フジオーゼックス株式会社（FUJI OOZX Inc.）

研究原文：FUJI OOZX Inc. (7299, TSE Standard) is Japan's largest engine-valve maker (company-claimed ~33% domestic and ~15% world share). It was founded in 1951, renamed in 1992 and has its head office and main plant in Kikugawa, Shizuoka. It has 10,279,750 shares issued and capital of JPY 3,019 million. It sells intake and exhaust valves, valve seats, cotters and retainers made to engine-maker specifications in Japan, Mexico, China and Indonesia. A small 'other' segment makes semiconductor-line control parts. The investable claim is a minority stake in a Daido Steel-controlled subsidiary whose economics sit mostly in the Japanese parent (76% of sales).

内部研究类型：company_identity_perimeter

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Every later judgment is read through the fact that minorities own a claim on a controlled subsidiary, not on an independent company.

不确定性（原始记录）：Share and position figures come from a secondary compilation and are used only for position, not calculation; no independent market-share audit exists.

研究原文：The question completed Research leaves is whether a company whose core product is tied to the combustion engine, but whose near-term demand is extended by the hybrid revival, can turn an 85% equity ratio, ~JPY 6 billion of net cash and a newly progressive dividend-and-buyback policy into durable per-share value before electrification reaches its core product. A second part is what Daido Steel's 54% control means for minorities along the way. After Coverage, the question is better framed as a race between two slow processes: compounding capital return at a deep discount, and the powertrain clock that shrinks the niche.

内部研究类型：core_investment_question

内部来源标记：OPEN_DISCOVERY

普通股东含义（原始记录）：It sets what the equity outcome depends on: allocation and delivery, not engine-market growth or solvency.

不确定性（原始记录）：Daido Steel's intentions are the irreducible unknown; the successor plan (May 2027) is the first dated evidence.

研究原文：H1 FY3/27 (to 30 Sep 2026) delivered a beat: sales JPY 14,542m (+2.6%), operating profit JPY 1,731m (+61.8%), ordinary profit JPY 1,693m (+47.1%), net income JPY 1,194m (+22.0%). On 27 Oct 2026 the FY3/27 forecast was raised to sales JPY 29,200m, operating and ordinary profit JPY 2,920m and net income JPY 1,850m (from JPY 2,600m / 2,600m / 1,700m). The dividend forecast was raised to JPY 60 from 54. Profit leverage came from cost reduction and price pass-through on near-flat sales. Net income also includes a JPY 101m asset-sale gain and the profit line includes an unquantified tariff refund.

内部研究类型：current_change_finding

内部来源标记：OPEN_DISCOVERY

普通股东含义（原始记录）：Proves first-half delivery and raises the cash return, but the pass-through reset is only one half old.

不确定性（原始记录）：The refund and asset-sale gain are not quantified in the release, so the clean run-rate cannot be isolated.

研究原文：The raised forecast is first-half loaded. Full-year ordinary profit JPY 2,920m less H1 JPY 1,693m implies H2 ordinary profit of about JPY 1,227m, roughly 23% below H2 FY3/26 (JPY 2,743m less 1,151m = JPY 1,592m). Implied H2 net income is about JPY 656m. The revision therefore proves delivery rather than a second-half acceleration, and the implied H2 level sits near the lower-middle of the five-year ordinary-profit band. It is either conservatism or an admission that the H1 margin level (including the tariff refund) is not the new run-rate.

内部研究类型：implied_expectation_gap

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The market's focus has moved to this bridge, not the full-year number.

不确定性（原始记录）：A derived inference from company guidance, not a disclosed H2 plan; the tariff refund size is unknown.

研究原文：FY3/26 was volume without spread. Sales rose 13.9% to a record JPY 29,093m (domestic +8.6% including the P&M consolidation; overseas +25.9%) but operating profit fell 4.2% to JPY 2,506m and automobile-parts segment profit fell 9.7%. Volume-response costs, tariffs, labour and materials outran price pass-through. Ordinary profit rose 17.2% to JPY 2,743m only with the help of a JPY 220m FX gain.

内部研究类型：historical_mechanism_finding

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Shows the operating-leverage downside: record sales did not create profit when costs led prices.

不确定性（原始记录）：The release does not split FY3/26 growth into volume, price and P&M consolidation.

研究原文：Ordinary profit, not net income, is the recurring anchor. FY3/26 net income (+38.6% to JPY 2,144m) was lifted by a Mexican deferred-tax credit from revaluing deferred tax assets, JPY 152m of insurance proceeds against a JPY 94m impairment, and the JPY 220m FX gain below the operating line, while operating profit fell 4.2%. H1 FY3/27 net income carries a JPY 101m Maruyoshi asset-sale gain and the tariff refund. One-off density is a recurring property of this register, so quality adjustment is standing procedure. Capitalising reported net income overpays in good-print years.

内部研究类型：earnings_quality_adjustment

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Prevents using flattered or depressed net income as the valuation base.

