4078 — Sakai Chemical Industry Co., Ltd.

Thesis

Update 2026-06-02 — the picture moved hard since the 23 May profile. (i) The stock re-rated +80%, ¥3,465 → ¥6,250. The "cheap ahead of the expansion" framing is largely spent: 22.2x forward P/E, 8.6x EV/EBITDA, 1.23x P/B, 2.56% yield (was 12.8x / 5.4x / 0.68x / 4.6%). Re-rate done at the same time the catalyst was being confirmed — the market did the work the no-coverage thesis assumed it wouldn't. (ii) FY3/26 results (13 May 2026) flipped the key negative. Electronic Materials operating profit, guided down 4.2%, came in +21.6% (¥1,816M, OPM 16.0%) on company-stated "AI-server-related demand." The upstream catalyst is now in the company's own numbers, not just the thesis. (iii) FY3/27 guidance now makes Electronic Materials the explicit profit engine: OP ¥2,400M, +32.2%, OPM 20.9% — 40% of consolidated OP (from 28% in FY3/26). The 30%→40% MLCC-OP-exposure claim from the new Archetype Capital note checks out exactly against primary segment data. (iv) The ¥130-dividend-cut flag below is wrong — actuals are ¥135 (FY3/25) → ¥145 (FY3/26, a raise) → ¥160 (FY3/27 forecast). Net read: the thesis was right on direction and wrong on entry timing — the asymmetry from a 0.68x P/B starting point is gone, and the stock now trades on whether the AI-server MLCC leg compounds. See the rebuilt Financials segment breakdown, the AI-server exposure section in Business, and the 2026-06-02 Decision-log entry. The low-conviction qualitative analysis below (minority of economics, Murata integration, named competition) still stands; only the price and the FY-segment numbers changed.

Verdict: HOLD / WATCH at ¥3,465. Low-to-medium conviction — the upstream MLCC thesis is real but governs a minority of the company, and the catalyst has not yet shown up in the company's own numbers. Sakai is a 90-year-old Japanese inorganic-chemicals conglomerate whose investable angle is one slice of one segment: hydrothermal-method barium titanate (BaTiO₃) dielectric powder, the binding ingredient inside every MLCC, where Sakai is a share leader on the high-end grade (mgmt-claimed 40–50% of the dielectric-materials market, self-reported). The bull framing — own the MLCC cycle upstream of the consensus crowding around Murata, Taiyo Yuden, and Yageo — survives contact with the financials only in diluted form.

Three facts keep conviction low. (1) The thesis governs a minority of economics. Electronic Materials is 11.9% of sales and 24.5% of segment operating profit before companywide expenses; the actual BaTiO₃ MLCC-powder revenue is most plausibly ¥5–7B (6–8% of company sales), not even the full ¥10B Electronic Materials line. (2) The largest historical customer is integrating away. Murata's MF Material JV (with Ishihara Sangyo + Fuji Titanium, Sept 2023) internalizes Murata's powder supply at the margin; Sakai is not a participant. The offset is that Samsung Electro-Mechanics, Taiyo Yuden, and Yageo remain merchant-dependent. (3) The catalyst is not in the numbers yet. Management's own FY3/26 guidance has Electronic Materials operating profit declining 4.2% (¥1,493M → ¥1,430M); mgmt states profit growth there "will be prohibitively difficult." The May-2026 industry MLCC price hikes (Murata 15–35%, Taiyo Yuden 6–13%, Yageo 10–20%) have not yet propagated to powder ASPs.

The valuation is genuinely cheap on conventional metrics — 12.8x forward P/E, 5.4x EV/EBITDA, 0.68x P/B, 4.62% dividend yield, ~14% trailing FCF yield, plus a ¥2.5B buyback (6.17% of shares) — and the capital-return signal is exactly what an undiscovered small-cap should send. But the cleaner comparison, Nippon Chemical Industrial (4092), is similarly cheap (10.1x fwd, 6.6x EV/EBITDA, 0.67x P/B, 3.07% yield) with materially faster near-term growth (+22% revenue vs Sakai +1.9% guided). The sector is broadly cheap; Sakai is not uniquely mispriced. The single highest-leverage next event is the FY3/26 full-year results pack (expected late May 2026, not yet released as of writeup) — the first read on whether powder ASPs have caught up to the finished-MLCC pricing cycle. Validation = FY3/27 commentary signaling double-digit Electronic Materials OP growth; continued flat-to-down = the upstream thesis is a slow burn at best.

(Note: this is a /profile-depth consolidation — no /deep-dive entry price, target, or DCF has been run yet. Verdict above is the profile's factual close reframed thesis-first; an explicit BUY/WATCH price ladder awaits a deep-dive.)

Snapshot

One-liner: Sakai makes the hydrothermal barium-titanate dielectric powder that goes inside high-cap MLCCs — the under-priced upstream slice of the AI-server / EV MLCC cycle — wrapped inside a cheap, no-coverage Japanese inorganic-chemicals conglomerate where that slice is only ~6–8% of sales.

