新日本電工株式会社（Nippon Denko Co., Ltd.）

研究原文：Reconciled core question after Research: is the FY12/26 reported-earnings hole (nine-month ordinary loss of JPY 1,246 million beside operating profit of JPY 7,315 million) the last act of a recurring manganese-cycle pattern, after which a safeguard-fed ferroalloy upcycle, a roughly 30%-margin ash-recycling engine and a formulaic rising shareholder return on a 74%-equity balance sheet at about 0.67x book re-rate the stock toward the FY12/27 plan? Or do policy-dependent prices, a back-loaded forecast and a strengthening Nippon Steel axis make the discount the correct price for the structure? Research resolves this into a race between two dated processes: delivery of the Q4 FY12/26 forecast and the FY12/27 plan under a new president, versus the finite life of the safeguard premium that created the current margin.

内部研究类型：investment_question

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Frames every other node: the shareholder outcome depends on whether the underlying run-rate is earned and policy-durable or borrowed from a reviewable trade measure.

不确定性（原始记录）：The answer is open until the February 2027 annual print and the 10th medium-term plan; both are unprinted at the 2026-11-09 cutoff.

研究原文：Nippon Denko (legal name Shin Nippon Denko), a Nippon Steel-group ferroalloy consolidator, earns through five engines under one roof: manganese ferroalloys (about 62% of sales, priced off international benchmarks), functional materials, incineration-ash recycling, aqua solutions and hydroelectric power, plus a small 'other'. Segment profit is measured at the ordinary-profit level and the segments plus adjustments sum to the consolidated ordinary result, so the ferroalloy segment carries the Malaysian equity-method affiliate inside it. The company therefore speaks in two profit languages that routinely differ by billions of yen: statutory ordinary profit (cycle, inventory valuation and one-offs) and management's 'underlying' ordinary profit (ex inventory effects and one-offs). Segment reads must use the underlying/reported split, never sales shares alone.

内部研究类型：business_model_and_profit_languages

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Reading reported EPS alone mispriced this stock in both directions across the cycle; the segment/underlying split decides what profit actually belongs to the ordinary shareholder.

不确定性（原始记录）：Underlying is management's own construction, consistent across years but unaudited as a series; no product-level revenue split is disclosed.

研究原文：The ordinary-profit record shows a manganese sawtooth: FY12/19 loss of JPY 6,426 million, FY12/20 JPY 3,492 million, FY12/21 JPY 6,870 million, FY12/22 peak of JPY 10,367 million, FY12/23 JPY 2,465 million (-76.2%), FY12/24 JPY 4,859 million, FY12/25 JPY 2,703 million (-44.4%), while underlying ordinary profit stayed in a band of about JPY 5.2-5.5 billion across FY12/23-25. History warns against extrapolating the top of the cycle: the reported series swings from -6.4 to +10.4 billion yen on the same asset base, with a smoother underlying band beneath it.

内部研究类型：cycle_history

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The market's discount partly reflects a long record of whipsaw in reported profit; only the underlying band has been stable.

不确定性（原始记录）：Cycle dating rests on the company's own price tables and profit record; no external cycle index was pulled.

研究原文：The FY12/26 affiliate write-down is the second time impairment, inventory write-down and affiliate-share valuation losses have blown a hole in reported profit: FY12/19 (ordinary loss JPY 6,426 million, notices of 30 January 2020) carried the same three-part anatomy seven years earlier, and the Pertama equity-method drag already existed in FY12/25 at JPY 1,392 million. The pattern is a recurring feature of the business model, not a one-off accident.

内部研究类型：recurring_pattern

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：If affiliate and inventory write-downs recur at cycle turns, the clean-up thesis must be priced as a recurring cost of the model; the 2026 charge is also described as the largest clean-up step.

不确定性（原始记录）：Whether the Pertama exit closes the recurrence channel depends on undisclosed cash exposure (see research debt) and on future affiliates or inventory turns.

研究原文：The FY12/26 operating surge is a price-and-valuation event, not a volume event. Nine-month operating profit rose 131.6% to JPY 7,315 million and gross margin jumped from 14.8% to 22.3% (gross profit JPY 13,132 million against JPY 8,574 million) while sales rose only 1.8%, because European high-carbon ferromanganese rose from a 2025 average of USD 1,175/t to USD 1,335/t in Q3 2026 under safeguard protection, the yen stayed near 159 per dollar, and inventory valuation was favourable. Q1 already showed the operating turn (operating profit +99.6%) a quarter before the affiliate charge hit below the line; ore eased from USD 5.4 to 4.8 per Mn% in Q3, widening the spread.

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）：Price-led profit converts at high incremental margin while it lasts and reverses the same way; it is more reversible than volume-led profit.

不确定性（原始记录）：Persistence depends on safeguard design and the yen; the split between price, FX and inventory valuation within quarters is disclosed only in aggregate.

研究原文：Pertama Ferroalloys (Malaysian equity-method affiliate inside the ferroalloy segment) was impaired in full in Q2 FY12/26: a parent-level special loss of JPY 7,877 million became, on consolidation, an equity-method loss of JPY 7,733 million in H1 (JPY 7,180 million of it the write-down beyond the running loss share). This turned reported ordinary profit to -JPY 2,913 million at H1 and -JPY 1,246 million at nine months and net profit negative (EPS -JPY 21.55 at H1), because taxes stay booked as equity-method losses are largely non-deductible. Underlying ordinary profit was about JPY 5.1 billion over nine months against JPY 3.2 billion a year earlier: the operating temperature and the reported temperature differ by one impairment.

内部研究类型：impairment_event

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Reported EPS collapsed by an accounting event that does not reflect the operating business; valuation on the statutory print overstates the multiple (about 94x forecast EPS) against about 6.3x on underlying.

不确定性（原始记录）：The underlying figure is management's own series; the split of the charge between parent and consolidated routing is clear but the cash consequence is not.

研究原文：The Pertama impairment is finished as a P&L item: Q3 added only about JPY 10 million (cumulative JPY 7,743 million), the carrying value is zero, the FY12/26 underlying forecast carries the overseas loss at only about JPY 0.4 billion, and the overseas operation is expected to return to underlying profit only in FY12/27. The interim and Q3 documents reviewed disclose no new loans, guarantees or capital injections for Pertama, so any further exposure would have to arrive through guarantees or new funding, which are not disclosed in the filings reviewed.

内部研究类型：exposure_closure_judgement

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The P&L channel through which the affiliate destroyed value is closed; whether a cash channel remains is the residual risk.

不确定性（原始记录）：Non-disclosure in reviewed filings is not proof of absence; the annual securities report may disclose guarantees or funding commitments.

