The batch ALD leader is poised to benefit simultaneously from global memory capacity expansion, Chinese demand, and high-value equipment upgrades

Overview

The report examines Kokusai's technological and market advantages in batch atomic layer deposition (ALD), focusing on how global memory capex, capacity expansion by domestic Chinese memory manufacturers, and mini-batch ALD upgrades translate into the company's revenue, product mix, margins, and valuation. Bernstein believes these drivers will have a compounding effect and maintains its Outperform rating.

Core views

Kokusai is a Japanese semiconductor equipment manufacturer specializing in deposition and thin-film processing. Its main products include large-batch and mini-batch deposition systems and single-wafer processing equipment. By sequentially introducing precursors, purging the chamber, and repeating self-limiting reactions, ALD achieves near-atomic-level film-thickness control and superior conformality, enabling more uniform coverage of the tops, sidewalls, and bottoms of deep trenches and holes. The trade-off is that repeated deposition and purging make it slower than CVD's continuous reactions. Batch ALD processes multiple wafers simultaneously, offsetting ALD's slower speed through productivity and cost-per-wafer advantages, making it particularly suitable for large-scale memory manufacturing with relatively standardized process flows. Batch and single-wafer architectures address different customer requirements. Large-batch equipment processes dozens of wafers at a time and emphasizes throughput. Mini-batch equipment uses more compact reaction chambers and exchanges smaller wafer loads for faster gas exchange, shorter purge times, and tighter process control. Single-wafer equipment offers greater flexibility and wafer-level control but generally entails a higher cost per wafer. Accordingly, the report argues that Kokusai's large-batch and mini-batch ALD systems are better suited to memory IDMs such as DRAM and NAND manufacturers, while single-wafer ALD suppliers such as ASMI are more oriented toward logic and foundry customers with complex product mixes and frequent recipe changes. ALD remains a relatively niche segment of the deposition market, but its share is increasing. In 2025, the ALD market was approximately $4.1bn, representing about 17% of the $24.9bn deposition market. ALD accounted for approximately 27% of the broader CVD market, while batch equipment represented approximately 13% of the ALD market. Bernstein expects the overall deposition market to grow at a 13% CAGR through 2031, with ALD's share of the deposition market rising from 16.5% in 2025 to 18.7% in 2031. The batch ALD market is primarily a duopoly between Kokusai and Tokyo Electron. Kokusai's share has steadily increased from 33% in 2016 to 59% in 2025, and the report views this leadership as a key foundation for capturing incremental demand. The memory capex cycle is the core earnings driver. Kokusai's memory exposure reached 61.6% in FY26, with DRAM accounting for 39.0% of equipment sales and NAND for 22.6%, significantly above the 16% memory exposure of single-wafer ALD peer ASMI. Bernstein expects global DRAM and NAND WFE spending to more than triple from 2025 to 2028, compared with 62% growth in logic and foundry WFE over the same period. NAND also has greater Kokusai equipment intensity: the company's equipment represents 3.1% of NAND WFE, nearly twice the 1.6% level in DRAM. NAND WFE spending was only $9bn in 2025, far below DRAM's $30bn, so Kokusai could benefit even more if the NAND investment gap narrows. The Chinese memory market represents a second structural growth avenue. China accounted for 39% of Kokusai's revenue in FY26/3. Management's FY27/3 outlook for China implies approximately 41% year-over-year growth, including multinational customers. The company expects overall growth of 32% among domestic Chinese customers, while domestic memory customers could grow by approximately 60%, driven by new capacity ramp-ups and orders from new fabs. Bernstein expects Chinese DRAM WFE to grow 5.5-fold over the next three years, significantly faster than overall Chinese WFE. It also expects Chinese DRAM capacity to increase to approximately five times its current level over the next five years, while NAND capacity rises to approximately 4.5 times its current level. Under a scenario assuming Kokusai's share of China's domestic memory market remains unchanged and extrapolating based on memory WFE growth, its domestic Chinese memory revenue could theoretically reach ¥151bn in FY29, equivalent to 4.6 times the FY26 trough. The report believes the near-term risk from Chinese equipment localization remains limited. The key criterion is not merely whether domestic suppliers offer ALD products, but whether they can simultaneously meet the requirements for film uniformity, conformality, and throughput in high-volume memory applications. Piotech's ALD revenue mainly comes from single-wafer PEALD, while its batch equipment has yet to be commercialized. Leadmicro has a more directly comparable batch architecture, but publicly available information is insufficient to demonstrate that it has achieved a meaningful share or a performance breakthrough in advanced high-volume memory processes. NAURA has a broader batch deposition platform and a strong historical track record, making it a credible long-term competitor, but the report has yet to see evidence of its entering Kokusai's core areas of strength in the near term. Bernstein therefore expects Kokusai to be a major beneficiary of Chinese memory batch ALD demand over the next several years. Mini-batch ALD is an additional catalyst for technology, selling prices, and margins. As NAND progresses from 200—300 layers to 300—400 layers and beyond, vertical channel holes become deeper, their aspect ratios and coating surface areas increase, and the requirements for uniformity and process control become more stringent. Kokusai's TSURUGI mini-batch platform improves film-thickness uniformity through a compact reaction chamber, faster process-gas purging, and reduced loading effects while retaining multi-wafer productivity. Mini-batch and large-batch processes currently each account for approximately half of the market. At the 200—300-layer stage, mini-batch penetration is expected to rise to 60%—70%. Beyond 300—400 layers, the company believes mini-batch becomes almost essential. Replacing large-batch equipment with first-generation mini-batch equipment can generate an approximately 2x ASP, while ASP rises by approximately 50% with each product generation. The third-generation product, which supports up to 400 layers, has an ASP approximately twice that of the first generation. Mini-batch equipment gross margin is approximately 10 percentage points above the equipment average, while large-batch products are only approximately 5 percentage points above it. High-value equipment accounts for 93%, 77%, and 62% of NAND, DRAM, and logic/foundry, respectively, so NAND growth and rising mini-batch penetration are expected to improve the product mix together. Based on memory capacity investment in China and elsewhere, scale benefits, and a rising share of high-value-added products such as mini-batch systems, Bernstein expects Kokusai's revenue to grow by 50% and 34% in FY27/3 and FY28/3, respectively, 6% and 13% above consensus. Operating margin is expected to increase from 17.8% in FY26 to 26.1%. EPS is forecast at ¥262.3 and ¥373.7 over the same period, representing year-over-year growth of 104% and 42%, respectively, and an approximately 72% two-year EPS CAGR. The ¥12,420 target price applies a 30x one-year forward P/E multiple to EPS of ¥413.8 for future Q5—Q8, also described as 24-month forward EPS, implying 53% upside. The firm therefore maintains its Outperform rating. Corporate governance and the shareholder structure provide additional context. Kazunori Tsukada became President and CEO in 2026 and was reappointed as Representative Director in June 2026. Following the June 2026 shareholders' meeting, six of the eight directors were outside directors, three of the four Audit and Supervisory Committee members were outside directors, and three directors were women, representing 37.5% of the board. The company has no controlling shareholder or large strategic shareholder group. Applied Materials Europe holds 4.98%, while former controlling shareholder KKR has fully exited.

