不二越 / NACHI-FUJIKOSHI (6474)

Signal panel

Investment thesis

1. Margin recovery is real and broad-based. Q3 FY2026 (Jun–Aug 2026) OP ¥4,688m, +94.1% YoY, OPM 6.9% (vs 4.1% LY); 9M OP ¥11,457m, +72.8% (source: filings/2026-11_Q3_短信_20261005.pdf p.1). The 9M OP bridge is price optimisation +¥4.2bn, operating rate +¥2.6bn, FX +¥1.9bn, cost-down +¥0.7bn, offset by materials −¥3.8bn and other costs −¥0.8bn — i.e. the lift is mostly self-help and volume, not a one-off (Q3 補足資料, 20261005_jp2.pdf).
2. Structural reform is the engine. Standard-bearing production is being consolidated, headcount is down (6,710 → 6,470 full-time staff in 9M), and the parts segment OP rose 60.8% on flat-ish sales. This is the same reform programme that produced FY2025's 47.3% OP recovery.
3. Orders are trending up and book-to-bill is above 1. Group quarterly orders rose from ¥57.8bn (FY2025 Q1) to ¥69.5bn (FY2026 Q3) against Q3 sales of ¥67.5bn (Q3 補足資料 chart). Order-driven segments (robots +24.4% YoY in Q3, hydraulic equipment +19.0%) are the growth engine; machine tools (−10.7%) and car hydraulics (−15.8%) are the laggards.
4. Guidance looks achievable with modest upside. FY2026 OP guidance of ¥15.3bn implies Q4 OP of ¥3.84bn, 18% below the Q3 run-rate; the H1/H2 plan (H1 OP ¥6.77bn, H2 ¥8.53bn) similarly assumes a Q4 slowdown (H1 presentation p.12). A repeat of the Q3 quarterly rate would put FY2026 OP near ¥16.1bn.
5. The value gap is in the assets, not the earnings. PBR 0.66 vs bearing peers 0.74–0.84; net cash-flow-positive; cross-shareholdings of ¥26.9bn (listed, end-FY2025) equal about 15% of equity and 20% of market cap — a potential source of buybacks/dividends and a lever on ROE. The company sold ¥3.7bn in FY2025 (¥3.1bn gain).

Why this is still only a Hold. The re-rating that these points justify has largely happened: the shares are 38% higher over 12 months and only 19.5% below the 52-week high. On the stock's own PER band the price sits slightly above the mean (z +0.36), and the implied EPS at the band-mean PER is ~¥351, i.e. essentially the company's own FY2026 forecast — no further revision is priced in, but neither is a re-rating. The PBR discount will not close until ROE clears the cost of equity, and management has set no such target.

Business and earnings

Nachi is a comprehensive machinery maker (総合機械メーカー): robots, cutting tools and machine tools (機械工具, 30.9% of FY2025 sales), bearings, hydraulic equipment and automotive hydraulics (部品, 61.1%), and special steel (その他, 8.1%) (feeds/segment_fy.json). Automotive is about 50% of sales, and overseas is 50.8% (Japan 49.2%, Asia 31.3%, Americas 14.8%, Europe 4.7%). The company's stated long-term milestones are an overseas sales ratio of 60% and a 10% operating margin (filings/2025-11_FY_有報_20260224.pdf, 経営方針, printed p.10).

Segment margins are recovering from a low base (FY2025 → 9M FY2026):

Special steel's jump is cyclical (higher volumes and price); the parts segment is the structurally important one, because it is where the standard-bearing consolidation and fixed-cost cuts land. Within 機械工具, robots are the growth story (+24.4% YoY in Q3) while work machine tools are shrinking (−10.7% YoY in Q3, and guided −10.7% for the full year). Within 部品, hydraulic equipment for construction machinery is recovering (+19.0% guided) while car hydraulics falls (カーハイドロリクス, −15.8% guided) — the visible EV/China drag.

Forecast test

Progress against the seasonal norm is only mildly ahead for operating profit and in line for sales and net profit: sales 75.4% vs 74.8% median (+0.65 pts), OP 74.9% vs 72.1% (+2.74 pts), RP 81.8% vs 70.1% (+11.68 pts), NP 88.4% vs 87.7% (+0.72 pts) (factsheet.md, stock_chart_prog_all.json). The RP flag is real but it reflects non-operating items (equity-method, FX), not operating strength.

Guidance was already raised once this year, in July, from OP ¥12.1bn to ¥15.3bn, and held at Q3 (guidance table in factsheet.md). The company's initial-guidance track record is poor in both directions — actual OP vs initial guidance: FY2020 −45.2%, FY2021 +84.0%, FY2022 +0.1%, FY2023 −32.1%, FY2024 −33.6%, FY2025 +13.6% — so a "conservative guidance" argument must be treated with care. The relevant read is that the July-raise level is being delivered and a modest Q4 beat is plausible from the run-rate.