不确定性（原始记录）：Deferred-tax credit size and tariff-refund amount are not separately disclosed; the deferred-tax asset is not cash and its usability is unproven.

研究原文：The balance sheet is a fortress. At 30 Sep 2026 the equity ratio was 85.3% on total assets of JPY 38,673m, net assets were JPY 33,546m, total liabilities only JPY 5,127m and short-term borrowings zero. FY3/26 net cash was about JPY 6.0bn (cash JPY 6,507m against debt JPY 473m), roughly 35% of the JPY 17.2bn market capitalisation. Liabilities are mostly trade payables and electronic obligations, so financial leverage is effectively absent. Land of about JPY 3.4bn is carried at historical cost. All residual risk is operating and allocative.

内部研究类型：balance_sheet_resilience

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Removes the distress scenario and replaces it with an opportunity-cost question about idle cash.

不确定性（原始记录）：Cash measures differ across sources (cash and equivalents JPY 7,429m vs cash and deposits JPY 5,801m at 30 Sep 2026), so the net-cash figure depends on definition; parent-available versus subsidiary cash is not separately disclosed.

研究原文：Capital return inflected in FY3/26. JPY 835m of buybacks (including the January 2026 ToSTNeT-3 purchase) took treasury shares from 78,745 (0.8%) to 446,777 (4.3%). Treasury shares were partly recycled to 379,355 for restricted-stock compensation and the employee shareholding association. Dividends moved JPY 52 - 54 - 60 planned (3.58% forecast yield). The payout ratio nonetheless fell from 34.3% to 25.4% as earnings rose, dividend-on-equity is only about 1.7-1.9%, and no payout, DOE or buyback framework is stated. FY3/26 returns of about JPY 1.39bn were roughly 8% of current market value and were funded from operating cash flow.

内部研究类型：capital_allocation_inflection

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The direction is shareholder-friendly, but with no commitment device the pace is discretionary and the cash pile still grows.

不确定性（原始记录）：One year is not yet a regime; the planned JPY 60 dividend is a forecast, not a declaration.

研究原文：The growth bet is geographic, not product. The Mexican subsidiary (FUJI OOZX MEXICO, Guanajuato) is investing US$10.175m (about JPY 1.5bn) to lift capacity to 2.3 million valves per month. Construction started in August 2026 and operations are scheduled for October 2027, aimed at firm North American hybrid demand. Construction in progress had already jumped from JPY 536m to JPY 1,237m at FY3/26, and capex (JPY 2,094m) ran below depreciation (JPY 2,447m), so investment is re-accelerating after years of harvesting. The same plant complex paid the tariffs now partly refunded. Expect a J-curve: capex trims FCF before it adds profit.

内部研究类型：growth_capex_bet

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The only volume-growth engine; its return decides whether capital deployment beats returning cash.

不确定性（原始记录）：Target return, customer volume commitments and total CIP programme size are undisclosed; effect on FY3/27 is described as minor.

研究原文：The only non-engine growth leg is small but real. The 'other' segment (semiconductor-line automatic-control parts, with P&M consolidated from July 2024) grew sales 36.1% in FY3/26 to JPY 1,146m, moving from a JPY 120m loss to JPY 34m profit. In H1 FY3/27 sales rose 42.0% to JPY 772m and profit rose from JPY 6m to JPY 91m. At about 4% of sales it is an option, not a driver.

内部研究类型：diversification_option

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：A non-engine profit leg of real size would rewrite the earnings band and the post-engine narrative, but it is not yet large enough.

不确定性（原始记录）：P&M's standalone earnings and purchase price are undisclosed; the segment may include acquisition effects.

研究原文：Control and supply both run through Daido Steel. It holds 47.95% of voting rights directly and 54.25% combined (31 Mar 2026). Fuji Oozex is the group's only engine-valve manufacturer. It bought JPY 3,320m of valve steel, die steel, cladding materials and processed parts in FY3/26 from Daido Kogyo, the group's trading arm, with prices negotiated periodically on overall cost and market prices (payables JPY 345m plus electronic obligations JPY 1,125m at year-end). Two of twelve directors hold concurrent Daido posts and the president has been a Daido career executive since June 2025. There are no loans or guarantees from the parent.

内部研究类型：control_and_supply_dependency

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The parent is simultaneously controller, principal materials channel and management source, so margin and capital policy are partly set in a related-party negotiation minorities cannot observe.

不确定性（原始记录）：The pricing formula with Daido Kogyo and the parent's long-term holding intention are undisclosed.

研究原文：Governance is parent-led with formal minority protections: a company-with-audit-committee structure, twelve directors of whom four are independent outsiders, and a governance committee for parent-subsidiary listed companies chaired by an independent outside director that reviews related-party transactions. Protection is procedural, not structural; the committee's review outcomes are not disclosed. Improvement would show in payout policy before board composition.