研究原文：The FY12/26 forecast was left unchanged at Q3: sales JPY 81,100 million (+5.0%), operating profit JPY 11,000 million (+113.1%), ordinary profit JPY 1,500 million (-44.5%), net profit JPY 500 million (EPS JPY 4.01). Subtracting nine-month actuals, Q4 must deliver ordinary profit of about JPY 2,746 million (about 2.1x the prior-year quarter) and operating profit of about JPY 3,685 million (against about JPY 2,002 million in Q4 FY12/25). Q3 standalone ordinary profit was about +JPY 1,667 million with no new one-offs, the cleanest read of run-rate. Derived cross-check (synthesis arithmetic on frozen figures): the underlying forecast of about JPY 7.5 billion less nine-month underlying of about JPY 5.1 billion implies roughly JPY 2.4 billion of underlying profit in Q4, against about JPY 0.6 billion in Q3, the seasonally softest quarter - so price and inventory effects, not run-rate alone, must carry Q4.

内部研究类型：forecast_structure

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The year's verdict is concentrated in one unprinted quarter: delivery validates the underlying-recovery framing; a cut revalidates the reported-hole narrative the market currently prices.

不确定性（原始记录）：Derived underlying Q4 arithmetic uses rounded 'about' figures and management's non-GAAP series; the quarter has not printed.

研究原文：Domestic ferroalloys is the swing earner: underlying profit of about JPY 1.4 billion in H1 (about JPY 2.4 billion reported) and about JPY 0.7 billion underlying in Q3 alone (about JPY 1.5 billion reported), against a reported segment loss (nine months -JPY 5,356 million on sales of JPY 36,621 million) created entirely by booking the Malaysian affiliate in the same segment. FY12/26 segment underlying plan: domestic about +JPY 2.0 billion, overseas about -JPY 0.4 billion; FY12/25 underlying ferroalloy profit was also positive once the equity-method charge and inventory effects are excluded.

内部研究类型：profit_pool

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The segment carrying most of the cycle also carries the affiliate noise; the domestic underlying earning power is what the market is not capitalising.

不确定性（原始记录）：Segment underlying splits are management briefing figures, not audited segment disclosures.

研究原文：Incineration-ash recycling is the under-rated second engine: nine-month sales of JPY 8,532 million and ordinary profit of JPY 2,589 million (30.3% margin; H1 JPY 1,521 million; FY12/25 JPY 2,074 million at 23.3%), already above the whole of FY12/25, driven by waste-treatment volumes and molten-metal prices. It is the highest-margin and least price-cyclical profit pool, underwritten by permits and waste volumes rather than alloy prices alone; FY12/26 underlying plan is about JPY 2.9 billion.

内部研究类型：profit_pool

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：This is the trough-earnings annuity: it earned through the alloy downturn and is the destination of growth capital; roughly matches the swing factor of the entire reported result.

不确定性（原始记录）：Molten-metal price normalisation could compress the margin; permit renewals and new entrants are not evidenced; capacity added by the Kashima expansion is not quantified.

研究原文：Functional materials is the soft spot: nine-month ordinary profit of JPY 966 million (margin 9.7%) against JPY 1,923 million for all of FY12/25 (margin 13.0%), after the November 2025 partial termination of a lithium-ion cathode OEM contract and weaker hydrogen-storage alloy demand, partly offset by electronic-component materials; FY12/26 underlying plan about +JPY 1.2 billion. Battery-materials investment continues in the plan despite the termination.

内部研究类型：profit_pool_weakness

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The segment meant to be the second growth engine is shrinking; the plan's step to JPY 10 billion needs it to recover or be replaced.

不确定性（原始记录）：The OEM exposure is known only through the termination notice; replacement products or volumes are not disclosed.

研究原文：At 30 September 2026 total assets were JPY 94,404 million and net assets JPY 70,119 million (equity ratio 74.3%; 76.0% at FY12/25). The balance sheet is conservative but not net cash: interest-bearing debt of about JPY 14.2 billion exceeds cash of JPY 5.7 billion; investment securities JPY 14.5 billion (JPY 19.2 billion at FY12/25 before the write-off flowed through), land JPY 5.3 billion at cost without appraisal in the evidence set, inventories about JPY 26.4 billion. Nine-month comprehensive income was +JPY 615 million against a net loss of JPY 1,380 million because unrealised securities gains cushioned equity, so BPS (about JPY 562) sits barely below FY12/25's JPY 569.17; market-value equity ratio was 47.8% against 76.0% book at FY12/25. Solvency is not the question in this register; allocation is.

内部研究类型：balance_sheet_structure

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Book value is sturdier than the P&L suggests; part of why 0.67x book is not 0.67x liquidation value. Securities gains are not cash and do not fund the dividend directly.

不确定性（原始记录）：Land hidden reserves are plausible but unquantified; securities are marked to market and will move with equities; year-end debt and securities balances not yet known.

研究原文：Cash conversion runs counter-cyclical to profit quality. In the upcycle, H1 FY12/26 operating cash flow fell to JPY 2,494 million from JPY 9,224 million even as operating profit rose 164.5%, because inventories and receivables absorbed cash as prices rose; investing outflow of JPY 3,607 million made free cash flow negative; construction in progress tripled to JPY 3,067 million by June and JPY 3,447 million by September on the Kashima ash-furnace expansion. In the trough, FY12/25 operating cash flow of JPY 14,569 million (more than ten times net profit of JPY 1,418 million, helped by depreciation, the non-cash equity-method loss and working-capital release) covered capex of about JPY 4,828 million, the JPY 4,031 million buyback and the dividend, with the equity ratio at 76.0%. Across a full cycle the machine is cash-positive; within a year it can be either.

内部研究类型：cash_conversion_mechanism

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Dividends and buybacks are safest exactly when reported profit looks worst (trough cash release); the upcycle tests cash cover, not profit.

不确定性（原始记录）：Half-year cash flows are seasonally loaded; the FY12/26 annual cash-flow statement against the JPY 11,000 million operating-profit forecast is the clean test.

研究原文：Working capital is structurally long: inventories of about JPY 26.4 billion (finished products JPY 13.3 billion, raw materials JPY 13.1 billion at September 2026; JPY 17.5 billion and JPY 10.1 billion at June) equal roughly a third of annual sales, which is why price turns reach reported profit through inventory valuation before they reach cash. Inventory is both the profit amplifier and the first loss absorber at a downturn.

内部研究类型：working_capital_structure

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：A price reversal hits inventory valuation and margin before it hurts cash; inventory book value is not liquid downside value without evidence of saleability.