One-offs: the 9M P&L includes ¥896m of structural-reform costs (special loss) and the Q3 segment note shows ¥331m of impairment in 部品 (filings/2026-11_Q3_短信_20261005.pdf pp.8–9). OP is clean: RP (¥10,873m) is below OP, so no large gain inflates the operating figure.

Consensus is stale and must not be treated as a forecast. IFIS shows two analysts, both neutral (3.00), with FY2026 OP ¥9.9bn / RP ¥8.9bn / NP ¥5.7bn and FY2027 OP ¥11.5bn / RP ¥10.5bn / NP ¥7.6bn (as of 2026-10-16, pages/report_est.html; feeds/cons_chart.json is a flat line, so the series carries no revision information). The FY2026 OP estimate is 35% below guidance and below 9M actual OP — arithmetically impossible as an updated number. The mean target of ¥4,570 (−16.9% vs price) is the average of a stale ¥3,300 and a ¥5,840 target dated 2026-07-29. Treat the analyst view as thin coverage, not as a signal of downside.

Financial strength and capital allocation

The balance sheet is sound. At end-Q3 FY2026 total assets were ¥348.75bn, equity ¥181.68bn, equity ratio 52.1% (from 51.5% at FY2025-end), interest-bearing debt ¥95.1bn, net D/E 0.33, current ratio 2.09 (FY2025). FY2025 operating cash flow was ¥17.9bn against net profit ¥5.25bn (OCF/NP 3.4x); OCF has been positive for five straight years (FY2021–FY2025). There is no liquidity or going-concern concern.

The weak spot is returns on capital. ROE was 3.2% in FY2025 and the company's own FY2026 plan is 4.3% (3M cumulative 5.0%); ROIC is 2.7% (IFIS definition). That is roughly half the ~8–9% cost of equity I assume for a large Japanese industrial. ROE decomposition shows the cause is margin, not leverage or asset turn: net margin ~2.2%, asset turnover ~0.71x, equity multiplier ~1.9x. Asset turnover has been flat at 0.71–0.72x for five years, so the improvement must come from margin or from shedding assets. There is no company ROE/ROIC target — the stated KPIs are overseas sales ratio and OPM (filings/2025-11_FY_有報_20260224.pdf, 経営方針).

Capital allocation:

• Dividends. Annual, year-end only; FY2026 forecast DPS ¥110 (from ¥100), yield 2.0%, payout ~32% of forecast EPS. No progressive-dividend or DOE policy.
• Buybacks. Intermittent, not programmatic: FY2024 ¥2.5bn, FY2025 ¥2.7bn, three ToSTNeT programs in 2025; none announced yet in FY2026. Treasury stock is 3.12m shares (12.5% of issued), which magnifies EPS but also signals weak share-demand management.
• Capex. FY2026 plan ¥16.0bn vs depreciation ¥18.5bn (capex/depreciation ~0.86x), after two years well below depreciation — a period of harvest, not expansion. The plan includes a bearing-plant expansion in India (2027 H2 start) and US/India sales offices.
• Cross-shareholdings. 24 listed holdings at ¥26.9bn and 30 unlisted at ¥0.47bn (end-FY2025). The company sold ¥3.67bn in FY2025 and says it will continue to trim holdings judged to be unbeneficial — a genuine value lever, but not distributable cash by itself.

IR and management credibility

• 2026: leadership change. The FY2025 短信 (2026-01-14) was signed by president 黒澤 勉; a board/officer change was announced 2026-01-22, and the Q3 FY2026 短信 (2026-10-05) is signed by president 中村 成利. The new management has kept the reform agenda and raised guidance once.
• Disclosure. Regular 短信; an H1 決算説明会 (2026-07-22) with presentation material; no Q3 説明会 (Q3 短信 states 決算説明会開催の有無：無). A 補足資料 is produced each quarter.
• Credibility. Guidance is not reliably conservative (FY2023 and FY2024 missed their initial OP targets by a third). The 買収防衛策 (poison pill) was renewed at the February 2026 AGM, a modest governance negative. Sensitivity on capital efficiency — no ROE/ROIC target, no published TSE cost-of-capital plan found — is the main credibility gap for a sub-1x-PBR company.

Valuation and scenarios

At ¥5,500 the stock is at 16.0x the company's FY2026 forecast EPS (¥344) and 0.66x book (BPS ~¥8,335). On the stock's own bands, PER is slightly above its mean (z +0.36) while PBR is a full standard deviation below (z −1.00) — the classic signature of a cheap-asset, poor-returns company. The IFIS theoretical price is ¥5,294 on mean PER and ¥5,825 on mean PBR; EV/EBITDA is 5.6x forecast. The price implies an EPS of ~¥351 at the band-mean PER, essentially the current company forecast.

Peer comparison (Yahoo! Finance Japan, 2026-10-16; peer PERs are on company forecasts for each stock's own next fiscal year, Nachi's on FY2026 — bases are not identical):

Nachi is the cheapest of the bearing peers on PBR and pays the lowest yield; its PER is mid-range. The low PBR with a low yield means the company retains earnings but does not earn much on them.

