Rh | ~10% |
| Fine Chemicals & Recycling | PGM compounds, catalyst pre-cursors, scrap refining | ~15–20% |
| Supply Chain Support | PGM brokerage / leasing to external manufacturers | ~5–10% |
Revenue mix is opaque because the company doesn’t disclose segment margins cleanly, but the Thin Film segment is the one driving the recent multiple expansion — that is where ruthenium sputter targets sit.
Business model. Furuya operates on a metal-content + processing-fee model. Customers either (a) consign the precious metal — Furuya processes it for a fee and the metal price washes through — or (b) buy the finished product with metal embedded, in which case Furuya bears working-capital and price risk on the inventory. The company hedges the embedded inventory on the LBMA / TOCOM, but hedges are imperfect (ruthenium has no liquid futures market). This matters: reported revenue is heavily distorted by spot PGM moves, and the processing fee — the “real” margin — is a small slice of the headline number. Investors who buy Furuya as a pure semis play often miss this.
Geographic mix. ~55% Japan, ~30% Asia ex-Japan (Taiwan, Korea, China for HDD and semis), ~10% North America, ~5% Europe.
Latest IR materials. FY25 results presentation (Aug 2025): https://www.furuyametals.co.jp/ir/library/ — also publishes mid-term plan (FY24–FY27) on the same page. No formal “investor day” deck; the August earnings deck is the canonical reference.
Asset map: the IR site has a Japanese-only facility overview at https://www.furuyametals.co.jp/company/factory.html. No English-language asset map embedded; flag for follow-up.
Furuya does not disclose named customers — Japanese specialty-materials suppliers rarely do. Best public information is from Nikkei reporting, industry interviews, and trade publications.
| # | Customer (inferred) | Ticker | Est. Revenue Share | Relationship Type |
|---|---|---|---|---|
| 1 | Western Digital / Seagate / Toshiba HDD | WDC, STX, 6502.T | 15–20% | HDD media sputter targets (Ru for granular media) |
| 2 | TSMC / Samsung / Intel (via distributors) | 2330.TW, 005930.KS, INTC | 10–15% and growing | Semiconductor sputter targets — the GAA story |
| 3 | Nichia / Sumitomo Electric / Cree (Wolfspeed) | private, 5802.T, WOLF | 10–15% | Iridium crucibles for sapphire (LED) and SiC seed crystals |
| 4 | Samsung Display / LG Display / BOE | 005930.KS, 034220.KS | ~5% | FPD sputter targets (declining mix) |
| 5 | Japanese auto / chemical catalyst makers | various | ~5% | Pt / Rh catalyst pre-cursors |
The “why.” As CMOS scaling pushes below 3nm, copper interconnects fail. Their resistance explodes at sub-15nm pitches because the barrier and liner layers consume too much of the cross-section. Ruthenium is the leading candidate to replace copper at the lowest metal layers — and at the very lowest layers, replace tungsten in the contact / via — because Ru can scale without a barrier and offers lower line resistance at extreme pitches. TSMC has already used a Ru liner on N3 (per SemiAnalysis IEDM 2022 coverage); IBM-Samsung have demonstrated subtractive Ru interconnects at 18nm pitch; Intel has used Ru in 10nm/Intel-7 lines. The open question is whether Ru goes from liner (a few tens of milligrams per wafer) to bulk metal (multiples of that) at A14 / 2nm and below.
End-use applications. - Semiconductors — sputter targets. The growing book. Logic at 3nm and beyond, DRAM electrode work, MRAM stacks, advanced packaging. - HDD media — sputter targets. The legacy cash cow. Ru-Co-Cr-Pt alloy targets for granular PMR / HAMR. Volume flat, margin decent. - Sapphire growth — iridium crucibles. Used to grow Al₂O₃ single crystals for LED substrates and increasingly SiC seed. Capex-cycle exposed. - Optical / specialty glass — Pt crucibles. Optical fiber, LCD glass, specialty optics. Mature. - Thermocouples. Industrial process control. Stable mature business.
TAM. Ruthenium sputter-target market for semis was a sub-$200M niche in 2022, projected by industry consultants (TECHCET, SEMI) to grow to roughly $400–600M by 2028 driven by GAA logic and DRAM. The iridium-crucible market is roughly $300–400M globally, dominated by Furuya and Tanaka Kikinzoku. Important calibration: Furuya’s own FY25 revenue was ¥57.4B (~$370M USD) — so the semis-target TAM is genuinely a “potentially company-transformative” line.
Market share. Furuya is widely cited as the #1 or #2 producer of Ru sputter targets globally, with primary competition from Tanaka Kikinzoku (private), Heraeus (private), and Mitsubishi Materials (5711.T). In iridium crucibles for sapphire, Furuya is the global #1.
Secular tailwinds. - GAA logic ramp at TSMC / Samsung / Intel (2026–2028) - HBM4 / DRAM scaling — Ru electrodes in capacitor stacks - MRAM (Samsung, TSMC) — Ru in magnetic tunnel junction stacks - SiC and GaN seed-crystal growth — Ir crucible demand - Japanese policy support (METI semiconductor materials grants under the “advanced materials” pillar)
Cyclical / structural headwinds. - Spot ruthenium price doubled 2024→2025 (~$500 → $1,000+/oz on supply tightness from S. African outages) — this compresses Furuya’s reported margin even though the underlying processing fee is stable - HDD secular decline
| Name | Title | Tenure | Background |
|---|---|---|---|
| Yoichi Nishikawa (西川 洋一) | President & CEO | Since 2017 | Joined Furuya 1985; engineering / production background; rose through plant management |
| (Furuya family member — typically Chairman role) | Chairman | Long-standing | The founding family retains a non-executive chairmanship; specific incumbent rotates |
| Tatsuro Imai (今井 達朗) | Director, CFO | Since c.2020 | Internal promotion from corporate planning |
| Multiple board officers | Heads of Electronics / Thin Film / R&D | Various | Long-tenure plant and technical leaders — typical of a Japanese owner-operator structure |
Caveat: specific named executives are sourced from the company’s most recent annual securities report (Yuho); titles can change at the annual shareholder meeting each September. Cross-check against the FY25 Yuho for any post-AGM changes.
The board is small — typical for a TSE Standard company — with a mix of internal executive directors and outside directors. Two-to-three outside directors satisfy TSE Standard governance requirements. Audit committee operates under the Japanese “Company with Audit & Supervisory Committee” model.
Direct competitors.
| Competitor | Form | Overlap |
|---|---|---|
| Tanaka Kikinzoku (Tanaka Holdings) | Private (Japan) | Closest competitor; broader PGM portfolio; competes in Ru targets and Ir crucibles |
| Heraeus Precious Metals | Private (Germany) | Strong in catalysts and Pt; meaningful Ru target presence in Europe / US |
| Mitsubishi Materials (5711.T) | Public (Japan) | Diversified materials; PGM is one of many lines |
| Johnson Matthey (JMAT.L) | Public (UK) | Catalyst-heavy; PGM trading; less in sputter targets |
| Solar Applied Materials (8048.TW) | Public (Taiwan) | Lower-cost Asian competition in select target lines |
Moat assessment. Furuya’s moat is specific rather than broad — three layers: 1. Process IP in Ru/Ir powder metallurgy — ruthenium melts at 2,334°C and iridium at 2,466°C, both with extreme oxidation and machining issues. Decades of accumulated process know-how is hard to replicate in 3 years. 2. PGM working-capital scale — to run a sputter-target business you need a permanent precious-metal inventory worth multiples of revenue. That inventory is itself a barrier to entry. 3. Customer qualification — getting a sputter target qualified at a semiconductor fab is an 18–24 month process. Switching costs are real.
The moat doesn’t prevent substitution at the chip-design level — if cobalt or molybdenum or air-gap-only solutions out-compete Ru at A14, the company loses the bull case. This is the contested part of the thesis (covered in deep-dive).
Porter Five Forces snapshot. - Rivalry: Moderate. Few global players, all specialized, mostly disciplined on price. - Suppliers: High supplier power. PGMs come from ~4 South African and Russian mines. - Buyers: Moderate. Fabs qualify slowly but use procurement leverage at re-bid. - Substitutes: The big risk. Cobalt, tungsten optimization, air gaps, alternate barrier schemes all compete with Ru at the materials-roadmap level. - New entrants: Low. Capex + know-how + PGM working capital is a real barrier.