不确定性（原始记录）：Inventory composition by product, ownership, costing basis and realizability haircut are not disclosed beyond aggregates; Q4 inventory direction unknown.

研究原文：Strategic investment is large relative to the P&L: JPY 45-50 billion planned through 2030 (roughly triple the previous plan's pace) against annual operating profit of JPY 5-11 billion, plus R&D of JPY 3.2 billion in the 9th plan and DX spend of about JPY 4 billion, concentrated on ferroalloy competitiveness, ash-recycling capacity and functional materials. Execution shows first in construction in progress (JPY 691 million at Dec 2025 to JPY 3,447 million at Sep 2026 on the Kashima ash-melting expansion announced November 2025) and a CO2 target of at least 45% reduction by FY2030 (Scope 1+2 vs FY2015) for an electric-furnace producer; the Tokushima gas-engine installation (December 2025) is one practical step.

内部研究类型：capital_investment

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：If recycling returns hold near the current 30% margin, today's CIP is tomorrow's trough protection; if not, the investment envelope strains cash and ROIC.

不确定性（原始记录）：Project-level returns and total programme cost are not disclosed; CIP conversion timing and CO2 cost are unquantified.

研究原文：Shareholder returns are formulaic: about 40% of 'underlying net profit' (underlying ordinary profit x 0.7), a minimum annual dividend of JPY 11 from FY12/26 (raised from JPY 10 because the share count shrank), and buybacks as a flexible complement. FY12/25 paid JPY 12 (6.00 + 6.00 including a JPY 1 centenary dividend), a 112.1% payout on reported EPS of JPY 10.70 but unremarkable on the underlying basis (about JPY 1.48 billion, broadly what JPY 12 costs on about 125 million shares). The FY12/26 forecast was raised to JPY 17.00 (interim 5.50 resolved, year-end 11.50 from an initial JPY 13.00) against forecast reported EPS of JPY 4.01, a payout above 400%, which is an artefact of the affiliate write-down, not a policy break. Cross-check (synthesis arithmetic): 40% of about JPY 5.25 billion (JPY 7.5 billion x 0.7) is about JPY 2.1 billion, close to JPY 17 x about 125.4 million shares (about JPY 2.13 billion). Forecast yield 4.53% at JPY 375.

内部研究类型：shareholder_return_policy

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Holders are paid to wait through the Q4 verification window; the underlying story reaches holders before the reported story catches up. The floor makes the minimum return durable while the variable part rides the cycle.

不确定性（原始记录）：The year-end JPY 11.50 remains a forecast until the February 2027 resolution; no formal DOE or payout ceiling is stated; the 10th plan may formalise or change the formula.

研究原文：Capital actions are material: the company bought back about JPY 4,031 million of stock between August and October 2025 (ToSTNeT-3, completed 28 October 2025), held 12,597,919 treasury shares (9.17% of 137,386,472 issued) at the FY12/25 year-end, and cancelled 11,948,808 shares (8.69% of issued) on 21 August 2026, leaving 125,437,664 shares issued; capital surplus fell from JPY 17,484 million to JPY 13,829 million as the cancellation was absorbed. Financing is debt-funded working capital with no equity issuance; per-share metrics have been repeatedly lifted by retiring roughly a tenth of the share count.

内部研究类型：capital_action

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The same dividend pool buys a higher per-share amount after cancellation; per-share outcomes compound even when the cycle does not cooperate, and the float is shrunk.

不确定性（原始记录）：Buyback price versus normalised value is not evaluated beyond the discount to book; future buyback cadence is discretionary.

研究原文：Nippon Steel is the largest shareholder at 24.29% (FY12/25 register ex-treasury; Master Trust Bank 11.41%, Japan Custody Bank 2.79%, business-partner association 2.00%, Nittetsu Mining 1.68%, employee association 1.50%). The stake triggers the annual TSE 'matters concerning controlling shareholders' notice (28 March 2024, 28 March 2025, 27 March 2026) and the controlling-shareholder framework is live, not theoretical. The takeover-defence plan was abolished on 29 January 2026, removing a minority-protection layer just as the parent axis strengthens; governance runs under an audit-and-supervisory-committee structure. The axis supplies governance and likely raw-material and customer channels, but commercial terms are not disclosed.

内部研究类型：control_structure

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Minorities ride on the parent axis's terms; alignment is assumed, not contracted; float scarcity cuts both ways (thin sponsorship up, thin bids down).

不确定性（原始记录）：Trust-account holdings are custodial; any change in Nippon Steel's stake is visible only through EDINET large-holder filings; the inference that the axis is strengthening is interpretive.

研究原文：Governance and leadership change together at the 2026 year-end: President Aoki Yasushi becomes director and advisor without representative rights on 1 January 2027; Kobayashi Jiro (born 1965, Nippon Steel since 1988, head of its European office in 2013 and of its raw-materials divisions 2017-2023, Nippon Denko director since March 2024, holds 25,900 shares) becomes representative director and president, timed to the 10th medium-term plan to be published within 2026. Group alignment and execution risk rise together.

内部研究类型：governance_event

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Succession is the plan-launch mechanism: execution credibility rises with a raw-materials specialist, but the governance discount may not fall with it.

不确定性（原始记录）：The 10th plan's content and any governance-framework changes are not yet public; the president's incentives and board deliberation detail are not disclosed.

研究原文：The 9th medium-term plan (FY2024-27) targets FY12/27 sales of JPY 95.0 billion, ordinary profit of JPY 10.0 billion and ROE 10% (2030 vision: sales of at least JPY 110 billion, ordinary profit at least JPY 13 billion, ROE at least 10%), from an FY2023 base of sales JPY 76.4 billion, ordinary profit JPY 2.5 billion (underlying JPY 5.5 billion) and ROE 6%. Against FY12/25 actuals (JPY 77.3 billion, JPY 2.7 billion, ROE 2.0%) and the FY12/26 forecast (sales JPY 81.1 billion, underlying ordinary profit about JPY 7.5 billion), the FY12/27 step requires sales +17% and about JPY 2.5 billion more ordinary profit; most of that step must come from the overseas turnaround, functional-materials recovery and ash-recycling expansion rather than price, and it implicitly assumes the safeguard premium largely survives. The 10th plan, due within 2026, will confirm or reset it.

内部研究类型：plan_target_gap

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The re-rating case built on plan delivery stands or falls with this target set; a cut would remove the main bridge from 6.3x underlying to a higher multiple.

不确定性（原始记录）：The 10th plan text is unpublished; the sales step of +17% is large against a flat-to-declining domestic steel pond.