内部研究类型：governance_mechanism

内部来源标记：COVERAGE35

普通股东含义（原始记录）：Governance quality is the discount; judge it by the pricing and capital decisions it produces.

不确定性（原始记录）：Committee outcomes and any engagement or voting evidence are not in the evidence set.

研究原文：Valuation at the 6 Nov 2026 close (research cut-off 9 Nov 2026): price JPY 1,675 (-4.23% on the day); market capitalisation JPY 17.22bn (JPY 1,675 x 10,279,750 issued shares); forecast PER 8.9x on revised FY3/27 forecast EPS JPY 187.70; actual PBR 0.51x on FY3/26-end net assets per share JPY 3,254.46; FY3/26 reported ROE 6.9%; forecast dividend yield 3.58% on the JPY 60 forecast dividend; EV/EBITDA 2.3x (EV JPY 11.42bn after net cash JPY 5.80bn at 30 Sep 2026; EBITDA JPY 4.95bn = FY3/26 operating profit JPY 2,506m + depreciation JPY 2,447m).

内部研究类型：valuation_snapshot

内部来源标记：OPEN_DISCOVERY

普通股东含义（原始记录）：Reference multiples for the discount-versus-delivery debate.

不确定性（原始记录）：Mixed bases: market cap uses issued shares including 379,355 treasury shares; PBR uses year-end rather than September book; EV combines a Sept net-cash measure with FY3/26 actual EBITDA; cash definitions differ. The multiples are indicative, not precise.

研究原文：At 0.5x book and under 9x raised-forecast earnings, the market prices the Daido-controlled thin float and the combustion-engine decline, not the balance sheet. PBR has stayed between 0.18x and 0.69x every year since 2010, so the discount is a structural feature of the register rather than a dislocation that will mean-revert unaided. Net cash is about 35% of market value, and returns near the cost of capital explain why the discount does not contradict arithmetic. Re-rating needs evidence: compounding capital return or a float/control event, not cheapness.

内部研究类型：valuation_interpretation

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Defines what must change for the discount to narrow.

不确定性（原始记录）：No analyst consensus is in the evidence set to compare against.

研究原文：The group has one operating segment, automobile parts (~96% of sales), plus 'other'. Parent-only sales of JPY 22,230m are 76% of the consolidated total, so the economics are those of a Japanese parent with an overseas production perimeter. Revenue is a derivative of engine build schedules: a fixed valve content per engine sold into negotiated model programmes at thin margins. The consolidated margin is therefore a direct read on valve pricing power, with no second engine to mask a failed pass-through. Operating leverage is high both ways.

内部研究类型：business_model

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Shareholders own a volume derivative; sales records do not guarantee profit records.

不确定性（原始记录）：No customer or programme split and no volume/price/consolidation decomposition of revenue.

研究原文：Demand is rotating rather than growing. Domestic sales rose 6.2% in H1 on Japanese production recovery and equipment demand, North America grew on firm hybrid demand and new orders, and China fell sharply as new-energy vehicles displaced gasoline cars, so overseas sales fell 4.1% and consolidated sales grew only 2.6% despite strong profit leverage. Three unrelated powertrain stories net into one growth rate, and revenue quality increasingly depends on one region's hybrid cycle.

内部研究类型：demand_structure

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Consolidated growth stays low-single-digit in good years unless North America accelerates; volume is not the equity story.

不确定性（原始记录）：Customer names, hybrid share of own valve volumes and demand by powertrain are undisclosed.

研究原文：The core product is bound to internal-combustion engines. Every point of battery-electric mix gain removes engine-valve demand entirely, and China, historically a growth market for the group, is already shrinking for exactly this reason. Management's outlook says BEV adoption is slowing in major markets while hybrid/PHEV demand rises and carbon-neutral fuels extend engine life, so the decline path looks longer and flatter than earlier scenarios. The range between a hybrid-extended decline measured in decades and a fast-BEV decline in years is the risk.

内部研究类型：structural_risk

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Sets the terminal-value ceiling: valuation should capitalise a long tail, neither a cliff nor growth.

不确定性（原始记录）：Management's outlook is company framing; no independent market-size series was pulled and BEV adoption pace is exogenous.

研究原文：The moat is qualification and process depth, not brand or patents visible in the accounts. Valves are failure-critical, specification-bound parts qualified per engine programme, made with materials and cladding know-how (sodium-filled hollow valves, carbon-neutral-fuel-compatible and high-function designs), so R&D expense is small. Switching means requalifying a safety-relevant part for negligible cost saving, and the niche is too small and specification-bound to invite entry. The moat defends share and spread for the life of an engine programme, is re-contested at each new programme, and has no content in BEV programmes.

内部研究类型：moat_mechanism

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Supports the downside case (share retention) more than the upside case.

不确定性（原始记录）：No contract terms, programme win/loss rates, patent data or hollow-valve price premium disclosed; competitor benchmarking was not pulled; the moat reading is structural inference.