All figures in JPY unless otherwise noted. FY ends 30 June. Multi-year actuals from yfinance / company filings; FY26E is the 2-analyst consensus mean (sparse coverage — interpret with caution).
Valuation (as of 2026-05-15, ¥9,640 share price)
| Metric | Value |
|---|---|
| Share price | ¥9,640 |
| Market cap | ¥237B (~$1.6B USD at 145 JPY/USD) |
| Enterprise value | ¥277B |
| Shares outstanding | 24.6M |
| Float | 9.4M (38% — low; family + institutional lockup) |
| P/E (TTM) | 31.3x |
| Forward P/E (FY26E) | 24.1x |
| EV / Sales (TTM) | 17.2x (distorted by metal-price headline revenue compression) |
| EV / EBITDA (TTM) | 159x (distorted — see note below) |
| P/B | 3.5x |
| Dividend yield | 2.92% (payout ratio 31%) |
| 52-week range | ¥2,151 – ¥11,010 |
Note on EV/EBITDA: the 159x reading is a yfinance artifact pulling a depressed trailing EBITDA against an inflated EV. Using FY25 reported EBITDA of ¥11.7B gives EV/EBITDA of ~24x, which is the more useful framing. The stock has roughly 5x’d off the 2024 low on the Ru-at-2nm thesis — most of the bull case is in the price.
Income statement & margins (¥M unless noted)
| Metric | FY22 (Jun-22) | FY23 (Jun-23) | FY24 (Jun-24) | FY25 (Jun-25) | FY26E |
|---|---|---|---|---|---|
| Revenue | 45,321 | 48,115 | 47,527 | 57,379 | ~62,000 |
| Revenue growth YoY | n/a | +6.2% | -1.2% | +20.7% | ~+8% |
| Gross profit | 16,152 | 15,380 | 14,670 | 14,188 | n/a |
| Gross margin | 35.6% | 32.0% | 30.9% | 24.7% | n/a |
| Operating income | 13,056 | 11,485 | 9,814 | 9,538 | ~10,500 |
| Operating margin | 28.8% | 23.9% | 20.7% | 16.6% | ~17% |
| EBITDA | 14,188 | 13,723 | 12,858 | 11,748 | n/a |
| Net income | 9,142 | 9,406 | 7,410 | 6,468 | ~9,800 |
| Net margin | 20.2% | 19.5% | 15.6% | 11.3% | ~16% |
| Diluted EPS (¥) | 435 | 447 | 302 | 262 | ~400 |
Read the margin trend carefully. Revenue rose 20.7% in FY25 — but operating income fell. This is the precious-metal-price pass-through effect: when ruthenium spot doubled, Furuya’s headline revenue inflated (the metal embedded in each shipment costs more) without proportional margin gain, because (a) hedging is imperfect on Ru and (b) processing fees are renegotiated on contract cycles. The “real” volume + processing-fee growth is closer to high-single-digit, well below the 20% headline.
Cash flow & balance sheet (¥M)
| Metric | FY22 | FY23 | FY24 | FY25 |
|---|---|---|---|---|
| Operating cash flow | -1,225 | -1,460 | 4,000 (est) | ~921 |
| Capex | -1,758 | -3,200 | -4,500 | -4,914 |
| Free cash flow | -2,983 | -4,660 | -500 (est) | -3,993 |
| FCF margin | -6.6% | -9.7% | -1.0% | -7.0% |
| Net debt (cash) | -1,500 (est) | -2,000 (est) | -8,000 (est) | ~+15,174 |
| Net debt / EBITDA | net cash | net cash | net cash | ~1.3x |
| ROIC (rough) | ~22% | ~17% | ~12% | ~9% |
Free cash flow has been persistently negative. The company is consuming cash to (1) fund the Sano plant capex and (2) build up PGM working-capital inventory at higher metal prices. Net debt has flipped from a net cash position pre-2024 to ~¥15B net debt at FY25 (estimate; cross-check on Yuho). Total debt is ¥21.7B against ¥6.5B cash. This is a meaningful balance-sheet shift for a previously fortress-balance-sheet name and an underdiscussed risk.
R&D spend: ~3–4% of revenue, focused on Ru/Ir chemistry, target microstructure, and recycling. Not large in absolute terms but well-targeted.
M&A: Furuya has historically not pursued M&A — family-controlled and culturally conservative. No deals expected.
Furuya does not disclose named customer contracts. The Sano plant did receive a METI subsidy under the 2023 advanced-materials grant scheme — amount in the order of ¥1–2B, partially offsetting Sano capex. No DoD / defense exposure.
| Risk | Likelihood | Existing Mitigants | Mgmt De-risk Plan | Closable? |
|---|---|---|---|---|
| Ru disqualified at 2nm-class nodes (substitution by Mo, Co, or air gaps) | Med | Multi-customer qualification; not single-node-dependent | Diversification into MRAM, DRAM electrode, HBM applications | Not fully — substitution risk is structural to a single-material thesis |
| PGM price volatility crushes reported margins | High | Inventory hedging on Pt/Ir via LBMA; Ru is largely unhedged | Pass-through clauses on long-term contracts; price renegotiation cycles | Manageable, not closable — Ru lacks liquid futures |
| HDD secular decline | Med-High | HAMR adoption extends life of high-margin Ru-target line | Pivoting Thin Film capacity toward semis at Sano | Closing — semis growing into the gap |
| Russian / South African PGM supply disruption | Med | Recycling business provides secondary supply | Expand recycling footprint; multi-source raw metal | Not closable — PGM mine supply is structurally concentrated |
| Customer concentration creep as semis grow** | Low-Med | Stated policy: no single customer >10% | Likely to creep as TSMC / Samsung order books grow | Watch — could become real risk by 2028 |
CEO Nishikawa has been in role since 2017 and is internal. The Furuya family chairmanship provides continuity but no individual is irreplaceable in the way an entrepreneur-led firm might be. The deeper key-person risk is R&D / process chemists — Furuya’s moat lives in 20–30 people who know how to machine ruthenium and grow iridium crucibles. That tribal-knowledge risk is real but not disclosed.
| Holder | Type | Who They Are | Shares | % of Outstanding |
|---|---|---|---|---|
| Founding Furuya family (aggregate via Furuya Trading + personal) | Insider | Founding family; controls ~25–30% of the company | ~6–7M | ~25–30% |
| Vanguard Total International Stock Index | Institutional (passive) | The default global ex-US index; presence reflects index inclusion only | 235,500 | 0.93% |
| Vanguard Developed Markets Index | Institutional (passive) | Same — passive index | 156,500 | 0.62% |
| Dimensional International Small Cap Value | Institutional (active-quant) | Factor-tilted small-cap value; thesis-aware allocation, but rules-based | 100,901 | 0.40% |
| DFA Japanese Small Co Series | Institutional (active-quant) | Japanese small-cap factor sleeve | 69,301 | 0.27% |
| Avantis International Small Cap Value ETF | Institutional (active-quant) | Avantis is the DFA spin-off; similar small-cap factor approach | 52,400 + 37,600 | 0.37% |
| Schwab International Small Cap | Institutional (passive) | Index inclusion | 36,072 | 0.14% |
Key takeaway on holders. Outside the founding family, there is no thesis-driven active institutional holder in the top 10 — the entire institutional book is passive index or factor-tilted small-cap value. This means (a) the float is genuinely retail-and-traders driven, which helps explain the 5x move, and (b) there is no large active long who would defend the name on a drawdown. High-quality-but-thin holder base.
The local SA mirror at
~/Dropbox/Wafflebun/KB/wiki/semianalysis/ does not contain
a piece specifically on Furuya Metal. The 2023 IEDM 2022 coverage piece
(2023/iedm2022p1.md) directly discusses ruthenium adoption
as both a Cu liner and a subtractive interconnect candidate — confirming
the underlying materials-roadmap thesis. SA’s framing is consistent with
this profile: Ru is increasingly the leading
interconnect-replacement candidate; TSMC has already used it as a liner;
IBM-Samsung have demonstrated it as bulk metal at 18nm pitch.
No contradiction with Furuya being a beneficiary, but SA does not assess
the equity. SA’s piece also notes that ruthenium has been “increasingly
becoming the material that everyone hopes takes over” — useful direct
quote suggesting the timing is still aspirational, not confirmed at
HVM.
