High-value-added MLCCs for AI servers are the key differentiator, with Murata Mfg seen as having the strongest advantage, while Ibiden's earnings are growing but expectations are too high

Overview

This report is Morgan Stanley's investor meeting note on Japan's electronic components sector, based on institutional investor meetings held in Singapore and Hong Kong from June 22 to 26, 2026. Discussions mainly centered on MLCCs and ABF package substrates, especially AI server demand for high-capacitance, miniaturized, high-value-added MLCCs, as well as Ibiden's earnings outlook for ABF package substrates related to NVIDIA Rubin and subsequent EMIB-T.

Core views

The core view is that the gap in corporate value among Japanese electronic components companies will depend more on continuously improving product competitiveness than on simply relying on price hikes for commoditized products. The report is positive on Murata Mfg's technology, mass-production, and product mix advantages in high-value-added MLCCs for AI servers, believing its ASP growth does not rely on raising prices for the same products, but rather comes from a higher share of high-priced, high-value-added products. By contrast, while Taiyo Yuden will benefit from AI/data center MLCC demand and higher utilization rates, its stable mass-production capability and capacity expansion are constrained. As for Ibiden, the report acknowledges that ABF package substrates will continue to drive earnings growth, but believes market expectations for its F3/28 and longer-term earnings are too high.

Analysis framework

The report forms its views through methods including investor feedback, industry supply-demand assessment, product generation upgrades, technical mass-production capability, ASP and product mix decomposition, and comparison of company earnings forecasts. The analytical focus is not on short-term price increase magnitude, but on the penetration of high-value-added products in AI server and data center demand, the stable mass-production capability of key manufacturers, and the gap between company guidance and market consensus expectations.

Methodology notes

• Investor meeting notesFeedback from one-on-one meetings and group lunch sessions

  Identification of institutional investor focus areas

  Based on 31 one-on-one meetings and group lunch sessions with institutional investors in Singapore and Hong Kong, the report summarizes investor focus on MLCCs, ABF package substrates, the positioning of Murata and Taiyo Yuden in AI server MLCCs, supply, demand and pricing of commoditized MLCCs, as well as Ibiden's technological competitiveness and valuation.

• Industry competition analysisComparison between product mix upgrades and commoditized product price hikes

  Quality of ASP growth

  The report distinguishes between ASP growth driven by price increases for the same product and ASP growth driven by a higher share of high-priced, high-value-added products, and believes the latter better reflects product competitiveness and medium- to long-term corporate value.

• Earnings forecast comparisonComparison of company guidance, FactSet consensus, and Morgan Stanley forecasts

  Gap versus market expectations

  For Ibiden, the report compares Morgan Stanley's OP forecasts with company guidance and FactSet consensus, concluding that while earnings growth is sustainable, market expectations for F3/28 and F3/31 are too high.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Murata Mfg

  The report maintains OW and views it as the leading beneficiary of high-value-added MLCCs for AI servers.

  Strengths

  It has technical capabilities in miniaturization, high capacitance, multilayering, and narrower external electrodes; it has stable mass-production capability for key AI server MLCC products; and product mix upgrades drive ASP growth.

  Weaknesses

  Investor discussion still focuses heavily on price hikes for commoditized products, which may underestimate the earnings contribution from growth in high-value-added products.

  Comparison

  Compared with Taiyo Yuden, the report believes Murata can more stably supply high-value-added products; compared with pure price-hike-driven manufacturers, Murata places more emphasis on product competitiveness.

  Risks

  If AI/data center demand growth is slower than expected, or if competitors achieve stable mass production more quickly, the premium and share advantage of high-value-added products may come under pressure.

• Taiyo Yuden

  The report believes it will benefit from growth in AI/data center MLCC sales and higher utilization rates, but with limited earnings contribution.

  Strengths

  AI/data center MLCC sales are expected to grow 82%-83% YoY in F3/27, with the share of related products rising to about 15%.

  Weaknesses

  The timing of stable mass production for some key high-value-added MLCCs remains unclear; the freeze on capital expenditures in 2H24 means F3/26 capacity growth is estimated at only about 5%.

  Comparison

  The report believes that even if Taiyo Yuden later achieves stable mass production of products similar to Murata's, Murata may already have progressed to a new generation of smaller, higher-capacitance products.

  Risks

  Insufficient capacity, lagging mass-production progress, or a widening generational product gap could limit its ability to win new business and earnings leverage.

• Ibiden

  The report maintains UW and believes earnings expansion exists but market expectations are too high.

  Strengths

  It is expected to have begun volume shipments of ABF package substrates for NVIDIA Rubin in F3/26 4Q, and Rubin-related sales may exceed Blackwell in F3/27 1Q.