研究原文：Margin equals negotiated prices minus steel, wage and tariff costs with a lag. Interim gross margin expanded from 22.5% to 25.7% (gross profit JPY 3,193m - 3,731m on sales +2.6%) and automobile-parts segment profit growth swung from -9.7% in FY3/26 to +54.2% in H1 FY3/27, driven by cost reduction and pass-through of wage increases rather than volume. The lag compresses margin on the way up the cost cycle and releases it on the way down. Pass-through is durable as a mechanism, but any single half's margin is not. Steel enters through the Daido channel, and the tariff refund sits inside cost of sales.

内部研究类型：causal_mechanism

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Profit quality is improving operationally, but each period needs one-offs stripped before the trend is read.

不确定性（原始记录）：Refund size is unknown, so the clean pass-through effect cannot be isolated; H2 margin bridge unresolved.

研究原文：Operating cash flow was JPY 5,028m in FY3/26 (pre-tax profit JPY 2,801m + depreciation JPY 2,447m), about twice net income on average over five years, and JPY 2,251m in H1 FY3/27. A mature, heavily depreciated asset base converts profit to cash at a high rate (capex JPY 2,094m versus depreciation JPY 2,447m), so free cash flow stays strongly positive and funds all shareholder returns without touching the balance sheet. The Mexico programme is the first material exception.

内部研究类型：cash_mechanism

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Cash generation, not earnings growth, funds the return story; rising capex will trim FCF first.

不确定性（原始记录）：Maintenance-versus-expansion capex split undisclosed; FCF here is OCF less tangible-asset purchases.

研究原文：Work in process rose JPY 494m in H1 to JPY 2,681m, the largest operating-cash-flow absorption of the half. Working capital runs at about two months of sales (FY3/26 receivables ~JPY 6.0bn, inventories JPY 6.3bn, payables ~JPY 3.9bn). The build is either production ahead of second-half deliveries (supporting the raised forecast) or the first sign of an inventory overhang if the implied weaker second half disappoints. Conversion by the Q3 balance sheet is the test.

内部研究类型：balance_sheet_watch_item

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：A cheap early read on whether the H2 forecast and demand are real; a build that persists would signal slippage.

不确定性（原始记录）：Whether the WIP is Japan- or Mexico-bound, and its ownership/realisability and costing basis, are undisclosed.

研究原文：There is no large hidden-asset unlock. Investment securities are only JPY 129m, so the gap between the JPY 17.2bn market value and JPY 33.5bn of net assets rests on land at historical cost (JPY 3.4bn book) and on the cash pile, not on a cross-shareholding portfolio. The Maruyoshi asset-sale gain is a small precedent for monetisation. Asset value supports a valuation floor but does not compound unless realised or deployed.

内部研究类型：asset_realisability

内部来源标记：COVERAGE35

普通股东含义（原始记录）：The discount is an allocation/governance discount, not an information gap about hidden assets.

不确定性（原始记录）：Appraised land values undisclosed; any monetisation of the controlled-entity assets would face minority-claim and tax friction.

研究原文：ROE has ranged from 2.2% to 6.9% over five years (3.8, 2.2, 6.8, 5.2, 6.9) and ROIC is around 6.0-6.5%, so returns hover around a plausible cost of capital. FY3/26 ROE of 6.9% is slightly above the auto-parts sector median (~6.1%) while PBR (0.5x) sits at the sector's discounted end. Idle cash at an 85% equity ratio dilutes ROE, so value must come from capital return and discount narrowing rather than compounding operations.

内部研究类型：capital_efficiency

内部来源标记：COVERAGE35

普通股东含义（原始记录）：Fundamental counterpart of a sub-book valuation.

不确定性（原始记录）：ROIC is an aggregator computation; the company's own WACC estimate is undisclosed; some of the ROE gain in FY3/26 came from tax and FX items.

研究原文：The free float of about 37.8% sits just above the TSE Standard continued-listing threshold. A large buyback that retires shares without a matching parent sell-down would tighten the float further, so capital return and float maintenance pull in opposite directions. Daido's holding intention is undisclosed. Treasury holding of 4.3% already reduces outstanding shares, and the float ratio figure is an aggregator's compiled number.

内部研究类型：mechanical_constraint

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：A mechanical limit on the pace of buybacks, and a potential catalyst if the parent restructures.

不确定性（原始记录）：The exact listing threshold arithmetic and how treasury shares are treated were not independently confirmed in the evidence set.

研究原文：Ordinary profit over FY3/22-FY3/26 was JPY 1,784 / 1,052 / 1,928 / 2,340 / 2,743m, and the revised FY3/27 forecast of JPY 2,920m sits at the top of that band. A normalised band of roughly JPY 1.0-2.9bn values the JPY 17.2bn market capitalisation at about 6-17x through-cycle earnings, with a mid-band of about JPY 1.9bn. The market is paying for the trough-to-mid part, not the forecast peak. Whether H1's margin resets the band upward is unproven until two consecutive halves are seen.

