ツガミ Tsugami (6101)

Signal panel

Investment thesis

1. Guidance is too low for H1. Q1 FY27 OP ¥13,466m against H1 guidance ¥19,500m leaves ¥6,034m for Q2 (短信 Q1 p.1). Q2/Q1 OP was 0.75 in FY26 and 0.71 in FY25 (fact sheet, quarters). Applying 0.73 gives Q2 OP about ¥9.8bn and H1 about ¥23.3bn, 19% above guidance (inference).
2. The revision calendar is predictable. FY26 guidance stayed at ¥19.5bn at Q1 (2025-07-30, with OP progress at 44.5%), went to ¥27.0bn at Q2 (2025-11-13), and to ¥36.0bn on 2026-04-21. FY25 followed the same pattern: unchanged at Q1, raised 2024-11-13 and 2025-04-22 (act_est_hist_n.json). The FY27 Q1 短信 again left guidance unchanged, citing "市場全体の不透明感" (p.3).
3. Demand is still accelerating, not rolling over. China segment Q1 sales were ¥38.0bn (+41% YoY) with segment profit ¥11.6bn (+50%) (短信 Q1 p.2). Contract liabilities (customer advances) rose ¥2.2bn in the quarter (p.2), and they rose ¥2.3bn in FY26 (有報 p.15). JMTBA total orders were +64.7% YoY in August and a monthly record +60% in September 2026, with China orders +61.7% (external reports of the JMTBA release).
4. The balance sheet funds a downturn. Net cash ¥32.9bn at FY26 end, group equity ratio 71.2% at Q1, OCF ¥28.6bn in FY26 (1.71x NP), and OCF positive every year in the fact sheet.
5. Valuation already discounts a peak. PER on company guidance is 14.2x. On our base EPS of about ¥475 it is 11.0x, below Okuma (14.0x) and DMG Mori (25.0x), which earn ROEs of 5 to 7% against Tsugami's 23.4%.

Business and earnings

Tsugami makes CNC machine tools, mainly Swiss-type automatic lathes (自動旋盤) for small precision parts used in smartphones and electronics, autos, hydraulics and medical devices. Automatic lathes are about 84% of sales (FY26 説明会 p.10: ¥1,086億 of ¥1,291億). It wins on price and productivity in the volume segment. It builds in China (Zhejiang and Anhui) at scale and keeps high-end development at the Nagaoka mother plant (capital-policy deck, 2026-05-13, p.4).

The profit engine is China. The China segment made ¥31.2bn of ¥33.9bn total segment profit in FY26 (92%) at a 28.1% margin. Japan made 8.4%, India 5.1% (fact sheet, segments). China customers were 79.9% of FY26 sales and 82.0% in Q1 FY27 (短信 Q1 p.13). The Zhejiang subsidiary alone earned ¥25.0bn net profit on ¥110.7bn sales, more than the group's ¥24.3bn total (有報 p.8).

Minority leakage matters. Tsugami owns 68.1% of the Cayman holding company of the China business, listed in Hong Kong as Precision Tsugami (China), 1651.HK (有報 p.7). Non-controlling interests took 31% of FY26 profit (¥7.6bn of ¥24.3bn) and 28% in Q1 FY27 (短信 p.6). The China subsidiary buying back its own shares raises Tsugami's stake (¥5.4bn spent in FY26, 説明会 p.4 and 有報 p.16), which is accretive to the parent.

Q1 FY27 bridge. Sales +35.3%, OP +55.2%, OP margin 31.3% (record). Gross margin rose from 36.9% to 38.7% and SG&A fell from 10.8% to 8.9% of sales, so most of the gain is operating leverage on volume (短信 p.6). Other income added ¥0.66bn, including ¥0.62bn of subsidies (p.10), worth about 5% of OP. The effective tax rate rose to 33.3% from 26.7% in Q1 FY26, which held NP growth to 55% despite pretax +60%. FY27 capex is ¥38億 in Zhejiang (+20% capacity from January 2028), ¥23億 at Nagaoka (September 2027), and the India plant opened in December 2025 at twice the old capacity (説明会 p.5).

Cyclicality is high. OP fell from ¥18.9bn (FY22) to ¥13.1bn (FY24), −31%, then rose to ¥36.1bn (FY26).