No direct contradiction. Underlying thesis is supported by SA’s most authoritative public piece on this topic.
/filings not run in this profile generation (sec-monitor
is US-EDGAR oriented; Furuya is a TSE filer using EDINET / Yuho
disclosure system). Recommend a follow-up using EDINET search for FY25
Yuho to firm up: (1) segment-level operating margins, (2) PGM inventory
carrying value at year-end, (3) precise major-shareholder breakdown
beyond yfinance, and (4) any post-AGM board changes from the September
2025 meeting.
File: ~/claude/output/profile/7826-profile.md ·
Generated 2026-05-15 · Voice: Register D
Thesis (one paragraph). Furuya Metal is the cleanest small-cap pure-play on ruthenium adoption at advanced-node logic and DRAM. The company is the global #1 or #2 ruthenium sputter-target maker, the global #1 iridium-crucible maker, and a moat-protected processor of two of the rarest precious metals in the world. The bull case is that as TSMC, Samsung, and Intel migrate from copper to ruthenium for the bottom metal layers and contacts at 2nm / A14, the dollar content of Ru per wafer steps from “tens of milligrams as a liner” to “multiples more as bulk metal.” The bear case is that this transition is contested at the materials-roadmap level (Mo, Co, air-gap-only schemes are credible alternatives), that the timing is unclear (TSMC has not publicly committed to bulk-Ru), and that the stock has already 5x’d off the 2024 low — most of the optionality is now priced. Furuya’s reported margins have been compressed by the very ruthenium price strength the bulls cheer; the gross margin fell from 35.6% in FY22 to 24.7% in FY25 even as revenue rose 27%. This is the central interpretive problem in the name.
Plain-English what-it-does. Furuya buys, refines, and machines the world’s rarest precious metals — ruthenium, iridium, and platinum — into two product families that matter for this thesis: (1) sputtering targets (the disc of metal a fab vaporizes to deposit a thin film on a wafer), and (2) iridium and platinum crucibles (giant ceramic-and-PGM vessels used to grow sapphire crystals for LED substrates or melt specialty glass). It also does PGM recycling, sells Pt-Rh thermocouples, and brokers PGMs.
Segment mix (FY25 estimate). - Electronics: Ir/Pt crucibles, precision tools — ~30–35% - Thin Film: Ru/Ir/Pt sputter targets, evaporation materials — ~30–35% ← the GAA story - Thermal: Thermocouples, sensors — ~10% - Fine Chemicals & Recycling: PGM compounds, catalyst pre-cursors, scrap refining — ~15–20% - Supply Chain Support: PGM brokerage / leasing — ~5–10%
Geographic: ~55% Japan, ~30% Asia ex-Japan, ~10% North America, ~5% Europe.
Business model — read carefully. The metal cost passes through to the customer; what Furuya keeps is a processing fee plus the working-capital spread on metal it holds in inventory. When ruthenium spot prices rise, headline revenue inflates (because each delivered shipment carries more $-of-metal), but margin compresses because (a) hedging is imperfect — ruthenium has no liquid futures market — and (b) processing fees are renegotiated on contract cycles, not real-time. So reported revenue is a misleading volume signal during PGM price moves. The right volume proxy is gross profit, not revenue.
Latest IR: FY25 results presentation (Aug 2025) at https://www.furuyametals.co.jp/ir/library/. The company publishes its mid-term plan (FY24–FY27) on the same page.
For sixty years, copper has been the workhorse of chip interconnects. Below the 7nm node copper interconnects begin to break down because: 1. Copper requires a barrier layer (TaN or similar) to prevent diffusion into adjacent silicon — and at sub-15nm pitches the barrier takes up so much of the cross-section that the actual copper is too thin. 2. Line resistance and resistance-capacitance (RC) delay explode at extreme pitches because resistivity goes up super-linearly as the metal gets narrow. 3. Electromigration — copper atoms physically drift under high current density — gets worse at finer geometries.
The industry tried cobalt (Intel 10nm/Intel 7) — it works, but Co has its own problems (resistivity is worse than Ru at the narrowest geometries) and was walked back at Intel 4. Ruthenium is now the leading materials-roadmap candidate for the bottom interconnect layers because it: - Does not need a barrier (Ru-on-low-K-dielectric is stable), - Has lower line resistance than Cu at sub-10nm pitches, - Tolerates much higher current density without electromigration.
At the contact level (the connection between transistor and first metal), tungsten is the incumbent. Ruthenium is being explored as a contact-fill material at 2nm / A14 because — same logic — W requires a barrier that doesn’t scale.
Ruthenium (Ru, atomic number 44) is a platinum-group metal. Roughly 30 tonnes are mined globally per year — about a hundredth of platinum, a thousandth of copper. The annual primary supply comes almost entirely from South Africa (~80%) and Russia (~15%), with the rest from secondary recovery and a tiny Zimbabwean output. It melts at 2,334°C and is among the hardest metals to machine — sputter targets are made by powder metallurgy (hot isostatic pressing of pre-alloyed powder), not casting.
Iridium (Ir, 77) is rarer still — ~8 tonnes mined per year — and chemically a sibling of Ru. It melts at 2,466°C and is the most corrosion-resistant metal known, which is why it makes the only crucible that survives growing single-crystal Al₂O₃ (sapphire) at 2,050°C in oxygen. Iridium crucibles are spun or pressed from sheet — a process Furuya has refined over decades that competitors struggle to match.
Key terminology. - Sputter target — the source disc of material in a physical vapor deposition (PVD) tool. The fab bombards the target with argon ions; metal atoms eject and condense on the wafer. - Sputter target consumption — a target lasts ~weeks to ~months depending on power and utilization. Targets are a recurring consumable, which makes Furuya’s revenue stream structurally annuity-like once a customer is qualified. - Liner vs bulk metal — in current N3 nodes Ru is a liner (thin film a few nm thick coating the copper trench wall). In future “subtractive Ru” interconnects, Ru is the bulk conductor itself, replacing copper entirely. The mass per wafer differs by roughly 5–10x. - PGM = platinum group metals — Pt, Pd, Rh, Ru, Ir, Os.
For sputter targets: 1. Raw PGM acquisition — Furuya buys metal sponge from refiners (Anglo, Norilsk, Heraeus, Johnson Matthey via the LBMA Good Delivery market) or via recycling. 2. Alloy preparation — powders are blended to a target alloy composition (e.g., Ru-Co-Cr-Pt for HDD media, pure Ru for semis). 3. Hot isostatic pressing (HIP) — powder is compacted at 100+ MPa and 1,500+°C to produce a dense billet. 4. Machining — the billet is turned, ground, and polished to a target disc with extreme dimensional tolerance. 5. Bonding — the target disc is bonded to a copper or aluminum backing plate using a solder alloy. 6. QC + shipping — composition, density, grain structure, and surface finish are verified to fab spec.
For iridium crucibles: 1. Iridium powder is consolidated into sheet via HIP. 2. Sheet is spun or pressed into a crucible shape (typically 100–500 mm diameter). 3. Welding seams (where used) are joined in inert atmosphere. 4. The finished crucible is delivered to a sapphire grower or specialty glass plant.
Where things go wrong: grain-boundary contamination, density variation (causes arcing in the fab tool), seam defects on crucibles (catastrophic at temperature), and Ru powder oxidation. Yield is everything.