• Ticker / exchange: 4078 on TSE Prime (Yahoo carries it as 4078.T)
• Legal name: Sakai Chemical Industry Co., Ltd. (堺化学工業); HQ Sakai City, Osaka; founded 1932; listed on the Tokyo Stock Exchange 1962; sakai-chem.co.jp
• Sector / industry: Basic Materials / Specialty Chemicals (inorganic chemicals conglomerate; investable angle = MLCC dielectric powder). Sits upstream of the passives-mlcc chain.
• Spot price: ¥6,250 (2 June 2026) — re-rated +80% since the 23 May profile's ¥3,465; near the 52-wk high ¥6,540 (52-wk range ¥2,559–6,540). (Historical: ¥3,465 at 22 May 2026 writeup, +44.2% trailing-12mo at that point.)
• Market cap: ¥95.5B (~USD 660M) | EV ¥88.1B (net cash position post-FY3/26: cash ¥15.5B vs debt ¥16.0B). (Was ¥53.1B mcap / ¥54.9B EV at writeup.)
• Valuation (2 June 2026): Trailing P/E 35.4x | Forward P/E 22.2x (FY3/27 EPS ¥282) | P/B 1.23x | EV/EBITDA 8.6x | dividend yield 2.56% (¥160/share FY3/27 forecast) | beta (5Y) 0.35. The re-rate has spent most of the cheapness — at writeup these were 22.0x / 12.8x / 0.68x / 5.4x / 4.62%. Sakai now trades above its own history and roughly in line with — no longer well below — peer Nippon Chemical (4092) (13.9x fwd, 8.9x EV/EBITDA, 0.92x P/B).
• Shares: 17.97M diluted (FY3/25); insiders 4.27%; institutional 56.14% across 53 institutions — passive/quant dominated (DFA, Vanguard, Avantis, Schwab); no concentrated active holder, no activist, no sell-side anchor
• Coverage: zero sell-side in yfinance — no mean PT, no consensus. Only continuous English coverage is Shared Research (company-commissioned, not consensus sell-side).
• Conviction: Low-to-medium — HOLD / WATCH at spot; the FY3/26 results pack is the gate

Business

What it does, plainly. A small-cap inorganic-chemicals house. It earns the right side of its income statement from one product family — high-purity ceramic dielectric powders, primarily barium titanate, used inside MLCCs — and the rest from a kitchen-sink portfolio: titanium dioxide (sunscreen, paint), zinc oxide (cosmetics, tires), plastic additives, catalysts, contract pharma, hygienic products. Management splits the book into "growth" (electronic materials, cosmetic materials, organic chemicals), "stable" (hygienic, contract processing), and "businesses under efficiency review" (titanium dioxide & zinc, plastic additives, catalysts, barium — the last meaning barium carbonate and sulfate, NOT the dielectric powder). The cleanup of the legacy book is the FY3/25 story; the dielectric-powder thesis is the FY3/27-and-beyond story.

Business model. Manufacturer, direct B2B sales, no licensing or recurring revenue. Electronic Materials pricing follows the broader MLCC cycle with a one-to-two-quarter lag — Sakai sells powder to capacitor makers who sell capacitors to OEMs. Margins are structurally higher in dielectric powders (segment OPM 14.9% vs corporate 7.2%) on process IP and qualification barriers; lower in titanium dioxide and catalysts where competition is commodified.

Segment mix — FY3/25 actuals (¥M; company segment disclosure, 27 May 2025 results pack p.10):

Segment	Sales	Operating Profit	OP Margin	Bucket
Electronic Materials	10,014	1,493	14.9%	Growth
Cosmetic Materials	2,676	293	11.0%	Growth
Organic Chemicals	6,638	770	11.6%	Growth
Hygienic Products	5,623	427	7.6%	Stable
Contract Processing	6,422	620	9.7%	Stable
Titanium Dioxide & Zinc	13,118	1,479	11.3%	Efficiency review
Plastic Additives	13,061	1,393	10.7%	Efficiency review
Catalysts	3,186	18	0.6%	Efficiency review
Barium (carbonate / sulfate)	5,175	826	16.0%	Efficiency review
Medical	8,321	(24)	(0.3%)	Standalone
Others	10,169	1,171	11.5%	—
Companywide expenses	—	(2,376)	—	—
Consolidated	84,409	6,093	7.2%	—

The headline read. Electronic Materials is 11.9% of sales and 24.5% of segment operating profit before companywide expenses. Even at the higher 14.9% OPM, this is not a barium-titanate pure play — it is an inorganic-chemicals conglomerate with a strategically positioned but financially modest electronic-materials slice. Size the thesis accordingly.

The share-vs-revenue reconciliation (the math to challenge mgmt on). Sakai claims 40–50% share of the "dielectric materials market" (self-defined). Against an externally-sourced powder market of plausibly USD 800M–1.2B, 40–50% implies ¥40–90B of MLCC-powder revenue at full external-market participation. But actual Electronic Materials segment revenue is ¥10B. Reconciliation: the share number is overstated, the market sizing is overstated, the powder is a small slice of segment revenue (Electronic Materials also includes non-MLCC dielectric, conductive pastes, other electronic chemicals), or captive/internal share dwarfs merchant share. Most likely reconciliation: actual BaTiO₃ MLCC-powder revenue is ~¥5–7B (6–8% of total company sales), not the full ¥10B line.

Geographic mix — now disclosed (FY3/26 brief, related-information note). FY3/26 external sales: Japan ¥67,313M (82.6%), Asia ¥10,993M (13.5%), North America ¥1,392M (1.7%), Europe ¥880M (1.1%), Middle East ¥725M, Other ¥141M. Overwhelmingly domestic-billed, but the Asia line (mostly China) is the swing factor — FY3/26/27 commentary repeatedly cites Chinese economic weakness. The earlier "split unverified" assumption flag is now resolved by primary disclosure. (Note: billing geography understates end-market exposure — Sakai sells powder to Japan-domiciled MLCC makers who export globally.)