研究原文：Capital efficiency is below target and near cost of equity on the best measure: ROE was 2.0% in FY12/25 and 4.5% in FY12/24 (EPS JPY 10.70 vs JPY 24.35; operating margin 6.7% vs 8.8%; ordinary profit to assets 3.0%), against the plan target of 10% (FY12/27 and 2030). On underlying profit the run-rate ROE is roughly 7-8%, below target and near the cost of equity. The earnings base halved in a year without balance-sheet stress. The company's own cost-of-capital estimate and ROIC are not disclosed (ROIC approximated through ROE).

内部研究类型：capital_efficiency_judgement

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Until ROE clears the cost of capital, trading below book is defensible; the 10% target requires the JPY 10.0 billion ordinary profit and no new equity drag.

不确定性（原始记录）：The underlying-profit ROE of 7-8% is a Research approximation; FY12/26 ROE print and the 10th plan's capital-efficiency language are pending.

研究原文：Valuation snapshot at the research cut-off (price at the 6 November 2026 close): share price JPY 375; market capitalisation JPY 47.04 billion (x 125,437,664 shares issued after the August 2026 cancellation); forecast PER 93.5x on FY12/26 forecast EPS of JPY 4.01 (aggregator shows 93.66x); actual PBR 0.66x on FY12/25 net assets per share of JPY 569.17 (0.67x on about JPY 562 at September 2026; aggregator PBR 0.67 with a 0.28-1.9x range since 2009); ROE 2.0% (FY12/25 actual); forecast dividend yield 4.53% on JPY 17.00; EV/EBITDA 5.9x (EV JPY 51.73 billion = market cap plus net interest-bearing debt of JPY 4.69 billion at 30 September 2026, being short-term borrowings JPY 5,000 million + current portion JPY 1,550 million + long-term JPY 3,887 million less cash JPY 5,747 million; EBITDA JPY 8.83 billion = FY12/25 operating profit JPY 5,161 million + depreciation JPY 3,665 million). Basis note: multiples mix FY12/25 actual denominators and FY12/26 forecast denominators; PER is distorted by the affiliate-loss-depressed EPS.

内部研究类型：valuation_snapshot

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The headline multiples are mutually inconsistent by construction (94x PER beside 0.66x PBR and 5.9x EV/EBITDA); each reflects a different period and profit basis.

不确定性（原始记录）：Point-in-time snapshot at the 6 November close; EBITDA uses FY12/25 actuals while FY12/26 operating profit is forecast near JPY 11,000 million; per-share denominators differ across multiples.

研究原文：Valuation depends on which profit language is used. At JPY 375 the market capitalisation of about JPY 47.0 billion is smaller than book equity (JPY 70.1 billion at September 2026), about 6.3x the JPY 7.5 billion underlying ordinary-profit forecast, about 31x the JPY 1.5 billion reported ordinary-profit forecast and about 94x forecast reported EPS; yield 4.53%. The market prices the reported-earnings hole and the governance axis, not the underlying run-rate or the plan. The discount has persisted for years (PBR range 0.28-1.9x since 2009; the 0.67x sits in the lower half), so re-rating needs delivery, not argument; the gap between market-value equity ratio (47.8%) and book (76.0%) is the standing discount in one number.

内部研究类型：valuation_assumption_set

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Downside is cushioned by book and yield; upside requires the February print and the 10th plan.

不确定性（原始记录）：Underlying-based multiples use management's non-GAAP series; the PBR anchor includes the securities cushion; no analyst consensus series was pulled.

研究原文：The market's first verdict on the Q3 print (6 November 2026) was a 5.3% fall to JPY 375: the document that carried an operating-profit beat (+131.6%) and an unchanged forecast also carried a nine-month ordinary loss, and the loss was what was priced. Credit data show retail positioning heavily one-sided into the autumn: margin longs of 1,750,300 shares against 26,600 short (65.8x, mid-July 2026). Retail believed early; institutions appear to be waiting for Q4. A Q4 beat is likely to be believed late and priced fast; a miss confirms the hole narrative.

内部研究类型：market_expectation_state

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Expectation asymmetry favours the patient holder if delivery comes; one clean Q4 print re-prices the underlying language.

不确定性（原始记录）：Expectations are inferred from price action and the margin balance (mid-July, stale by November); no analyst consensus was pulled.

研究原文：Normalised earnings power is best read on the underlying ordinary-profit series: about JPY 5.5 billion (FY12/23), 5.2 (FY12/24), 5.3 (FY12/25) and about 7.5 billion forecast for FY12/26 (initially about 6.0 billion; H1 about 4.5 billion vs 2.2 billion; nine months about 5.1 billion vs 3.2 billion) - a normalised band of roughly JPY 5-7.5 billion before plan delivery, against reported prints from -6.4 to +10.4 billion. The FY12/26 forecast already includes upcycle pricing, so the top of the band is not a trough number. At about 6.3x the top of the band the market pays for delivery of the band, not the plan.

内部研究类型：normalized_earnings_judgement

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Separates structural earning power from price-rent: the lower half of the band is the trough-tested part, the upper half rests on safeguard pricing.

不确定性（原始记录）：Underlying is management's non-GAAP construction (ex inventory effects and one-offs); the company does not publish a full quarterly bridge; continuation of underlying disclosure under the 10th plan is not assured.

研究原文：Downside protection is asset- and annuity-based, not earnings-based: equity ratio 74.3%, JPY 14.5 billion of securities, land at cost, the JPY 11 dividend floor, ash-recycling and hydro profits that persist through alloy troughs, and a demonstrated trough anatomy (FY12/19) that the balance sheet absorbed. Protection has been tested twice in a decade and held; the realistic downside is a return to the JPY 225-300 base zone (0.3-0.5x book trough framework), not distress. Counter-weight: a safeguard lapse combined with a price collapse has no exact precedent in the company's record.

内部研究类型：downside_structure

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：If any thesis leg fails, the correct action is to re-price to the trough framework rather than to average down on narrative.

不确定性（原始记录）：Protection is constructed from the company's own trough history; the 0.3-0.5x book trough range is a Research framing, not a disclosed fact.

研究原文：Where the accounts mislead (each anomaly pushes reported profit below the operating business in 2026; the reverse held at the FY12/22 peak): the parent special loss (JPY 7,877 million) differs from the consolidated equity-method addition (JPY 7,180 million, beyond the running loss share, JPY 7,733 million in total at H1); inventory valuation sits inside gross profit; non-deductible equity-method losses leave the tax line heavy against a pre-tax loss; other comprehensive income absorbs equity-level securities gains (nine-month comprehensive income +JPY 615 million vs net loss -JPY 1,380 million); and ordinary-level segment profit hides the affiliate inside ferroalloys. This is a quality-of-evidence catalogue, not a fraud presumption.

