Home/Reports/JP Stock DD — GPT Validation Request

JP Stock DD — GPT Validation Request

2026-04-12 19:55 · 71.4 KB

Time: 19:55

Generated: 2026-04-12

Companies: 48

Themes: nuclear, ev, ai, semiconductor, robotics, vtuber

Instructions for GPT

Please validate the following equity research findings. For each company:

1. Verify fundamentals — Are PE, forward PE, ROE, operating margin, FCF figures plausible? Flag any that seem wrong.

2. Check thesis — Is the investment thesis logically sound? Any obvious flaws?

3. Validate supply chain — Are the customer/supplier relationships accurate? Any outdated or incorrect claims?

4. Risk assessment — Are the identified risks complete? Any major risks missing?

5. Score reasonableness — Does the conviction level (HIGH/MEDIUM/LOW) match the fundamentals?

Output a table of issues found, with columns: Ticker | Issue Type | Severity (CRITICAL/HIGH/MEDIUM/LOW) | Finding


Rankings Summary

RankTickerNameScoreConvPEFwd PEROEOp MarginFCFThemes
16857.TAdvantest89.1HIGH63.5136.6149.3%39.7%+¥304Bai, semiconductor
25032.TANYCOLOR (Nijisanji)81.4HIGH19.8x24.9x74.2%40.2%N/Avtuber
38035.TTokyo Electron76.4HIGH40.2431.126.5%25.3%+¥398Bai, semiconductor
46368.TOrgano Corporation75.9HIGH26.4224.4723.5%21.4%+¥19Bnuclear
57735.TSCREEN Holdings75.5HIGH24.0818.719.8%19.0%+¥45Bai, semiconductor
64063.TShin-Etsu Chemical72.5HIGH25.7125.0411.5%25.6%+¥346Bai, semiconductor
75253.TCover Corp (Hololive)72.2HIGH18.6xN/A33.7%11.3%+¥2Bvtuber
86920.TLasertec71.8MEDIUM43.0445.0643.0%48.7%+¥96Bai, semiconductor
96841.TYokogawa Electric70.1HIGH24.8220.7312.2%14.6%+¥88Bnuclear
106504.TFuji Electric69.5HIGH19.6217.1311.8%10.3%+¥51Bnuclear, ev
116762.TTDK Corp67.3HIGH23.4120.949.5%10.1%+¥169Bev
126981.TMurata Manufacturing66.7HIGH40.2525.667.3%15.1%+¥239Bev
136525.TKokusai Electric66.4HIGH48.9641.3216.4%18.6%+¥35Bsemiconductor
145201.TAGC Inc66HIGH17.8914.74.7%6.3%+¥65Bev
156506.TYaskawa Electric64.2HIGH36.0626.567.8%8.7%+¥6Brobotics
166806.THirose Electric58.3MEDIUM23.6221.448.0%20.3%+¥26Bev
177013.TIHI Corporation58.2MEDIUM28.9428.0223.6%9.1%+¥68Bnuclear
186954.TFanuc56HIGH36.0831.149.4%21.1%N/Arobotics
196370.TKurita Water Industries55.3MEDIUM38.2221.676.8%13.7%+¥25Bnuclear
206723.TRenesas Electronics54.6MEDIUMLoss17.78-2.1%17.9%+¥364Bai, semiconductor
215713.TSumitomo Metal Mining53MEDIUM28.6316.535.0%5.0%+¥87Bev
223407.TAsahi Kasei53MEDIUM14.1213.3N/A7.3%+¥134Bev
236302.TSumitomo Heavy Industries52.1MEDIUM20.4618.634.7%4.8%+¥11Bnuclear
246645.TOmron50.8MEDIUM39.2521.182.3%6.1%+¥6Brobotics
256501.THitachi50.4MEDIUM26.6425.37N/A11.1%+¥1,390Bnuclear
266503.TMitsubishi Electric47.2MEDIUM31.8629.62N/A6.5%+¥317Bev
276861.TKeyence46.9MEDIUM36.2232.31N/A50.9%N/Aai, robotics
286758.TSony Group46.1MEDIUM15.9x18.4x14.9%9.7%N/Avtuber
297012.TKawasaki Heavy Industries45.9MEDIUM25.7732.17N/A5.3%+¥47Brobotics
306594.TNidec Corp45.5LOW21.9613.746.3%5.0%+¥174Bev, robotics
316146.TDISCO Corp42.7MEDIUM55.4444.98N/A41.7%N/Aai, semiconductor
325631.TJapan Steel Works42.6HIGH35.2834.3110.5%8.4%-¥19.7Bnuclear
336324.THarmonic Drive Systems42.5HIGHN/AN/AN/AN/AN/Arobotics
345802.TSumitomo Electric42MEDIUM32.0724.1211.0%7.8%N/Aev
357832.TBandai Namco42MEDIUM22.2x18.6x13.8%N/AN/Avtuber
364751.TCyberAgent40.7MEDIUMN/AN/A16.8%9.3%N/Avtuber
374062.TIbiden39.8HIGH72.1144.917.6%13.4%-¥8Bsemiconductor
384004.TResonac Holdings39.4LOW83.5328.84.4%4.3%+¥24Bev, semiconductor
394023.TKureha Corp34.3MEDIUM17.8915.735.0%7.6%-¥4Bev
407004.TKanadevia33MEDIUM19.6214.095.8%2.0%-¥12.8Bnuclear
416526.TSocionext31.9MEDIUM40.3722.446.0%6.2%-¥4Bai, semiconductor
429984.TSoftBank Group31.5LOW5.8922.0926.0%7.5%-¥1,396Bai
436268.TNabtesco30.5MEDIUMN/AN/AN/AN/AN/Arobotics
449468.TKadokawa25.4LOW25.9xN/A4.1%6.0%N/Avtuber
453632.TGREE Holdings (REALITY)24LOW45.7xN/AN/A6.7%N/Avtuber
463668.TColopl21.3LOW40.7xN/A-0.4%N/AN/Avtuber
473436.TSUMCO12.6LOWLossN/A-1.5%-0.8%-¥11Bai, semiconductor
486963.TRohm Co.7.5LOWLoss55.75Loss-4.1%-¥44Bev

Detailed Company Reports

Advantest (6857.T)

Score: 89.1 | Conviction: HIGH | Layer: Testing Equipment | Weight: 12%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: Monopoly-level share in SoC testers for AI accelerators. NVIDIA is ~40%+ of revenue. Revenue +51% TTM, net income +112%. ATE lead times >6 months — supply-constrained. V93000 platform dominates AI chip testing. FCF margin 29%.

Risk: >40% revenue concentration in NVIDIA. A NVIDIA capex pullback would be material. P/B 27x is extreme.

Monitoring Trigger: Watch NVIDIA quarterly capex guidance and TSMC N3/N2 wafer starts. If NVIDIA cuts capex >15%, reduce position.

Fundamentals:

  • PE: 63.51 | Fwd PE: 36.61 | P/B: 27.11
  • ROE: 49.3% | Op Margin: 39.7% | D/E: 14%
  • Div Yield: 0.2% | FCF: +¥304B | Mkt Cap: ¥18.28T

Segments:

  • Semiconductor Test Systems: ¥876B (85%) — V93000 SoC testers, T2000 memory testers (HBM)
  • Mechatronics: ¥103B (10%) — Handler systems
  • Services: ¥52B (5%) — Support, maintenance

Supply Chain:

  • [confirmed] NVIDIA: V93000 SoC testers for AI GPUs — Primary AI chip tester; est. 40%+ revenue
  • [probable] TSMC and its customers: Logic chip test systems — Volume tied to TSMC advanced nodes
  • [probable] SK Hynix / Samsung: HBM memory test systems — T2000/T5000 for HBM2e/HBM3 testing

ANYCOLOR (Nijisanji) (5032.T)

Score: 81.4 | Conviction: HIGH | Layer: VTuber Agency / IP Platform | Weight: 20%

Themes: vtuber | Data Date: 2026-04-12

Thesis: Dominant VTuber agency operating Nijisanji with 200+ talents across JP, KR, and EN. Exceptional profitability — 40% operating margin and 74% ROE on an asset-light IP model. Revenue diversifying from streaming to merch, events, and game collaborations with Square Enix and Bandai Namco.

Risk: Talent departure risk and concentration in YouTube ecosystem. EN segment engagement declining vs Cover Corp's Hololive.

Monitoring Trigger: If top-10 talent leaves or YouTube changes monetization rules, reassess position.

Fundamentals:

  • PE: 19.8x | Fwd PE: 24.9x | P/B: N/A
  • ROE: 74.2% | Op Margin: 40.2% | D/E: 0.1%
  • Div Yield: 1.2% | FCF: N/A | Mkt Cap: ¥184B

Segments:

  • Commerce: ¥20B (39%) — Merchandise, events, physical goods
  • Streaming: ¥15B (29%) — YouTube/platform donations and ad revenue
  • IP / Promotions: ¥16B (32%) — IP licensing, brand collaborations, game integrations

Supply Chain:

  • [confirmed] YouTube: Primary distribution platform — All Nijisanji talents stream on YouTube
  • [confirmed] Bandai Namco: Game IP collaboration — Official game integration partnerships
  • [confirmed] Square Enix: Game collaboration — Strategic game partnership

Tokyo Electron (8035.T)

Score: 76.4 | Conviction: HIGH | Layer: Semiconductor Equipment | Weight: 12%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: #3 global WFE. ~90% global share in coater/developer tracks — every advanced chip passes through TEL equipment. FCF ¥398B with zero debt. AI-related equipment rising to 40% of sales. Record WFE market expected in CY2026.

Risk: US export controls on China (currently ~30% of revenue). BIS rule changes targeting Japanese equipment.

Monitoring Trigger: TSMC and Samsung capex announcements. BIS export rule changes. If China revenue drops below 20%, thesis strengthens.