Forecast test

• Guidance (2026-05-13, unchanged 2026-07-30): Sales ¥145.0bn, OP ¥36.5bn (+1.1%), NP ¥17.0bn, EPS ¥370.28, DPS ¥98. The split is H1 OP ¥19.5bn and H2 ¥17.0bn, against FY26 H2 actual ¥20.8bn (説明会 p.2). Guidance therefore assumes H2 OP falls 18% YoY.
• Consensus: none in IFIS for FY27 or FY28. There is no market estimate to beat, and no evidence a raise is priced in.
• Seasonal progress (fact sheet): OP 36.9% vs a 33.2% prior median (+3.7 pts). NP 38.4% vs 35.1% (+3.3 pts). Neither crosses the +8 pt threshold, because the priors are also measured against lowballed guidance (FY26 Q1 was 44.5%). Against final actuals, Q1 has averaged 26.0% of the year's OP. Do not annualise. The test that matters is the Q2 gap in thesis point 1.
• Track record: actual vs initial OP was +189%, +71%, −12%, −13%, +55%, +85% for FY21 to FY26. The company is conservative in upturns and late to cut in downturns.
• Orders and backlog: not disclosed as numbers. The 説明会 p.11 chart is unlabelled in the extracted text, and the 有報 omits orders (p.16). Contract liabilities (+¥2.2bn in Q1) and the JMTBA China series are the proxies.
• One-offs: subsidies of ¥0.6 to 0.7bn per quarter run through other income. No sale gain in guidance. EPS needs no restatement.
• Revision probability: high at the Q2 results, expected around 2026-11-12 to 11-13 (2025-11-13 and 2024-11-13 precedent, date not yet confirmed). Our base case is FY OP about ¥47bn (H1 ¥23.3bn plus H2 ¥23.7bn, +14% on FY26 H2). A cut would require China orders to turn negative in October and November.

Financial strength and capital allocation

• Liquidity: cash ¥44.7bn, debt about ¥9bn, current ratio 3.08 (FY26). No going-concern note (短信 p.11). Much of the cash sits in the 68%-owned China group and reaches the parent as dividends (a subsidiary dividend was announced 2026-09-11).
• Cash conversion: OCF/NP was 1.38, 0.66, 1.08, 2.21, 0.81 and 1.71 over FY21 to FY26. Q1 FY27 OCF was ¥4.8bn after a ¥5.7bn payables outflow and ¥3.8bn receivables build. Receivable days are high at 140 (FY26), up from 86 in FY23. Watch this: Chinese customers settle slowly, and a downturn would test collection.
• Capital efficiency: ROE 23.4%, ROA 11.9%, IFIS ROIC 27.7% (FY26). The margin drove the ROE rise: NP margin went from 6.4% (FY24) to 13.0% (FY26), asset turnover from 0.71 to 0.84, while leverage fell. Peer ROE (Yahoo, actual): Okuma 5.3%, DMG Mori 7.3%, Star Micronics 4.0% (stale March 2026 page). The company states a cost of equity of 9 to 11% and targets an ROE "stably above" it, with no numeric ROE goal (capital-policy deck p.6).
• Returns: DPS ¥85 to ¥98 (10th straight rise, payout 26.5%). FY26 total payout 35.7% (説明会 p.7). Buybacks are small: ¥2.0bn in FY26, and the current programme is up to ¥1.5bn through 2027-05-12, 0.6% of market cap. Policy is a cash target of about 3 months' sales and a group equity ratio of 60% or more (deck p.7). At 71% the group is above target, so payout can rise, but management has not committed to that.
• Hard gates: all pass. No restatement or audit issue found (FY21 and FY22 有報 corrections on 2023-04-21 appear in the disclosure list; their nature was not checked). No liquidity distress. OCF positive in all six years. No theme claim is made: JMTBA attributes the industry's records to AI demand, but Tsugami's mix is small precision parts and we do not count AI exposure.

IR and management credibility

• 2026-05-13: first 資本コストや株価を意識した経営 statement, three years after the TSE request. Its targets are qualitative.
• Results briefings run twice a year. Q1 and Q3 have none (短信 Q1 p.1: 決算説明会 無).
• English translations are published, and the HK subsidiary refiles the parent's China segment (HKEX 2026-07-30).
• Guidance behaviour is the main credibility issue. Guidance is low and is raised twice a year in upturns, and it lagged the downturn in FY23 and FY24. Read it as a floor in upturns, not as a forecast.
• KPI continuity: segment, product-line and region tables are consistent across the FY26 and Q1 FY27 filings.

Valuation and scenarios

FY2027/03 parent NP. We assume parent share of NP at 47% of OP (46.4% FY26, 48.5% Q1 FY27) and 46.5m shares outstanding.

At 25/55/20 weights the expected return is about +15% before the 1.9% dividend.

