日本汽车与半导体：电装的下一个半导体收购目标是谁

Japan Autos & Semiconductors: Who could be Denso's nextsemiconductor acquisition target? Since Denso's acquisition proposal for Rohm was withdrawn at the end of April, a number ofinvestors continue to show interest and raise questions about Denso’s next steps. We believeM&A remains the most realistic and effective option to achieve the growth targets outlined inits mid-term plan. In this report, we examine Denso's semiconductor business strategy, andassess the strategic options potentially available to the company. Masahiro Akita+81 3 6777 6998masahiro.akita@bernsteinsg.com David Dai, CFA+852 2918 5704david.dai@bernsteinsg.com Denso to expand into industrial and consumer semis:Denso's semi business is entirelycaptive and centered on Power & Analog, including Si-IGBT, SiC, and battery managementASICs. Semiconductor revenue was equivalent to ~JPY 420 bn in FY3/22, with Discreteaccounting for ~74% of revenue and ASICs for ~16%. The mid-term plan highlightsexpansion of its semi business into industrial/consumer electronics. As Denso’s minorityinvestments and partnerships with power semi majors have focused on strengthening itsauto business, we believe a target with industrial and consumer exposure is key. Tomohiro Kashimoto+81 3 6777 6975tomohiro.kashimoto@bernsteinsg.com Seunghyeok Kim+81 3 6777 6974seunghyeok.kim@bernsteinsg.com Denso’s target segments look attractive:While overall semi market growth is driven byAI-related memory and ASIC/ASSP, Denso’s target segments still are appealing. Across9 devices × 7 applications combinations, Denso’s target 6 segments (Discrete/Analog ×Automotive/Industrial/Consumer) offer reasonable scale, relatively high and stable growth,and room for consolidation. The combined market size is expected to grow from USD 50 bn(2025) to USD 80 bn (2030), implying a CAGR of 10%. Jack Lin+852 2123 2683jack.lin@bernsteinsg.com Juho Hwang+81 3 6777 6980juho.hwang@bernsteinsg.com Who could be the best fit?We view Fuji Electric (NC) as the most feasible option in termsof acquisition cost, device/application portfolio and management fit. It has the lowestmarket cap among the major power semi candidates at ~JPY 2 tn, with a long relationship inIGBT/SiC development and meaningful industrial exposure. For the Rohm/Toshiba/MELCOconsortium (All NC), Denso might be able to take a partial stake, but a full acquisition wouldlikely be operationally difficult, in our view. A hostile acquisition of Rohm would also bedifficult given the reputational risk. While Onsemi (NC) has an attractive product portfolio,we think fully acquiring Onsemi, Renesas, Infineon/STM (NC) looks financially difficult. Carmine Milano, CFA+44 20 7762 1857carmine.milano@bernsteinsg.com Implications for Japan Autos:Automakers are internalizing H/W-linked S/W layerswhere Denso has been strong amid the shift to SDVs making expansion in semiconductorsimportant. Cost reduction of power semi is critical for improving Toyota’s xEV profitability,which would potentially make Denso’s M&A reasonable. But using Toyota stake proceedsfor M&A could disappoint investors expecting shareholder returns, while buybacks wouldbe positive for Toyota. ReiterateMarket-Perform/OutperformonDenso/Toyota. Implications for Japan/EU Semiconductors:We are constructive on power semi giventhe rise of AI power (Link). We think a Fuji Electric transaction would likely have a limitedimpact on Renesas, as Denso’s benefit would be concentrated in power semis. A Rohm/Toshiba structure could be more relevant for Renesas by strengthening Denso’s domesticpower semi platform and reducing external sourcing. However, Renesas’s MCU leadershipand broad customer base would likely limit core business risk. The impact on Infineon wouldlikely also be limited. ReiterateOutperformonRenesasandInfineon. BERNSTEIN TICKER TABLE INVESTMENT IMPLICATIONS Japan Autos and Auto Parts: As the industry shifts toward SDVs, automakers are increasingly seeking control over hardware-linked upstream softwarelayers, where suppliers such as Denso have historically captured value through integrated hardware and embedded softwareofferings. This shift could reduce the scope for supplier-led system design over time, making Denso’s expansion into upstreamhardware such as semiconductors, as well as non-auto businesses, strategically important. For Toyota, cost reduction of powersemiconductors, particularly SiC, is important for improving its xEV competitiveness, which would make a potential acquisitionby Denso’s of a power semiconductor company strategically reasonable. However, if Denso were to use proceeds from apotential sale of its Toyota stake to fund semiconductor M&A, this could disappoint Denso’s investors expecting shareholderreturns. Conversely, if Toyota were to buy back those shares, it would be positive for Toyota. We remainMarket-PerformonDenso(PT = 2,050.00) andOutperformonToyota(PT=4,200.00). Japan and European Semiconductors: We are constructive on power semi given the r