·Dashboard·Research·Work·Archive
/
← wiki
briefing briefingdeep-divemlccpassive-componentsai-infrastructure updated 2026-05-23

2026-05-23-mlcc-sector

TLDR

Five independent sources converged on the same MLCC sector read this week — BofA, the sell-side Murata-vs-Yageo desk note, a six-dimension global passives basket framework, Goldman Sachs on Yageo, and Banyan Lane Capital's Substack thesis on Murata. The structural setup is real: only four qualified suppliers for ultra-high-CV MLCCs (Murata, TDK, Samsung Electro-Mechanics, Kyocera AVX), hyperscalers locked 60-70% of high-CV capacity, AEC-Q200 qualification at 12-18 months, lead times stretched from 8 weeks to 24-40 weeks, and Murata's +89.5% capacitor backlog surge confirming demand is 2× supply at the high end. The cycle is happening, not forecast. The problem is positioning: every Taiwan passives name went limit-up at +10% on 23 May, Taiyo Yuden +12%, Samsung E-M +11%, while every name in the complex except Kingboard now trades above its sell-side price target. The right answer to a top-2 question is a structural anchor plus a crowding-discounted differentiator.

INVESTMENT THESIS — 3 BULLETS

  • The AI MLCC supply/demand setup is structurally different from a normal MLCC cycle. Capacity additions take 24-30 months from capex commitment, customer qualification adds 6-12 months, and the dielectric powder (barium titanate) comes from 3-4 qualified suppliers globally. Murata's Izumo plant started shipping April 2026 — 25 months from announce. SEMCO's larger Korean expansion is not until mid-2027. The shortage cannot be fixed inside an 18-24 month thesis horizon.

  • Bifurcation drives most of the margin story. Strategic MLCCs (high-voltage, polymer electrolytic, ultra-high-CV) are 35-40% of market value but 60-70% of industry profit, growing 12-17% CAGR with 30-80% price premiums and 24-47 week lead times. Commodity MLCCs face -10-20% pricing pressure from Chinese capacity. Mix shift toward the strategic tier is why Murata's FY27 operating profit guide is +34.8% on only +7.1% revenue growth.

  • Crowding asymmetry across the basket is the real signal today. The May 21 Global MLCC & Passives Basket scored Murata 21/30 Buy (Moat 5/5, Crowding 3/5) and Kingboard 19/30 Watch (Crowding 3/5) — both moderate. Yageo, Walsin, Holy Stone, Samsung E-M, Taiyo Yuden all scored Crowding 1-2/5 on limit-up tape and 4-9× moves over 12 months on flat-to-cut earnings revisions. Taiyo Yuden's Revision Velocity 1/5 (estimates cut from ¥130 to ¥90 while stock rallied 77%) is the rubric's worst-case scenario.

THE PICKS

The choice is a structural anchor plus a crowding-discounted differentiator, deliberately drawn from different sub-industries and different exchanges.

Rank Ticker Conviction One-Line Thesis Biggest Risk
1 6981.T Murata High (scale-in) Cleanest moat in AI MLCC at 60% high-CV share within a 4-supplier oligopoly; FY28E EPS ¥307 vs consensus ~¥210; ¥150B buyback retires 4% of float Samsung Electro-Mechanics design-win momentum (CL31X227 + CL32X337 products are direct competition); long-term eDTC / HD-MiM on-chip substitution
2 **0148.HK 0148/0148 Kingboard Holdings** Starter + add Upstream CCL/PCB at 9.88x fwd P/E and 3.81% div yield; ONLY name in universe with mean PT (HK$64) ABOVE spot HK$58.30; consensus strong_buy; Collyer Bridge actively posting on the subsidiary Prosperity Dielectrics — all Taiwan limit-up parabolic at fresh 52-week highs.