Sole-source / hydrothermal positioning. Mgmt uses the hydrothermal synthesis method, "ideally suited for miniaturization" (FY3/25 pack p.26), producing particles down to ~100–200 nm with the spherical, narrow-distribution, high-crystallinity morphology high-cap MLCC requires. The solid-state method used by older suppliers and captive producers (Murata, Samsung) is cheaper but cannot reliably hit the same particle-size envelope, and qualifying a new supplier into a tier-one line is a multi-year process. The original swarm-brief "sole-supplier on sub-100nm" framing should be downgraded to "share leader on the hydrothermal-grade high-end with named competition" — mgmt's own slide acknowledges "one European company, one Japanese company, several Chinese companies."

AI-server MLCC exposure — the Archetype thesis, attributed. A 1 June 2026 note from Archetype Capital ("Sakai Chemical & Nippon Chemical MLCC Capacity Crunch," archetype-research.com — the author explicitly flags it as incomplete / DYOR and cites a Zephyr X post for market sizing) frames Sakai as a leveraged way to own the AI-server MLCC cycle upstream. Its core claims, with my own check against primary data:

• "~30% of Sakai's operating income is exposed to high-end MLCC, rising to ~40% on management's 2027 estimate." This checks out exactly. Electronic Materials OP is 28.1% of consolidated OP in FY3/26 actual (¥1,816M / ¥6,452M) and 40.0% in the FY3/27 forecast (¥2,400M / ¥6,000M). Archetype is using the right denominator (consolidated OP, after companywide expenses); the numbers land precisely on 28→40%. Note the 40% is inflated partly by the denominator shrinking — FY3/27 consolidated OP is guided down 7% as pigment TiO₂ winds off — not purely by Electronic Materials growing. But Electronic Materials genuinely is the growth engine: +32.2% OP guided, the only large segment guided meaningfully up.
• "Management is sandbagging the 2027 Electronic Materials estimate." Plausible, and the FY3/26 print is direct evidence for it: a segment guided down 4.2% delivered +21.6%. If that pattern repeats, the ¥2,400M FY3/27 OP figure is a floor, not a ceiling. I'd weight the sandbag read as credible but unproven for FY3/27 specifically.
• "Almost all of our electronic materials are for MLCC applications" (Integrated Report 2024) — effectively ~99% MLCC. Consistent with the company's own segment description (高純度誘電体材料・誘電体粉末 — high-purity dielectric materials and dielectric powders). I have not independently verified the literal ~99% figure against the 2024 report; [VERIFY] the exact wording, but it is directionally consistent with all prior disclosure. If true, it tightens the earlier reconciliation: the ¥11,377M Electronic Materials line is mostly MLCC dielectric, narrowing the gap with the earlier ¥5–7B "pure BaTiO₃ powder" estimate.
• Inputs: BaTiO₃ + other perovskite dielectrics; high-purity BaCO₃, Ba/Sr salts, high-purity TiO₂; electrode materials. Consistent with Sakai's product catalog (the 無機材料 / Inorganic Materials segment supplies the BaCO₃ / SrCO₃ feedstock chemistry in-house — a vertical-integration positive the earlier write-up under-weighted).
• Three BaTiO₃ routes by market share (Archetype): hydrothermal 13.5% (Sakai's focus — finest, most uniform particles, best for the thinnest / highest-end layers), oxalate 21.4% (Nippon Chemical's route — see Nippon Chemical (4092)), solid-state 61.8% (cheapest, scale, less fine). AI-server MLCC (high-capacitance, high-reliability) skews toward hydrothermal + oxalate, which is structurally favorable to Sakai and 4092 vs the solid-state majority. This route-share split is Archetype's number (sourced to Zephyr); I have not found an independent corroboration, so treat the exact percentages as one analyst's estimate, not established fact — but the direction (hydrothermal/oxalate favored by high-end mix-shift) is consistent with the physics and with Sakai's own miniaturization framing.

Market sizing — Zephyr (via Archetype), 29 May 2026 X post, single-source. Total MLCC ~$15B; server MLCC $1.3B in 2025 ($600M AI servers + $700M general servers); AI-server MLCC ~80%+ CAGR, general-server ~30–40% CAGR (agentic AI lifting CPU demand); smartphone/mobile MLCC in negative growth 2026–27; humanoids a future growth leg; book-to-bill >1 for most MLCC suppliers. This is one X post's framing, unverified against a primary market-research source — useful for shape, not for precision. The negative-mobile / positive-server split is the key qualitative claim and is consistent with the broader passives-mlcc cluster reads. The capacity-crunch thesis (Archetype's title) is that hydrothermal + oxalate capacity is the binding constraint on high-end AI-server MLCC, and that Sakai + Nippon Chemical are the merchant beneficiaries — directionally coherent with book-to-bill >1, but the "crunch" is asserted, not demonstrated with capacity-vs-demand figures in the note.