Fundamentals:

  • PE: 40.24 | Fwd PE: 31.1 | P/B: 9.99
  • ROE: 26.5% | Op Margin: 25.3% | D/E: 0%
  • Div Yield: 1.4% | FCF: +¥398B | Mkt Cap: ¥20.03T

Segments:

  • Semiconductor Production Equipment: ¥2,271B (95%) — Coaters, etch, CVD, thermal, cleaning
  • FPD Equipment: ¥119B (5%) — Display equipment

Supply Chain:

  • [confirmed] TSMC: Coater/developer tracks, etch systems — Leading logic customer; est. 30%+ of WFE revenue
  • [confirmed] Samsung: Logic and DRAM equipment — Major customer
  • [probable] Intel Foundry: Equipment for Intel 18A — Resumed investment

Organo Corporation (6368.T)

Score: 75.9 | Conviction: HIGH | Layer: Nuclear Operations / Water | Weight: 10%

Themes: nuclear | Data Date: 2026-04-12

Thesis: Near-monopoly on BWR condensate water treatment in Japan — IFD technology adopted in ALL BWR condensate demineralizer vessels. Every reactor restart = Organo revenue. Best quality: ROE 23.5%, op margin 21.4%. Also serves semiconductor fabs (TSMC Kumamoto confirmed customer).

Risk: PE 26x and +202% 52-week run have priced in significant restart upside. P/B 5.7 is expensive.

Monitoring Trigger: If P/B expands >7x without earnings support, trim. Watch each reactor restart announcement — direct revenue event.

Fundamentals:

  • PE: 26.42 | Fwd PE: 24.47 | P/B: 5.7
  • ROE: 23.5% | Op Margin: 21.4% | D/E: 24%
  • Div Yield: 1.2% | FCF: +¥19B | Mkt Cap: ¥745B

Segments:

  • Water Treatment Engineering: ¥105B (60%) — Nuclear, power plant, semiconductor ultrapure water systems
  • Performance Products: ¥70B (40%) — Ion exchange resins, chemicals, maintenance services

Supply Chain:

  • [confirmed] All Japanese BWR operators: IFD condensate demineralizers (monopoly) — IFD technology in ALL BWR condensate vessels in Japan since 1970
  • [confirmed] TSMC Kumamoto (JASM): Ultrapure water systems — Semiconductor fab customer
  • [probable] PWR operators (KEPCO, etc.): Makeup water, waste treatment — Organo serves both BWR and PWR per product pages

SCREEN Holdings (7735.T)

Score: 75.5 | Conviction: HIGH | Layer: Cleaning Equipment | Weight: 10%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: #1 globally in wafer cleaning. Advanced packaging (CoWoS, HBM stacks) requires intensive cleaning steps. Forward PE 18.7x — cheapest semi-cap equipment name vs peers at 30-45x. 20% capacity expansion (S3-6 facility, Feb 2025). ROE 19.8%.

Risk: Revenue dipped 4.3% TTM — cyclical softness. Memory investment pause pressures near-term.

Monitoring Trigger: HBM capacity additions at SK Hynix/Samsung. TSMC CoWoS expansion announcements.

Fundamentals:

  • PE: 24.08 | Fwd PE: 18.7 | P/B: 4.58
  • ROE: 19.8% | Op Margin: 19.0% | D/E: 1%
  • Div Yield: 1.3% | FCF: +¥45B | Mkt Cap: ¥2.03T

Segments:

  • Semiconductor Equipment: ¥443B (75%) — Wafer cleaning (#1 global), track systems
  • Graphic Arts: ¥59B (10%) — Printing equipment
  • Display/Coater: ¥47B (8%) — FPD, advanced packaging coater
  • PCB Equipment: ¥41B (7%) — PCB processing

Supply Chain:

  • [probable] SK Hynix: Cleaning for HBM advanced packaging — Dominant HBM producer needs SCREEN cleaning
  • [probable] TSMC: Cleaning for CoWoS packaging — CoWoS for NVIDIA GPUs
  • [inferred] Samsung: DRAM and HBM cleaning — Major memory producer

Shin-Etsu Chemical (4063.T)

Score: 72.5 | Conviction: HIGH | Layer: Silicon Wafers / Materials | Weight: 10%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: Most defensive AI infrastructure play — ~30% global silicon wafer share (#1). Every chip requires their wafers or specialty chemicals. Fortress balance sheet: ¥1.25T net cash, D/E 0.05. Also EUV photomask blanks and photoresists. Developing 300mm GaN-on-QST for AI data center power.

Risk: PVC/chlor-alkali is ~30% of revenue — cyclical commodity exposed to Chinese oversupply.

Monitoring Trigger: Silicon wafer ASP disclosures in quarterly earnings. 300mm contract pricing uptick signals AI demand.

Fundamentals:

  • PE: 25.71 | Fwd PE: 25.04 | P/B: 2.71
  • ROE: 11.5% | Op Margin: 25.6% | D/E: 5%
  • Div Yield: 1.6% | FCF: +¥346B | Mkt Cap: ¥12.17T

Segments:

  • Electronics Materials: ¥800B (40%) — Silicon wafers (#1 global), photoresists, photomask blanks
  • Infrastructure Materials: ¥600B (30%) — PVC resin, caustic soda
  • Functional Materials: ¥400B (20%) — Rare earth magnets, LED, encapsulants
  • Processing & Services: ¥200B (10%) — Wafer processing

Supply Chain:

  • [confirmed] TSMC: 300mm silicon wafers, EUV photomask blanks — Leading-edge logic customer
  • [confirmed] Samsung: Silicon wafers for NAND + DRAM + logic — Major customer
  • [confirmed] SK Hynix: Advanced 300mm wafers for HBM3 — HBM requires most advanced wafers

Cover Corp (Hololive) (5253.T)

Score: 72.2 | Conviction: HIGH | Layer: VTuber Agency / IP Platform | Weight: 20%

Themes: vtuber | Data Date: 2026-04-12

Thesis: #1 VTuber agency globally by brand power. Revenue surging 44% YoY to ¥43.4B. Diversifying into music (Hololive original songs), anime, and merchandise. Sony mocopi partnership validates their tech integration for 3D live performances.

Risk: High P/B (5.75x) pricing in growth. Talent dependency — top performers drive outsized revenue share.

Monitoring Trigger: If revenue growth declines below 20% YoY or major talent graduation wave, reduce conviction.

Fundamentals:

  • PE: 18.6x | Fwd PE: N/A | P/B: 5.75x
  • ROE: 33.7% | Op Margin: 11.3% | D/E: N/A
  • Div Yield: 0% | FCF: +¥2B | Mkt Cap: ¥99B

Segments:

  • Streaming: ¥18B (41%) — YouTube Super Chat, memberships, ad revenue
  • Commerce: ¥15B (35%) — Merchandise, events, concerts
  • Licensing & Music: ¥10B (24%) — IP licensing, original music, anime projects

Supply Chain:

  • [confirmed] Sony: mocopi motion capture hardware — Hololive talents use mocopi for 3D streams and concerts
  • [confirmed] YouTube: Primary distribution platform — All Hololive talents stream on YouTube
  • [confirmed] Bushiroad: Trading card game collaboration — Hololive TCG product line

Lasertec (6920.T)

Score: 71.8 | Conviction: MEDIUM | Layer: EUV Inspection | Weight: 6%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: Pure monopoly — sole provider of actinic EUV mask blank inspection. Non-negotiable step for every chip at 5nm and below. 48.7% operating margin, 38% FCF margin. Co-dependent with ASML.

Risk: Forward PE 45x higher than trailing 43x — growth expectations front-loaded. TSMC N2 delay would defer orders. Scorpion short-seller report tail risk.

Monitoring Trigger: TSMC N2 ramp schedule. ASML EUV tool shipment volumes (each needs Lasertec support).

Fundamentals:

  • PE: 43.04 | Fwd PE: 45.06 | P/B: 16.52
  • ROE: 43.0% | Op Margin: 48.7% | D/E: 0%
  • Div Yield: 0.8% | FCF: +¥96B | Mkt Cap: ¥3.73T

Segments:

  • Semiconductor Inspection: ¥226B (90%) — EUV actinic blank + mask inspection (monopoly)
  • LED/Display Inspection: ¥25B (10%) — Non-semiconductor inspection

Supply Chain:

  • [confirmed] ASML: Co-dependent EUV ecosystem — ASML directs customers to use Lasertec inspection
  • [confirmed] TSMC: EUV reticle inspection for 3nm/2nm — Only foundry at volume EUV; requires Lasertec at every reticle stage
  • [inferred] Intel / Samsung: EUV mask inspection — Any EUV user must use Lasertec

Yokogawa Electric (6841.T)

Score: 70.1 | Conviction: HIGH | Layer: Controls / Instrumentation | Weight: 18%

Themes: nuclear | Data Date: 2026-04-12

Thesis: Signed exclusive DPCS contract with Rolls-Royce SMR (Feb 2026) for global fleet control systems. CENTUM VP is the standard DCS platform for Japanese nuclear plants. Near-zero debt (D/E 0.05), 14.6% op margin, 19.2% ROIC — highest quality business in the group.

Risk: Nuclear I&C is a small share of total revenue — incremental impact. Stock +107% in 52 weeks.

Monitoring Trigger: Watch Rolls-Royce SMR UK licensing progress. If approved, major revenue recognition catalyst.

Fundamentals:

  • PE: 24.82 | Fwd PE: 20.73 | P/B: 2.78
  • ROE: 12.2% | Op Margin: 14.6% | D/E: 5%
  • Div Yield: 1.1% | FCF: +¥88B | Mkt Cap: ¥1.44T

Segments:

  • Industrial Automation & Control: ¥500B (85%) — DCS, SCADA, nuclear I&C, process control
  • Measuring Instruments: ¥32B (5%) — Precision test & measurement
  • New Businesses: ¥56B (10%) — Life sciences, new energy

Supply Chain:

  • [confirmed] Rolls-Royce SMR: DPCS for global SMR fleet — Exclusive contract Feb 2, 2026 — UK + Czech Republic delivery
  • [confirmed] Saudi Aramco: CENTUM VP DCS + AI control agents — CPA since 2012. AI agents deployed at Fadhili Oct 2025
  • [confirmed] Shell: Robotics + AI plant maintenance — Long-term agreement Jun 2025
  • [confirmed] ADNOC: ICSS, ESD, fire & gas systems — AED 2.6B ($708M) framework agreement
  • [probable] Japanese nuclear utilities: CENTUM VP DCS for reactor control — Standard platform for J-nuclear plants

Fuji Electric (6504.T)

Score: 69.5 | Conviction: HIGH | Layer: Nuclear Operations / Power Semi | Weight: 15%

Themes: nuclear, ev | Data Date: 2026-04-12

Thesis: 95% domestic market share in nuclear radiation monitoring — every operating J-reactor is a customer. Dual theme: nuclear instrumentation + top-5 global IGBT/SiC for EVs. Confirmed customers: Honda (IGBT), Bosch (SiC JV), Vestas (wind), JNFL Rokkasho (300-channel panel). Both Claude and GPT rated 'agree; keep.'