内部研究类型：normalized_economics

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Paying about 9x the top of the band is paying about 13-17x trough earnings.

不确定性（原始记录）：The band is the five disclosed years, not a fitted model; the tax-rate normalisation and net-income bridge are not modelled.

研究原文：The downside case is stagnation, not distress. Even the FY3/23 trough produced JPY 1,052m of ordinary profit at an 80% equity ratio, and with net cash of ~JPY 6bn the realistic loss scenario is years of low-single-digit ROE at a persistent discount (0.3-0.5x book) while cash accumulates. A faster BEV transition or a China-plant impairment (plausible but unquantified) would compress volumes or add write-downs. The balance sheet converts downside into time and makes buybacks at low prices possible.

内部研究类型：downside_scenario

内部来源标记：COVERAGE35

普通股东含义（原始记录）：Capital loss risk comes from multiple stagnation and opportunity cost, not impairment of the claim.

不确定性（原始记录）：Built from the company's own trough and valuation history, not a stress model; China plant impairment not quantified.

研究原文：The dated catalysts are the Q3 FY3/27 release around January 2027 (the WIP build must convert to sales and the H2 bridge tested), the FY3/27 year-end release in May 2027 with the successor to the 2026 medium-term plan, and the Mexico plant start in October 2027. Each is evidentiary, resolving a named open question rather than creating momentum. The next two prints can each move the implied full year by the size of the one-offs they reveal.

内部研究类型：dated_catalyst_path

内部来源标记：COVERAGE35

普通股东含义（原始记录）：Defines when the thesis can be confirmed or falsified.

不确定性（原始记录）：The successor plan's timing is an expectation, not a company commitment, and whether it will contain a capital framework is unknown.

研究原文：The market's verdict on the raise is visible: after the 27 Oct 2026 forecast and dividend raise the shares fell sharply on profit-taking the next day and closed 6 Nov at JPY 1,675, about 10% below the October close of JPY 1,870. Good news is faded at 0.5x book, and the marginal holder's question is the H2 bridge and capital policy, not the FY3/27 number. Expectations now require delivery, not announcements.

内部研究类型：expectation_signal

内部来源标记：COVERAGE35

普通股东含义（原始记录）：Near-term returns depend on delivery against a sceptical register.

不确定性（原始记录）：No analyst consensus; sellers' identity unknowable; the market report is a single secondary source plus aggregator price data.

研究原文：The final two years of the 2026 medium-term plan (FY3/25-FY3/27, formulated March 2024) concentrate change: the January 2024 stock split and shareholder-benefit programme, P&M consolidation (July 2024), the first large buyback (January 2026), the Maruyoshi liquidation decision (February 2026; dissolved September 2026), the Jatos-Tetos merger (April 2026) and the Mexico expansion (July 2026). They share one direction: concentrate on the valve core and North America, seed a non-engine leg and start returning the cash pile. Whether this is a regime change or a plan-period flourish is decided by the successor plan in May 2027.

内部研究类型：strategic_turning_point

内部来源标记：COVERAGE35

普通股东含义（原始记录）：The capital-policy turn is the one that matters most to minorities and the least contractually committed.

不确定性（原始记录）：Reading these actions as a closing programme is this research's inference.

研究原文：Thesis: a niche-leading, parent-controlled engine-valve maker with a fortress balance sheet is priced (0.5x book, ~9x peak-band earnings) as if decline and cash-trapping were permanent. Operating leverage, a first large buyback, a rising dividend and North American hybrid capacity support a slow narrowing of that discount through capital return and delivery, not growth. Confidence is moderate. Falsifiers: (1) the successor plan (May 2027) arrives with no capital framework while cash keeps accumulating; (2) H2 FY3/27 shows the H1 margin was refund and pass-through timing rather than a reset; (3) China moves from decline to impairment without a North American offset. If falsified, the position becomes a value trap with a strong balance sheet: not distress, but no reason to own.

内部研究类型：thesis_and_falsification

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：States what would make the discount narrow and what would kill the case.

不确定性（原始记录）：Daido Steel's intentions remain the irreducible unknown.

研究原文：US tariff policy on Mexican production is a recurring two-way P&L line. Tariffs paid by the Mexican subsidiary were a profit headwind in FY3/26, and a partial refund supported H1 FY3/27 profit, sitting inside cost of sales. This widens the range of reported outcomes around the operating trend. It is a different channel from emissions policy, which acts through powertrain mix. The refund size and the remaining tariff run-rate are undisclosed.

内部研究类型：policy_transmission

内部来源标记：COVERAGE35 / OPEN_DISCOVERY

普通股东含义（原始记录）：Policy is a recurring P&L and valuation factor, so the margin level cannot be read without it.

不确定性（原始记录）：Amount of the refund and forward tariff exposure are not disclosed.