Plants & footprint:

Plant	Location	Status	Notes
Sakai Plant	Sakai City, Osaka	Operating (HQ)	Primary complex; multiple segments
Yumoto Plant	Iwaki, Fukushima	Operating	TiO₂ and zinc products
Onahama Plant	Iwaki, Fukushima	Operating	Inorganic chemicals
Multipurpose plant (Sakai)	Sakai City, Osaka	Completion Feb 2026	Cosmetic materials (zinc oxide); earnings from FY3/27

Pigment-grade TiO₂ production scheduled to terminate by end-2025 (efficiency review); freed cash rotates to the cosmetic-materials zinc-oxide plant. The medium-term plan flags ¥5.7B of MLCC-related investment over FY3/24–FY3/27, but no explicit capex announcement for a new BaTiO₃ / sub-100nm dielectric facility appears in public materials, and the specific grades targeted aren't broken out in English.

JVs / partnerships. None of strategic significance disclosed in English. Sakai Trading Co. is a wholly-owned distribution arm (recategorized into reportable segments FY3/25). The material competitive event is the MF Material Co. JV — Murata + Ishihara Sangyo + Fuji Titanium, formed Sept 2023 to produce/sell barium titanate, internalizing Murata's powder supply ("from BaTiO₃ powder through finished product"). Sakai is not a participant. The brief's "Murata is a top customer with sole-supplier dynamics" framing is at minimum incomplete, at maximum stale.

Customers & concentration — partial answer from the FY3/26 brief. The FY3/26 segment note discloses no single customer reaches 10% of consolidated sales (主要な顧客ごとの情報 — "記載を省略" because none ≥10%). That bounds the concentration risk at the company level: even the largest MLCC customer is <¥8.1B of total sales, i.e. <10% consolidated. Within the ¥11.4B Electronic Materials segment specifically, concentration is plausibly higher (a tier-one or two anchoring most powder revenue) but is still not separately disclosed — the segment-level customer table remains a research gap (pull EDINET Yuho). Mgmt cites "high customer stickiness — once materials are adopted, switching away typically involves a lengthy process" (FY3/25 p.26). Tier-one MLCC makers collectively (Samsung Electro-Mechanics, Taiyo Yuden, Yageo, Walsin, TDK, Holy Stone, PDC) are material buyers; Murata's share is declining as MF Material ramps (new BaTiO₃ plant in Nobeoka, Miyazaki, targeting capacity by 2027 — Fuji Titanium 70% / Murata 20% / Ishihara 10%).

Financials

Income statement & margins (¥M; FY3/26 now ACTUAL per 13 May 2026 results brief; FY3/27 is mgmt forecast):

Metric	FY3/23	FY3/24	FY3/25	FY3/26 actual	FY3/27E (mgmt)
Net sales	83,725	82,105	84,409	81,447	81,700
Revenue growth YoY	—	−1.9%	+2.8%	−3.5%	+0.3%
Gross margin	~20%	~20%	25.4%	~24.9%	n/a
Operating profit	4,407	2,942	6,093	6,452	6,000
OP margin	5.3%	3.6%	7.2%	7.9%	7.3%
Ordinary profit	—	—	6,279	6,545	6,100
Net income	2,344	(7,092)	5,013	2,752	4,400
EPS (¥)	144.85	(437.65)	309.21	176.42	282.00
Diluted EPS (¥)	—	—	278.91	157.55	—
Shares issued (M)	16.19	16.21	17.00	16.00	—

FY3/26 actuals — what changed vs the guidance the 23 May profile worked off. Sales fell 3.5% (the planned pigment-TiO₂ wind-down dragging the top line), but consolidated OP still rose 5.9% to ¥6,452M (OPM 7.9%). Net income fell 45.1% to ¥2,752M — that drop is a one-off cosmetic-materials impairment, not operating deterioration (operating and ordinary profit both rose). The FY3/24 net loss similarly reflected ¥6.6B of efficiency-review impairment; FY3/25 normalized; FY3/26 took a fresh cosmetics writedown.

The key reversal: Electronic Materials did the opposite of guidance. The segment was guided down 4.2% (¥1,493M → ¥1,430M) on "sluggish sales volume and fixed-cost pressure." It actually came in +21.6% (¥1,816M OP, sales ¥11,377M +13.6%, OPM 16.0%) — the company attributes the beat directly to "AI-server-related demand" (本決算説明会, "電子材料：AIサーバー関連の需要を取り込み、誘電体・誘電体材料ともに好調に推移"). The May-2026 MLCC upcycle the profile flagged as "not yet in Sakai's numbers" is now in them. This is the single most important change since the profile, and it validates the upstream thesis on direction.

Segment financial breakdown — FY3/26 actual (¥M; external-customer sales; segment OP is before companywide expenses of −¥2,500M; 13 May 2026 results brief, segment note p.15–16. The legacy "Barium" segment is now reported as Inorganic Materials (無機材料) = BaSO₄ / SrCO₃; this is the BaCO₃/SrCO₃ feedstock chemistry, NOT the dielectric powder):

Segment	Sales	OP	OP margin	% of pos. segment OP	MLCC / AI exposure
Electronic Materials	11,377	1,816	16.0%	19.2%	Direct — ~99% MLCC dielectric; AI-server driven
Cosmetic Materials	1,720	(437)	(25.4%)	—	None (UV-care ZnO/TiO₂; impaired FY3/26)
Organic Chemicals	7,185	721	10.0%	7.6%	None (sulfur chems, pharma intermediates)
Hygienic Products	5,353	457	8.5%	4.8%	None
Contract Processing	6,677	805	12.1%	8.5%	None
Titanium Dioxide & Zinc	10,244	1,222	11.9%	12.9%	None (pigment TiO₂ winding down)
Plastic Additives	11,507	1,068	9.3%	11.3%	None
Catalysts	3,462	648	18.7%	6.9%	None
Inorganic Materials (BaSO₄/SrCO₃)	5,100	1,200	23.5%	12.7%	Indirect — feedstock chemistry into MLCC inputs
Medical	8,391	(48)	(0.6%)	—	None
Others	10,427	1,498	14.4%	15.8%	Mixed
Companywide expenses	—	(2,500)	—	—	—
Consolidated	81,447	6,452	7.9%	—	—