Risk: Nuclear I&C is a portion of Energy segment, not standalone. SiC competitors (Rohm, Infineon) may erode IGBT share.

Monitoring Trigger: Watch May earnings for nuclear instrumentation order flow and power semi backlog.

Fundamentals:

  • PE: 19.62 | Fwd PE: 17.13 | P/B: 2.12
  • ROE: 11.8% | Op Margin: 10.3% | D/E: 13%
  • Div Yield: 1.6% | FCF: +¥51B | Mkt Cap: ¥1.68T

Segments:

  • Energy: ¥354B (32%) — Nuclear monitoring, power plants, geothermal turbines
  • Industry: ¥400B (36%) — Drives, inverters, FA components
  • Semiconductors: ¥237B (21%) — IGBT, SiC power modules
  • Food & Beverage: ¥112B (10%) — Vending machines (70% J-market share)

Supply Chain:

  • [confirmed] JNFL Rokkasho: 300-channel radiation monitoring panel — Largest in Japan — Fuji Electric product page names JNFL
  • [confirmed] Honda: 14-in-1 IGBT power module for Accord Hybrid — Production since Dec 2012
  • [confirmed] Robert Bosch: SiC power module co-development — Agreement Dec 2025
  • [confirmed] Vestas Wind Systems: Large-capacity IGBT for wind turbines — Partnership since 2017, HPnC X series
  • [confirmed] All Japanese nuclear utilities: Radiation monitoring (95% market share) — Near-monopoly position

TDK Corp (6762.T)

Score: 67.3 | Conviction: HIGH | Layer: Passive Components / Magnets / Sensors | Weight: 15%

Themes: ev | Data Date: 2026-04-12

Thesis: Uniquely positioned with MLCC + NdFeB magnets + MEMS sensors + DC link caps — three separate EV tailwinds in one stock. EUR 150M Hungary NdFeB plant (2024) for European auto OEMs. Forward PE 20.9x reasonable for the breadth.

Risk: HDD magnetic segment faces secular decline. TDK cut dividend -60% YoY signaling cash prioritization. No named production vehicle OEM customers confirmed.

Monitoring Trigger: If automotive + energy combined share exceeds 50% of revenue. Watch GM-CATL battery deal status.

Fundamentals:

  • PE: 23.41 | Fwd PE: 20.94 | P/B: 2.08
  • ROE: 9.5% | Op Margin: 10.1% | D/E: 33%
  • Div Yield: 1.6% | FCF: +¥169B | Mkt Cap: ¥4.38T

Segments:

  • Passive Components: ¥837B (35%) — MLCC, film caps, HF components
  • Energy Application: ¥598B (25%) — Batteries, DC-DC
  • Magnetic Application: ¥478B (20%) — NdFeB magnets, HDD
  • Sensor Application: ¥359B (15%) — MEMS, temperature, pressure

Supply Chain:

  • [confirmed] Porsche (Formula E): MLCCs, sensors, capacitors — Official Technology Partner Oct 2025
  • [confirmed] LEM International: TMR dies for EV current sensors — Development agreement Oct 2023
  • [probable] European Auto OEMs: NdFeB magnets for traction motors — EUR 150M Hungary plant targeting EU OEMs

Murata Manufacturing (6981.T)

Score: 66.7 | Conviction: HIGH | Layer: Passive Components / MLCC | Weight: 15%

Themes: ev | Data Date: 2026-04-12

Thesis: ~50% global automotive MLCC share. EVs use 9,000+ MLCCs (3-5x ICE). Structural volume tailwind regardless of which OEM wins. Near-zero debt (D/E 0.02), ¥239B FCF. Forward PE 25.7x reasonable for dominant franchise.

Risk: Cyclical MLCC pricing. 'Worst semiconductor downturn in a decade' could hit demand short-term.

Monitoring Trigger: If automotive segment revenue share crosses 40%. Watch MLCC ASP recovery in quarterly results.

Fundamentals:

  • PE: 40.25 | Fwd PE: 25.66 | P/B: 2.89
  • ROE: 7.3% | Op Margin: 15.1% | D/E: 2%
  • Div Yield: 1.5% | FCF: +¥239B | Mkt Cap: ¥7.55T

Segments:

  • Components (MLCC): ¥1,070B (60%) — Automotive/industrial MLCC, world #1
  • Devices & Modules: ¥530B (30%) — SAW/BAW filters, RF modules, connectivity
  • Other: ¥180B (10%) — Batteries, sensors

Supply Chain:

  • [confirmed] Tesla: Automotive-grade MLCC (9,000+ per Model 3) — WSJ/Benzinga cited; primary MLCC supplier
  • [confirmed] Apple: Consumer + automotive grade MLCC — Major customer
  • [probable] Toyota/Honda: Automotive MLCC — Murata supplies most Japanese OEMs

Kokusai Electric (6525.T)

Score: 66.4 | Conviction: HIGH | Layer: Deposition Equipment | Weight: 10%

Themes: semiconductor | Data Date: 2026-04-12

Thesis: ~70% global share in batch ALD — essential for GAA transistors and 3D DRAM. Structural tailwind: GAA increases ALD steps per wafer. Won TSMC 2025 Excellent Production Support Award. IPO'd Oct 2023 from KKR.

Risk: China = 47% of revenue. US/Japan export control escalation could trigger immediate order cancellations. Beta 2.12 (highest volatility).

Monitoring Trigger: Japan METI export control list updates. Quarterly China revenue disclosure for pull-forward or cancellation signs.

Fundamentals:

  • PE: 48.96 | Fwd PE: 41.32 | P/B: 7.52
  • ROE: 16.4% | Op Margin: 18.6% | D/E: 26%
  • Div Yield: 0.5% | FCF: +¥35B | Mkt Cap: ¥1.60T

Segments:

  • Batch Film Deposition: ¥164B (69%) — LP-CVD, batch ALD, oxidation, diffusion (70% global share)
  • Parts & Services: ¥73B (31%) — Aftermarket, maintenance

Supply Chain:

  • [confirmed] TSMC: Batch ALD/CVD for logic and advanced nodes — TSMC 2025 Excellent Performance Award
  • [confirmed] Samsung Electronics: DRAM/logic deposition tools — Named in IPO prospectus, top-3 customer
  • [confirmed] Micron Technology: Memory deposition tools — Named in IPO prospectus, top-3 customer

AGC Inc (5201.T)

Score: 66 | Conviction: HIGH | Layer: Automotive Glass | Weight: 8%

Themes: ev | Data Date: 2026-04-12

Thesis: World's #1 automotive glass maker. Every car needs their glass. EV premium products (smart glass, PHUD, ultra-thin) carry higher ASP. P/B 0.71 is deeply undervalued for global #1. Confirmed Tesla supplier. 3.6% yield.

Risk: Cyclical auto production volumes. Chinese glass competitors (Fuyao) aggressive on pricing. EV premium products still small share.

Monitoring Trigger: Auto production data; smart glass design win announcements; AGC automotive margin trend.

Fundamentals:

  • PE: 17.89 | Fwd PE: 14.7 | P/B: 0.71
  • ROE: 4.7% | Op Margin: 6.3% | D/E: 37%
  • Div Yield: 3.6% | FCF: +¥65B | Mkt Cap: ¥1.24T

Segments:

  • Automotive Glass: ¥650B (32%) — World #1. Tesla confirmed supplier
  • Architectural Glass: ¥550B (27%) — Building materials
  • Electronics: ¥470B (23%) — Display glass, substrates
  • Chemicals & Ceramics: ¥370B (18%) — Chlor-alkali, ceramics

Supply Chain:

  • [confirmed] Tesla: Solar control glass, windshields — Confirmed by industry sources, including Cybertruck
  • [confirmed] Toyota/Honda/Nissan: OEM automotive glass — #1 global share = all major OEMs
  • [probable] Mercedes-Benz: Electrochromic smart glass — AGC-compatible tech for Vision V concept

Yaskawa Electric (6506.T)

Score: 64.2 | Conviction: HIGH | Layer: Servo Motors / Industrial Robots | Weight: 15%

Themes: robotics | Data Date: 2026-04-12

Thesis: World's largest servo motor manufacturer (~30% global share) — the 'muscles' of every robot. Humanoid demand multiplier: 20-30 servos per humanoid vs 6-12 for industrial arms. Motoman robots are top-3 global. NVIDIA partnership for AI-enabled robot intelligence.

Risk: Operating margin 8.7% is thin. ~25% China revenue. FCF only ¥5.9B (high capex year).

Monitoring Trigger: Motion Control quarterly order intake. Any humanoid OEM supply agreements.

Fundamentals:

  • PE: 36.06 | Fwd PE: 26.56 | P/B: 2.57
  • ROE: 7.8% | Op Margin: 8.7% | D/E: 25%
  • Div Yield: 1.4% | FCF: +¥6B | Mkt Cap: ¥1.27T

Segments:

  • Motion Control: ¥298B (55%) — Servo motors (~30% global), drives, inverters
  • Robotics: ¥190B (35%) — Motoman industrial robots
  • System Engineering: ¥54B (10%) — Solutions integration

Supply Chain:

  • [confirmed] GM / Toyota / VW: Motoman welding/assembly robots — Major automotive OEM supplier
  • [confirmed] NVIDIA: AI-enabled robot intelligence — Isaac integration partnership
  • [inferred] Humanoid OEMs (Figure, Unitree, etc.): Servo motors for humanoid joints — Every humanoid needs Yaskawa-class servos

Hirose Electric (6806.T)

Score: 58.3 | Conviction: MEDIUM | Layer: Connectors / HV | Weight: 4%

Themes: ev | Data Date: 2026-04-12

Thesis: Highest quality business by margin (20.3% op margin, 42.6% gross margin) with near-zero debt and ¥80B net cash. HVH-280 series for 400V/800V EV battery systems. Every EV needs more connectors than ICE.

Risk: PEG 7.14 — market has priced in strong growth. Thin liquidity (130K shares/day). No confirmed HV powertrain OEM customer named publicly.

Monitoring Trigger: Any 800V platform design win from European OEMs. Automotive connector segment revenue disclosure.