The base case carries the recovery one more year, then values the stock near its own mean PER. The skew is mildly positive (probability-weighted ~+5%), which is why this is a Hold with an upside lean rather than a Sell — but the base case offers no margin of safety, and the bull case depends on a re-rating that needs ROE to move.

Bull case

• Structural reform keeps compounding: OPM reaches the company's 10% milestone within the plan horizon, lifting ROE toward 8%.
• Robots and hydraulic equipment for construction/industrial machinery sustain double-digit growth; the US and India build-out (bearing capacity from 2027 H2) adds volume.
• A Q4 beat forces a further FY2027 guidance raise, and the ¥26.9bn cross-shareholding block funds a larger buyback.
• PBR re-rates from 0.66x toward the peer 0.74–0.84x.

Bear case

• The recovery is cyclical. Auto (~50% of sales) and construction machinery turn, and the reform gains plateau at an OPM of around 6–7%, leaving ROE stuck at 4–5% and the PBR discount intact.
• Car hydraulics (−16%) and machine tools (−11%) keep shrinking; China and domestic demand stay weak; raw-material costs outrun price pass-through (materials were a ¥3.8bn drag already).
• Yen appreciation: guidance assumes USD 154; 9M average was 158, so the FX tailwind reverses if the yen firms.
• A further restructuring charge, or capital tied up in ¥26.9bn of cross-shareholdings, keeps returns below the cost of equity.

Catalysts

What would change the rating

• Upgrade if FY2027 guidance shows OP above ¥16.5bn with OPM ≥6.5%, and management publishes an ROE/ROIC target or a cost-of-capital plan that credibly lifts ROE toward 8%, ideally alongside an increased buyback.
• Downgrade if FY2026 Q4 misses guidance, if FY2027 guidance implies flat-to-down OP, if the reform gains stall at an OPM below 6%, or if working capital (inventory days 136) keeps building into a demand slowdown.
• Thesis invalidated if auto and construction demand roll over before ROE clears the cost of equity, which would make the PBR discount a permanent feature rather than an opportunity.

Monitoring

Scorecard

Fact-sheet flags

Computed by factsheet.mjs. Each flag is a prompt to investigate, not a conclusion.

• iinfoprogress-ahead-rp 2026/11 Q3 rp progress 81.8% vs median 70.1% (+11.7 pts): ahead of seasonal norm
• ▲amberconsensus-below-guidance 2026/11 consensus OP 9,900 is -35.3% vs company guidance

Sources and limitations

Primary:

• data/6474/2026-10-17/filings/2026-11_Q3_短信_20261005.pdf (2026-10-05), pp.1–9 — Q3 FY2026 results, guidance, segments, structural-reform costs.
• Q3 決算補足資料 (2026-10-05), https://www.nachi-fujikoshi.co.jp/dcms_media/other/20261005_jp2.pdf — quarterly orders/sales, OP bridge, segment and financial detail.
• data/6474/2026-10-17/filings/2025-11_FY_短信_20260114.pdf (2026-01-14) — FY2025 results.
• data/6474/2026-10-17/filings/2025-11_FY_有報_20260224.pdf (2026-02-24), pp.2, 10, 18–20, 30–31, 33, 51–52 — financial history, 経営方針, 事業等のリスク, 大株主, 配当政策, 政策保有株式.
• H1 FY2026 決算説明資料 (2026-07-22), https://www.nachi-fujikoshi.co.jp/dcms_media/other/2611_20260724.pdf, pp.9–19 — FY2026 plan, OP bridge, segment plan, financial plan.
• 株式分割のお知らせ (2026-10-05), https://www.nachi-fujikoshi.co.jp/dcms_media/other/20261005_Notice_of_Stock_Split.pdf.
• IFIS snapshot data/6474/2026-10-17/ — feeds, pages and tables as cited (feeds/cons_chart.json, stock_chart_prog_all.json, segment_fy.json; pages/report_est.html, report_topix.html, report_disclose.html).
• Yahoo! Finance Japan quote pages for 6471, 6472, 6473, 6481 (2026-10-16) — peer PER/PBR/ROE/yield.

Limitations:

• The IFIS consensus and target price are thin (2 analysts) and stale — no update since late July 2026 — so they are reported but not relied on.
• Peer multiples come from a different source and, for NSK/NTN/JTEKT, a different fiscal year (FY ending March 2027) than Nachi's (FY ending November 2026); bases are stated.
• The 大量保有報告書 lead (individuals 内山彰博, 小松一成) is unverified — 四季報 online and EDINET require JavaScript and could not be fetched.
• No medium-term management plan with an ROE/ROIC target was found; the company's stated KPIs are overseas sales ratio (60%) and OPM (10%).
• Cost of equity of 8–9% is my analyst assumption, not a company figure.

This is research, not individualised investment advice.