Analysis framework

Bernstein first compares the technical characteristics of ALD, CVD, and different wafer-processing architectures to explain why batch ALD is suitable for high-volume memory production. It then combines deposition market size, ALD penetration, and Kokusai's historical market share to assess its competitive position. The report subsequently maps global and Chinese memory WFE, DRAM and NAND equipment intensity, and capacity expansion to the company's revenue, and estimates the potential of its Chinese business under a constant-market-share scenario. Finally, it incorporates changes in mini-batch equipment penetration, ASP, and gross margin into its earnings forecasts and uses a forward P/E multiple to determine the target price.

Methodology notes

• Industry/Sector Analysis FrameworkUpstream, Midstream, and Downstream Value-Chain Transmission

  Transmission of memory capex into WFE, deposition equipment, and company revenue

  Starting with DRAM and NAND capacity investment, the report sequentially estimates WFE spending, the intensity of Kokusai's equipment within WFE, and the company's market share to explain how industry capacity expansion translates into company orders and revenue.

• Industry/Sector Analysis FrameworkSupply-demand framework

  Memory capacity expansion and equipment demand forecasts

  The report uses global and Chinese DRAM and NAND capacity and equipment spending forecasts to assess incremental equipment demand, with particular analysis of the potential catch-up opportunity created by NAND investment remaining low relative to DRAM.

• Industry/Sector Analysis FrameworkVolume-price decomposition

  Equipment demand growth, product upgrades, and structural price increases

  The earnings analysis considers both equipment volume growth driven by memory investment and the pricing and product-mix improvements generated by the shift from large-batch to high-ASP, high-margin mini-batch equipment.