内部研究类型：accounting_translation_map

内部来源标记：COVERAGE35

普通股东含义（原始记录）：Readers who screen on reported EPS misprice this stock in both directions across the cycle.

不确定性（原始记录）：Catalogued from the filings' own notes; no forensic re-audit; the FY12/26 securities report (March 2027) may add new one-off categories.

研究原文：European safeguard measures on ferroalloy imports are currently the group's largest single profit variable, larger than any operating lever management controls: management attributes the price rise mainly to safeguards, which lifted Japanese export demand and domestic prices (European HC-FeMn USD 1,175/t in 2025 to USD 1,335/t in Q3 2026; US benchmark USD 1,345-1,361/t). Safeguards raise the import price ceiling; when they lapse the domestic premium compresses toward import parity. The measures have a reviewable, finite life and their post-2026 design is not yet public. Waste law supports ash-recycling volumes, a policy floor under the second engine.

内部研究类型：policy_rent

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Part of the cycle's premium is a policy rent, and policy rents expire; it is a plan-level exposure, not a market footnote.

不确定性（原始记录）：EU safeguard review outcomes and import-price data are not available; no named-peer benchmarking or import-price series was pulled.

研究原文：Alloy demand is derived from crude-steel output and safeguard-shaped trade flows into Europe and the US, not from final consumption; domestic steel demand is structurally flat-to-declining, so growth must come from consolidation, capacity and mix rather than market beta. Management's commentary attributes the H1 and Q3 profit mix more to price and valuation than to volume. Ash demand follows waste volumes, far steadier than steel; aqua solutions deliveries are project-timed; hydropower generation is stable.

内部研究类型：end_market_dependency

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The same steel tonnage produces alloy profit only while safeguard premiums hold domestic prices above import parity; the blend with waste demand is the stabiliser.

不确定性（原始记录）：Customer names and volumes are not disclosed; FY12/27 steel-demand outlook awaits the 10th plan.

研究原文：All ferroalloy production depends on imported manganese ore priced off the USD per Mn% benchmark (USD 4.5 in 2025, 5.4 in Q2 2026, 4.8 in Q3), on furnace electricity and on the yen (157-160 realised in H1; H2 assumption 155 per dollar; about 159 in Q3). Ore and power costs enter COGS with a lag while alloy prices float; the lag is the margin cycle. A weak yen inflates both alloy yen-prices and ore yen-costs, with the price side winning while safeguards hold. A simultaneous ore spike and yen strength is the classic margin squeeze; FX reversal compresses the same margin it expanded.

内部研究类型：input_dependency

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The spread between the ore price and the safeguard-supported alloy price is the group's core alloy margin; hedging and supplier terms are not disclosed.

不确定性（原始记录）：Supplier identities, contract terms and FX sensitivities are not disclosed; sensitivity is inferred from market tables.

研究原文：The century-old franchise (founded October 1925, centenary marked October 2025) has survived by consolidation: it absorbed listed peer Chuo Denki Kogyo (share exchange 2014, merged 1 July 2024), stopped the Kashima high-carbon ferromanganese furnace in 2021-22 to concentrate capacity, and now concentrates manganese alloys at Takaoka and Kashima with the ex-Chuo works at Myoko and Wakayama (plus Tokushima and Koriyama). It is the domestic reference ferroalloy supplier; smelting know-how, waste-treatment permits and in-house hydro assets are the structural moat, while the alloy moat is cost-curve position only.

内部研究类型：industry_position

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Consolidation gives pricing resilience in troughs and operating leverage in upcycles, which is why the upcycle lands on this P&L with full force.

不确定性（原始记录）：No independent market-share series was pulled; position is inferred from consolidation history; any new capacity or exit moves by remaining rivals are unknown.

研究原文：The dated catalyst path is dense for two quarters, then reverts to cycle-watching: the 10th medium-term plan (within 2026); the presidency change (1 January 2027); the FY12/26 annual print with the implied-Q4 verdict (February 2027); the year-end dividend resolution (JPY 11.50 forecast); the March 2027 controlling-shareholder disclosure. Each catalyst resolves one layer of the two-language problem: plan (targets), print (delivery), dividend (formula).

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）：Position sizing should respect the February binary; catalyst sizes are not estimable in advance and the Q4 implied numbers define the surprise axis.

不确定性（原始记录）：The 10th plan publication date is not announced beyond 'within 2026'.

研究原文：Thesis: the FY12/26 reported hole is a one-layer accounting event over an improving underlying business, and the discount (about 0.67x book, about 6.3x underlying ordinary profit) overprices it; the thesis stands on three legs - underlying delivery of the back-loaded year, plan continuity under the new president, and a zeroed affiliate - each with its own dated test. Falsifiers: Q4 ordinary profit far below the implied JPY 2,746 million without a ferromanganese price collapse (the 'underlying recovery' framing fails); the 10th plan cutting the FY12/27 JPY 10.0 billion ambition (the plan-delivery re-rating case fails); new cash support to Pertama; a monthly close back below the 2025 low band. If any leg fails, re-price to the trough framework (0.3-0.5x book) rather than average down on narrative.

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）：Gives the shareholder a short evidence fuse: all three legs are tested within two quarters.

不确定性（原始记录）：Confidence about 0.75 as stated by Research; Q4 print, 10th plan text and FY12/27 Q1 underlying disclosure are all pending.

研究原文：Strongest counter-evidence to the 'underlying recovery' framing, retained distinct from principal risk: (1) the FY12/26 margin is partly policy rent from European safeguards with a finite, reviewable life; (2) the reported series whipsaws and the FY12/19 anatomy has recurred, and FY12/22's JPY 10,367 million peak fell 76% the next year; (3) the forecast is back-loaded, requiring about JPY 2.4 billion of underlying profit in Q4 against about JPY 0.6 billion in Q3; (4) H1 free cash flow was negative as working capital absorbed the upcycle; (5) functional materials is shrinking after the cathode OEM termination; (6) the takeover-defence plan was abolished as the parent axis tightens; (7) the underlying series is management's own construction; (8) the plan's FY12/27 step needs +17% sales in a flat pond. Offsetting: Q3 standalone profit was clean, the affiliate P&L channel is closed and the balance sheet absorbed two trough events.

内部研究类型：counter_evidence_stack

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：The discount is partly the correct price for structure if these items cluster; they are why the stock is not simply 'cheap on underlying'.

不确定性（原始记录）：Items are not independent (several share the safeguard and price-cycle driver), so they should not be counted as eight separate negatives.