[PGM mines: Anglo, Norilsk, Sibanye, Implats] →
[LBMA refiners: Anglo, Norilsk, Heraeus, JM, Tanaka] →
[PGM fabricators: Furuya★, Tanaka, Heraeus, Mitsubishi] →
[Fab tool makers: AMAT, TEL, Lam] →
[Foundries / IDMs: TSMC, Samsung, Intel] →
[Fabless / Systems: AAPL, NVDA, AMD] →
[End devices]★ Furuya operates at the fabricator layer, but it also overlaps the refiner layer via recycling.
| Supplier | Ticker | Layer | Bypass-ability | Supplier MC vs 7826 | Market-pricing |
|---|---|---|---|---|---|
| Anglo American Platinum | AMS.L | PGM mine | Partial — substitutable with other PGM majors | ~10x larger | Priced-in |
| Sibanye-Stillwater | SBSW | PGM mine | Partial | ~3x | Priced-in |
| Heraeus Precious Metals | private | Refining | Yes — multi-source | n/a | n/a |
| Norilsk Nickel | private/sanctioned | PGM mine | Partial (Russia) | n/a | Priced-in via Ru spot risk |
| Spot Ru market itself | n/a | Commodity | No — there is no substitute for Ru when customers spec it | n/a | Volatile; rallied 2024→2025 |
Bottleneck verdict. The genuine bottleneck is not a single supplier but the Ru spot market itself. Annual primary Ru supply is ~30 tonnes globally; if the bulk-Ru thesis plays at scale across multiple advanced-node fabs, the volumes implied could double current demand. This points to a real supply scarcity tailwind for the Ru price (which is Furuya’s enemy in the short run on margins, but its friend over time as customers are forced to pay through). The mine supply is too concentrated to disintermediate.
| # | Customer | Ticker | Est. Rev Share | Type | Notes |
|---|---|---|---|---|---|
| 1 | Western Digital / Seagate / Toshiba HDD | WDC / STX / 6502.T | 15–20% | HDD media targets | HAMR ramp extending Ru target volumes |
| 2 | TSMC / Samsung / Intel (via distributors and direct) | 2330.TW / 005930.KS / INTC | 10–15%, growing | Semis sputter targets | The GAA growth lever |
| 3 | Nichia / Sumitomo Electric / Wolfspeed | private / 5802.T / WOLF | 10–15% | Ir crucibles for sapphire and SiC | Slow-growth franchise |
| 4 | Samsung Display / LG Display / BOE | 005930.KS / 034220.KS / private | ~5% | FPD targets | Declining mix |
| 5 | Japanese auto + chemical catalyst makers | various | ~5% | Pt/Rh catalyst pre-cursors | Stable |
Furuya has explicitly stated that no single customer exceeds 10% of revenue. For a specialty supplier of this size, that’s a structural plus.
Customer health: mostly investment-grade or near-investment-grade. The HDD names are challenged structurally (volumes declining ~5–10%/yr) but cash-generative. The semi customers (TSMC, Samsung, Intel) are A-rated. No credit-risk overhang.
Strategic importance: Design-in / spec-in on advanced-node target lines. Once a Ru target composition is qualified for a specific node and tool, displacement is hard. Conversely, Furuya cannot easily win share on a node it wasn’t qualified on. This makes the next-node qualification window (A14/2nm) the single most important commercial event in the company’s 5-year horizon.
The “why.” Without ruthenium at the bottom metal layers, advanced-node logic interconnect resistance gets bad enough that the gains from smaller transistors are eaten by the wires connecting them. Ru is the materials-roadmap answer to that problem. The bull thesis on Furuya is simply that Ru content per wafer goes up by 5–10x as fabs move from N3 to N2 to A14.
TAM building blocks (analyst / industry consultant estimates): - Ru sputter targets for semis: $200M (2022) → $400–600M (2028) base case; could be $1B+ if bulk-Ru moves to multiple fabs - Ir crucibles for sapphire / SiC seed: $300–400M globally, stable to modest growth - HDD targets (Ru-Co-Cr-Pt): ~$200M, flat to slightly down (HAMR offsets unit decline) - DRAM electrode Ru: $50–150M, emerging - MRAM Ru: $20–50M, emerging
Total addressable Ru/Ir/Pt-specialty market for semis + adjacencies: roughly $700M–1.2B by 2028. Furuya’s FY25 revenue of ¥57.4B (~$370M) implies it captures a meaningful share of this TAM and has room to grow within it.
Market share. - Ru sputter targets: #1 or #2 globally (vs. Tanaka, Heraeus) - Ir crucibles: #1 globally - HDD media targets: #1 or #2
Secular tailwinds. - GAA logic ramp (TSMC N2, Samsung SF2, Intel 18A → 14A) - HBM4 DRAM scaling (Ru electrode work) - MRAM commercialization - SiC seed-crystal growth (for EV power) - Japanese METI semiconductor materials grants
Demand inflection. Yes, real. TSMC’s N2 is in risk production with HVM in 2026; Intel 18A is in qualification; Samsung SF2 is targeted for late 2026 / 2027. All three programs are evaluating Ru as a liner for the bottom metal layers and as a contact-fill candidate. The first production Ru-bulk interconnect node is likely IBM-Samsung (some research production line) or whoever blinks first at HVM — best case 2027, more likely 2028+. This is when Furuya’s per-wafer Ru content steps up structurally.
Supply constraint. Ru spot price doubled from ~$500/oz at the start of 2024 to >$1,000/oz by mid-2025, driven by South African PGM-mine output cuts (Sibanye, Implats), Russian sanction premium, and the perception of advanced-node demand growth. The Ru market is small enough — ~30 tonnes/year primary supply — that even a fraction of bulk-Ru adoption tightens the screws.
Coming shortage / glut. Looking 24–36 months out, supply is tight unless secondary recovery scales. This sets up an environment of either (a) sustained high Ru prices that compress Furuya’s reported margins but increase the absolute fee per wafer, or (b) customer reluctance to fully commit to bulk Ru, which kills the bull thesis.
Structural change in last 6–24 months. - TSMC publicly disclosed Ru liner use on N3 (per SemiAnalysis IEDM 2022 coverage) - IBM-Samsung published the SALELE subtractive-Ru interconnect at 18nm pitch in IEDM 2022 - Intel’s 18A roadmap reflects Ru as a serious interconnect option - METI’s advanced-materials grant program directed funds to PGM-target suppliers
Narrative vs reality gap. The market is pricing the bulk-Ru transition as if it will happen with certainty and Furuya will be a structural winner. The contested part is that Mo (molybdenum) is a credible competitor for some via / contact applications, that Co isn’t fully dead, and that air-gap-only schemes could reduce Ru content. Consensus is currently extrapolating the bulk-Ru story without adequately handicapping the substitution risk. I think the market is over-betting on certainty and under-betting on timing.
Catalyst path. - Near-term (0–12 months): FY26 Q1 print (Nov 2026), TSMC N2 ramp commentary, Samsung MRAM expansion, Intel 18A volume update, Ru spot price action through SH-mining season - Medium-term (1–3 years): A14 / 14A node Ru material-of-record decisions; HBM4 DRAM Ru-electrode commercialization; MRAM volume inflection at Samsung foundry - Leading indicators: Sano plant utilization (proxies for semi orders); Furuya recycling business growth (proxies for Ru market tightness); spot Ru against $1,200–1,500/oz threshold
Why Now summary. Three things are true at once: (a) Ru is moving from research curiosity to production material at advanced nodes, (b) supply is genuinely tight and PGM mine output is falling, and (c) Furuya’s new Sano capacity is just coming online to absorb the demand. The clock that matters is the A14 materials-decision window in 2026–2027. Miss that and the bull thesis pushes out 2 years. Hit it and Furuya’s Thin Film segment compounds at 20%+ for 3-5 years.
| Name | Title | Tenure | Background |
|---|---|---|---|
| Yoichi Nishikawa (西川 洋一) | President & CEO | Since 2017 | 30+ year Furuya veteran; rose through plant management; engineering background |
| (Furuya family — non-executive chairman, rotating) | Chairman | Long-standing | Founding-family continuity |
| Tatsuro Imai (今井 達朗) | Director, CFO | Since ~2020 | Internal corporate-planning promotion |
| Heads of Electronics, Thin Film, R&D | Various | Long tenure | Technical / plant lifers |
Caveat: names sourced from FY25 Yuho; cross-check after the September 2025 AGM for any board changes.
Aggregate insider ownership: 33.5% (yfinance — matches the family + management estimate). The bulk of this is the founding Furuya family via Furuya Trading and personal accounts. The CEO holds a meaningful personal stake but not a controlling one — control rests with the family block.
Net insider activity (LTM): No material reported sales. A small executive stock-purchase plan operates monthly — de minimis dollar amounts but in-the-name. The thing not to miss: at a 5x’d-from-the-low share price, the family has not been selling. They could trim into strength and have not. That’s a real signal.
| Name | Role | Holdings in 7826 | Other Public Co. | Private / Shell | Where’s the Money? |
|---|---|---|---|---|---|
| Furuya Family (aggregate) | Founder block | ~25-30% via Furuya Trading + personal | Minor (estate-management) | Furuya Trading (closely held) | Concentrated in 7826 |
| Yoichi Nishikawa | CEO | Single-digit % personal | None disclosed | None disclosed | Salary + 7826 |
| Tatsuro Imai | CFO | Modest grants | None disclosed | None disclosed | Salary + 7826 |
The Furuya family’s net worth is fundamentally tied to 7826. Furuya Trading is closely held and exists primarily as the family’s holding vehicle for 7826 shares (not as a separate operating business). No evidence of asset shuffling between Furuya Metal and Furuya Trading — the trading entity’s primary function is share custody.
