· 8 stocks
Three key electronics material sectors are entering a destocking → restocking → price increase cycle in 2026. Glass fiber cloth (玻纖布) faces acute shortage with 6-month CCL lead times and 15-30% price hikes — Nitto Boseki holds 90% monopoly in T-glass for AI chip packaging. Power semiconductors (功率半導體) have 20-45 week lead times with SiC critically tight. Passive components (被動元件) see Taiyo Yuden initiating May 2026 MLCC price increases, Murata expected to follow — AI servers use 30,000 MLCCs each (6x normal). This is a structural supply cycle, not speculative.
Portfolio Overview
| # | Company | Conv | Wt | PE | Fwd PE | PB | ROE | OpMar | D/E | DY | FCF |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Nitto Boseki3110.T | HIGH | 15% | 29.00 | 22.00 | 5.00 | 13% | 13% | N/A | 0.5% | N/A |
| 2 | Murata Manufacturing6981.T | HIGH | 15% | ~49.7x TTM (fwd ~37x on FY2026 guidance) | 18.23 | 2.80 | 8.09% | 15.1% | N/A | 1.39% | N/A |
| 3 | Fuji Electric6504.T | HIGH | 12% | 20.60 | 18.00 | 2.50 | 13% | 11% | N/A | 1.48% | N/A |
| 4 | TDK Corporation6762.T | MEDIUM | 10% | 21.00 | 12.30 | 2.00 | 8.32% | 10% | N/A | 1.5% | N/A |
| 5 | Taiyo Yuden6976.T | MEDIUM | 10% | 65.00 | 28.00 | 1.50 | 3.92% | 5% | N/A | 2.5% | N/A |
| 6 | Rohm6963.T | MEDIUM | 8% | Loss (FY2025 NI -¥158.4B SiC impairment) | 15.00 | 2.56 | Loss | 1% | N/A | 3% | N/A |
| 7 | Asahi Kasei3407.T | MEDIUM | 8% | 13.40 | 12.00 | 1.00 | 7.8% | 7% | N/A | 3.5% | N/A |
| 8 | Arisawa Manufacturing5208.T | MEDIUM | 5% | 14.85 | 14.54 | 1.00 | 6.85% | 7% | N/A | 5.93% | N/A |
Portfolio Construction
Stock-by-Stock Analysis
Nitto Boseki
3110.THIGHMaterialsWeight: 15%
Why this stock
World sole T-glass supplier (90% share) for AI chip packaging substrates — this monopoly is SAFE through 2027+. HOWEVER, their NE/NER glass for PCB motherboards faces disruption: NEZ glass (Df 0.001) cannot meet M9 grade spec (Df ≤0.0007) required by Nvidia Rubin-era products. Asahi Kasei Q Glass wins M9. Long-term risk: glass core substrates (Intel, Samsung, TSMC CoPoS) could bypass glass cloth entirely by 2028-2030. Near-term bull: T-glass capacity tripling, JPY 15B expansion, pricing power intact.
What could go wrong
NEZ glass lost M9 CCL specification to Asahi Kasei Q Glass — impacts NE/NER segment (PCB motherboards). Glass core substrates (Intel patents, Samsung Electro-Mechanics, TSMC CoPoS pilot June 2026) could structurally displace glass cloth by 2028-2030. PE 29x already prices in some AI premium. Single factory concentration (Fukushima).
Monitoring trigger
Watch M9 CCL adoption in H2 2026 — if Q Glass takes share from NER, NE segment revenue at risk. Watch TSMC CoPoS pilot (June 2026) for glass panel vs glass cloth direction. T-glass remains safe but monitor capacity utilization after tripling.
Murata Manufacturing
6981.THIGHPassive Components / MLCCWeight: 15%
Why this stock
40% global MLCC share — the toll-gate on AI infrastructure. Every Nvidia GB300 needs 30,000 MLCCs. Full AI cabinet needs 440,000. Orders at 2x capacity. Price increases 15-35% effective April 1, 2026 — Murata SETS the price, Samsung/Taiyo Yuden follow. 10-year technology lead vs China (1,000 vs 300 ceramic layers). New revenue: AI server power modules targeting JPY 50B by FY2027. Book-to-bill >1.0. AI MLCC CAGR projected 30% to 2030.
What could go wrong
China 48% of revenue — geopolitical escalation risk. Samsung EMCO closing AI gap (40% vs 45% share) with vertical BaTiO3 integration. Silver/nickel cost inflation (60% of production cost). If AI capex cycle reverses, high-end could correct as in 2019 post-2018 super cycle.
Monitoring trigger
Watch Q1 FY2026 book-to-bill ratio (current 1.01). Samsung AI share trajectory. China rare earth export controls impact on BaTiO3 supply. AI capex guidance from hyperscalers (Microsoft, Google, Amazon).
Fuji Electric
6504.THIGHPower Semiconductors / IGBTWeight: 12%
Why this stock
Most focused power semiconductor play among large-cap Japanese industrials. Investing JPY 200B to establish SiC production and expand IGBT capacity through FY2026. Operating profit guided to JPY 128.5B. Well-positioned for EV adoption and renewable energy power conversion. Lead times 20-45 weeks indicate structural tightness.