Segment financial breakdown — FY3/27 forecast (¥M; 本決算説明会 p.14, "セグメント別予想"):

Segment	Sales	OP	OP margin	YoY OP	Read
Electronic Materials	11,500	2,400	20.9%	+32.2%	The profit engine — guided up hardest, highest margin
Cosmetic Materials	2,000	(400)	(20.0%)	loss	Still loss-making
Organic Chemicals	8,000	1,000	12.5%	+38.7%	Second growth leg
Hygienic Products	5,200	400	7.7%	−12.5%	Stable, softening
Contract Processing	7,100	900	12.7%	+11.8%	Stable, up
Titanium Dioxide & Zinc	6,800	100	1.5%	−91.8%	Pigment-TiO₂ wind-down drag
Plastic Additives	11,800	1,200	10.2%	+12.4%	Up
Catalysts	4,000	700	17.5%	+8.0%	Up
Inorganic Materials	5,100	700	13.7%	−41.7%	One-off FY3/26 strength reverting
Medical	8,700	(200)	(2.3%)	loss	Still loss-making
Others	11,500	1,500	13.0%	+0.1%	Flat
Companywide expenses	—	(2,300)	—	—	—
Consolidated	81,700	6,000	7.3%	−7.0%	TiO₂ wind-down swamps EM growth

The headline read on the segment economics. Two facts now coexist. (1) Electronic Materials is the most valuable real estate in the company on a forward basis — 20.9% OPM, +32.2% OP growth, the only large segment guided meaningfully up, and 40.0% of consolidated FY3/27 OP (¥2,400M / ¥6,000M) vs 28.1% in FY3/26. This is exactly the 30→40% Archetype flags, and it is the company's own forecast, not an analyst's. (2) Consolidated OP still falls 7% because the pigment-TiO₂ termination (−¥1,122M, −91.8% on the TiO₂/Zinc line) more than offsets Electronic Materials' +¥584M. So the "40% of OP" figure is half EM growth and half denominator shrinkage — read it as the MLCC slice becoming the company's center of gravity as the legacy book is pruned, not as a clean 40%-and-rising compounding line. The strategic signal is real; the FY3/27 consolidated number is still down.

Vertical-integration note (under-weighted in the profile). The Inorganic Materials segment (BaSO₄/SrCO₃, OPM 23.5% in FY3/26) supplies the high-purity barium/strontium carbonate feedstock chemistry that the Electronic Materials BaTiO₃ process consumes. Sakai therefore owns more of the MLCC-input value chain in-house than a powder-only read implies — a structural cost/quality advantage over merchant-feedstock competitors that the earlier write-up did not credit.

Cash flow & balance sheet (¥M):

Metric	FY3/23	FY3/24	FY3/25	FY3/26 actual
Operating cash flow	n/a	6,866	12,005	14,479
Investing CF	n/a	n/a	(5,714)	(4,645)
Capex (PP&E + intangibles)	2,658	4,024	6,948	5,731
Free cash flow (OCF − capex)	(1,885)	2,842	5,057	8,748
Financing CF	n/a	n/a	(6,879)	(10,592)
Cash & equivalents (end)	n/a	n/a	16,153	15,444
ROE	n/a	(9.2%)	6.6%	3.5%

FY3/26 is the strongest cash year in the set: OCF ¥14.5B (up 21% YoY), FCF ¥8.7B, comfortably funding the ¥10.6B financing outflow (the ¥2.5B buyback + ¥2.2B dividend + debt paydown). The balance sheet is now roughly net-cash (cash ¥15.5B vs total debt ¥16.0B per yfinance, vs ¥5.7B net debt at FY3/25). ROE optically fell to 3.5% — but that is the cosmetics-impairment hit to net income, not operating deterioration; the FY3/27 forecast net income ¥4,400M (+59.8%, helped by a ~¥1B asset-sale gain) restores ROE toward the 8% plan target. (FY3/22 column dropped to fit the FY3/26 actual; FY3/22 detail remains in the archived profile.)

(Note: the profile's trailing FCF-yield ~14% line cites "FY3/25 FCF ¥7.5B / mcap ¥53.1B," while the cash-flow table shows FY3/25 FCF of ¥5,057M (¥5.1B). The ¥7.5B figure appears to be OCF-less-maintenance-capex or a different basis — discrepancy retained, both numbers kept; verify against the FY3/26 pack. On the table's ¥5.06B, trailing FCF yield is ~9.5%, not 14%.)