Fundamentals:

  • PE: 23.62 | Fwd PE: 21.44 | P/B: 1.84
  • ROE: 8.0% | Op Margin: 20.3% | D/E: 2%
  • Div Yield: 2.3% | FCF: +¥26B | Mkt Cap: ¥695B

Segments:

  • Multi-Pin Connectors: ¥121B (60%) — Automotive, industrial, EV HV connectors
  • Coaxial Connectors: ¥50B (25%) — RF, antenna, 5G
  • Other: ¥30B (15%) — Custom solutions

Supply Chain:

  • [confirmed] Marelli (Tier 1): Automotive connectors — Sole recipient of Marelli Supplier Excellence Co-Innovation Award 2024
  • [confirmed] Unnamed leading EV OEM: AU1 USB Type-C connector — Press release Mar 2026 — adopted for production
  • [confirmed] Nissan: GT17HN connector in Leaf TCU — Teardown confirmed

IHI Corporation (7013.T)

Score: 58.2 | Conviction: MEDIUM | Layer: Equipment / SMR | Weight: 10%

Themes: nuclear | Data Date: 2026-04-12

Thesis: Japan's most direct SMR supply chain play: X-energy partnership (Mar 2026) positions IHI to supply fabricated nuclear components into the US market. NuScale investor ($20M) and preferred manufacturer. Aero engine segment (GE/P&W JVs) provides high-margin earnings base while nuclear optionality is largely unpriced.

Risk: SMR timelines uncertain. Stock +145% in 52 weeks — good news may be priced in. Aero segment exposed to PW1100G engine inspection costs.

Monitoring Trigger: If X-energy or NuScale reach construction milestones, increase. If stock corrects >20%, consider adding.

Fundamentals:

  • PE: 28.94 | Fwd PE: 28.02 | P/B: 5.92
  • ROE: 23.6% | Op Margin: 9.1% | D/E: 108%
  • Div Yield: 0.6% | FCF: +¥68B | Mkt Cap: ¥3.5T

Segments:

  • Aero, Space & Defense: ¥556B (34%) — Jet engines (GE/P&W JVs), GCAP fighter engine, rocket motors
  • Resources, Energy & Environment: ¥411B (25%) — Nuclear components, boilers, carbon solutions
  • Industrial Systems: ¥485B (30%) — Engines, industrial equipment
  • Social Infrastructure: ¥162B (10%) — Bridges, water gates

Supply Chain:

  • [confirmed] X-energy: SMR component fabrication — MOU signed March 2026 for US-Japan SMR supply chain
  • [confirmed] NuScale Power: SMR preferred manufacturer — $20M investment + JNI JV with JGC
  • [confirmed] GE Vernova: Aero engine JV partner — Long-standing CF34/GEnx engine programs
  • [probable] Japanese nuclear utilities: Reactor vessels, containment — Historical domestic nuclear supplier

Fanuc (6954.T)

Score: 56 | Conviction: HIGH | Layer: CNC / Industrial Robots | Weight: 20%

Themes: robotics | Data Date: 2026-04-12

Thesis: Picks-and-shovels monopolist of global manufacturing: ~60% CNC global share, ~25% industrial robot share. Zero debt (¥697B net cash). Apple iPhone machining via Robodrill confirmed. NVIDIA AI robotics initiative. CRX collaborative robots piloted in care facilities for 570K worker gap.

Risk: ~25% China revenue. Capex cycle lumpiness. Forward PE 31x on thin growth.

Monitoring Trigger: Monthly China PMI > 52 = bullish signal. Watch quarterly China revenue %.

Fundamentals:

  • PE: 36.08 | Fwd PE: 31.14 | P/B: 3.22
  • ROE: 9.4% | Op Margin: 21.1% | D/E: 0%
  • Div Yield: 1.6% | FCF: N/A | Mkt Cap: ¥5.83T

Segments:

  • FA (CNC, lasers, Robodrill): ¥334B (40%) — CNC controllers ~60% global share
  • Robots: ¥251B (30%) — Industrial arms, SCARA, collaborative (CRX)
  • Robomachines: ¥251B (30%) — Robodrill, Robocut wire EDM

Supply Chain:

  • [confirmed] Apple: Robodrill for iPhone machining — Extensively documented
  • [confirmed] Toyota / Honda / Denso: Welding, assembly robots — Major auto OEM customer base
  • [confirmed] NVIDIA: AI robotics initiative (Isaac integration) — Joint announcement

Kurita Water Industries (6370.T)

Score: 55.3 | Conviction: MEDIUM | Layer: Nuclear Operations / Water | Weight: 8%

Themes: nuclear | Data Date: 2026-04-12

Thesis: Dual nuclear + semiconductor water chemistry exposure. Nuclear: coolant chemistry, corrosion inhibitors, radioactive waste minimization. Semiconductor: ultrapure water for Samsung/SK Hynix (confirmed) and TSMC Japan (probable). 16 consecutive years of dividend growth signals discipline.

Risk: Trailing PE 38x is stretched. ROE 6.8% is low. Depends on semiconductor + nuclear materializing simultaneously.

Monitoring Trigger: If fwd PE expands >25x without earnings revision, trim to 5%.

Fundamentals:

  • PE: 38.22 | Fwd PE: 21.67 | P/B: 2.52
  • ROE: 6.8% | Op Margin: 13.7% | D/E: 32%
  • Div Yield: 1.4% | FCF: +¥25B | Mkt Cap: ¥888B

Segments:

  • Electronics Industry: ¥200B (49%) — Ultrapure water chemicals for semiconductor fabs
  • General Industry: ¥212B (51%) — Power plant water chemistry, industrial water treatment

Supply Chain:

  • [confirmed] Samsung Electronics / SK Hynix: Ultrapure water chemicals for Korean fabs — Kurita Korea major supplier per geographic disclosures
  • [probable] Japanese nuclear utilities: Coolant chemistry programs, corrosion inhibitors — Serves all major power generation verticals
  • [probable] TSMC Japan fabs: Ultrapure water systems — Electronics Industry segment primary growth driver

Renesas Electronics (6723.T)

Score: 54.6 | Conviction: MEDIUM | Layer: Chip Design / MCU | Weight: 8%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: De facto embedded intelligence for cars globally — auto digitization requires more Renesas MCUs per vehicle. R-Car X5H is industry's first 3nm automotive SoC. Expanding into AI data center power (800V GaN, confirmed NVIDIA partnership). Operating FCF ¥364B despite trailing IFRS loss.

Risk: ~55% automotive revenue = sensitive to global auto production. Trailing loss from Intersil/IDT/Dialog acquisition goodwill impairment — IFRS item, not operating.

Monitoring Trigger: R-Car Gen 5/X5H design wins at Tier-1s (Bosch, Denso). These signal 3-5 year revenue visibility.

Fundamentals:

  • PE: Loss | Fwd PE: 17.78 | P/B: 1.97
  • ROE: -2.1% | Op Margin: 17.9% | D/E: 50%
  • Div Yield: 1.1% | FCF: +¥364B | Mkt Cap: ¥4.82T

Segments:

  • Automotive: ¥770B (55%) — RH850 MCU, R-Car ADAS SoC, body control
  • Industrial/IoT: ¥630B (45%) — RA/RX MCU, analog, power management, DDR5 RCD

Supply Chain:

  • [confirmed] Toyota / Toyota Group: RH850 MCU for ECUs — Dominant MCU in Toyota vehicles
  • [confirmed] Honda / Stellantis / BMW: R-Car ADAS SoC — Disclosed customer wins
  • [confirmed] NVIDIA: 800V DC AI data center power (GaN) — Partnership announced

Sumitomo Metal Mining (5713.T)

Score: 53 | Conviction: MEDIUM | Layer: Battery Materials / Cathode | Weight: 8%

Themes: ev | Data Date: 2026-04-12

Thesis: Sole Japanese NCA cathode source at scale, directly in Panasonic→Tesla chain. Capacity tripling: 60K→120K→180K tonnes by 2031 ($424M invested). Closed-loop nickel recycling JV with Panasonic (Mar 2025). Forward PE 16.5x attractive at nickel trough.

Risk: Nickel price volatility is primary P&L driver — net income fell from ¥248B to ~¥95B on metals prices alone. +316% 52W run.

Monitoring Trigger: LME nickel price + Panasonic battery shipments to Tesla. Any US capacity announcements.

Fundamentals:

  • PE: 28.63 | Fwd PE: 16.53 | P/B: 1.28
  • ROE: 5.0% | Op Margin: 5.0% | D/E: 30%
  • Div Yield: 2.4% | FCF: +¥87B | Mkt Cap: ¥2.70T

Segments:

  • Smelting & Refining: ¥900B (55%) — Copper, nickel, cobalt refining
  • Mineral Resources: ¥430B (26%) — Gold, copper mines
  • Materials: ¥320B (19%) — NCA cathode, nickel chemicals

Supply Chain:

  • [confirmed] Panasonic Energy: NCA cathode active material — Long-term supply + joint recycling JV (Mar 2025)
  • [confirmed] Tesla (via Panasonic): NCA cathode for 2170/4680 cells — Panasonic cells for Tesla use SMM NCA

Asahi Kasei (3407.T)

Score: 53 | Conviction: MEDIUM | Layer: Battery Materials / Separator | Weight: 7%

Themes: ev | Data Date: 2026-04-12

Thesis: Classic trough-to-recovery: Hipore separator EBITDA collapsed from $170M to <$20M due to Chinese pricing. But $1B+ North American plants (Canada + NC) with confirmed Toyota Tsusho offtake (Jul 2025) provide 2027 recovery catalyst. PE 14.1x at trough.

Risk: Hipore near breakeven through 2026. Chinese separators could follow into North America. Homes/Healthcare segments mask separator weakness.

Monitoring Trigger: Charlotte, NC facility commissioning timeline. Additional capacity-rights agreements. Separator segment EBITDA recovery.

Fundamentals:

  • PE: 14.12 | Fwd PE: 13.3 | P/B: 1.07
  • ROE: N/A | Op Margin: 7.3% | D/E: 55%
  • Div Yield: 2.5% | FCF: +¥134B | Mkt Cap: ¥2.22T

Segments:

  • Materials: ¥1,520B (50%) — Hipore separator, resins, fibers
  • Homes: ¥760B (25%) — Hebel Haus residential
  • Healthcare: ¥760B (25%) — Zoll medical, pharma

Supply Chain:

  • [confirmed] Toyota Tsusho / Toyota battery chain: Coated Hipore separator — Signed capacity rights Jul 2025 for North American supply from 2027
  • [probable] Korean battery makers (LG, Samsung SDI): Hipore wet-process separator — Historical supply, North America pivot

Sumitomo Heavy Industries (6302.T)

Score: 52.1 | Conviction: MEDIUM | Layer: Nuclear Medicine / Fusion | Weight: 10%

Themes: nuclear | Data Date: 2026-04-12

Thesis: Only company with confirmed ITER fusion cryocooler contracts — world's largest cryocooler manufacturer. Also sole Japanese PET cyclotron maker (100+ deliveries) and world-first BNCT cancer treatment (NeuCure). Trading at 0.92x book is a value anomaly given the cryocooler technology moat.