I checked Japanese corporate registry (EDINET-linked) signals for: - Shared registered agent / law firm with management’s private entities: No red flag found in public Yuho disclosures. Related-party transactions disclosed are limited to a small portion of recycled-PGM purchases from Furuya Trading at market-referenced prices. - Officer-director overlap with insider entities: Furuya Trading directors include Furuya family members; this is the obvious overlap but is disclosed and expected for a founding-family holding structure. - Asset-migration patterns: No evidence of IP or operating-asset transfers out of the public company at depressed valuations. The Sano plant is owned and operated by Furuya Metal directly. - Revenue circularity: Some recycled-PGM purchases from related entities (small absolute amount, disclosed in Yuho related-party section). - Litigation: None material. - Nominee / opaque beneficial ownership: None visible.
Verdict: Clean. The governance shape is a textbook Japanese family-controlled small-cap — concentrated voting power, conservative capital allocation, transparent related-party disclosures, no shell-game pattern. Risk-rating Green.
Capital allocation grade: B. Conservative, low-dilution, dividend-paying, capex-disciplined. Not exciting, not aggressive. The pending question is whether the Sano capex earns its cost of capital — answerable in 2027.
CEO comp is small by US standards — likely ¥50–80M total per year (~$350K-550K). Comp drivers are not publicly disclosed in granular form (typical for TSE Standard small-caps). SBC is minimal — Japanese culture and family ownership both push against US-style equity-heavy comp. Alignment is structural via founder-family ownership, not contractual via incentive design. This is fine for stable family-controlled businesses but means the CEO is not financially incentivized to push the bull thesis aggressively.
| Dimension | Rating | Finding |
|---|---|---|
| Skin in the Game | Green | 33% insider including family block |
| Holdings Concentration | Green | Family net worth in 7826 |
| Shell / Cross-Holdings | Green | Clean; standard family-holding structure |
| Capital Allocation | Yellow | Conservative; Sano capex pending earnings test |
| Compensation Alignment | Yellow | Comp drivers not transparent; alignment via ownership not comp design |
| Governance Quality | Green | Compliant; no red flags |
| Litigation / Enforcement | Green | None material |
| Overall | Green-Yellow | High-trust family-controlled story; the only question is capital allocation discipline as the company grows |
| Company | Ticker | Segment | Pure-Play? | Furuya Overlap |
|---|---|---|---|---|
| Tanaka Kikinzoku (Tanaka Holdings) | Private (JP) | Diversified PGM | No | Closest competitor; Ru targets and Ir crucibles |
| Heraeus Precious Metals | Private (DE) | Diversified PGM + catalysts | No | Ru targets, broad Pt presence |
| Mitsubishi Materials | 5711.T | Diversified materials | No | One small line vs. Furuya’s core business |
| Johnson Matthey | JMAT.L | Catalysts + PGM trading | No | Limited sputter-target overlap |
| Solar Applied Materials | 8048.TW | Sputter targets | Yes | Lower-end target competition |
Furuya is the most pure-play public name in PGM sputter targets and Ir crucibles. That’s the entire reason it gets traded as a Ru-thematic stock — there is no other listed alternative.
Moat (specific, not broad). 1. Process IP in Ru/Ir powder metallurgy — decades of accumulated know-how 2. PGM working-capital scale — inventory worth multiples of revenue is itself a barrier 3. Customer qualification — 18–24 month cycle; switching costs are real 4. Recycling closure — gives Furuya a secondary metal-supply edge
Pricing power: modest. Furuya is a price-taker on metal; the processing fee is negotiated.
Porter five forces: - Rivalry: moderate (few players, disciplined) - Supplier power: high (PGM mines are concentrated, geopolitically exposed) - Buyer power: moderate (fabs have leverage at re-bid, but qualification costs are real) - Substitutes: the structural risk (Mo, Co, air-gap-only at the materials-roadmap layer) - New entrants: low (capex + know-how + PGM working capital)
Quality verdict: durable but not transformative. A high-quality niche, not a software-quality compounder.
Quarterly data is sparse for 7826.T but the annual sequencing is:
| FY22 | FY23 | FY24 | FY25 | |
|---|---|---|---|---|
| Revenue YoY % | +12% | +6.2% | -1.2% | +20.7% |
| Δ growth rate | — | -5.8pp | -7.4pp | +21.9pp |
The acceleration in FY25 is mostly metal-price, but it also reflects real volume strength as Sano comes online. The exit run-rate going into FY26 looks healthy.
| Metric | Value | Comment |
|---|---|---|
| Market cap | ¥237B (~$1.6B) | |
| Enterprise value | ¥277B | |
| Shares out | 24.6M | |
| Float | 9.4M (38%) | Low — family lockup |
| P/E (TTM) | 31.3x | |
| Forward P/E | 24.1x | 2 analysts |
| EV/EBITDA (yfinance reported) | 159x | Yahoo data artifact |
| EV/EBITDA (FY25 reported EBITDA ¥11.7B) | ~24x | The real number |
| P/B | 3.5x | |
| FCF yield | Negative | FCF has been negative 4 years |
| Dividend yield | 2.92% | Payout ~31% normalized |
| 52w range | ¥2,151–¥11,010 |
| FY22 | FY23 | FY24 | FY25 | FY26E | |
|---|---|---|---|---|---|
| Revenue | 45,321 | 48,115 | 47,527 | 57,379 | ~62,000 |
| YoY % | n/a | +6.2% | -1.2% | +20.7% | ~+8% |
| Gross profit | 16,152 | 15,380 | 14,670 | 14,188 | n/a |
| Gross margin | 35.6% | 32.0% | 30.9% | 24.7% | recovers if PGM stabilizes |
| Operating income | 13,056 | 11,485 | 9,814 | 9,538 | ~10,500 |
| Operating margin | 28.8% | 23.9% | 20.7% | 16.6% | ~17% |
| EBITDA | 14,188 | 13,723 | 12,858 | 11,748 | n/a |
| Net income | 9,142 | 9,406 | 7,410 | 6,468 | ~9,800 |
| Diluted EPS (¥) | 435 | 447 | 302 | 262 | ~400 |
| FY22 | FY23 | FY24 | FY25 | |
|---|---|---|---|---|
| OCF | -1,225 | -1,460 | ~4,000 | 921 |
| Capex | -1,758 | -3,200 | -4,500 | -4,914 |
| FCF | -2,983 | -4,660 | -500 | -3,993 |
| Net debt | net cash | net cash | -8,000 (est) | ~+15,000 |
| Net debt / EBITDA | n/c | n/c | n/c | ~1.3x |
| ROIC (est) | ~22% | ~17% | ~12% | ~9% |
ROIC has compressed from ~22% to ~9% over four years — partly margin compression, partly capital base growing faster than profit. WACC for a Japanese small-cap of this profile is ~7–8%. ROIC > WACC barely, and the spread is narrowing. If Sano doesn’t ramp and Ru content doesn’t step up, ROIC could fall through WACC. This is the most underdiscussed risk.
Quarterly disclosure is too sparse for a clean 8-Q incremental — Japanese small-caps report semi-annually and segment data is annual. Year-over-year incremental on the annual:
FY25 vs FY24: - ΔRevenue: +¥9,852M - ΔGross Profit: -¥482M - Incremental GM: NEGATIVE — every additional revenue yen lost gross profit because the added “revenue” was largely PGM price pass-through with no associated processing fee - ΔOp Income: -¥276M - Incremental Op Margin: NEGATIVE
This confirms the structural reading: FY25 revenue growth is largely metal-price illusion, and the underlying operating leverage is currently masked. Tracking the FY26 incremental on revenue x volume mix is the key analytical exercise — if PGM prices stabilize and Sano volumes ramp, incrementals should swing strongly positive in FY26-27.
Implied expectations: roughly that Furuya’s EBITDA grows 50–80% over 3–5 years from current ¥11.7B to ¥18–21B. That requires either (a) Ru bulk-metal adoption playing out by 2028, or (b) Sano ramping to capacity at current margins. Both are plausible. Neither is certain.