• Own bands: PBR 2.80 vs a 3.32 mean (z −0.78), so it sits below its average while ROE sits at a 10-year high. The theoretical price at mean PBR is ¥6,223 (fact sheet). No IFIS PER band exists, because there is no consensus. EV/EBITDA 5.7x.
• What the price implies: ¥5,240 at 13x is EPS of about ¥403, i.e. FY27 OP of about ¥40bn. That means the market expects only 10% more OP than guidance despite the Q1 run-rate.
• Sum of parts (inference): 1651.HK market cap was about HK$18.3bn on 2026-09-02 (StockAnalysis). Tsugami's 67 to 68% stake is HK$12.3bn, about ¥250 to 260bn at ¥20 to 21 per HKD. That is roughly Tsugami's entire ¥251.5bn market cap, so Japan, India and parent cash come for free. Caveat: the October HK price was not verified, and the HK stock would fall too in a China downturn.
• Why the market might be right: the OP margin of 31% is double the FY24 level, China is 80% of sales, and machine-tool cycles turn quickly. The stock fell from ¥8,630 (07-01) to ¥5,720 by 08-03, around the Q1 release, then −7.2% on 08-18, −8.3% on 09-08 and −6.2% on 10-15. We found no company-specific news for those days (Tsugami IR lists only buyback notices since July), so we read it as de-rating on peak-cycle and China concerns after a +114% year (inference).

Bull case

• Q2 OP near ¥10bn and an FY raise to ¥45bn or more on about 2026-11-13, followed again by a late-April raise.
• The Zhejiang capacity (+20% from January 2028) and India (2x) extend the cycle.
• Higher payout or a larger buyback, since the group equity ratio is 71% against a 60% target.
• The SOTP discount closes: the Tokyo listing trades at or below the value of its HK stake.

Bear case

• China small-parts demand (smartphones, EVs) peaks. JMTBA China orders −2.5% MoM in September is the first soft print.
• Receivable days of 140 and rising, and a tax rate that jumped to 33%.
• NCI takes about 30% of profit. In the parent's look-through it is a China-listed company's earnings at a Japanese multiple.
• Geopolitics and export controls between Japan, China and the US. The 有報 risk section (p.14) flags RMB and country-regulation risk.

Catalysts

What would change the rating

• Upgrade (higher confidence) if the Q2 短信 raises FY OP to ¥45bn or more and contract liabilities keep rising.
• Downgrade to Hold if the Q2 guide stays at ¥36.5bn with H1 OP below ¥22bn, or if JMTBA China orders turn negative YoY for two months.
• Thesis invalidated if Q2 standalone OP falls below the ¥6.5bn of Q2 FY26, i.e. the cycle is turning now and the guidance was not conservative.

Monitoring

Scorecard

Fact-sheet flags

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

• iinfonet-cash Net cash of 32,868 in 2026/03 (BS_NET_DEBT -32,868)
• iinfofar-below-52w-high Price 5,240 is -39.3% from the 52-week high 8,630

Sources and limitations

• Snapshot data/6101/2026-10-17/ (no capture errors). Fact sheet data/6101/analysis/2026-10-17/factsheet.md.
• 2027年3月期 第1四半期決算短信, 2026-07-30 (filings/2027-03_Q1_短信_20260730.pdf), pp.1, 2, 3, 6, 10, 12, 13 (printed pages 1 to 13 match the PDF index minus 2 after p.2).
• 2026年3月期 決算説明資料, 2026-05-13 (filings/2026-03_FY_説明会_20260513.pdf), pp.3 to 12 (printed 2 to 11).
• 有価証券報告書 FY2026/03, 2026-06-16 (filings/2026-03_FY_有報_20260616.pdf), pp.5, 7, 8, 14, 15, 16.
• 資本コストや株価を意識した経営の実現に向けた対応について, 2026-05-13, tsugami.co.jp, pp.3 to 9.
• HKEX announcement by Precision Tsugami (China), 2026-07-30: hkexnews (67.15% ownership).
• Tsugami IR news list, checked 2026-10-17: tsugami.co.jp/ir. No revision or results disclosure after 2026-07-30.
• JMTBA orders as reported: JC-NET, September 2026, kikai-news, August 2026, JMTBA statistics. The September breakdown in the secondary report looks partly mixed with August figures. Confirm with the JMTBA final release.
• Peers (Yahoo! Finance Japan, company-forecast PER, actual PBR and ROE, 2026-10-16): Okuma 6103, DMG Mori 6141. Star Micronics 7718 returned a stale 2026-03-12 page and is used for ROE only.
• 1651.HK market cap: StockAnalysis, 2026-09-02. Not verified for October. The HKD/JPY rate is our assumption.
• Price history: Nikkei, minkabu, Kabutan.
• Missing: company order and backlog numbers, analyst consensus, a dated October HK price, and the reason for the 2023-04-21 corrected 有報.
• This is research, not individualised investment advice.