KNOWN vs. UNKNOWN

Verified:

  • Murata Apr 1 2026 industry-leading price hike of 15-25% on AI/auto-grade MLCC; Taiyo Yuden May 2026 +6-13%; Samsung E-M signaling 5-10%
  • Lead times stretched from 8 weeks to 24-40 weeks through 2026 (TrendForce, Murata customer notices)
  • Murata's capacitor backlog +89.5% to ¥269.2B, computers/servers application +28.4% YoY
  • Morgan Stanley NVL72 BoM: MLCC content $1,530 (GB300 Blackwell) → $4,320 (VR200 Rubin), +182% generation-on-generation
  • Annual Capacitor Dossier (04/26): High-CV MLCC (220µF/4V, AI GPU rail) prices +29% from 2024 to Q1 2026
  • Only 4 qualified suppliers for ultra-high-CV at production scale: Murata, TDK, Samsung E-M, Kyocera AVX
  • Hyperscalers (Google, Meta, Microsoft, NVIDIA) locked in 60-70% of available high-CV capacity
  • Murata's Izumo plant started shipping April 2026; SEMCO's Korean expansion not until mid-2027

Open:

  • Murata's exact share of AI-server-specific MLCCs — Banyan Lane uses 60%, industry sources range 50-70%, downside scenario at 50% weakens thesis materially
  • Whether the 23 May complex-wide limit-up move is the cycle's next leg or the top — depends on Q2 earnings prints in late July
  • How fast Chinese mid-spec CCL capacity additions arrive and whether they undercut Kingboard's pricing power
  • Kingboard's exact mix between high-spec AI server CCL vs mid-spec — disclosure is opaque on the segment split
  • Collyer Bridge's full thesis on Kingboard Laminates (subscriber chat thread bodies not fully extracted)

Watch for:

  • Murata FY27 Q1 earnings expected late July 2026 — first print showing whether the Apr 2026 price hikes flow through to reported financials with the assumed 1-2 quarter lag
  • Prosperity Dielectrics Q1 2026 earnings 11 June — asymmetric to the downside given 53x trailing P/E
  • Samsung Electro-Mechanics design-win announcements on the CL31X227MRKNNW# and CL32X337MSVN4S# products (220µF and 330µF in 1206/1210 sizes) — direct competition for Murata equivalents

METRICS TO TRACK

The five metrics that matter for the structural thesis, in priority order:

  1. AI MLCC ASP (high-CV, 220µF/4V reference part) — TrendForce projects +30-40% for full-year 2026 on high-end parts. Already +29% from 2024 through Q1 2026. The single best read on whether the pricing discipline is holding.
  2. Murata capacitor backlog and computers/servers application growth — currently +89.5% backlog and +28.4% computers/servers YoY. A sustained sub-+20% reading on either would suggest the AI MLCC demand spike is normalizing.
  3. Murata's AI MLCC share trajectory — Samsung E-M's design-win cadence with hyperscalers is the leading indicator. Banyan Lane downside scenario at 50% share (vs working 60%) weakens FY28 EPS materially.
  4. Lead times across the high-end product stack — 24-40 weeks today vs 8 weeks normal. A return toward 12-16 weeks confirms supply catching up; further extension confirms the cycle has more room.
  5. Kingboard's high-spec CCL pricing actions — four announced price hikes through April 2026. Watch for a fifth hike (or Resonac / Mitsubishi Gas Chemical extending their 30%+ bonding sheet / copper foil resin sheet hikes) as confirmation that upstream margin support is durable.

NEXT STEPS

Priority Command Why
High /profile 0148 Kingboard Holdings is net-new to the vault — needs a canonical entity hub before sizing
High /profile 4078 [[4078/4078
Medium /profile 1888 Kingboard Laminates is the pure-play CCL subsidiary of 0148.HK — avoids the property / chemicals overhang
Medium /mgmt-dd 6981 Murata governance not yet covered post the May 21 buyback announcement and Izumo plant ramp
Medium Browser read of Collyer Bridge subscriber chat threads on "MLCCs" and "Kingboard Laminates" (21-23 May) opencli's eval pattern couldn't extract React-virtualized chat panel bodies — the qualitative analyst take needs a 5-min in-browser read
Low /profile CTR Holdings Flagged by Collyer Bridge 22 May (8 replies) for follow-up — no current vault coverage
Low Re-check entry math Monday 26 May Asia desk closes for the weekend; +5-12% moves across the complex on 23 May may continue or reverse on Monday
Pages that link here (1)
Passives & Components (MLCC)
Open in Obsidian