Risk: Nuclear/fusion is a minority of revenue. Low ROE (4.7%) and margins (4.8%) from industrial machinery segments dilute returns.

Monitoring Trigger: Track BNCT installation announcements and ITER construction milestones. If no progress in 12 months, reduce to 5%.

Fundamentals:

  • PE: 20.46 | Fwd PE: 18.63 | P/B: 0.92
  • ROE: 4.7% | Op Margin: 4.8% | D/E: 37%
  • Div Yield: 2.8% | FCF: +¥11B | Mkt Cap: ¥629B

Segments:

  • Mechatronics: ¥200B (19%) — Cryocoolers (ITER, MRI), injection molding machines
  • Industrial Machinery: ¥350B (33%) — Cyclotrons, accelerators, forging presses
  • Logistics & Construction: ¥280B (26%) — Cranes, excavators
  • Energy & Lifelines: ¥240B (22%) — Power turbines, water treatment

Supply Chain:

  • [confirmed] ITER Organization: Cryocoolers for fusion reactor cryogenic system — Specified components in ITER procurement
  • [confirmed] KEK (J-PARC): Cryocoolers for particle accelerators — SuperKEKB and J-PARC facilities
  • [probable] GE HealthCare / Siemens MRI: Cryocoolers for MRI superconducting magnets — SHI is world's largest cryocooler maker; GE/Siemens are dominant MRI OEMs

Omron (6645.T)

Score: 50.8 | Conviction: MEDIUM | Layer: Automation Components / PLCs | Weight: 10%

Themes: robotics | Data Date: 2026-04-12

Thesis: Quality Japanese FA franchise trading at P/B 0.94 — priced for distress. IAB segment (45% rev) has best-in-class sensors, PLCs, safety. China FA recovery from -22.6% trough would drive margin recovery from 6% toward historical 12-14%. Healthcare (blood pressure) levered to aging demographics.

Risk: Payout ratio 87% on thin FCF. Z-Score 2.54 (borderline). China demand recovery timing uncertain.

Monitoring Trigger: IAB quarterly order intake turning positive YoY (May 2026 earnings). China PMI > 51.

Fundamentals:

  • PE: 39.25 | Fwd PE: 21.18 | P/B: 0.94
  • ROE: 2.3% | Op Margin: 6.1% | D/E: 24%
  • Div Yield: 2.2% | FCF: +¥6B | Mkt Cap: ¥919B

Segments:

  • Industrial Automation (IAB): ¥376B (45%) — PLCs (Sysmac), sensors, safety, cobots, AMRs
  • Healthcare: ¥209B (25%) — Blood pressure, cardiac monitoring
  • Social Systems: ¥125B (15%) — Rail, energy management
  • Devices & Modules: ¥125B (15%) — Relays, switches, PCBs

Supply Chain:

  • [confirmed] Toyota: Safety systems for manufacturing — Long-standing supplier
  • [confirmed] Panasonic: Manufacturing line automation — Sensors, PLCs
  • [probable] TSMC / Samsung fabs: Fab automation sensors — Vision, proximity sensors for fab equipment

Hitachi (6501.T)

Score: 50.4 | Conviction: MEDIUM | Layer: Equipment / SMR | Weight: 10%

Themes: nuclear | Data Date: 2026-04-12

Thesis: Holds the most commercially advanced SMR asset globally through GE Hitachi Nuclear Energy (50/50 JV) developing the BWRX-300 — the only SMR with contracted projects (Canada, Poland, US). $40B Japan-US deal earmarks tens of billions for GEH reactors. Digital segment provides growth while nuclear is the optionality.

Risk: Nuclear is a small sub-segment of a ¥10T conglomerate — limited direct earnings leverage. Large-cap (¥21.6T) means nuclear catalysts move the needle slowly.

Monitoring Trigger: Watch BWRX-300 construction progress at Darlington (Canada). If first SMR connects to grid, major re-rating catalyst.

Fundamentals:

  • PE: 26.64 | Fwd PE: 25.37 | P/B: 3.3
  • ROE: N/A | Op Margin: 11.1% | D/E: 16%
  • Div Yield: 1.0% | FCF: +¥1,390B | Mkt Cap: ¥21.6T

Segments:

  • Digital Systems & Services: ¥3,500B (34%) — IT, cloud, consulting — highest growth
  • Green Energy & Mobility: ¥3,000B (29%) — Power grids, GEH nuclear JV, rail
  • Connective Industries: ¥3,800B (37%) — Industrial IoT, manufacturing automation

Supply Chain:

  • [confirmed] GE Vernova (50/50 JV): BWRX-300 SMR development and deployment — March 2026 MOU for Southeast Asia expansion
  • [confirmed] Ontario Power Generation (Canada): BWRX-300 reactor — Contracted — Darlington site construction started May 2025
  • [confirmed] SaskPower (Canada): BWRX-300 reactor — LOI for SMR deployment
  • [probable] Southeast Asian utilities: BWRX-300 exploration — March 2026 MOU to explore deployment

Mitsubishi Electric (6503.T)

Score: 47.2 | Conviction: MEDIUM | Layer: Power Semiconductors / IGBT | Weight: 7%

Themes: ev | Data Date: 2026-04-12

Thesis: Top-5 global IGBT maker. MAJOR CATALYST: talks to merge power semiconductor business with Rohm and Toshiba (Mar 2026) — could create world's #2 behind Infineon. Massive FCF (¥317B). Near-zero debt.

Risk: Low operating margin (6.5%) from conglomerate overhead. PE 30x reflects premium. Merger talks could fail.

Monitoring Trigger: Official merger announcement for power semi unit. Semiconductor segment revenue in quarterly results.

Fundamentals:

  • PE: 31.86 | Fwd PE: 29.62 | P/B: 2.77
  • ROE: N/A | Op Margin: 6.5% | D/E: 8%
  • Div Yield: 1.0% | FCF: +¥317B | Mkt Cap: ¥11.9T

Segments:

  • Industry & Mobility: ¥1,700B (30%) — FA systems, EV powertrain
  • Life: ¥1,420B (25%) — HVAC, elevators
  • Infrastructure: ¥1,140B (20%) — Rail, power grid
  • Semiconductor & Device: ¥570B (10%) — IGBT, SiC power modules

Supply Chain:

  • [confirmed] Toyota: Powertrain systems — Long-standing supplier
  • [confirmed] Mitsubishi Motors: EV/PHEV power modules — Keiretsu relationship
  • [probable] Stellantis/Renault-Nissan: IGBT modules for inverters — Global module supply

Keyence (6861.T)

Score: 46.9 | Conviction: MEDIUM | Layer: Factory Automation / AI Vision | Weight: 8%

Themes: ai, robotics | Data Date: 2026-04-12

Thesis: Most profitable industrial company globally — 50.9% operating margin. AI vision/sensing embedded across factory floors building AI hardware. IV3 AI vision system, XG-X deep-learning QC. Diversified AI infrastructure exposure without direct chip-cycle volatility. ¥1.33T net cash.

Risk: PEG 4.03 is expensive for growth rate. Only +19% 52W vs peers at +150-380% — market prices it as compounder, not AI growth name.

Monitoring Trigger: Japan/Korea/Taiwan factory automation capex trends. China FA revenue exposure.

Fundamentals:

  • PE: 36.22 | Fwd PE: 32.31 | P/B: 4.54
  • ROE: N/A | Op Margin: 50.9% | D/E: 0%
  • Div Yield: 0.9% | FCF: N/A | Mkt Cap: ¥15.14T

Segments:

  • Sensors/Switches: ¥336B (30%) — Proximity, photoelectric, fiber optic
  • Vision Systems: ¥280B (25%) — AI machine vision, deep learning QC
  • Laser Markers: ¥224B (20%) — CO2/fiber/UV lasers
  • Measurement: ¥168B (15%) — 3D profilometers, CMM
  • Safety: ¥112B (10%) — Light curtains, safety controllers

Supply Chain:

  • [confirmed] Toyota / major auto OEMs: FA vision and sensor systems — Largest industrial automation customer base in Japan
  • [probable] Semiconductor fabs globally: Vision systems for wafer/PCB inspection — Widely used in fab QC
  • [inferred] Electronics manufacturers: AI defect detection on production lines — Samsung, Foxconn installed base

Sony Group (6758.T)

Score: 46.1 | Conviction: MEDIUM | Layer: Motion Capture Hardware / Anime IP | Weight: 15%

Themes: vtuber | Data Date: 2026-04-12

Thesis: Only major TSE-listed company manufacturing consumer motion capture hardware (mocopi). Vertically integrated: hardware (mocopi/XYN Motion Studio) + content (Aniplex, Crunchyroll) + distribution (PlayStation). Mocopi at $450 democratizes VTuber and anime production.

Risk: Mocopi is a tiny segment of Sony's ¥13T business. Conglomerate discount. AI-based software mocap solutions could commoditize hardware approach.

Monitoring Trigger: If mocopi unit sales reported or XYN studio adoption by major anime studios, upgrade conviction.

Fundamentals:

  • PE: 15.9x | Fwd PE: 18.4x | P/B: 2.1x
  • ROE: 14.9% | Op Margin: 9.7% | D/E: 19%
  • Div Yield: 0.7% | FCF: N/A | Mkt Cap: ¥14.4T

Segments:

  • Game & Network Services: ¥4.3T (34%) — PlayStation, game publishing
  • Music: ¥1.8T (14%) — Sony Music, Aniplex anime
  • Pictures: ¥1.6T (13%) — Sony Pictures, Crunchyroll
  • I&SS: ¥1.6T (13%) — Image sensors, mocopi hardware
  • ET&S / Financial: ¥3.3T (26%) — Electronics, financial services

Supply Chain:

  • [confirmed] Cover Corp (Hololive): mocopi motion capture sensors — Hololive talents showcased mocopi integration
  • [probable] Anime studios: XYN Motion Studio — Used in Solo Leveling anime production
  • [confirmed] Global consumers: mocopi (6-sensor kit) — Retail at $450, consumer mocap

Kawasaki Heavy Industries (7012.T)

Score: 45.9 | Conviction: MEDIUM | Layer: Industrial Robots / Humanoid | Weight: 10%

Themes: robotics | Data Date: 2026-04-12

Thesis: Defense rerate story with robotics option. Japan defense budget doubling to ¥10T+ by FY2027. Kaleido humanoid (fire/rescue/heavy labor) is most advanced Japanese humanoid. Also makes submarines, aircraft. +162% 52W.

