JPMorgan raises its focus on Japanese electronic components: AI server MLCC and substrate demand are the core themes

Overview

This is a JPMorgan summary report on valuations, target prices, and ratings for Japan's electronic components industry. The report covers areas including MLCCs, capacitors, connectors, ABF substrates, automotive components, discrete semiconductors, and power semiconductors, and uses AI server demand, automotive electronics, smartphones, and PCs as the main analytical threads.

Core views

The core view is that high-power GPUs/ASICs driven by AI servers, rising total TDP, and growing demand for high-capacitance MLCCs are changing the earnings elasticity of MLCCs and related high-end components; Murata, with its leading side-forming MLCC capabilities, is viewed as the company best able to capture this demand slope. At the same time, Ibiden benefits from AI/general server substrate demand and changes in its non-Intel customer mix, while TDK, Taiyo Yuden, and Rohm are also covered positively; however, the sector is not a blanket bullish call, as Nidec's target price implies a negative return and it is rated UW.

Analysis framework

The report mainly uses multidimensional valuation and earnings forecast indicators such as implied return to target price, rating distribution, P/E, P/B, EV/EBITDA, ROE, ROIC, FCF, FCF yield, and dividend yield, and combines them with MLCC supply-demand models, capacity utilization, AI server shipments and TDP trends, and estimates of MLCC value content and ASP per device across different application scenarios to assess relative company attractiveness.

Methodology notes

• Valuation methodsforward valuation multiples

  forward valuation multiples

  The report compares FY26E-FY28E P/E, P/B, and EV/EBITDA, and references 12-month and 24-month forward valuations as well as relative TOPIX valuation to assess sector pricing.

• fundamental_forecastearnings and cash flow forecast

  earnings and free cash flow forecast

  The report presents EPS, ROE, ROIC, FCF, FCF yield, net debt, and dividend-related metrics to assess earnings quality, capital returns, and cash flow elasticity.

• industry_modelMLCC supply-demand model

  MLCC supply-demand and capacity utilization model

  The report tracks MLCC revenue, market share, application mix, capacity, capacity utilization, and the impact of price declines on operating profit to assess cyclical turning points and margin elasticity.

• technology_demandAI server component content analysis

  AI server component value-content analysis

  Using GPU/ASIC shipments, total server TDP, average TDP, MLCC counts per rack, and capacitance changes, the report estimates incremental AI server demand for high-end MLCCs and related components.

Asset mapping & comparison

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

• Murata Manufacturing (6981)

  Core beneficiary, rated OW

  Strengths

  Leader in side-forming MLCCs, with strong elasticity to high-capacitance MLCC and AI server demand; target price implies 81.0% return.

  Weaknesses

  Still affected by MLCC price declines, capacity utilization, and end-demand cycles.

  Comparison

  Relative to peers, the report believes Murata is better able to capture incremental demand for higher-priced specifications such as 1005/47µF.

  Risks

  AI server ramp weaker than expected, deterioration in MLCC supply-demand, and larger-than-expected price declines.

• Taiyo Yuden (6976)

  MLCC-related beneficiary, rated OW

  Strengths

  Target price implies 79.6% return, benefiting from MLCC demand and margin recovery.

  Weaknesses

  Earnings are sensitive to MLCC industry capacity utilization and pricing cycles.

  Comparison

  Like Murata, it is a major covered MLCC name, but there may be differences in technology and high-end product positioning.

  Risks

  MLCC price pressure, volatility in consumer electronics demand, and weaker-than-expected capacity utilization.

• TDK (6762)

  Electronic components and magnetic/battery/sensor-related beneficiary, rated OW

  Strengths

  Target price implies 78.3% return, with business exposure to multiple growth areas including sensors, batteries, and magnetic applications.

  Weaknesses

  Its multi-business portfolio may create mixed impacts from different end-market cycles.

  Comparison

  Compared with pure-play MLCC companies, TDK has more diversified sources of benefit.

  Risks

  End-demand decline, deterioration in product mix, and FX volatility.

• Ibiden (4062)

  AI and server substrate beneficiary, rated OW

  Strengths

  Target price implies 63.7% return; the report focuses on SAP demand, ABF substrate customer revenue, and non-Intel capacity allocation.

  Weaknesses

  FCF is negative in some forecast years, and capacity expansion and customer mix changes may create execution risk.

  Comparison

  Compared with MLCC companies, Ibiden maps more directly to AI/general server substrate loading and SAP demand.