研究原文：Research Debt: does the successor medium-term plan (expected with FY3/27 results in May 2027) commit to a payout ratio, DOE or buyback framework, and how does the revised FY3/27 forecast compare with the 2026 plan's final-year numerical targets (which this research did not obtain)? Currently dividend-on-equity is only ~1.7-1.9% and no framework exists to hold management to the new trajectory. If no framework appears while cash keeps accumulating, the thesis weakens. Resolving evidence: the successor plan document and plan-target disclosure.

内部研究类型：research_debt

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Decides whether capital return is a commitment or a discretion.

不确定性（原始记录）：Future disclosure; no reasonable current public path exists, and the plan's numerical targets were not obtained.

研究原文：Research Debt: how much of the H1 FY3/27 profit beat came from the Mexican tariff refund and the JPY 101m Maruyoshi asset-sale gain? The release does not quantify the refund. The answer separates a pass-through reset from timing noise and conditions the credibility of the new margin level. Resolving evidence: the Q3 and year-end special-items lines or any quantified management commentary.

内部研究类型：research_debt

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Determines whether the normalised earnings band shifts up.

不确定性（原始记录）：May never be separately disclosed.

研究原文：Research Debt: did the JPY 494m interim WIP build convert into second-half sales, or did it remain as inventory at Q3? Resolving evidence: the Q3 FY3/27 balance sheet (around January 2027), which is future disclosure.

内部研究类型：research_debt

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：First test of the production-ahead reading versus an overhang.

不确定性（原始记录）：Post-cutoff evidence; belongs to the next Review.

研究原文：Research Debt: what is the disclosed demand path for engine valves by powertrain and region, and has the company quantified how hybrid volumes offset BEV substitution? Today only the regional sales split and qualitative commentary exist, and the hybrid share of own volumes is undisclosed. This bounds the length of the decline tail that valuation should capitalise.

内部研究类型：research_debt

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Sets the range between a decades-long tail and a years-long decline.

不确定性（原始记录）：No reasonable public source identified beyond management commentary.

研究原文：Research Debt: what is the status and scale of the China operation as Chinese gasoline-car volumes shrink, and is it a restructuring or impairment candidate? Falsifier (3) of the thesis depends on this; subsidiary-level financials are not disclosed beyond segment and note data.

内部研究类型：research_debt

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：A plausible but unquantified write-down item and a thesis falsifier.

不确定性（原始记录）：Subsidiary-level data absent.

研究原文：Research Debt: what return does the Mexico expansion (US$10.175m, 2.3m valves per month) target and what North American volume commitments support it, and how large is the total construction-in-progress programme (JPY 1,237m at FY3/26) and which plants and products does it serve? Without these, the growth capex cannot be tested against the cost of capital.

内部研究类型：research_debt

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Determines whether growth capex creates value or just dilutes ROIC.

不确定性（原始记录）：Company has not disclosed target returns or customer commitments.

研究原文：Research Debt: does Daido Steel intend to hold, raise or reduce its stake over time, and could TSE float expectations pressure that decision? This is the irreducible unknown at the centre of the thesis and governs buyback pace, float and any control event.

内部研究类型：research_debt

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The binding constraint on re-rating through float or control events.

不确定性（原始记录）：Undisclosed and not reasonably resolvable by public research now.

研究原文：Daido Steel is parent with 47.95% direct and 54.25% combined voting rights; no loans or guarantees from the parent. Control is settled and stable, so the minority question is the terms of coexistence rather than the risk of a control change.

内部研究类型：controls

内部来源标记：OPEN_DISCOVERY

研究原文：Daido Kogyo, the Daido group's trading arm, supplied JPY 3,320m of valve steel, die steel and materials in FY3/26 at periodically negotiated prices, so steel-price cycles and group pricing practice transmit directly into the valve margin.

内部研究类型：supplies_cost_input_to

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Valve volumes follow combustion and hybrid engine production in Japan, North America and China. End demand is a derivative of powertrain mix, not vehicle demand alone, so the same vehicle parc yields valve demand only while engines are combustion or hybrid.

内部研究类型：derived_demand_constraint

内部来源标记：OPEN_DISCOVERY

研究原文：The wholly owned Mexican subsidiary is the site of the US$10.175m expansion to 2.3m valves per month from October 2027. It is simultaneously the North American growth engine and the point where US trade policy enters the P&L.

内部研究类型：owns_production_subsidiary

内部来源标记：OPEN_DISCOVERY

研究原文：US tariffs on Mexican production reduced FY3/26 profit and a partial refund supported H1 FY3/27 profit. Trade policy is now a recurring P&L line in both directions, widening reported outcomes around the operating trend.

内部研究类型：policy_to_producer_profit

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：P&M, acquired July 2024 inside the non-automotive segment, grew segment sales +36.1% in FY3/26 and profit from JPY 6m to JPY 91m in H1 FY3/27. It is the only growth leg not tied to combustion engines, but is about 4% of sales.