Risk: D/E 1.19 is high. Fwd PE 32x above trailing 25.8x — defense ramp priced in. Execution risk on government programs.

Monitoring Trigger: Japan defense supplemental budget. Kaleido commercial deployment contracts. Boeing production rate increases.

Fundamentals:

  • PE: 25.77 | Fwd PE: 32.17 | P/B: 3.23
  • ROE: N/A | Op Margin: 5.3% | D/E: 119%
  • Div Yield: 0.9% | FCF: +¥47B | Mkt Cap: ¥2.83T

Segments:

  • Aerospace Systems: ¥627B (28%) — Boeing substructures, defense aircraft
  • Energy & Marine: ¥456B (20%) — LNG ships, submarines, hydrogen
  • Precision Machinery & Robot: ¥251B (11%) — Kawasaki Robotics, Kaleido humanoid
  • Defense: ¥365B (16%) — P-1 patrol, submarines
  • Powersports: ¥365B (16%) — Motorcycles, jet ski
  • Rolling Stock: ¥205B (9%) — Trains

Supply Chain:

  • [confirmed] Boeing: Fuselage sections — Long-standing substructure supplier
  • [confirmed] Japan MOD: P-1 patrol aircraft, submarines — Primary defense contractor
  • [probable] Auto/food/pharma OEMs: Industrial robots — Welding, painting, palletizing

Nidec Corp (6594.T)

Score: 45.5 | Conviction: LOW | Layer: Motors / E-Axle | Weight: 5%

Themes: ev, robotics | Data Date: 2026-04-12

Thesis: E-Axle traction motor leader with strategic position in highest-value EV drivetrain component. Forward PE 13.7x with PEG 0.9 prices in no growth — asymmetric if governance resolves. Revenue +5.3%, net income +26.4%.

Risk: CRITICAL: Active securities fraud investigation (Mar 2026). Dividend cancelled. ROIC 4% below WACC 7.7%. Altman Z-Score 2.75. Highest risk in the group.

Monitoring Trigger: Corrective action plan post-investigation. Any financial restatement. Oasis governance proxy outcome. Do NOT increase until resolved.

Fundamentals:

  • PE: 21.96 | Fwd PE: 13.74 | P/B: 1.48
  • ROE: 6.3% | Op Margin: 5.0% | D/E: 40%
  • Div Yield: 0% | FCF: +¥174B | Mkt Cap: ¥2.63T

Segments:

  • Automotive Components: ¥786B (30%) — E-Axle, EPS motors, thermal
  • Medium/Large Motors: ¥655B (25%) — Industrial motors
  • Small Precision Motors: ¥524B (20%) — HDD, consumer
  • Appliance/Commercial: ¥655B (25%) — HVAC, appliance motors

Supply Chain:

  • [inferred] Chinese NEV OEMs (BYD etc.): E-Axle traction motors — Nidec announced China supply but no named OEM confirmed
  • [probable] Stellantis: E-Axle for European EVs — Announced partnership

DISCO Corp (6146.T)

Score: 42.7 | Conviction: MEDIUM | Layer: Dicing / Grinding | Weight: 8%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: Near-monopoly in precision dicing/grinding (~70%+ global share). Every advanced AI package (CoWoS, HBM, chiplet) requires DISCO equipment. Operating margin 41.7% reflects pricing power. No credible at-scale competitor.

Risk: Forward PE 45x is expensive. Small company (5,256 employees) — capacity scaling constraints. FCF data unavailable.

Monitoring Trigger: TSMC CoWoS capacity additions (each line needs DISCO tools). HBM unit shipment volumes.

Fundamentals:

  • PE: 55.44 | Fwd PE: 44.98 | P/B: 13.37
  • ROE: N/A | Op Margin: 41.7% | D/E: 0%
  • Div Yield: 0.7% | FCF: N/A | Mkt Cap: ¥7.26T

Segments:

  • Dicing Saws: ¥234B (55%) — Die singulation for advanced packaging
  • Grinding/Polishing: ¥127B (30%) — Wafer thinning for 3D IC stacking
  • Blades/Consumables: ¥64B (15%) — High-margin recurring

Supply Chain:

  • [confirmed] TSMC: Dicing for CoWoS packaging — Est. 70%+ global dicing share; NVIDIA H100/B100 packages
  • [confirmed] SK Hynix / Samsung: HBM die singulation — All HBM production requires DISCO dicing
  • [inferred] ASE / Amkor: OSAT advanced packaging — Leading OSATs use DISCO equipment

Japan Steel Works (5631.T)

Score: 42.6 | Conviction: HIGH | Layer: Materials / Forgings | Weight: 17%

Themes: nuclear | Data Date: 2026-04-12

Thesis: Near-monopoly on ultra-large nuclear-grade steel forgings (reactor pressure vessels, steam generator shells). Only manufacturer globally with this capability. Backlog >¥135B. METI confirmed as BWRX-300 supplier. ¥28B CAPEX for SMR capacity expansion. 18-month manufacturing cycle = multi-year visibility.

Risk: FCF turned negative (-¥19.7B) in FY2025 due to capex. PE 35x on negative cash flow is expensive. Industrial machinery segment exposed to weak China demand.

Monitoring Trigger: If operating CF turns positive in May 2026 earnings, increase to 20%. If negative for second year, reduce to 12%.

Fundamentals:

  • PE: 35.28 | Fwd PE: 34.31 | P/B: 3.53
  • ROE: 10.5% | Op Margin: 8.4% | D/E: 34%
  • Div Yield: 0.9% | FCF: -¥19.7B | Mkt Cap: ¥724B

Segments:

  • Industrial Machinery: ¥160B (58%) — Plastic processing machines, extruders
  • Materials & Engineering: ¥117B (42%) — Nuclear forgings, defense, energy plant equipment

Supply Chain:

  • [confirmed] GE Hitachi / BWRX-300: Reactor pressure vessel forgings — METI named JSW as BWRX-300 supplier (Mar 2026)
  • [probable] Westinghouse / AP1000: RPV and steam generator forgings — Historical AP1000 forging supplier
  • [probable] Japanese nuclear utilities: Replacement forgings for restarts — All J-reactors historically sourced from JSW

Harmonic Drive Systems (6324.T)

Score: 42.5 | Conviction: HIGH | Layer: Precision Gears (Humanoid Critical) | Weight: 10%

Themes: robotics | Data Date: 2026-04-12

Thesis: Makes precision strain-wave gears — THE critical joint component for ALL humanoid robots. ¥2.5B humanoid orders in FY3/26, doubling to ¥5B in FY3/27. Every humanoid OEM globally (Figure, Unitree, Tesla Optimus, Kawasaki Kaleido) needs Harmonic Drive gears. Pure component play on humanoid volume.

Risk: Small company. Humanoid market still nascent (50-100K units globally). If humanoid adoption delays, orders won't materialize.

Monitoring Trigger: Quarterly humanoid order intake. Any named humanoid OEM supply agreements. Global humanoid shipment forecasts.

Fundamentals:

  • PE: N/A | Fwd PE: N/A | P/B: N/A
  • ROE: N/A | Op Margin: N/A | D/E: N/A
  • Div Yield: N/A | FCF: N/A | Mkt Cap: ~¥300B

Segments:

  • Harmonic Drive Gears: N/A (~80%) — Strain-wave precision gears for robot joints (monopoly-like)
  • Other Motion Control: N/A (~20%) — Actuators, servo gears

Supply Chain:

  • [probable] Figure AI: Humanoid joint gears — All humanoid OEMs use Harmonic Drive or Nabtesco gears
  • [probable] Unitree Robotics: Precision gears for humanoid joints — Chinese humanoid OEM
  • [confirmed] Industrial robot OEMs (Fanuc, Yaskawa, ABB): Joint gears for industrial arms — Supplied to all major robot OEMs

Sumitomo Electric (5802.T)

Score: 42 | Conviction: MEDIUM | Layer: Wiring / HV Cables | Weight: 4%

Themes: ev | Data Date: 2026-04-12

Thesis: Dominant Japanese wire harness supplier — EV content ~30% higher value than ICE due to HV shielding. Confirmed Toyota, Honda, Nissan OEM supply. ROE 11%.

Risk: +466% 52W run is extreme. Piotroski F-Score 3 (weakest). FCF not reported. Net debt ¥454B. Mean-reversion risk.

Monitoring Trigger: EV harness order intake commentary. FCF reporting resumption. Japan-China tariff impacts on harness production.

Fundamentals:

  • PE: 32.07 | Fwd PE: 24.12 | P/B: 3.1
  • ROE: 11.0% | Op Margin: 7.8% | D/E: 30%
  • Div Yield: 1.1% | FCF: N/A | Mkt Cap: ¥8.25T

Segments:

  • Automotive: ¥1,972B (40%) — Wire harness, HV cables for EV
  • Infocommunications: ¥740B (15%) — Fiber optic, telecom
  • Environment & Energy: ¥740B (15%) — Power cable, EV charging cable
  • Electronics: ¥740B (15%) — PCB, electronic wires
  • Industrial Materials: ¥740B (15%) — Metals, alloys

Supply Chain:

  • [confirmed] Toyota: Primary wire harness supplier — Long-standing OEM relationship
  • [confirmed] Honda: Wire harness — Long-standing OEM
  • [confirmed] Nissan: HV harness for Leaf/Ariya — EV-specific HV harness supply

Bandai Namco (7832.T)

Score: 42 | Conviction: MEDIUM | Layer: Anime IP / Game Ecosystem | Weight: 12%

Themes: vtuber | Data Date: 2026-04-12

Thesis: Largest anime IP portfolio in Japan (Gundam, Dragon Ball, One Piece games). Natural partner for VTuber collaborations — already partners with ANYCOLOR for game integrations. IP monetization flywheel across games, toys, and virtual content.

Risk: IP licensing revenue cyclical with anime popularity. Metaverse strategy execution risk.

Monitoring Trigger: If VTuber-specific game or metaverse product launches with meaningful revenue, upgrade conviction.