  Risks

  AI server substrate demand below expectations, customer concentration, capacity ramp timing, and capital expenditure pressure.

• Nidec (6594)

  Negative or cautious name, rated UW

  Strengths

  Has scale and a multi-business foundation.

  Weaknesses

  The table shows target-price implied return of -30.3%, one of the weakest risk-reward profiles in the coverage list.

  Comparison

  Compared with multiple OW-rated names, Nidec's relative attractiveness in the report is clearly lower.

  Risks

  Slower-than-expected earnings recovery, valuation derating, and business transformation execution risk.

• Rohm (6963)

  Power semiconductor and discrete semiconductor-related name, rated OW

  Strengths

  Target price implies 60.0% return; the report provides FY3/27 sales and OP estimates.

  Weaknesses

  The power semiconductor cycle and product demand may fluctuate.

  Comparison

  Used together with Toshiba and Mitsubishi Electric for power/discrete semiconductor market estimates.

  Risks

  Power semiconductor demand below expectations, intensified competition, and margins missing expectations.

Key data

• Number of covered names16 companiesThe disclosure page lists Alps Alpine, Hirose Electric, Ibiden, JAE, Kyocera, MinebeaMitsumi, Murata Manufacturing, NISSHA, Nichicon, Nidec, Nippon Chemi-Con, Niterra, Rohm, TDK, Taiyo Yuden, and Wacom.
• Simple average implied sector returnabout 36.5%From the Simple Average Return in the valuation table on page 2.
• Murata Mfg. target-price implied return81.0%Share price ¥8,400, target price ¥15,200, rating OW.
• Taiyo Yuden target-price implied return79.6%Share price about ¥12,525, target price about ¥22,500, rating OW.
• TDK target-price implied return78.3%Share price ¥3,085, target price ¥5,500, rating OW.
• Ibiden target-price implied return63.7%Share price ¥17,405, target price ¥28,500, rating OW.
• Nidec target-price implied return-30.3%Share price ¥2,583, target price ¥1,800, rating UW.
• Incremental AI server MLCC ASP estimateIncremental ASP per unit could exceed US$0.06The report compares demand per rack for GB300 and VR200, and believes the incremental specifications mainly come from 1005/22µF and 1005/47µF.
• FY25 AI server MLCC average ASPabout US$0.01The main text on page 26 mentions that the average ASP of FY25 AI server MLCCs is about US$0.01.
• Rohm and Toshiba semiconductor FY3/27 combinedsales about ¥1trn, OP about ¥90bnEstimates related to discrete semiconductors and power semiconductors on page 46.

Impact & implications

The investment implication is that opportunities in Japan's electronic components sector are concentrated in AI servers and rising value content of high-end MLCCs/substrates, rather than relying purely on a traditional consumer electronics recovery. High-capacitance MLCCs, side-forming processes, server substrate capacity, and customer mix may determine differences in earnings elasticity across companies; at the same time, automotive, smartphones, PCs, magnetic applications, and power semiconductor businesses will still affect valuation divergence.

Risks

• AI server GPU/ASIC shipments or capital expenditure coming in below expectations, weakening incremental demand for MLCCs and substrates.
• New capacity additions in the MLCC industry or weaker end demand leading to lower capacity utilization and greater pricing pressure.
• High-capacitance MLCCs, side-forming processes, or new-spec yield ramps falling short of expectations, affecting cost and margins.
• Recovery in traditional end demand such as automotive electronics, smartphones, and PCs falling short of expectations.
• JPY exchange rates, raw material costs, inventory cycles, and customer order adjustments may affect earnings forecasts.
• Company ratings and target prices are based on J.P. Morgan estimates; if FY26E-FY28E earnings, ROIC, FCF, or valuation assumptions change, the investment conclusions may also change.

What to watch

• AI server GPU/ASIC shipment volume, total TDP, and average TDP trends.
• The pull from new platforms such as GB300 and VR200 on MLCC unit counts, capacitance, and higher-priced specifications.
• Capacity utilization, ASP, and price-decline magnitude at MLCC makers such as Murata and Taiyo Yuden.
• Ibiden's SAP demand, ABF substrate customer mix, non-Intel capacity allocation, and factory utilization.
• Automotive MLCC value content per vehicle, market share of the top four suppliers, and automotive business margins.
• FY3/27 revenue and OP delivery for power/discrete semiconductor businesses such as Rohm, Toshiba, and Mitsubishi Electric.
• Changes in ROE, ROIC, FCF, and net debt for each covered company over FY26E-FY28E.