Industry landscape

An upstream BaTiO₃ dielectric-powder play riding the AI-server / EV / miniaturization MLCC cycle. Industry-wide detail lives on the sector page — see passives-mlcc (the MLCC peer cluster: Murata, Taiyo Yuden, TDK, Yageo, Walsin, and the dielectric-powder upstream). Why it matters: barium titanate is the dielectric (insulating) layer between metal electrodes inside an MLCC; as MLCCs shrink (008004 sizing, sub-micron layers, 1,000+-layer high-cap parts) the powder must be finer, more uniform, more spherical, higher-crystallinity — the hydrothermal envelope. Global MLCC market ~USD 14–17B (2025–26), ~7–9% CAGR to 2030; the dielectric-powder addressable slice is ~USD 1.5–2.5B, structurally growing faster than finished MLCC because high-cap mix-shift raises dielectric-quality requirement per unit. Secular tailwinds: AI-server MLCC content (30,000+ MLCCs per high-end AI server vs hundreds in legacy enterprise), EV electrification (10,000+ MLCCs per EV vs ~1,000+ per ICE vehicle), 5G/6G densification, and miniaturization driving mix toward the high-end grades where Sakai's advantage is most binding.

Management

Data limitation: English-language management bios are thin; treat the table as directional pending an EDINET/TDnet pull or the FY3/26 annual report.

Name	Title	Tenure	Background
Takeshi Yano	Representative Director, President	Multi-year (per IR site)	Long-tenured Sakai engineer/executive, internal-promotion path
[CFO not separately disclosed in English IR materials]	—	—	—

Disclosure quality — governance-positive. Commentary throughout the FY3/25 deck is candid and self-critical. Page 26 grades the year-one "Dielectrics" sub-target as △ (not achieved / in progress): "No significant progress achieved toward ensuring customer adoption within high-end and mid-range markets." Unusually direct for a Japanese mid-cap, where the pattern is to claim universal "achieved" status. Weight the honesty.

Ownership & governance. Insider 4.27% (standard Japanese mid-cap, not founder-controlled). Institutional 56.14% across 53 institutions — passive/quant dominated (top holder DFA International Small Cap Value 1.37%; Vanguard Total International 0.98%; Japan Smaller Capitalization Fund 0.93%). No concentrated active holder, no activist, no sell-side anchor — the textbook "uncovered Japanese small-cap" structure, and the structural reason it hasn't re-rated alongside consensus MLCC names: no active manager is doing the work. Single share class, no poison pill, no staggered-board issue identified in English filings.

Capital allocation — the strongest governance positive. Committed FY3/27 ROE target 8%. Medium-term plan includes a ¥2.5B share buyback executed in FY3/26 — share count fell from 17.0M to 16.0M issued (a real ~6% reduction, confirmed in the 13 May 2026 brief), bought back at a sub-book multiple. Dividend correction (the earlier "¥130 cut" flag is wrong against actuals): the payout is ¥135 (FY3/25) → ¥145 (FY3/26, a RAISE) → ¥160 (FY3/27 forecast, +15), not a cut to ¥130. The 3-year medium-term-plan total shareholder return is now guided at ~¥9.4B, well above the ≥¥8B plan target. Both dividend and buyback are rising — an unambiguously positive capital-return signal, and management is explicit that returns over the plan period overshot. The 23 May profile's "reduced per-share payout" framing was based on a stale ¥130 guidance figure and should be disregarded.

Key-person risk: Low — long-tenured team, no single founder-CEO dependency.

Catalysts & risks

Catalysts — the near-term gate has now passed (and cleared). The FY3/26 results pack (13 May 2026) was the gate the profile flagged. It cleared: Electronic Materials OP +21.6% vs guided −4.2%, company-attributed to AI-server MLCC demand; FY3/27 guides the segment up a further +32.2%. The "is the MLCC rally in the numbers yet?" question is answered yes. The new open question is no longer whether the catalyst exists but whether it's now priced — at 22.2x forward P/E and 1.23x P/B the answer is "substantially." Medium-term, the out-year call options (FY3/27+):

0. AI-server MLCC compounding (the live thesis). Per Zephyr/Archetype, AI-server MLCC is the fastest-growing MLCC end-market (~80%+ CAGR off a $600M-2025 base) and skews to hydrothermal/oxalate powder — Sakai's grade. If management is sandbagging FY3/27 Electronic Materials (the FY3/26 beat says they might be), ¥2,400M segment OP is a floor. Bear offset: smartphone/mobile MLCC in negative growth 2026–27, so the server leg has to carry the segment alone.
1. MLCC cycle catch-up (the single binary catalyst). Sakai's powder pricing lags finished-MLCC pricing by 1–2 quarters. If the hikes hold through calendar Q3 2026, FY3/27 Electronic Materials margin should expand as contract repricing catches up.
2. Cosmetic-materials zinc-oxide capacity. New multipurpose plant online Feb 2026, earnings from FY3/27; contribution not quantified by mgmt.
3. Pharmaceutical CDMO scale-up. Medical segment break-even/slightly negative; long-dated, low conviction.

Near-term restructuring tailwind already in the FY3/26 guide: Catalysts swing ¥18M → ¥430M (+2,288%) on full-year flow-through of FY3/25 selling-price revisions. R&D ~¥3–4B/yr (~3.5–4.5% of sales). No active M&A deal disclosed.