Fundamentals:

  • PE: 22.2x | Fwd PE: 18.6x | P/B: N/A
  • ROE: 13.8% | Op Margin: N/A | D/E: ~0%
  • Div Yield: 1.5% | FCF: N/A | Mkt Cap: ¥2.5T

Segments:

  • Digital: ¥450B (35%) — Mobile and console games
  • Toys & Hobby: ¥380B (29%) — Gunpla, figures, collectibles
  • IP Creation: ¥280B (22%) — Anime production, IP licensing
  • Amusement: ¥180B (14%) — Arcades, location-based entertainment

Supply Chain:

  • [confirmed] ANYCOLOR (Nijisanji): VTuber game collaboration — Official game integration partnerships
  • [confirmed] Toei Animation: Anime IP co-production — Dragon Ball, One Piece game rights

CyberAgent (4751.T)

Score: 40.7 | Conviction: MEDIUM | Layer: Mocap Studio / Streaming Platform | Weight: 12%

Themes: vtuber | Data Date: 2026-04-12

Thesis: Cygames subsidiary operates one of Asia's largest motion capture studios, rivaling Hollywood facilities. ABEMA streaming platform provides distribution for virtual content. Digital advertising expertise enables VTuber marketing integrations.

Risk: ABEMA continues to burn cash. Gaming revenue is cyclical. Mocap studio is a cost center, not a direct revenue driver.

Monitoring Trigger: If ABEMA reaches breakeven or Cygames announces VTuber-specific product, upgrade conviction.

Fundamentals:

  • PE: N/A | Fwd PE: N/A | P/B: N/A
  • ROE: 16.8% | Op Margin: 9.3% | D/E: N/A
  • Div Yield: 1.0% | FCF: N/A | Mkt Cap: ¥721B

Segments:

  • Internet Advertising: ¥430B (41%) — Digital advertising agency
  • Game (Cygames): ¥310B (30%) — Mobile/console games, mocap studio
  • Media (ABEMA): ¥300B (29%) — Streaming platform, virtual content

Supply Chain:

  • [confirmed] Cygames: In-house mocap studio for game/anime — One of Asia's largest mocap facilities
  • [confirmed] ABEMA: Streaming distribution — Platform for virtual content and events

Ibiden (4062.T)

Score: 39.8 | Conviction: HIGH | Layer: IC Package Substrates | Weight: 10%

Themes: semiconductor | Data Date: 2026-04-12

Thesis: Critical AI chip chokepoint — one of only 2-3 companies globally making high-end ABF substrates. Without Ibiden substrates, NVIDIA cannot manufacture GPUs. AI server substrates >15% of revenue and growing. 2x capacity expansion (Ono Plant) underway. +526% 52W.

Risk: FCF negative from massive ¥420B capex program. If AI demand disappoints, stranded overcapacity. Ceramics segment (auto DPF) declining from EV adoption.

Monitoring Trigger: NVIDIA quarterly GPU shipments (each H100/B200 needs one Ibiden substrate). Ono Plant ramp updates at May earnings.

Fundamentals:

  • PE: 72.11 | Fwd PE: 44.91 | P/B: 5.02
  • ROE: 7.6% | Op Margin: 13.4% | D/E: 55%
  • Div Yield: 0.2% | FCF: -¥8B | Mkt Cap: ¥2.76T

Segments:

  • Electronics: ¥343B (86%) — ABF IC package substrates (NVIDIA, AMD, Intel), PCBs
  • Ceramics: ¥55B (14%) — DPF (diesel), insulation, graphite

Supply Chain:

  • [confirmed] NVIDIA: ABF substrates for H100/H200/B100 GPUs — All NVIDIA AI GPUs use Ibiden substrates
  • [confirmed] Intel: ABF substrates for Xeon/Core — Historical anchor ~30% of substrate revenue
  • [confirmed] AMD: ABF substrates for EPYC/Instinct — Primary supplier alongside Unimicron
  • [probable] Hyperscaler ASICs (Google TPU, Amazon): Custom ASIC substrates — Rapid growth in ASIC orders per management

Resonac Holdings (4004.T)

Score: 39.4 | Conviction: LOW | Layer: SiC Substrates | Weight: 5%

Themes: ev, semiconductor | Data Date: 2026-04-12

Thesis: Upstream chokepoint for entire non-Chinese SiC supply chain — every SiC inverter chip starts as a Resonac substrate. Wolfspeed bankruptcy created supply anxiety. 200mm mass production starting 2025. Soitec JV for bonded SiC.

Risk: Highest leverage in basket (D/E 1.33). Altman Z-Score 1.51 (distress). Trailing PE 83x. Chinese SiC competitors (TanKeBlue, SICC) holding 17%+ share each.

Monitoring Trigger: SiC substrate pricing (TrendForce monthly). Semiconductor segment quarterly margin. Named 8-inch supply agreements.

Fundamentals:

  • PE: 83.53 | Fwd PE: 28.8 | P/B: 3.34
  • ROE: 4.4% | Op Margin: 4.3% | D/E: 133%
  • Div Yield: 0.5% | FCF: +¥24B | Mkt Cap: ¥2.43T

Segments:

  • Semiconductor & Electronic Materials: ¥499B (35%) — SiC substrates, CMP, photoresist
  • Mobility: ¥285B (20%) — Auto chemicals
  • Innovation Enabling: ¥285B (20%) — Advanced materials
  • Chemicals: ¥356B (25%) — Petrochemicals, industrial gases

Supply Chain:

  • [confirmed] Rohm: SiC epitaxial wafers for power chips — Primary Japanese SiC substrate customer
  • [probable] Infineon / onsemi / STMicro: SiC substrates — Ships to automakers in NA and EU
  • [confirmed] Soitec (France): Bonded SiC wafer JV — Joint development Sept 2024

Kureha Corp (4023.T)

Score: 34.3 | Conviction: MEDIUM | Layer: Battery Materials / PVDF Binder | Weight: 8%

Themes: ev | Data Date: 2026-04-12

Thesis: Global #1-2 in PVDF cathode binder — present in every NMC/NCA battery globally. P/B 0.86 (below book). 5.3% yield. Only Japanese-domestic PVDF producer at scale.

Risk: Negative FCF (-¥4B) from capex cycle. Altman Z-Score 1.75 (distress signal). Chinese PVDF producers aggressively undercutting. Dividend funded by debt.

Monitoring Trigger: PVDF spot pricing China vs Japan. If FCF negative 2+ more quarters, reduce. If dividend cut, exit.

Fundamentals:

  • PE: 17.89 | Fwd PE: 15.73 | P/B: 0.86
  • ROE: 5.0% | Op Margin: 7.6% | D/E: 64%
  • Div Yield: 5.3% | FCF: -¥4B | Mkt Cap: ¥158B

Segments:

  • Advanced Materials: ¥68B (43%) — PVDF binder (#1 globally), PPS, carbon fiber
  • Specialty Chemicals: ¥45B (29%) — Renal drugs, agrochemicals
  • Specialty Plastics: ¥25B (16%) — Engineering plastics
  • Other: ¥20B (12%) — Construction, services

Supply Chain:

  • [probable] Major LIB manufacturers (CATL, Panasonic, Samsung SDI, LG): PVDF cathode binder — Quasi-commodity but Kureha-grade is specialty qualified
  • [probable] Japanese battery makers: PVDF binder — Domestic qualification history

Kanadevia (7004.T)

Score: 33 | Conviction: MEDIUM | Layer: Decommissioning / Waste | Weight: 10%

Themes: nuclear | Data Date: 2026-04-12

Thesis: Only pure-play spent nuclear fuel logistics company on TSE. 40+ year track record in transport/storage casks (since 1978). TEPCO JV for Fukushima decommissioning. US subsidiary NAC International provides dollar-denominated revenue. Decarbonization segment rebranding captures ESG capital.

Risk: Negative FCF (-¥12.8B) and high D/E (0.88). Low op margin (2%). Contract slippage risk on EPC projects.

Monitoring Trigger: If D/E exceeds 1.0 or FCF negative 2+ more periods, cut to 5%. If TEPCO JV ships first cask, increase.

Fundamentals:

  • PE: 19.62 | Fwd PE: 14.09 | P/B: 1.09
  • ROE: 5.8% | Op Margin: 2.0% | D/E: 88%
  • Div Yield: 2.1% | FCF: -¥12.8B | Mkt Cap: ¥204B

Segments:

  • Environment: ¥280B (45%) — Waste-to-energy, water treatment
  • Machinery & Infrastructure: ¥180B (29%) — Marine, process equipment
  • Decarbonization: ¥120B (19%) — Nuclear casks, pressure vessels, fuel cycling
  • Other: ¥42B (7%) — Real estate, services

Supply Chain:

  • [confirmed] TEPCO: Fukushima decommissioning casks (Toso Mirai JV) — JV announced Apr 2024, plant targeting FY2025 startup
  • [confirmed] US nuclear utilities (via NAC International): Spent fuel transport/storage casks — Wholly-owned US subsidiary
  • [probable] Deep Isolation: Deep borehole nuclear waste disposal — Kanadevia/NAC invested in Deep Isolation technology

Socionext (6526.T)

Score: 31.9 | Conviction: MEDIUM | Layer: Custom SoC Design | Weight: 6%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: Highest-margin fabless way to play custom AI silicon in Japan. Designs SoCs on TSMC 5nm/3nm for hyperscaler customers. 48.7% gross margin on design-only revenue. Near-zero debt. As hyperscalers design proprietary ASICs, Socionext's services grow in relevance.

Risk: Revenue -14.8% in FY2025 from customer inventory digestion. Revenue concentration in 1-2 large unnamed customers. Design win cancellation risk.

Monitoring Trigger: New multi-year design contracts or tape-out starts at TSMC N3. Tape-out revenues are 2-3 year leading indicators.

Fundamentals:

  • PE: 40.37 | Fwd PE: 22.44 | P/B: 2.45
  • ROE: 6.0% | Op Margin: 6.2% | D/E: 1%
  • Div Yield: 2.8% | FCF: -¥4B | Mkt Cap: ¥316B

Segments:

  • SoC Design (single segment): ¥210B (100%) — Custom SoCs: networking, automotive radar, AI inference, imaging

Supply Chain:

  • [confirmed] Fujitsu: Custom SoC design (legacy parent) — Anchor customer since 2014 spinout
  • [probable] Major US hyperscaler: AI inference SoC / SmartNIC — Disclosed 'major US data center customer' — unnamed
  • [probable] Japanese automotive Tier-1s: Radar sensor SoC for ADAS — Used in Japanese auto supply chain

SoftBank Group (9984.T)

Score: 31.5 | Conviction: LOW | Layer: AI Platform / Investment | Weight: 6%

Themes: ai | Data Date: 2026-04-12

Thesis: Leveraged call option on AI infrastructure. ~90% of Arm (every AI chip uses Arm IP). $64B invested in OpenAI. Project Izanagi $67B AI infra pledge. Arm royalties compound as AI chips proliferate.