内部研究类型：diversification_seed_against_concentration

内部来源标记：OPEN_DISCOVERY

研究原文：With about 62% of votes in the parent orbit and a ~38% float, minorities receive the dividend (JPY 60 planned) and were the counterparties to the FY3/26 buyback. The float's return depends on the parent's tolerance for a rising payout and buyback pace.

内部研究类型：conditions_minority_return

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：The automobile-parts segment profit (JPY 1,640m in H1 FY3/27) is nearly the consolidated operating profit it feeds (JPY 1,731m). The group result is the spread between steel-driven costs and negotiated prices, with no second engine to mask a failed pass-through.

内部研究类型：segment_spread_is_group_result

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Net cash (~35% of market capitalisation), land at old cost and negligible investment securities set the valuation floor by assets rather than earnings momentum. The debate is the size of the discount applied to assets controlled by a parent, not whether the assets exist.

内部研究类型：sets_valuation_floor

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：A float of about 37.8% just above the continued-listing threshold quietly caps buyback pace and keeps the parent-shareholding question alive. Capital return and float maintenance pull in opposite directions.

内部研究类型：constrains_pace

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Mexican capacity of 2.3m valves per month from October 2027 is being built against firm North American hybrid demand. This fuses the demand thesis and the tariff thesis into one plant-level question, and the group's only volume-growth engine is North American hybrids produced in Mexico.

内部研究类型：capacity_built_against_demand

内部来源标记：OPEN_DISCOVERY

研究原文：FY3/26's cost-led compression (operating profit -4.2% on +13.9% sales) reversed into H1 FY3/27's pass-through-led expansion (operating profit +61.8% on +2.6% sales). This is consistent with a lagged pass-through catching up, so the swing may reflect timing as much as a permanent reset.

内部研究类型：lagged_reversal

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：The raised full-year guidance minus delivered H1 implies a weaker H2 than H2 FY3/26, so the revision proves first-half delivery but not acceleration.

内部研究类型：arithmetic_implication

内部来源标记：OPEN_DISCOVERY

研究原文：WIP conversion by Q3 tests whether H2 is conservative guidance (build converts to sales) or a genuine slowdown (inventory overhang).

内部研究类型：early_test_of

内部来源标记：COVERAGE35

研究原文：The tariff refund sits inside cost of sales and so inflates the 25.7% interim gross margin by an unknown amount, limiting how much of the pass-through gain is clean.

内部研究类型：contaminates_margin_read

内部来源标记：COVERAGE35

研究原文：High cash conversion from a mature asset base funds the dividend and buyback entirely from operating cash flow, so returns can rise without financial strain. Mexico capex will absorb part of that cash from FY3/27.

内部研究类型：funds

内部来源标记：COVERAGE35

研究原文：Net cash and an 85% equity ratio let the company fund the Mexico transition and shareholder returns without asking the market for money, so dilution risk is nil and the risk is cash trapping.

内部研究类型：funds_without_external_capital

内部来源标记：COVERAGE35

研究原文：Returns near the cost of capital are the fundamental counterpart of the 0.5x PBR: the market is not mispricing assets so much as pricing the low spread on them.

内部研究类型：explains

内部来源标记：COVERAGE35

研究原文：Parent control, related-party supply and procedural-only minority protection keep a structural governance and float discount (PBR below 0.7x for sixteen years) that only compounding capital return can narrow.

内部研究类型：contributes_to

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Combustion dependence caps how long the normalised ordinary-profit band can persist and is why mid-band, not peak-band, earnings are the sober anchor.

内部研究类型：bounds_durability_of

内部来源标记：COVERAGE35

研究原文：The sell-the-news response to the raise is consistent with the market focusing on the implied weaker H2 and the quality of the margin level rather than the headline raise.

内部研究类型：reprices_toward

内部来源标记：COVERAGE35

研究原文：The successor plan's capital framework is the first dated falsifier or confirmation of the thesis.

内部研究类型：decisive_unresolved_evidence_for

内部来源标记：COVERAGE35

研究原文：The fundamental event record (27 Oct 2026 beat, raise, dividend raise) and the observed post-event close are cross-referenced to the protected Technical Synthesis as a whole. This relation does not alter, extend or reinterpret that Technical truth; the Technical State keeps its own non-confirming status (continuation unconfirmed), and positive fundamental delivery does not rewrite it.

内部研究类型：fundamental_event_context_for_price_response

内部来源标记：COVERAGE35 / TECHNICAL

研究原文：Protected Technical Synthesis (RESEARCH_TRUTH.technical_state)

内部研究类型：protected_technical_synthesis

内部来源标记：TECHNICAL

不确定性（原始记录）：受保护的 Technical Research 结构；公开边界以其正文路由为准。

公开边界：受保护 Technical Research 仅保留可追踪身份，本索引不重写其内部语义。

研究原文：Durable Business Portrait: a niche share-holder with a fortress balance sheet inside a parent group｜The company earns a specification-bound niche spread on engine production through qualification depth and a position inside the Daido group, converts it to cash at a high rate on a depreciated asset base, keeps no debt, and earns near its cost of capital. What is durable is position, balance sheet and the pass-through mechanism. What is not durable is any year's profit level, which is set by engine volumes and a negotiated pass-through lag, nor the niche itself, which is bounded by combustion-engine production. Governance sits with the parent.