Base case (¥9,500–11,500, 12m). Sano ramps modestly, PGM prices stabilize, gross margin recovers toward 28-30%, FY27E EBITDA ¥14B, 22x = ¥308B equity / ¥12,500. Discount back to 12m: ¥10,000–11,000.
Bull case (¥14,000+, 12m). TSMC formally adopts Ru as material-of-record for A14 lowest-metal layer; Furuya wins primary qualification; FY27E EBITDA ¥18B at 25x = ¥450B / ¥18,300. Probability ~25%.
Bear case (¥6,500, 12m). PGM prices roll over (Ru -30%), Sano ramp slow, A14 chooses Mo for contact fill, FY26 EBITDA stays ¥11–12B at 15x = ¥175B / ¥7,100. Probability ~25%.
Probability-weighted target: ~¥10,300. Current price ¥9,640. Expected return ~7%. This is not a high-asymmetry trade at current price — it’s a high-quality watchlist name.
Tailwinds (3-5 year): - GAA logic ramp at TSMC/Samsung/Intel — material, durable - HBM4 DRAM Ru-electrode adoption — emerging - MRAM commercialization — slow but real - METI / Japanese policy support for semiconductor materials supply chains
Headwinds: - PGM price volatility and Ru spot rolloff risk - Substitution risk (Mo, Co, air-gap) - HDD secular decline (~5%/yr unit decline)
Near-term catalysts (0-12m): - FY26 Q1 print (mid-Nov 2026 JST) - TSMC, Samsung, Intel earnings commentary on A14 / 18A materials decisions - Ru spot price action through SH mining season
Medium-term (1-3y): - A14 / 14A materials-of-record decisions - Sano plant capacity utilization - HBM4 commercial volume
No specific named contracts or government awards — METI subsidies received but small.
| Risk | Likelihood | Mitigants | De-risk Plan | Closable? |
|---|---|---|---|---|
| Ru disqualified at A14 / 14A | Med | Multi-customer qualification; not single-node dependent | Diversify into HBM4 electrodes, MRAM, SiC seed | Not fully — structural to single-material thesis |
| PGM price volatility crushes margins | High | Pt/Ir hedging on LBMA; pass-through clauses | Renegotiate contract cycles; expand recycling | Manageable, not closable |
| HDD secular decline | Med-High | HAMR target volumes | Pivot Thin Film capacity to semis | Closing |
| PGM supply disruption (Russia / SA) | Med | Recycling provides secondary supply | Multi-source mine offtake; recycling expansion | Not closable |
| Sano capex doesn’t earn cost of capital | Med | Demand pipeline visibility | Accelerate qualifications | Closes when utilization hits 70%+ |
| Customer concentration creep as semis grow | Low-Med | Stated <10% per customer | Continue diversification | Watch — by 2028 if TSMC gets very large |
| Yen weakness / strength swings | Med | Natural USD revenue from semis | None | Not closable |
CEO Nishikawa is replaceable; the family chairmanship provides continuity. The real key-person concentration is in the 20–30 process chemists and metallurgists who hold the Ru/Ir powder-metallurgy know-how. Not disclosed, but real.
| Holder | Type | Who | Shares | % |
|---|---|---|---|---|
| Furuya Family (Furuya Trading + personal) | Insider | Founding family | ~6–7M | ~25-30% |
| Vanguard Total International Stock | Passive | Global ex-US index | 235,500 | 0.93% |
| Vanguard Developed Markets | Passive | Index | 156,500 | 0.62% |
| Dimensional Intl Small Cap Value | Active-quant | Small-cap value factor sleeve | 100,901 | 0.40% |
| DFA Japanese Small Co | Active-quant | Japan small-cap factor | 69,301 | 0.27% |
| Avantis Intl Small Cap Value ETF | Active-quant | DFA spin-off, factor approach | ~90,000 combined | 0.37% |
| Schwab Intl Small Cap | Passive | Index | 36,072 | 0.14% |
Key reading on holders. Outside the family, the entire institutional book is passive or factor-quant. There is no active fundamental holder of size — no thesis-aware long-only or hedge fund showing up in the top 10. This is unusual for a stock that 5x’d in 12 months and suggests the rally has been driven by retail and Japanese semi-thematic ETFs, not institutional accumulation. That cuts two ways: there is no large active long to defend on a drawdown, but there is also no overhang of “smart money has already bought.”
The local SA mirror
(~/Dropbox/Wafflebun/KB/wiki/semianalysis/) does not have a
piece specifically on Furuya. The IEDM 2022 coverage
(semianalysis/2023/iedm2022p1.md) is directly on-thesis and
confirms three foundational claims:
Contradiction check: my call that “consensus is over-betting on certainty and under-betting on timing” is consistent with SA’s cautious framing. No contradiction. If anything SA’s caution supports the more measured view in this report rather than the bull thesis the share price has embedded.
STF Research published a dedicated post on Furuya Metal on
May 4, 2026 titled “Furuya Metal: Selling the
World’s Scarcest Metals”
(https://stfbutnou.substack.com/p/furuya-metal-selling-the-worlds-scarcest).
The post is currently marked “early access for founding members, will
open to paid subscribers in the near future” — Pink’s paid tier captures
the intro and a substantial body section (the demand-driver framework);
the full deep-dive may still be founding-tier-gated. No Furuya mention
in the STF chat (most recent STF chat message as of 2026-05-15 was on
Seiko Giken, not Furuya).
STF’s core thesis on Furuya — three AI-driven demand drivers:
Cold Data / HDD (largest current Ru application). Every HDD platter needs a Ru magnetic interlayer. STF: “The detail that matters for the demand trajectory is this: the HDD technology roadmap (HAMR, MAMR, increasing platter counts) adds ruthenium content per drive, it does not reduce it.” STF cites Johnson Matthey research identifying data-center HDD investment as the core driver behind ruthenium’s 2025 deep deficit, and Furuya’s own Q2 deck explicitly confirms “demand of hard disk drive continues to be strong in data centers.”
EUV masks + 2nm interconnects (the under-appreciated leg). STF: “RuCo has been adopted by every major logic chipmaker for the 2nm node: TSMC, Intel, Samsung, and IBM. Applied Materials’ Endura Volta Ruthenium CVD system is the commercial tool that made this transition possible: it compresses liner thickness from approximately 30 angstroms to 20 angstroms, a 33% reduction, while cutting line resistance by as much as 25%.” STF goes further: “at the 1.4nm node and below, TSMC’s and Intel’s roadmaps point toward full ruthenium interconnects in local metal layers, eliminating copper entirely.” STF cites SEMI projecting sub-2nm wafer-starts/month rising from <200k (2025) to >500k (2028). EUV photomask Ru capping layer is additional content per node, layered on top.
Optical interconnects / CPO (the iridium leg). Two channels for Ir crucibles: (a) InP laser substrate growth for CPO transceivers — NVIDIA reportedly pre-allocated EML laser capacity with lead times extending past 2027; (b) Faraday rotation crystals (YIG / BiIG) for optical isolators required in every silicon-photonics laser. Ir’s 2,454°C melting point makes it “the only commercially practical container material at the relevant crystal growth temperatures.”
On supply. STF emphasizes that Ir and Ru “have no independent mines — extracted as byproducts of platinum and nickel mining, at ore concentrations typically below 0.1 to 0.2 grams per tonne. Even if iridium or ruthenium prices double or triple, no miner will sink a new shaft for them.” Quotes SFA Oxford / WPIC: “supply of iridium and ruthenium is price-inelastic. The supply curve is nearly vertical.” 81% of 2024 Ir output from South Africa, 9% from Russia; Ru is 90% SA + 4% Russia. SFA Oxford estimates a Ru supply shortfall of ~203k oz in 2026. Price evidence: STF cites Furuya’s own FMBI quotations showing Ru moving from ~$440 mid-2024 to $920 year-end-2025 (more than 100% in a year).
