日本半导体：小盘股的宏大思考（二）：美科尼克——DRAM扩产的纯标的

Japan Semi: Big Thinking on Small Caps (2): Micronics - Pure playfor DRAM expansion This report is part of a series of reports on Japan semi small cap companies, which wedon’t cover but occupy unique positions in the supply chain and potentially benefit from thetechnology shifts. Investors may find them interesting as the small caps generally trade atlower multiples than large caps, but the growth profile is not necessarily inferior.The focus ofthe report today is Micronics (MJC; 6871.JP, not covered). Also seeJapan Semi: Big Thinkingon Small Caps (1): Tokyo Seimitsu. As we mentioned in ASML and Japan SPE: DRAM shortage fuels capacity rush, the DRAMcapacity expansion is accelerating, and Micronics (MJC) as one of the most heavily DRAMexposed companies should benefit from it. Based on the growth drivers below, we believethe underlying probe card demand potentially can grow growth 32% / 27% in CY26/27,respectively, accelerating from 27% growth in 2025. We note four factors shaping Micronics's revenue outlook: (1) increased HBM contributiondriving up probe card content per wafer, (2) DRAM capacity expansion across all threemajor players, (3) Samsung HBM exposure as a concentrated but large revenue driver, and(4) Micronics's own manufacturing capacity expansion as the primary historical constraint. Micronics is a probe card manufacturer focused on memory DRAM testing, competingwith FormFactor (FORM.US), Technoprobe (TPRO.IT), and Japan Electronic Materials (JEM;6855.JP, all not covered). Micronics is a major beneficiary of DRAM capacity expansion andhigher HBM testing intensity. Probe cards account for roughly 85-90% of total revenue,of which memory DRAM constitutes approximately 85-88%, making Micronics one ofthe most concentrated listed plays on DRAM probe card demand globally. Also, in termsof customer mix, Micronics is heavily exposed to Samsung, which we expect to have thefastest HBM capacity growth in the near term, which should also benefit Micronics. BERNSTEIN TICKER TABLE INVESTMENT IMPLICATIONS We rate Disco (TP=¥70,800) and Tokyo Electron (TP=¥56,800) Outperform, and Advantest (TP=¥26,000) Market-Perform. DETAILS As we mentioned in ASML and Japan SPE: DRAM shortage fuels capacity rush, the DRAM capacity expansion is accelerating dueto surging AI HBM and DDR demand. Micronics is a semiconductor probe card maker focusing on DRAM, making it a pure playfor the DRAM capacity expansion theme. Based on the growth drivers below, we believe the underlying probe card demand potentially can grow growth 32% / 27% inCY26/27, respectively, accelerating from 27% growth in 2025. We lay out our detailed analysis below. Micronics is focused on DRAM probe card. In FY2025, Micronics reported record-high ¥70.2bn sales (Exhibit 1), as well asoperating profit of ¥16.5bn (Exhibit 2), and for FY2026, company guided to at least ¥88bn in sales, with a growth rate of >26%.The company generates most of their revenue from probe cards, which account for >90% of total revenue, and around 90% ofprobe card sales comes from DRAM, of which 40-50% is for HBMs. We focus on the probe card segment for the scope of this report. The test equipment (TE) segment includes testers, testsockets and other FPD related tools and has less than 10% of annual revenue, but we do not identify this as a primary growthdriver. MICRONICS GROWTH DRIVERS We believe the main growth drivers for Micronics are as follows: 1. HBM testing intensity, 2. DRAM capacity expansion 3.Samsung capacity ramp, and 4. Micronics’s capacity expansion. HBM testing intensity:HBM probe card demand has grown disproportionately relative to HBM's share of total DRAM waferoutput (Exhibit 3). In CY2025, HBM wafers represented approximately 16% of total DRAM wafer output by volume, yet HBM-related revenue accounted for approximately 40% of Micronics's DRAM probe card revenue. This ~2-3x ratio reflects severalfactors: •Higher wafer count per stack: As HBM migrates from 8-hi to 12-hi to 16-hi stacking, the number of wafers increases andyield becomes lower. Probing happens pre-stack, so more stacks simply means more wafer tests, and the number of DRAMcore wafers tested per final HBM package increases proportionally. •Higher testing complexity per wafer: Each generation of HBM requires tighter temperature control and more stringentelectrical specifications — each change typically requires a new probe card specification, and offsets the decreasing pincounts at migration. DRAM capacity expansion:Beyond just HBMs, all three major DRAM makers — Samsung, SK hynix, and Micron — are inactive capacity expansion cycles (Exhibit 4, Exhibit 5). Management confirmed that probe card demand is correlated withDRAM wafer starts rather than testers alone; as new fab capacity comes online, new probe cards are required for each newproduct configuration. Management also noted that the migration from DDR4 to DDR5, and within DDR5 to higher-densityconfigurations (16Gb → 24Gb → 32Gb), tends to ge