Risks:

1. Catalyst not yet in the numbers (High) — FY3/26 Electronic Materials OP guided DOWN 4.2%; the upstream thesis is unconfirmed by the company's own forecast. The rally may be a slow burn.
2. Murata vertical integration via MF Material (Medium, structural) — Murata is ~34% of global MLCC share, so even partial in-housing of powder removes a meaningful slice of the addressable market; erodes wallet share at the largest historical customer; manageable, not closable. Offset: Samsung Electro-Mechanics (~22% MLCC share), Taiyo Yuden (~12%), Yageo (~13%) remain merchant-dependent.
3. Chinese new-entrant competition on mid-range grades (Medium-High, structural) — Hayna (北矿), Guoci (300285.SZ) credible at mid-range; defend high-end via R&D, cede mid-range.
4. Pigment-grade TiO₂ termination fixed-cost overhang in 2H FY3/26 (High, mgmt-admitted) — closes by FY3/27.
5. Customer concentration (Medium, undisclosed) — the single biggest research gap; no names or concentration in English filings.
6. US reciprocal tariffs (Medium, mgmt-flagged) — excluded from guidance pending clarity; not company-controllable.
7. Single-product (BaTiO₃) dependence within Electronic Materials (Medium) — closes only via M&A or organic diversification, multi-year.
8. Cyclicality (High) — Electronic Materials fully cyclical with the MLCC up-cycle; TiO₂/zinc tied to construction/auto; catalysts/additives tied to general industrial. Beta 0.35 is misleading — it reflects low historical TOPIX correlation, not low underlying cyclicality.

Dilution risk: Low — diluted shares 16.19M → 17.97M (FY3/23→25) reflects convertible-bond treasury movement; the ¥2.5B FY3/26 buyback is net-negative dilution (taking shares out at 0.68x P/B). No ATM, no warrant overhang. Execution risk: Low — 90-year-old group running known processes; the only open execution items (cosmetic-plant ramp, catalysts restructuring) are well-disclosed and on schedule. Regulatory: Low company-specific.

Valuation / DCF

No DCF modeled yet (this is a /profile-depth consolidation).

2026-06-02 — the cheapness is largely gone. At writeup (¥3,465) the stock was genuinely cheap: 12.8x fwd P/E, 5.4x EV/EBITDA, 0.68x P/B, 4.62% yield. At ¥6,250 (+80%) it is 22.2x fwd P/E, 8.6x EV/EBITDA, 1.23x P/B, 2.56% yield — above its own history and no longer at a discount to the cleaner comp. Does the live valuation support "cheap ahead of the expansion"? No — that window has closed. The market re-rated the stock as the catalyst confirmed (FY3/26 beat + FY3/27 AI-server guide), which is precisely the re-rating the no-coverage thesis was waiting for; it happened faster than a HOLD/WATCH stance captured. What remains investable is not a cheap-multiple bet but a durability bet: whether the AI-server MLCC leg compounds enough to grow into a 22x forward multiple, and whether mgmt is sandbagging FY3/27 (the FY3/26 beat says possibly). At 1.23x P/B the margin of safety the original thesis leaned on is spent.

The market-share leadership on the hydrothermal high-end is real and durable; that did not change. What changed is the price you pay for it.

The cross-camp check (refreshed 2 June 2026): Nippon Chemical Industrial (4092) — the closest direct Japanese comp, and the oxalate-route BaTiO₃ player to Sakai's hydrothermal — now trades at ¥5,350 / mcap ¥46.3B / forward P/E 13.9x, EV/EBITDA 8.9x, P/B 0.92x, dividend yield 2.15%, on revenue ¥40.2B. Both have re-rated; the gap narrowed but did not invert — 4092 is still cheaper on every multiple (13.9x vs 22.2x fwd P/E, 0.92x vs 1.23x P/B), while Sakai now carries the premium. Per the Archetype route-share framing, both win from AI-server mix-shift (hydrothermal 13.5% + oxalate 21.4% of BaTiO₃ supply, the high-end-favored routes, vs solid-state 61.8%). The pair still says the sector re-rated on the MLCC/AI story, with Sakai leading and now the more expensive of the two. The open pairs question: does Sakai's hydrothermal positioning + buyback justify a premium to Nippon Chemical, or is the cleaner growth and lower starting multiple at 4092 the better Japanese-specialty-chemicals-with-MLCC-exposure pick? Unresolved — flagged for a deep-dive / /compare.

Market-share cross-checks: Chemical Research Insight (2025) and persistencemarketresearch.com both rank Sakai top-three. The 40–50% mgmt claim is share of external merchant supply (the upper bound); one LinkedIn industry note's 18–20% figure likely includes captive (Murata/Samsung) in the denominator. True number unverifiable from public sources.

Decision log

2026-05-23 — Profile (factual close, no entry price set). HOLD / WATCH at ¥3,465. Triggered because three sub-agents (Taiyo Yuden, TDK, Walsin) in the May 2026 passives-mlcc peer swarm independently flagged Sakai as the under-priced upstream alpha play (BaTiO₃ as the binding MLCC ingredient; few qualified suppliers; sub-micron grade the bottleneck). Tested whether the upstream framing survives the financials. Conclusion: real but smaller than the swarm brief implied. (a) Electronic Materials is only 11.9% of sales / 24.5% of segment OP — actual powder revenue ~¥5–7B (6–8% of sales); (b) Murata's MF Material JV weakens the largest historical customer; (c) mgmt's own FY3/26 guide has Electronic Materials OP DOWN 4.2% — the 2026 MLCC rally is not yet in Sakai's numbers; (d) valuation cheap (12.8x fwd, 0.68x P/B, 4.6% yield, 6.17% buyback) but Nippon Chemical (4092) is similarly cheap with far better near-term growth (+22% vs +1.9%). Several "sole-supplier" framings downgraded to "share leader with named competition." Highest-leverage next action: FY3/26 full-year results pack (the powder-ASP catch-up read). Validation = FY3/27 commentary signaling double-digit Electronic Materials OP growth; continued flat-to-down = slow-burn thesis.