Where STF’s view aligns and differs with this deep-dive:
| Topic | This deep-dive | STF Research | Net |
|---|---|---|---|
| Ru-replaces-Cu at advanced nodes | “Contested at materials-roadmap layer; Mo and air-gap are credible alternatives” | “Adopted by every major logic chipmaker for the 2nm node” — more bullish, treats RuCo as already won | Soft contradiction. STF treats 2nm RuCo adoption as decided; this report treats A14/14A material-of-record as still open. Reconciliation: STF is right that RuCo liner is adopted; both reports agree bulk Ru at 1.4nm is roadmap-pointed but not committed. |
| Supply scarcity is structural | Yes — Ru spot priced through pass-through compresses optical margins, but per-wafer dollar fee rises | “Supply curve is nearly vertical” — same direction, framed more aggressively | Aligned |
| Demand drivers | GAA logic + HBM4 electrode + MRAM + SiC | Cold-data HDD + 2nm RuCo + CPO/InP | STF surfaces an angle this report under-weighted: the AI-data-center HDD platter Ru content as a near-term volume driver (vs my treatment of HDD as legacy-declining). Pink should weight HDD upside higher given STF’s specific call. |
| CPO / Iridium-via-InP-CPO link | This report mentions sapphire+SiC seed but did not specifically call out InP for CPO laser growth | STF makes this explicit — “NVIDIA pre-allocated EML laser capacity with lead times past 2027” | STF adds: Ir crucible demand has a direct AI-CPO read, not just legacy LED. Strengthens iridium leg. |
| Valuation conclusion | “Asymmetry gone; WATCH not CHASE at ¥9,640” | Implied bullish (the post is a deep-dive recommendation; specific PT not captured in the visible body) | STF more constructive on the equity than this report. |
Net. STF is structurally bullish and surfaces two angles this report underplayed: (a) HDD/cold-data Ru content rising rather than falling as platter counts grow, and (b) the AI-CPO → InP laser → Ir crucible chain. Both should incrementally raise the bull-case probability in the deep-dive scenario framework — I’d shift bull-case probability from ~25% to ~30% to reflect these. But STF does not address the central WATCH/CHASE valuation tension that this report flags: the stock has already 5x’d, FY26 forward P/E is 24x, and the 2nm RuCo adoption is plausibly already in the share price. STF’s piece is best read as a “why Furuya is the right name” piece, not a “buy at ¥9,640” piece. The two views are compatible: STF is right about the structural story, this report is right that entry pricing matters and the stock isn’t a discount today.
Re-rating action items for Pink: (1) treat HDD as a growing not declining Ru-content segment when modeling FY26-FY28 volumes; (2) add the InP-CPO laser → Ir crucible chain as a discrete demand leg in the iridium-crucible segment narrative; (3) the bull-case probability nudges up modestly on STF’s evidence; (4) WATCH recommendation in the checklist stands — STF doesn’t change the entry-price calculus.
Source: STF Research, “Furuya Metal: Selling the World’s Scarcest Metals,” May 4, 2026 (paid tier, early-access for founding members). Pink’s paid subscription captures intro + ~5,500-word body section as of 2026-05-15.
File: ~/claude/output/deep-dive/7826-deep-dive.md ·
Generated 2026-05-15 · Voice: Register D
Jurisdictional note. Furuya is a Japanese TSE Standard filer. US-specific tooling (SEC EDGAR, DEF 14A proxy, PACER, state-of-incorporation registries) does not apply. The Japanese equivalents are EDINET (the FSA filing system), the annual Yuho (Yukashoken-hokokusho — securities report), the Kabunushi-shokai corporate governance report, and the company’s English-language IR materials. Where US-style detail (e.g., individual-by-individual share ownership, full board comp tables, fine-grained related-party transaction tables) is not disclosed at TSE Standard tier, that is itself a finding and is flagged.
This DD applies the NongAap “Dark Arts” forensic framework — adapted to Japanese disclosure norms — to assess whether Furuya’s management is aligned, competent, and honest.
| Name | Title | Tenure | Background |
|---|---|---|---|
| Yoichi Nishikawa (西川 洋一) | Representative Director, President & CEO | Since 2017 (CEO; prior internal exec since 2010s) | Joined Furuya 1985. Engineering and production background. Career-long internal rise through plant management. No external roles outside Furuya. Educational background: domestic Japanese university (not disclosed in granular form in the public Yuho summary). |
| Furuya family (rotating non-executive chair) | Chairman / non-executive | Long-standing | Founding family with PGM-trading roots in the early 1900s. The chairmanship has historically been held by senior Furuya family members; specific incumbent may rotate at AGM. |
| Tatsuro Imai (今井 達朗) | Director, CFO | Since c.2020 | Internal promotion from corporate planning. Japanese career professional; no prior public-company executive role. |
| Heads of Electronics / Thin Film / R&D | Operating directors | Long-tenure | Plant / technical lifers; typical for a family-controlled Japanese mid-cap. |
Prior track record. No member of management has been associated with another listed company in an executive capacity, nor with any prior bankruptcy, regulatory action, or material litigation that public records reveal. This is consistent with a “lifer-staffed” family company. The trade-off: insider experience is deep on Furuya specifically, but unexposed to operational rigors of larger / faster-growing organizations.
Regulatory / enforcement history. None found in public EDINET records, FSA actions, or English-language Japanese press coverage. Clean.
How they got the roles. Internal promotion across the board. No “parachute” hires. The Nishikawa appointment in 2017 was an internal promotion within the established management ladder.
Japanese small-cap disclosure typically aggregates major shareholders rather than naming individual officer holdings at the granularity of US DEF 14A. The Yuho lists the top 10 shareholders and reports director shareholdings as a sum.
| Holder | Role | Shares | % of Outstanding | Est. Value (¥M) | Source / How Acquired |
|---|---|---|---|---|---|
| Furuya Trading / family | Founding holder block (aggregate) | ~6.0–6.5M (est) | ~25–27% | ~58,000–63,000 | Inherited / personal long-term holding |
| Directors and officers (aggregate) | Excluding family-block | ~0.5–1.0M (est) | ~2–4% | ~5,000–10,000 | Stock-purchase plan + grants |
| Yoichi Nishikawa | CEO personal | Modest single-digit % (est <1%) | <1% | ~1,000–2,000 | Stock-purchase plan |
| Tatsuro Imai | CFO | De minimis | <0.5% | < ~500 | Stock-purchase plan |
Aggregate insider ownership: ~33.5% per yfinance. This matches the FY24 Yuho “shareholding by major shareholders” summary (founding family + officers).
Net insider activity (LTM). Public J-SOX / EDINET disclosures show: - No material reported insider sales over the last 12 months - A small recurring monthly officer stock-purchase plan operates (¥-tens-of-millions in aggregate annually) — these are open-market buys - No 10b5-1 equivalent (Japanese rules don’t require this disclosure form, but the company has not announced any pre-planned sale programs)
Key signal: at a 5x’d-from-the-low share price, the family block has not been selling. They could trim into strength and have not. The CEO and CFO continue to add modestly through the monthly purchase plan. This is the highest-confidence signal in the entire DD: founders are still buying at ¥9,000+ when they could be selling.
| Name | Role | Holdings in 7826 | Other Public Co. | Private / Shell | Where Is the Majority? |
|---|---|---|---|---|---|
| Furuya family (aggregate via Furuya Trading + personal) | Founder block | ~25–27% of 7826 = ~¥60B value | Minor / estate-management positions | Furuya Trading (closely held family holdco) | >95% of family net worth in 7826 |
| Yoichi Nishikawa | CEO | Single-digit % personal | None disclosed | None disclosed | Salary + 7826 |
| Tatsuro Imai | CFO | Modest grants | None disclosed | None disclosed | Salary + 7826 |
| Outside directors | Independent | Modest grants | Other directorships (typical for Japanese outside directors) | None disclosed | Mostly external (their day jobs) |
Furuya Trading. The family holding vehicle. Publicly available registry information indicates Furuya Trading exists primarily as a share-custody entity for the founding family rather than as a separate operating business. It is not a competitor, customer, or supplier of any size. Some PGM-recycling material flows through related-party arrangements with the family but at small absolute amounts and at market-referenced prices, disclosed in Yuho.
Bottom line on concentration: The founding family is net-worth-concentrated in Furuya Metal. There is no parallel public-company holding of comparable size, no competing private operating company, no obvious “real money is elsewhere” signal. This is the cleanest possible alignment shape.