Risk: CRITICAL: ¥19.8T net debt. Negative FCF (-¥1.4T). Entirely dependent on asset monetization. Credit/liquidity event possible if AI sentiment reverses. Not a traditional operating company.

Monitoring Trigger: Arm quarterly royalty revenue growth. If below 15% YoY, AI thesis weakens. Watch SoftBank credit spreads.

Fundamentals:

  • PE: 5.89 | Fwd PE: 22.09 | P/B: 1.18
  • ROE: 26.0% | Op Margin: 7.5% | D/E: 132%
  • Div Yield: 0.3% | FCF: -¥1,396B | Mkt Cap: ¥21.53T

Segments:

  • SoftBank (telecom): ¥3,200B (40%) — Domestic telecom — stable cash flow engine
  • Arm Holdings: ¥500B (6%) — Chip architecture licensing — every AI chip
  • Vision Funds: N/A (N/A) — SVF1 + SVF2 tech investments
  • Investment/Other: ¥4,300B (54%) — T-Mobile, PayPay, direct holdings

Supply Chain:

  • [confirmed] Apple: Arm architecture licensing — Largest single Arm revenue driver
  • [confirmed] NVIDIA: Arm Neoverse for Grace Hopper SoC — CUDA on Arm
  • [confirmed] OpenAI: $64B equity investment + Stargate infra — Major strategic partner

Nabtesco (6268.T)

Score: 30.5 | Conviction: MEDIUM | Layer: Precision Gears (Robot Joints) | Weight: 10%

Themes: robotics | Data Date: 2026-04-12

Thesis: RV-series reduction gears are in 60%+ of global industrial robot joints. Like Harmonic Drive, a critical component for humanoid robots. Also makes aircraft flight control systems (defense angle). Adapting RV gears for humanoid joint applications.

Risk: Industrial robot demand cyclical. China ~30% of robot orders. No confirmed humanoid contracts yet.

Monitoring Trigger: Quarterly robot-related gear orders. Humanoid OEM supply announcements.

Fundamentals:

  • PE: N/A | Fwd PE: N/A | P/B: N/A
  • ROE: N/A | Op Margin: N/A | D/E: N/A
  • Div Yield: N/A | FCF: N/A | Mkt Cap: ~¥500B

Segments:

  • Precision Equipment: N/A (~50%) — RV gears for robot joints (60%+ global share)
  • Transport Equipment: N/A (~25%) — Rail doors, marine equipment
  • Aerospace: N/A (~25%) — Flight control actuators (defense)

Supply Chain:

  • [confirmed] Fanuc / Yaskawa / ABB / KUKA: RV reduction gears for robot joints — 60%+ of global industrial robot joints use Nabtesco RV gears
  • [confirmed] Boeing / defense: Flight control actuators — Aircraft actuation systems
  • [inferred] Humanoid OEMs: Precision gears for humanoid joints — Adapting RV gears for humanoid applications

Kadokawa (9468.T)

Score: 25.4 | Conviction: LOW | Layer: Virtual Events / IP Publisher | Weight: 8%

Themes: vtuber | Data Date: 2026-04-12

Thesis: Owns Dwango (Niconico), Japan's pioneering streaming platform that birthed VTuber culture. Massive anime/manga/light novel IP library provides content pipeline. Virtual events platform could be leveraged for VTuber concerts.

Risk: Low profitability (ROE ~4%). Niconico losing share to YouTube. Sony acquisition attempt adds uncertainty.

Monitoring Trigger: If Sony acquisition completes or Niconico VTuber events show growth metrics, reassess.

Fundamentals:

  • PE: 25.9x | Fwd PE: N/A | P/B: N/A
  • ROE: 4.1% | Op Margin: 6.0% | D/E: 4%
  • Div Yield: 0.9% | FCF: N/A | Mkt Cap: ¥461B

Segments:

  • Publication: ¥90B (35%) — Light novels, manga, books
  • Animation & Live Action: ¥70B (27%) — Anime production, IP licensing
  • Game: ¥45B (18%) — FromSoftware (Elden Ring), game publishing
  • Web Services (Niconico): ¥50B (20%) — Niconico streaming, virtual events

Supply Chain:

  • [confirmed] Dwango/Niconico: Streaming platform for VTuber content — Pioneered VTuber streaming in Japan
  • [confirmed] FromSoftware: Game development subsidiary — Elden Ring, Armored Core

GREE Holdings (REALITY) (3632.T)

Score: 24 | Conviction: LOW | Layer: VTuber Platform / Metaverse | Weight: 8%

Themes: vtuber | Data Date: 2026-04-12

Thesis: REALITY platform democratizes VTuber creation — anyone can create an avatar and livestream from their smartphone. Platform play vs. talent agency model of Cover/ANYCOLOR. Metaverse segment is a distinct growth driver.

Risk: High PE (45x) on low margins (6.7%). REALITY faces competition from established platforms. Legacy mobile gaming business declining.

Monitoring Trigger: If REALITY MAU exceeds 5M or metaverse segment becomes profitable, upgrade conviction.

Fundamentals:

  • PE: 45.7x | Fwd PE: N/A | P/B: N/A
  • ROE: N/A | Op Margin: 6.7% | D/E: N/A
  • Div Yield: 1.9% | FCF: N/A | Mkt Cap: ¥67B

Segments:

  • Game: ¥30B (45%) — Legacy mobile gaming
  • Metaverse (REALITY): ¥15B (22%) — VTuber avatar platform, livestreaming
  • IP / DX / Investment: ¥22B (33%) — IP licensing, digital transformation, investments

Supply Chain:

  • [confirmed] Smartphone users: REALITY avatar creation app — Free app for VTuber-style livestreaming
  • [confirmed] App Store / Google Play: Distribution platform — Primary distribution channels

Colopl (3668.T)

Score: 21.3 | Conviction: LOW | Layer: VR Gaming | Weight: 5%

Themes: vtuber | Data Date: 2026-04-12

Thesis: Pure-play VR gaming exposure on TSE. If VTuber and virtual entertainment drive VR headset adoption, Colopl's VR titles benefit. High dividend yield (4.3%) provides downside cushion.

Risk: Currently unprofitable (negative net margin). High PE on minimal earnings. Mobile gaming revenue declining. VR adoption slower than expected.

Monitoring Trigger: If quarterly earnings turn positive or VR title achieves breakout success, reassess.

Fundamentals:

  • PE: 40.7x | Fwd PE: N/A | P/B: N/A
  • ROE: -0.4% | Op Margin: N/A | D/E: 0.5%
  • Div Yield: 4.3% | FCF: N/A | Mkt Cap: ¥59B

Segments:

  • Mobile Games: ¥20B (60%) — Shironeko Project, other mobile titles
  • VR Games: ¥8B (24%) — VR titles for Meta Quest
  • Other: ¥5B (16%) — Blockchain gaming, new ventures

Supply Chain:

  • [confirmed] Meta: Quest VR platform — Primary distribution for VR titles
  • [confirmed] App Store / Google Play: Mobile game distribution — Mobile gaming distribution

SUMCO (3436.T)

Score: 12.6 | Conviction: LOW | Layer: Silicon Wafers | Weight: 4%

Themes: ai, semiconductor | Data Date: 2026-04-12

Thesis: Leveraged recovery play on silicon wafer upcycle. Currently loss-making but 300mm wafer demand from AI is structurally growing. +161% 1Y run reflects market pricing 2026 inflection.

Risk: No product diversification (100% wafers). Loss-making with negative FCF. ¥353B debt with -1.2x interest coverage. If upcycle delays, cash burn severe.

Monitoring Trigger: Quarterly wafer shipment volume guidance. TSMC long-term supply agreement announcements.

Fundamentals:

  • PE: Loss | Fwd PE: N/A | P/B: 1.1
  • ROE: -1.5% | Op Margin: -0.8% | D/E: 55%
  • Div Yield: 1.0% | FCF: -¥11B | Mkt Cap: ¥711B

Segments:

  • Silicon Wafers (single segment): ¥410B (100%) — 300mm polished, epitaxial, SOI wafers

Supply Chain:

  • [confirmed] TSMC: 300mm silicon wafers — Major disclosed customer
  • [confirmed] Samsung / SK Hynix: Wafers for NAND, DRAM, HBM — Disclosed in annual reports
  • [probable] Intel: Silicon wafers — Historical key customer

Rohm Co. (6963.T)

Score: 7.5 | Conviction: LOW | Layer: SiC Power Devices | Weight: 2%

Themes: ev | Data Date: 2026-04-12

Thesis: Only pure-play Japanese SiC company in global top-5. Schaeffler inverter partnership + Infineon MOU (Sept 2025) for SiC packaging. Current loss reflects heavy SiC fab investment — forward recovery bet.

Risk: Currently loss-making (op margin -4.1%). Negative FCF (-¥44B). Forward PE 55.8x. SiC demand slowed. Chinese competitors emerging. NOT a confirmed Tesla SiC supplier (common misconception — that's STMicro).

Monitoring Trigger: Return to operating profitability. Named 800V design wins. SiC wafer pricing trends.

Fundamentals:

  • PE: Loss | Fwd PE: 55.75 | P/B: 1.52
  • ROE: Loss | Op Margin: -4.1% | D/E: 43%
  • Div Yield: 1.4% | FCF: -¥44B | Mkt Cap: ¥1.41T

Segments:

  • Discrete (SiC, MOSFET): ¥210B (44%) — SiC power devices — core EV play
  • LSI / ICs: ¥175B (37%) — Analog, power management
  • Modules: ¥88B (19%) — Power modules, LED

Supply Chain:

  • [confirmed] Schaeffler: SiC inverter technology — Partnership announced
  • [confirmed] Infineon: SiC package co-development — MOU Sept 2025
  • [inferred] Tesla: SiC for traction inverter — CORRECTION: Tesla Model 3 uses STMicro, NOT Rohm. Common misconception.