内部研究类型：durable_business_portrait

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Downside is stagnation at a discount, not distress; upside depends on allocation and delivery rather than growth.

不确定性（原始记录）：Programme-level, customer-level and powertrain-level data are undisclosed; competitor benchmarking was not obtained.

研究原文：Two slow clocks: compounding capital return versus the powertrain transition｜What is durable here is position rather than growth: the largest share of a specification-bound niche, a materials and production position inside the Daido group, an unlevered balance sheet and a price structure whose declines have been arrested higher. The investment question resolves into a race between compounding capital return at a deep discount and the powertrain transition shrinking the niche. Both are observable and neither is fast. Falsifiers are dated: the May 2027 successor plan, the two prints before it, and the monthly price sensors recorded in the protected Technical Synthesis.

内部研究类型：structural_durability_race

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The balance sheet buys time for the capital-return clock to compound; if cash merely accumulates, the discount cannot narrow without heroic engine-market growth.

不确定性（原始记录）：Pace of BEV transition is exogenous; Daido's intentions undisclosed.

研究原文：From cost-led compression to pass-through-led expansion, with one-offs in between｜Steel, wage and tariff costs enter cost of sales immediately while prices move by negotiation, so margin follows the cost cycle with a lag. FY3/26 shows the squeeze (record sales, lower operating profit, FX-assisted ordinary profit, tax-flattered net income); H1 FY3/27 shows the release (higher margin on flat sales), partly helped by a tariff refund and asset-sale gain. The clean trend sits in ordinary profit, and the band of JPY 1.0-2.9bn is the sober anchor.

内部研究类型：causal_chain

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Valuing on reported net income or on the forecast peak overpays; the margin reset is unproven until two halves at the new level.

不确定性（原始记录）：Refund and deferred-tax credit sizes undisclosed; causal attribution between pass-through timing and reset is inferential.

研究原文：The second-half bridge: implied slowdown, WIP build and a market that sold the raise｜The raise shows H1 delivery, yet implies H2 ordinary profit about 23% below the prior-year H2, the interim WIP build awaits conversion, and the market sold the news. Together these frame the January 2027 print as the first evidentiary test of whether the margin level is a reset.

内部研究类型：expectation_delivery_test

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Near-term outcomes depend on the Q3 bridge; confirmation supports the delivery narrative, a persistent WIP build or a weak H2 would revert the margin reading to timing.

不确定性（原始记录）：Derived from guidance arithmetic and a single market-reaction report; tariff refund unquantified.

研究原文：What the market pays for: 0.5x book, ~9x peak-band earnings, 35% net cash｜The valuation prices control and decline, not assets or current earnings. Assets set a floor (net cash ~35% of market value, no hidden cross-holdings), normalised earnings of JPY 1.0-2.9bn give 6-17x, returns near cost of capital justify part of the discount, and parent control and float explain the rest. Re-rating requires evidence of compounding return or a float or control event.

内部研究类型：valuation_bridge

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Entry price already assumes mid-band delivery; the downside is multiple stagnation and opportunity cost.

不确定性（原始记录）：Mixed share-count, cash and period bases in the snapshot; no consensus comparison.

研究原文：Capital allocation under parent control: cash, buybacks, Mexico and the float ceiling｜Operating cash flow and an unlevered balance sheet let the company choose between returning cash and building Mexican capacity. Parent control and a float just above the listing threshold bound how far buybacks can go, and there is no stated framework to commit management. The plan-end cluster of actions suggests a policy turn, but only the successor plan can show whether it is a regime.

内部研究类型：capital_allocation_mechanism

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：If the turn compounds, per-share asset value rises mechanically at a discount; if cash accumulates, it cannot.

不确定性（原始记录）：No formal capital policy; Daido's holding intention undisclosed.

研究原文：Research Debt carried to the next Review｜All carried questions are bounded by future disclosure or company non-disclosure; none has a reasonable current public path likely to change the Thesis, so Research Stop holds and no reopen is required.

内部研究类型：research_debt_set

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Each debt conditions confidence in a specific mechanism: capital policy, margin quality, H2 bridge, demand tail, China exposure, Mexico return and parent intention.

不确定性（原始记录）：Several (Q3 balance sheet, successor plan) are post-cutoff evidence belonging to later Reviews.

研究原文：The balance sheet bridges the two clocks｜The powertrain clock shrinks the niche over years, while about JPY 6bn of net cash and an 85% equity ratio let the company return capital and fund the Mexico transition without asking the market for money in the meantime.

内部研究类型：cross_domain_bridge

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Time is financed internally; the question is whether it is used.

不确定性（原始记录）：Cash definitions differ across sources.