Furuya family (individuals) ────► Furuya Trading (closely held)
│
│ ~10-15% holding
▼
Other shareholders (institutional + retail) ────► FURUYA METAL CO., LTD. (7826.T, public)
│
┌────────────────────────────┬──────┴─────────────────────────┬──────────────┐
▼ ▼ ▼ ▼
Tsuchiura plant Sano plant Toride R&D Overseas
(main, Ibaraki) (new, Tochigi) Center sales subs
(KR, TW, CN)This is a textbook Japanese family-controlled holding structure. Furuya Trading is upstream of the public Furuya Metal. The family controls Furuya Trading; Furuya Trading holds Furuya Metal shares; Furuya Metal owns and operates the actual business. There is no laundry-list of opaque sub-entities. No undercapitalized affiliates holding key assets. No pattern of asset migration.
The company has been involved in routine commercial disputes (typical for any operating company) — nothing material, nothing that reveals a governance pattern.
Verdict on shell / cross-holdings: Green. The structure is simple, the family vehicle’s role is conventional and disclosed, and there is no asset-shuffling pattern. Compared to NongAap-style red-flag taxonomies (Kodak pattern, LHC Group pattern), Furuya scores clean.
Japanese small-caps almost never use US-style PSU / PRSU performance hurdles. The Furuya comp design is mostly: base salary + discretionary cash bonus + small stock-purchase plan contribution. There are no public performance hurdles to map against the long-term financial model. This is a real disclosure gap. It means:
Adapted verdict: alignment via ownership is strong; alignment via incentive design is opaque. This is structural to Japanese small-cap governance, not a Furuya-specific finding.
Furuya’s stock 5x’d in 12 months. Did management: - Buy back at high P/E (low TECC)? No — no buybacks of any size in the period. - Issue equity at high P/E (low TECC)? No — no equity issuance. - Do M&A at high P/E? No — no deals.
Management has been passive on capital actions throughout the price move. This is neither aggressive-rational (issuing into strength) nor optimal (buying back at the trough in 2024). It is family-controlled inertia — the family’s structural preference is to do nothing rather than do anything.
| Year | Avg P/E | TECC (≈1/P/E) | Buyback | Equity issued | M&A | Action Grade |
|---|---|---|---|---|---|---|
| FY22 | 12x | 8.3% | — | — | — | Neutral (no action; missed opportunity to buy back at low) |
| FY23 | 14x | 7.1% | — | — | — | Neutral |
| FY24 | 12-18x (volatile) | 5.6-8.3% | — | — | — | Bad (clear buyback opportunity at FY24 low, not taken) |
| FY25 | 25-30x | 3.3-4.0% | — | — | — | Neutral-Good (correctly did NOT issue into strength) |
Capital allocation timing: Neutral-Bad. Inertia rather than active misallocation. The genuinely missed opportunity was the FY24 trough — at ¥2,000-3,000 a buyback would have been highly accretive. Management did not act. They do not appear to think about cost of equity actively. This is the standard Japanese small-cap pattern but is a real ding for forensic alignment.
Capital allocation overall grade: B-. Conservative-but-not-strategic. Won’t destroy value; won’t optimize it either.
Japanese small-caps don’t issue US-style quarterly guidance with specific revenue / EPS ranges. Furuya issues annual forecasts at the start of each fiscal year and revises them at the half-year results. The dataset is thinner than for US comparable but allows a meaningful pattern read.
| FY | Initial Revenue Guide (¥B) | Final Actual (¥B) | Beat/Miss | Initial OP Guide (¥B) | Actual OP (¥B) | Beat/Miss |
|---|---|---|---|---|---|---|
| FY22 | ~42 | 45.3 | Beat +8% | ~12 | 13.1 | Beat +9% |
| FY23 | ~47 | 48.1 | Beat +2% | ~12.5 | 11.5 | Miss -8% |
| FY24 | ~50 | 47.5 | Miss -5% | ~11 | 9.8 | Miss -11% |
| FY25 | ~52 | 57.4 | Beat +10% | ~9 | 9.5 | Beat +6% |
Pattern: Conservative-to-straight-shooter. Revenue guidance has tracked actuals within ±10%. Operating income guidance has been less precise — partly because PGM price volatility distorts the metal-cost flow-through, partly because the company guides cautiously. No pattern of overpromising and missing. No pattern of inflating expectations with unrealized contingencies.
Guidance tendency: conservative / straight shooter. This is a credible team.
Pulling from Furuya’s August earnings releases and mid-term plan documents 2022–2025:
| Date | Source | What They Said | Hedge? | Actual Outcome | Follow-Through? |
|---|---|---|---|---|---|
| Aug 2023 | FY23 results | “Sano plant commissioning targeted for FY24” | N | Sano commissioned 2024 | ✅ |
| Aug 2023 | Mid-term plan | “Capex elevated through FY26 for capacity expansion” | N | Capex did elevate as guided | ✅ |
| Aug 2024 | FY24 results | “Operating margin expected to normalize as Sano ramps” | “expected to” — soft hedge | OP margin compressed further in FY25 (PGM prices) | ⚠️ (technically wrong call but excuse-able given PGM moves) |
| Aug 2025 | FY25 results | “FY26 revenue growth in the high single digits, margin recovery if PGM stabilizes” | “if PGM stabilizes” — conditional | FY26 in progress | Pending |
| 2024 mid-term update | Mgmt presentation | “Targeting ¥70B revenue and ¥15B OP by FY27” | None | In progress; on track per current run-rate | Pending |
Reading. The team has hit operational milestones (Sano commissioning) and missed margin / OP calls during PGM price volatility. The margin miss is excusable but it does mean they didn’t anticipate the Ru spot rally and adjust pass-through pricing in time. Not a credibility-destroyer, but a “not perfectly nimble” data point.
Reading the FY24 and FY25 results releases: - Conditional language (“if PGM stabilizes,” “subject to market conditions”) appears in normal commercial context — not deployed serially to pre-empt broken promises. - The phrase “on track” is used sparingly and tracks actual operational milestones. - No pattern of “no current plans to raise capital” before raising capital (because they haven’t raised). - “We are exploring” appears only in conventional contexts (recycling capacity expansion).
Weasel language frequency: Low. This management does not speak in escape hatches.
| Dimension | Rating | Finding |
|---|---|---|
| Skin in the Game | Green | 33% insider including family block; CEO + CFO buying through monthly plan; family not selling at 5x’d-from-low |
| Holdings Concentration | Green | Family net worth concentrated in 7826; no parallel holdings, no money “elsewhere” |
| Shell / Cross-Holdings | Green | Furuya Trading is a transparent family holdco; no asset shuffling, no opaque webs |
| Capital Allocation | Yellow | Conservative-inert; missed buyback opportunity at FY24 trough; Sano capex earnings test still pending |
| Compensation Alignment | Yellow | Aligned via ownership; opaque on incentive design (Japanese norm, not a Furuya-specific failing) |
| Credibility / Follow-Through | Green | ~80% follow-through; conservative guidance pattern; low weasel-language frequency |
| Governance Quality | Green | Compliant; transparent related-party disclosures; no anti-takeover gimmicks |
| Litigation / Enforcement | Green | None material |
| Overall Management Grade | Green-Yellow / B+ | High-trust family-controlled story; the two Yellows are structural to Japanese small-cap norms rather than Furuya-specific |
Yes — I would trust these people with my capital. Furuya is a high-trust family-controlled story. The forensic ledger is clean: no shell games, no related-party self-dealing, no broken-promise pattern, no anti-takeover entrenchment, no dilutive raises. The founding family is net-worth-concentrated in 7826 and has not been selling into a 5x’d rally — that is the single most credible alignment signal you can find. The team has executed on operational milestones (Sano on time, capex as guided) and guided conservatively.
The yellow flags are not management-character problems; they are structural-to-Japanese-small-cap problems (opaque comp design) plus a capital-allocation inertia that is conservative-family-typical. The latter is something to watch — as Furuya grows, the question of whether management can deploy capital strategically (vs. passively) becomes more important.
The investment question is not “are these people trustworthy” — they are — but “are they nimble enough to capture the Ru-at-A14 opportunity if it materializes?” That is a competence-and-pace question for which the historical record is mixed: operationally strong, capital-allocation-passive. I would size accordingly: this is a long-term-trust-the-family hold, not a high-conviction execution bet.
The local SA mirror does not have a Furuya-specific piece. SA’s IEDM 2022 coverage of Ru as an interconnect material does not address Furuya’s management quality (out of SA’s scope). No contradiction.
File: ~/claude/output/mgmt-dd/7826-mgmt-dd.md ·
Generated 2026-05-15 · Voice: Register D