4368 — Fuso Chemical Co., Ltd.

Deep-value analysis — Fuso Chemical Co., Ltd. (扶桑化学工業), TSE Prime 4368

Analysis date: 08/25/2026 · reference price … (close 08/25/2026, 15:30 JST) · market cap …–323.2 bn · EV … · single currency JPY (flows and price) · analysis, not investment advice.

Fiscal year convention. The fiscal year ends March 31. Throughout this report “FY2026” = the year ended 03/31/2026 (the company calls it 2025年度), “FY2027” = the current year, ending 03/31/2027. The “Leap 2030” plan (飛躍2030) covers FY2027–FY2031 (in the company’s notation, 2026–2030年度).

Split adjustment. On 03/30/2026 a 3:1 forward split took place (source: stockanalysis.com/statistics, Stock Splits section, verified 08/25/2026). All per-share figures in this report are post-split. This alone resolves the discrepancy flagged in the research brief (dividend “…” vs “…”) — see the “Accounting red flags” chapter, point 6.

Executive summary

The thesis. Fuso Chemical is, almost certainly, the best niche business analyzed in recent months and, at the same time, one of the most expensively priced. The company makes ultra-pure colloidal silica (“Quartron”), the raw material of CMP slurries used to planarize silicon wafers — a critical input, individually qualified at each fab, with a small cost within the customer’s structure and catastrophic consequences in case of defect. The electronic materials segment delivered FY2026 revenue of … with an operating margin of 38.3% and an EBITDA margin of 60.2% (“Leap 2030” plan, p. 6, figures in 億円: 415 / 159 / 250). This isn’t a chemical manufacturer; it’s a tollgate business with a chemical manufacturer’s cost structure.

The problem isn’t quality, it’s the price and the timing of cash flow. At …/share, the stock trades at 18.1× trailing profit (TTM EPS …) and at 19.5× FY2027 guidance profit cleaned of the government subsidy (EPS …). The 5-year P/E band, calculated on 20 quarters from the S&P Global Market Intelligence series, has its median at 13.96× and its historical maximum at 26.61× — so today’s price sits in the top quartile of its own valuation history, after a … rise over 52 weeks. P/S is 3.97× against a 5-year median of 2.34×.

The analysis’s central finding. The management plan published 05/15/2026 contains a capital-allocation bridge that says, in the company’s own numbers, what will be left for the shareholder over the next five years: of … of cumulative operating cash flow (FY2027–FY2031), … go to investment (… maintenance/consolidation, … R&D, … the new Kyoto line, … additional growth investment), … to debt repayment and … to dividends. OUT exactly equals IN. The plan’s implicit post-investment free flow is … over five years, i.e. …/year — or …/year after debt repayment — not the … reported as FCF in FY2026. FY2026 was a pause year between two capex programs, not the normal level.

Confirmation already comes from Q1 FY2027: quarterly FCF was −…, net PP&E rose … in a single quarter on paid capex of only …, and “other current liabilities” jumped … — 169 mil off from the PP&E gap. The investment program is running, in commitments, at almost double the payment pace.

Estimated value. I used as the owner-earnings base … — below the FY2026 reported FCF (…), but above the figure implicit in the company’s own plan (…/year) — with 8% phase-1 growth, a 9% discount rate (identical to the … WACC S&P Global estimates) and 1.5% terminal growth. Result, triangulated across five models:

Range … … …, median …. The Monte Carlo simulation across 20,000 scenarios gives a median intrinsic value of … and a probability of undervaluation of …%.

Verdict. Excellent business, fortress balance sheet (net cash …, Altman Z 5.74, interest coverage 175×), earnings quality HIGH, but a price that already pays for full execution of the 2030 plan. Not a deep-value position at … Becomes interesting below ~… (the simulation’s median intrinsic value) and clearly attractive below ~… a level last reached in December 2025. The divergence from consensus target price (… …, 4 analysts) is the largest in recent analyses and is fully explained by what happens to the cash: consensus values the FY2030 plan profit; I value the cash that reaches the shareholder in the meantime, and the plan says that’s nearly zero.