What could go wrong
Semiconductor segment ~40% of profits (energy ~60%). SiC ramp execution risk. EV demand slowdown in China could reduce near-term orders.
Monitoring trigger
Watch SiC production line commissioning and FY2026 earnings. If EV demand in China weakens further, reassess conviction.
TDK Corporation
6762.TMEDIUMPassive Components / Magnets / SensorsWeight: 10%
Why this stock
Raised full-year guidance to JPY 2.37T revenue, JPY 245B operating profit, JPY 180B net profit. Forward PE 12.3x is cheap for quality. Diverse passive portfolio (ceramic caps, film caps, inductors, ferrites). Managing nickel/palladium costs via copper electrode substitution.
What could go wrong
Passive segment revenue down 3.4% YoY from auto softness. Battery segment volatility. Less pure-play passive vs Murata.
Monitoring trigger
Watch for passive component price increase announcements following Taiyo Yuden/Murata. Auto recovery in H2 would lift passive segment.
Taiyo Yuden
6976.TMEDIUMPassive Components / MLCCWeight: 10%
Why this stock
The catalyst company — initiated May 2026 price increases across MLCCs, inductors, ferrite beads. Mizuho upgraded to Buy. Morningstar raised fair value 25% citing tight MLCC supply and margin expansion. Highest-beta play on MLCC restocking cycle. Operating profit guided to JPY 18B (recovering from trough).
What could go wrong
Trailing PE ~65x (trough earnings). If price increases don't stick, margins won't recover. Smaller scale = less diversification vs Murata.
Monitoring trigger
May 1 price increase effective date — watch customer acceptance and order flow. FY2025 results in May for earnings confirmation.
Rohm
6963.TMEDIUMSiC Power DevicesWeight: 8%
Why this stock
SiC MOSFET specialist anchoring Japan's power semiconductor consolidation JV (Rohm + Toshiba TDSC + Mitsubishi Electric MoU March 27, 2026). FY2025 (Mar 2026): Revenue ¥481.1B, OP ¥10.8B (beat ¥5B guide), NI -¥158.4B (kitchen-sink SiC impairment — depreciation falls sharply from FY2026 forward). FY2026 guided: Revenue ¥510B (+6%), OP ¥30B (+176%) — mechanically credible as depreciation burden falls. Gen 5 SiC (~30% lower on-resistance) sample availability July 2026. Toyota bZ5, Schaeffler, Infineon confirmed. CAUTION: JV talks slower than expected (Rohm president warned May 15, 2026); Denso bid withdrawn April 28, 2026; P/B 2.56x vs historical 1.3x; analyst consensus PT ¥2,944 vs market ¥5,400+ (-47%).
What could go wrong
1) JV EXECUTION: Three-party JV (Rohm+Toshiba+Mitsubishi) delayed — Rohm president acknowledged talks more complex than expected (May 15, 2026). No formal equity structure yet. 2) VALUATION: P/B 2.56x vs pre-crisis 1.3-1.4x — stock 3x'd from lows on M&A speculation; analyst consensus PT ¥2,944 vs price ¥5,400+ = -47% implied downside. 3) SiC market: Chinese competition (BYD, CATL vertical integration) structural headwind; BEV slowdown caused the ¥158.4B impairment. 4) Net loss FY2025 — ROE still negative. First clean profitability quarter not yet achieved. 5) Denso bid withdrawn — clearest value-unlock path now gone.
Monitoring trigger
FY2025 results in May — need to see operating profit turn positive. Watch SiC automotive design win announcements.
Asahi Kasei
3407.TMEDIUMMaterialsWeight: 8%
Why this stock
Q Glass (quartz, Df 0.0005) meets M9 CCL spec for Nvidia Rubin — but Asahi Kasei is ONE OF SEVERAL Q glass suppliers, not a monopolist. Competitors: Glotech (Taiwan, 5475.TWO) makes higher-purity quartz yarn (99.995%), Feilihua (China, 300395.SZ) has full vertical chain with Nvidia cert and volume production. Still worth owning at PE 13.4x with 3.5% yield — priced for zero Q glass premium. Target 3x glass cloth sales by 2030.
What could go wrong
NOT a monopoly like Nittobo T-glass — multiple Q glass competitors globally. Feilihua (China) already has Nvidia certification and 50,000 m/month production. Glotech (Taiwan) may be upstream quartz yarn supplier. Conglomerate dilution persists.
Monitoring trigger
Watch M9 CCL mass production ramp in H2 2026. If Q Glass adoption confirmed by major CCL makers (TUC, EMC), thesis materially strengthens. Track segment profit disclosure for glass materials.
Arisawa Manufacturing
5208.TMEDIUMMaterialsWeight: 5%
Why this stock
Small-cap pure-play on electronic materials including glass cloth for PCB substrates. PE 14.85x with nearly 6% dividend yield provides downside protection. Benefits directly from CCL price upcycle. Ranked #1 in Japan glass cloth attention rankings.
What could go wrong
Smaller scale, less pricing power than Nittobo. Lower growth trajectory. Less direct AI exposure.
Monitoring trigger
Watch CCL price pass-through to margins. Potential acquisition target given small market cap.
AI-generated for research purposes only. NOT investment advice. Generated .