  Weaknesses

  Morgan Stanley's OP forecasts for F3/28 and F3/31 are below company forecasts, indicating a more cautious view on the realization of long-term earnings.

  Comparison

  Compared with market consensus, the report believes Ibiden's earnings growth has been priced with excessive optimism; meaningful contribution from EMIB-T is not expected to appear until after F3/29.

  Risks

  Although EMIB-T products may be priced significantly above existing ABF package substrates, the report believes it will be difficult for their margins to reach the level of NVIDIA's current products.

• Commoditized MLCC suppliers

  The report uses them as a sector comparison against the route of competing through high-value-added product competitiveness.

  Strengths

  Higher customer prices for commoditized MLCC distributors may bring short-term profit improvement.

  Weaknesses

  If they rely on undifferentiated price increases while underinvesting in R&D and manufacturing technology, their medium- to long-term product competitiveness and market share may be constrained.

  Comparison

  The report believes some manufacturers in China, Korea and Taiwan are more inclined toward short-term profit maximization, whereas Japanese manufacturers such as Murata place greater emphasis on technology and cost competitiveness.

  Risks

  Pure price increases may lower entry barriers and trigger more intense competition, ultimately eroding medium- to long-term share and corporate value.

Key data

• Number and locations of meetings31 one-on-one meetings; group lunch sessions with institutional investors in Singapore and Hong Kong; June 22 to 26, 2026The meetings mainly discussed MLCCs and ABF package substrates.
• Estimated 2025 MLCC market shareMurata Manufacturing 40.8%; SEMCO 22.5%; Taiyo Yuden 11.3%The report believes Murata ranks first in high-value-added MLCCs for AI servers.
• Murata AI/data center MLCC sales share and growth10%-15% of MLCC sales in F3/26; expected YoY growth of 85%-90% in F3/27, with share rising to 20%-25%Growth assumptions include about 40% volume growth and about 50% ASP growth driven by a high-value-added product mix.
• Taiyo Yuden AI/data center MLCC sales share and growth5%-10% of MLCC sales in F3/26; expected YoY growth of 82%-83% in F3/27, with share at about 15%The report believes the degree of benefit is limited by stable mass-production capability and capacity growth.
• Key MLCC demand for AI servers1608 size 100µF, 1005 size 47µF, 0603 size 10µFThe report says demand for these products is growing significantly, but at present only Murata appears capable of stably mass-producing all of them.
• Ibiden OP forecast comparisonF3/27: Morgan Stanley ¥94.7bn, company forecast ¥90bn, FactSet consensus ¥99.7bn; F3/28: Morgan Stanley ¥128.5bn, company forecast ¥150bn, FactSet consensus ¥151.5bn; F3/31: Morgan Stanley ¥242.6bn, company forecast ¥300bnBased on this, the report believes Ibiden's earnings will grow, but market expectations are too high.

Impact & implications

In terms of investment implications, the report prefers companies with technological roadmaps and stable mass-production capabilities for high-value-added products, especially Murata Mfg; it is more cautious on manufacturers that rely on price hikes for commoditized products, because short-term price increases may lift profits but could also lower entry barriers, attract Chinese, Korean and Taiwanese manufacturers, and lead to medium- to long-term share loss. For Ibiden, although NVIDIA Rubin-related ABF package substrates may drive earnings expansion, the market has already priced in high expectations, so it will be necessary to watch the timing and margin level of earnings contribution from EMIB-T products.

Risks

• Demand growth from AI servers and data centers for high-value-added MLCCs comes in below expectations.
• Competitors achieve stable mass production of miniaturized, high-capacitance MLCCs more quickly, weakening Murata's leading advantage.
• Short-term profit improvement from price increases in commoditized MLCCs may mask the risk of medium- to long-term share loss.
• Volume ramp, pricing, or margins for Ibiden's Rubin- or EMIB-T-related ABF package substrates fall short of expectations.
• Excessively high market consensus expectations may leave Ibiden facing valuation pressure even if earnings continue to grow.

What to watch

• Whether Murata's F3/27 AI/data center MLCC sales achieve 85%-90% YoY growth.
• Whether Murata's ASP growth mainly comes from high-value-added product mix upgrades rather than price increases for the same product.
• When Taiyo Yuden will be able to stably mass-produce key AI server MLCCs such as 1608 size 100µF, 1005 size 47µF, and 0603 size 10µF.
• Whether rising prices for commoditized MLCCs attract more manufacturers from China, Korea and Taiwan and change the supply-demand landscape.
• Whether Ibiden's Rubin-related ABF package substrate sales exceed Blackwell as expected, and whether EMIB-T begins to generate meaningful earnings contribution from F3/29.
• The gap between Ibiden's actual OP and company forecasts, FactSet consensus, and Morgan Stanley forecasts.