Open items for a deep-dive: explicit entry/target price ladder + DCF; resolve the ~14% vs ~9.5% FCF-yield discrepancy against the FY3/26 pack; pull EDINET Yuho for customer-concentration table, segment capex, named-officer comp, top-10 shareholders; settle the Sakai-vs-Nippon-Chemical (4092) pairs question.

2026-06-02 — Financials deepened + FY3/26 results + Archetype AI-server thesis integrated. No verdict change in label, but the basis shifted. Triggered by integrating the Archetype Capital note (1 Jun 2026, archetype-research.com — one analyst's view, flagged DYOR, citing a 29 May 2026 Zephyr X post for market sizing). Three things established against PRIMARY data (Sakai 13 May 2026 results brief + presentation, EDINET-equivalent TDnet): (1) The FY3/26 results landed and reversed the key negative — Electronic Materials OP came in +21.6% (¥1,816M, OPM 16.0%) vs the −4.2% it was guided to, company-attributed to AI-server MLCC demand. FY3/27 guides the segment +32.2% to ¥2,400M OP (20.9% OPM). (2) Archetype's headline numbers check out exactly: Electronic Materials = 28.1% of consolidated OP FY3/26 → 40.0% FY3/27 forecast (their "~30%→~40%"). The 40% is half EM growth, half denominator shrinkage (pigment-TiO₂ wind-down, −¥1,122M) — consolidated OP is actually guided down 7%. (3) The stock re-rated +80% to ¥6,250, taking it from 12.8x→22.2x fwd P/E, 0.68x→1.23x P/B, 4.6%→2.6% yield — the "cheap ahead of the expansion" framing is spent; the market priced the catalyst as it confirmed. Two corrections to the 23 May profile: the dividend is a RAISE (¥135→¥145→¥160 forecast), not the "¥130 cut" previously flagged; and customer concentration is now bounded (no customer ≥10% of consolidated sales, per the FY3/26 note). Net: thesis right on direction, wrong on entry timing. What's left is a durability/sandbag bet at a full multiple, not a cheap-multiple bet. Highest-leverage next action: a /deep-dive with DCF to test whether AI-server MLCC compounding justifies 22x forward, and whether mgmt is sandbagging FY3/27 (the FY3/26 beat is evidence they might be). Route to inv-q for review.

Sources

Fragments folded into this canonical page (consolidated 2026-06-02; original archived to _migration-archive/2026-06-02/4078/): 4078-t-profile.md (profile, 2026-05-23).

Related vault pages: passives-mlcc · Nippon Chemical Industrial (4092) · Murata (6981) · Taiyo Yuden (6976) · TDK (6762) · Walsin (2492) · Prosperity Dielectrics (6173) · MLCC/mlcc-peer-swarm-2026-05-20

Key external sources (FY3/26 primary, added 2026-06-02): yfinance (live price/valuation 4078.T + 4092.T, accessed 2 June 2026; financials FY3/22–FY3/26); Sakai FY3/26 決算短信 / Results Brief, 13 May 2026 — authoritative segment table p.15–16 (FY3/25 + FY3/26 actual), consolidated p.1, FY3/27 forecast p.2, geographic + customer-concentration note (no customer ≥10%); Sakai FY3/26 本決算説明会 / Results Presentation — FY3/27 segment forecast p.14 (Electronic Materials ¥11,500M / ¥2,400M OP), AI-server commentary p.3 ("電子材料：AIサーバー関連の需要を取り込み").

New analyst source integrated 2026-06-02 (attributed, not authoritative): Archetype Capital, "Sakai Chemical & Nippon Chemical MLCC Capacity Crunch," archetype-research.com, 1 June 2026 — one analyst's view; author flags it as incomplete / DYOR. Source for the 30%→40% OP-exposure framing (verified against primary segment data: 28.1%→40.0%), the ~99%-MLCC electronic-materials claim ([VERIFY] exact 2024 IR wording), and the three-route BaTiO₃ share split (hydrothermal 13.5% / oxalate 21.4% / solid-state 61.8% — Archetype's figures, sourced to Zephyr, not independently corroborated). Market sizing within it is attributed to a Zephyr X (Twitter) post, 29 May 2026 (total MLCC ~$15B; server MLCC $1.3B 2025; AI-server ~80%+ CAGR, general-server 30–40%; mobile negative 2026–27; book-to-bill >1) — single-source, unverified against primary market research, used for shape not precision.

Prior-year + cross-check sources: Sakai FY3/25 Financial Results Presentation, 27 May 2025; Integrated Report 2025; Medium-Term Management Plan; Murata MF Material JV (passive-components.eu) + Murata JV establishment notice (new BaTiO₃ plant, Nobeoka, Miyazaki; Fuji Titanium 70% / Murata 20% / Ishihara 10%; capacity targeted by 2027); Chemical Research Insight 2025; Shared Research 4078; kabutan 4078. SemiAnalysis cross-check: no SA coverage on Sakai / barium titanate / MLCC dielectric powders in the local mirror — coverage gap, not a contradiction; the AI-server MLCC-powder thesis still has zero SA validation. Filings note: Sakai is TSE-only; primary disclosures are TDnet (決算短信, used here) + EDINET (annual Yuho — still the pull for the segment-level customer table, segment capex, named-officer comp).