研究原文：What is durable here is not any year's profit but the position: the consolidated domestic manganese-alloy franchise built over a century, a permit-based recycling annuity earning near 30% ordinary margins, hydro assets, a 74%-equity balance sheet with a listed-securities cushion, and a shareholder-return formula that pays on underlying profit and has already survived a reported-loss year. What is not durable is the reported profit level itself: the statutory series has printed -6.4, +10.4 and +1.5 billion yen (forecast) within seven years on the same asset base, because manganese prices, inventory valuation and equity-method affiliates all strike the same line. The investment question resolves into a race between two dated processes: delivery of the back-loaded FY12/26 forecast and the FY12/27 plan (JPY 10.0 billion) under a new president, against the finite life of the safeguard premium that created the current margin. Both resolve within two quarters; the falsifiers are dated (the February 2027 print, the 10th plan within 2026, and the monthly price sensors at JPY 559, JPY 400 and JPY 334).

内部研究类型：structural_durability_judgement

内部来源标记：OPEN_DISCOVERY / COVERAGE35 / TECHNICAL

普通股东含义（原始记录）：Tells the shareholder what can be underwritten (position, annuity, balance sheet, formula) and what cannot (reported profit level, safeguard premium).

不确定性（原始记录）：Race outcome unknown; the price sensors named belong to the protected Technical structure and are cited here only as dated falsifier references.

研究原文：Claim-level price-record observations retained as Research Truth (the protected Technical structure remains the authoritative Technical source; this node does not interpret or restate Technical State): on the monthly record the stock lived mostly between JPY 100 and 400, and the 2025-26 advance to a May 2026 high of JPY 559 is the first sustained trade above the JPY 400 shelf since 2008; the retracement since June 2026 has so far held above the rising 24- and 60-month averages. The May 2026 spike and June reversal (close JPY 434 from JPY 547) on volume of 23.3 million shares (about 2.4x the 2024 monthly norm) over a June-October base (closes JPY 390-463). At October 2026: close JPY 412 against MA12 JPY 424, MA24 JPY 357, MA60 JPY 334, with the 6 November daily close of JPY 375 testing the MA24 zone. About 45% of monthly closes since 2001 are below JPY 300 and about 9% above JPY 500.

内部研究类型：price_record_claim_observations

内部来源标记：TECHNICAL

普通股东含义（原始记录）：Gives the price context within which the fundamental delivery race will be read; sponsorship above JPY 500 is historically thin.

不确定性（原始记录）：Prices are raw unadjusted prints from a mirror with MISSING_TAIL status accepted through 27 October 2026; November is covered only by the 6 November daily close.

研究原文：Will Q4 FY12/26 deliver the implied ordinary profit of about JPY 2,746 million (inventory effects and ferromanganese prices permitting), or will the full-year forecast be cut at the annual print? Priority 1. Opposite answer: a miss without a price collapse falsifies the underlying-recovery framing. Resolving evidence: the FY12/26 annual release (February 2027), Q4 inventory balance and Q4 market-table prints.

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 the reported hole is verified as accounting noise.

不确定性（原始记录）：Resolves only with the unprinted February 2027 annual release; no current public path exists.

研究原文：With the Pertama stake written down to zero, is the exposure truly closed, or can the affiliate still consume cash through loans, guarantees or further funding? Priority 1. Opposite answer: new cash support would reopen the channel and falsify the 'zeroed affiliate' leg. Resolving evidence: FY12/26 securities report notes (March 2027) on guarantees and commitments; any new TDnet notice.

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 residual cash risk of the largest value destroyer in the reported P&L.

不确定性（原始记录）：Not disclosed in filings reviewed; the annual report is not yet public.

研究原文：If European safeguard measures lapse or are redesigned after 2026, how much of the domestic ferroalloy price premium, and of the FY12/26 gross-margin jump from 14.8% to 22.3%, survives? Priority 1 (raised twice in Research). Opposite answer: a lapse compresses margin toward import parity and undercuts the FY12/27 plan. Resolving evidence: EU safeguard review announcements, company risk disclosures and import-price 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）：The single largest external risk to the plan and to normalised margin.

不确定性（原始记录）：The post-2026 design is not public; magnitude cannot be estimated from company disclosure alone.

研究原文：What will the 10th medium-term plan (due within 2026) set for FY12/27-29, and does it keep the JPY 10.0 billion ordinary-profit ambition and the current dividend formula? Priority 1-2 (stated twice in Research). Opposite answer: a cut removes the plan-delivery re-rating case; a formula change alters the shareholder-return anchor. Resolving evidence: the plan publication.

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 bridge from underlying run-rate to a higher valuation.

不确定性（原始记录）：Unpublished at cutoff; publication date only 'within 2026'.

研究原文：Can the ash-recycling segment's roughly 30% ordinary margin persist as molten-metal prices normalise, and what does the Kashima melting-furnace expansion add to capacity and profit? Priority 2. Opposite answer: margin normalisation would remove the trough annuity and weaken the capex case. Resolving evidence: segment results through FY12/27 and disclosure on furnace capacity.

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 highest-margin pool and the destination of growth capital.

不确定性（原始记录）：Project returns and capacity increment are not disclosed.

研究原文：Will functional materials recover from the lithium-cathode OEM contract termination and weak hydrogen-storage alloy demand, or does the segment shrink toward its electronic-materials core? Priority 2. Opposite answer: shrinkage would raise the share of the plan step that must come from ash recycling and the overseas turnaround. Resolving evidence: FY12/26 annual segment note (February 2027) and 10th plan segment strategies.

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 a second growth engine exists besides recycling.

不确定性（原始记录）：Replacement product or volume disclosure absent.

研究原文：Does Nippon Steel's 24.29% stake, the abolished takeover-defence plan and the incoming president from its raw-materials division change capital policy or the company's eventual group positioning? Priority 3. Opposite answer: a stake or positioning change would reprice the governance discount in either direction. Resolving evidence: March 2027 controlling-shareholder disclosure, EDINET large-holder filings and the 10th plan's positioning language.

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）：Governance is part of why the multiple stays below book.

不确定性（原始记录）：Interpretive; no stated intention by Nippon Steel in the evidence set.

研究原文：Will the break above the JPY 400 shelf hold through the February 2027 annual print, or fade like the 2017 spike back into the lifelong base? Priority 3. This is the price-side question posed by the protected Technical Question and is answered only by future monthly price behavior and the results-day reaction; it must not be answered by fundamental narrative.

内部研究类型：research_debt

内部来源标记：OPEN_DISCOVERY / TECHNICAL

普通股东含义（原始记录）：Determines whether the altitude is accepted by the market.

不确定性（原始记录）：No fundamental evidence can resolve it; only subsequent price behavior.

研究原文：Evidence-role limits that qualify judgments: no named-peer benchmarking or independent market-share series was pulled; ore supplier identities and contract terms, customer names and volumes, and product-level revenue and R&D pipeline are not disclosed; ROIC and the company's own cost-of-capital estimate are not published; no analyst consensus series was pulled; land is carried at cost without appraisal; the underlying-profit series and segment underlying splits are management constructions; the market-data snapshot comes from an aggregator (price, margin balance) while price history is a mirror with MISSING_TAIL status; part of the interim-briefing evidence comes from a transcript service rather than primary filings.

内部研究类型：evidence_boundaries

内部来源标记：COVERAGE35

普通股东含义（原始记录）：Prevents non-disclosure from being read as absence of risk or strength; several judgments (moat, competitive position, ROIC) are structural characterisations only.

不确定性（原始记录）：Each limit becomes resolvable only with annual securities report disclosures or third-party data not obtained.

研究原文：Policy to producer: safeguards lifted European high-carbon ferromanganese and pulled Japanese exports, raising domestic prices and the FY12/26 gross margin; trade policy is currently the largest single profit variable and has a reviewable, finite life.

内部研究类型：causes

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Input dependence: ore priced off the USD/Mn% benchmark (USD 5.4 in Q2 2026, 4.8 in Q3) feeds all production and, with the yen, sets the cost floor under every alloy tonne; the spread between ore cost and safeguard-supported alloy price is the core margin.

内部研究类型：conditions_margin_spread

内部来源标记：OPEN_DISCOVERY

研究原文：End-market relation: ferroalloy volumes and prices follow crude-steel production and safeguard-shaped trade flows; derived demand turns into alloy profit only while the premium holds domestic prices above import parity.

内部研究类型：derived_demand_conditions

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Price rises reach reported profit through inventory valuation before they reach cash; rising prices make inventories and receivables absorb cash.

内部研究类型：transmits_through_inventory

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Long working capital converts price falls into cash release (FY12/25) and price rises into cash absorption (H1 FY12/26), making cash quality counter-cyclical to profit quality.

内部研究类型：causes_cash_divergence

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Booking the affiliate inside the ferroalloy segment makes the segment and consolidated ordinary result diverge from the underlying business: the two-layer result of operating surge over a reported loss.

内部研究类型：creates_profit_language_divergence

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Equity-method affiliate relation: the stake written down in full in Q2 FY12/26 (equity-method loss JPY 7,733 million in H1) leaves a zero carrying value; the P&L channel is closed and only undisclosed cash channels (loans, guarantees) could reopen it.

内部研究类型：closes_pnl_channel

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Control sits with the Nippon Steel axis (24.29%, annual controlling-shareholder disclosures) and is tightening; strategy, succession and capital policy operate inside the group frame, and the minority question is the terms of coexistence - a structural component of the 0.67x book valuation.

内部研究类型：contributes_to_discount

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Capital relationship: Nippon Steel and stable association holders anchor the register; minorities receive the formula dividend (JPY 17 forecast) and were counterparties to the 2025 buyback of about JPY 4,031 million; the float's return depends on the formula's survival and buyback cadence, both of which held through a reported-loss year.

内部研究类型：anchors_register_for

内部来源标记：OPEN_DISCOVERY

研究原文：Customer relationship: municipalities and industrial waste generators supply incineration-ash treatment volumes under waste law with molten-metal recovery; treatment demand does not follow metal prices, which is why the segment earned JPY 2,589 million in nine months while the alloy segment reported a loss.

内部研究类型：underwrites

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Forecast concentration: about JPY 2,746 million of ordinary profit and JPY 3,685 million of operating profit must land in Q4 FY12/26; delivery validates the underlying-recovery framing, a miss revalidates the reported-loss narrative the market prices.

内部研究类型：concentrates_verdict

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Capacity investment: construction in progress rose from JPY 691 million (Dec 2025) to JPY 3,447 million (Sep 2026) on the Kashima ash-melting expansion; if it earns at the segment's current 30% ordinary margin, trough earnings rise structurally.

内部研究类型：expands_capacity_of

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Policy horizon: the domestic ferroalloy premium underpinning the FY12/26 margin jump depends on measures whose next design is not yet public; the FY12/27 plan's JPY 10.0 billion ambition implicitly assumes the premium largely survives, making the safeguard review a plan-level event.

内部研究类型：implicit_assumption_of

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Succession: a Nippon Steel raw-materials executive (Nippon Denko board since March 2024) becomes president on 1 January 2027, launching the 10th medium-term plan; the same event deepens parent alignment and delivers the document that confirms or resets the FY12/27 targets minorities are pricing.

内部研究类型：launches_and_owns

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：The dividend is anchored to the underlying-profit formula, so the JPY 17.00 forecast coexists with reported EPS of JPY 4.01; the apparent payout above 400% is an artefact of the write-down.

内部研究类型：anchors

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：Buyback-and-cancel and the JPY 11 dividend floor are two halves of one per-share policy: cancelling 8.69% of shares raises the per-share payout from the same dividend pool.

内部研究类型：reinforces_per_share

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：FY12/25 trough operating cash flow of JPY 14,569 million funded the JPY 4,031 million buyback and the dividend while capex was about JPY 4,828 million and the equity ratio stayed at 76.0%.

内部研究类型：funds

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：The securities cushion kept book value near JPY 562 per share through three loss-making reported quarters, so a 0.67x book multiple is not 0.67x liquidation value; securities gains are not cash.

内部研究类型：cushions_book_value_behind

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：A shrinking functional-materials segment raises the share of the FY12/27 profit step that must come from the overseas turnaround and ash-recycling expansion.

内部研究类型：weakens_delivery_path

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：The overseas operation returning to underlying profit in FY12/27 (from about -JPY 0.4 billion in FY12/26) is one of the steps the JPY 10.0 billion ordinary-profit target needs.

内部研究类型：is_precondition_for

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：The counter-evidence stack conditions the thesis: confidence rests on three dated legs, and clustering of policy, cycle and forecast-shape items could jointly break them.

内部研究类型：qualifies

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：The repeated impairment/inventory/affiliate anatomy (FY12/19 and FY12/26) sits on top of the price cycle and amplifies its reported-profit swings.

内部研究类型：reinforces_pattern

内部来源标记：OPEN_DISCOVERY / COVERAGE35

研究原文：The Q4 delivery question is the open uncertainty on the forecast shape.

内部研究类型：resolves_when_printed

内部来源标记：OPEN_DISCOVERY

研究原文：Undisclosed guarantees or funding qualify the closure judgment.

内部研究类型：qualifies

内部来源标记：OPEN_DISCOVERY

研究原文：The unknown post-2026 safeguard design qualifies the durability of the premium.

内部研究类型：qualifies

内部来源标记：OPEN_DISCOVERY

研究原文：The unpublished 10th plan is the open reset of the plan targets and formula.

内部研究类型：qualifies

内部来源标记：OPEN_DISCOVERY

研究原文：Margin persistence and furnace capacity are unresolved for the second engine.

内部研究类型：qualifies

内部来源标记：OPEN_DISCOVERY

研究原文：Recovery or shrinkage of functional materials is unresolved.

内部研究类型：qualifies

内部来源标记：OPEN_DISCOVERY

研究原文：Whether the group axis changes capital policy or positioning is unresolved.

内部研究类型：qualifies

内部来源标记：OPEN_DISCOVERY

研究原文：The base-hold question corresponds to the Technical Question held in the protected Technical structure; it is referenced as a whole and not answered or redefined here.

内部研究类型：poses_question_to_whole

内部来源标记：OPEN_DISCOVERY / TECHNICAL

研究原文：The fundamental expectation finding (operating beat and unchanged forecast sold 5.3% on the reported loss) is available as event context for the price-response and timing study; it reuses existing evidence identities and does not enter or alter the Pure Technical State, which a non-confirming or confirming fundamental reading may never rewrite.

内部研究类型：event_context_for_response_study

内部来源标记：OPEN_DISCOVERY / TECHNICAL

研究原文：The price-record claim observations defer in interpretation and completeness to the protected Technical structure as a whole.

内部研究类型：claim_identity_defers_to

内部来源标记：TECHNICAL

研究原文：Protected Technical State (RESEARCH_TRUTH.technical_state, dereference only)

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 consolidated manganese franchise with a recycling annuity, a strong balance sheet and a formula that pays on underlying profit｜Foundation before change: the position (domestic alloy consolidation, recycling permits, hydro, a 74%-equity balance sheet and securities cushion, a return formula tested through a reported-loss year) is durable; the reported profit level is not. The current upturn, the affiliate write-down and the plan are increments overlaid on this base, not substitutes for it. Self-repair has been demonstrated twice (FY12/19, FY12/26) by balance-sheet absorption, annuity profits and the dividend floor; irreversible-break risk would come from a safeguard lapse coinciding with a price collapse, which has no precedent in the record. Structural durability: what is durable is the position, not any year's profit; the investment question is a race between delivery and the finite safeguard premium.

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）：Holders own a durable position at below-book pricing with a formula-backed return, but must accept cyclical, policy-dependent reported earnings and capital efficiency near the cost of equity.

不确定性（原始记录）：The durability of the alloy moat is cost-position only; the recycling moat evidence is structural rather than contract-level; return on the investment envelope is unproven.

研究原文：Two-layer profit chain: policy-fed price rise to operating surge, inventory amplification, affiliate charge below the line｜European safeguards lift ferromanganese prices; a weak yen and easing ore widen the spread; inventory valuation gains amplify gross margin (14.8% to 22.3%) and operating profit (+131.6%); the same price rise absorbs cash into inventories and receivables, so operating cash flow falls while profit rises; below the operating line the full Pertama write-down turns reported ordinary profit negative while underlying profit roughly doubles. The unchanged forecast then back-loads the year into Q4.

内部研究类型：main_causal_chain

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Valuation must be run on the underlying language (about 6.3x) rather than the statutory print (about 94x EPS), but the underlying profit rests on a policy rent and valuation effects and is verified only when Q4 prints.

不确定性（原始记录）：Split of the profit rise among price, FX and inventory valuation is not disclosed by quarter; underlying series is non-GAAP.

研究原文：The balance sheet bridges the two profit languages｜The balance sheet is the bridge between the two profit languages: a 74% equity ratio, a JPY 14.5 billion securities portfolio whose unrealised gains kept book value near JPY 562 a share through three loss-making reported quarters, and inventories of about JPY 26.4 billion that translate price turns into reported profit before they translate into cash.

内部研究类型：cross_domain_mechanism

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Book value and the dividend floor protect holders while reported profit is distorted; cash, not profit, is the test in the upcycle.

不确定性（原始记录）：Securities values are market-sensitive; year-end balances unknown.

研究原文：Ash recycling fuses the margin, capex and downside theses｜The ash-recycling segment fuses the margin thesis, the capex thesis and the downside thesis: it is the highest-margin profit pool (30.3%), the destination of the growth capex now visible in construction in progress, and the annuity that protected trough earnings in FY12/23-25 - its Kashima expansion is the single most informative capital decision in the register.

内部研究类型：cross_domain_mechanism

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：If the expansion earns near the current margin, trough earnings rise structurally; if molten-metal prices normalise, the annuity and the capex case weaken together.

不确定性（原始记录）：Capacity and return of the expansion are undisclosed.

研究原文：The dividend formula is where the underlying language binds cash｜The dividend formula is the one place the underlying language is legally binding on cash: a JPY 17 forecast paid inside a reported-loss year, on a share count cut by 8.69%, is the mechanism by which the underlying story reaches holders before the reported story catches up.

内部研究类型：cross_domain_mechanism

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Holders receive formula returns (4.53% forecast yield) while waiting; the return survives only if underlying profit and cash cover hold.

不确定性（原始记录）：Year-end dividend resolution and any 10th-plan formula change are pending.

研究原文：The race: dated delivery legs versus the finite safeguard premium｜Three delivery legs (Q4 underlying delivery, plan continuity under the new president, zeroed affiliate) are tested within two quarters against a policy premium that has no disclosed design beyond 2026. Counter-evidence is retained as a contradiction rather than generic risk; valuation at about 0.67x book and 6.3x underlying prices the reported hole, so delivery re-rates it while failure re-prices toward the trough framework.

内部研究类型：thesis_architecture

内部来源标记：OPEN_DISCOVERY / COVERAGE35

普通股东含义（原始记录）：Binary-leaning payoff around the February 2027 print with book, yield and annuity cushioning the downside.

不确定性（原始记录）：All three legs are unresolved at the cutoff.

研究原文：Research Debt carried into the next Review｜Open questions each with materiality, opposite-answer consequence and resolving evidence; duplicated Research statements (10th plan, safeguard premium) are merged without losing either wording. Most resolve with the February 2027 annual release, the 10th plan and the March 2027 securities report; no reasonable current public path exists to resolve them earlier.

内部研究类型：research_debt_register

内部来源标记：OPEN_DISCOVERY / COVERAGE35 / TECHNICAL

普通股东含义（原始记录）：These determine whether the thesis legs hold; the first three carry the largest valuation consequences.

不确定性（原始记录）：No reopen of Research is warranted: remaining questions depend on future disclosures after the cutoff.