• Competition and Strategy FrameworkMoat / competitive advantage

  Competitive barriers formed by batch ALD technological capabilities and market share

  The report assesses the difficulty of replacing Kokusai relative to domestic and international competitors based on its 59% market share, long-term share gains, and the requirements of high-volume memory processes for uniformity, conformality, and throughput.

• Valuation MethodPE/PEG valuation

  Forward P/E target valuation

  Bernstein applies a 30x one-year forward P/E multiple to EPS of ¥413.8 for future Q5—Q8, also described as 24-month forward EPS, to derive a target price of ¥12,420.

Asset mapping & comparison

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

• Kokusai Electric Corp

  The company is a direct beneficiary of rising batch ALD demand and advanced-memory capex, as well as memory capacity expansion in China and mini-batch equipment upgrades.

  Strengths

  A 59% share of the batch ALD market, relatively high memory exposure among WFE peers, and mini-batch equipment combining high ASPs with high gross margins.

  Weaknesses

  Revenue has high exposure to memory and Chinese investment cycles, while earnings forecasts depend on sustained capex growth and product-mix improvement.

  Comparison

  Compared with single-wafer ALD peer ASMI, where memory accounts for only 16% of the business, Kokusai's memory exposure is 61.6%. The batch ALD market is primarily a duopoly with Tokyo Electron.

  Risks

  An earlier-than-expected peak in Chinese investment, faster domestic substitution, and adverse currency fluctuations could reduce revenue or profit.

Key data

• 2025 ALD market size$4.1bnApproximately 17% of the $24.9bn deposition market
• ALD share of the deposition market16.5% to 18.7%Expected to rise from 16.5% in 2025 to 18.7% in 2031
• Kokusai's batch ALD market share59%Increased from 33% in 2016 to 59% in 2025
• FY26 memory business exposure61.6%DRAM accounted for 39.0% and NAND for 22.6%; ASMI's memory exposure was 16%
• Kokusai equipment intensityNAND 3.1%; DRAM 1.6%NAND equipment intensity is nearly twice that of DRAM
• 2025 memory WFE spendingNAND $9bn; DRAM $30bnNAND investment remains significantly below DRAM
• Global memory WFE forecastMore than triple from 2025—2028Both DRAM and NAND are expected to grow by more than threefold, while logic and foundry WFE grows by 62% over the same period
• China's share of FY26/3 revenue39%China is already an important source of revenue for Kokusai
• FY27/3 China revenue outlookApproximately +41% year-over-yearIncludes multinational customers; domestic Chinese memory customers are expected to grow by approximately 60%
• China DRAM WFE forecast5.5-fold growth over the next three yearsExpected to significantly outpace overall Chinese WFE growth
• Domestic Chinese memory revenue scenarioReaches ¥151bn in FY29Assuming unchanged market share, equivalent to 4.6 times the FY26 trough
• Mini-batch ALD penetrationCurrently approximately 50%; 60%—70% at 200—300 layersThe company believes it becomes almost essential beyond 300—400 layers
• Mini-batch ALD margin premiumApproximately +10 percentage pointsRelative to the company's average equipment gross margin; large-batch products are approximately 5 percentage points higher
• Revenue growth forecastFY27/3 +50%; FY28/3 +34%6% and 13% above consensus, respectively
• Operating margin forecastRises from 17.8% to 26.1%From the FY26 level to the forecast-period level
• EPS forecast¥262.3; ¥373.7Year-over-year growth of 104% and 42% in FY27/3 and FY28/3, respectively, with an approximately 72% two-year CAGR
• Target price¥12,420A 30x forward P/E multiple applied to EPS of ¥413.8, implying 53% upside

Impact & implications

The report believes Kokusai's high exposure to memory equipment investment gives it greater revenue sensitivity than WFE peers with more diversified businesses or greater exposure to logic applications. A catch-up in NAND investment relative to DRAM and continued expansion of domestic Chinese memory capacity could both amplify this sensitivity. Meanwhile, the shift from large-batch equipment to mini-batch ALD driven by increasing NAND layer counts means growth will come not only from equipment volumes but also from a product mix with higher ASPs and better margins. The firm therefore forecasts that Kokusai's earnings growth will significantly outpace its revenue growth.

Risks

• Chinese investment may be weaker than expected, or capex may peak earlier than anticipated.
• Substitution by domestic Chinese equipment suppliers may proceed faster than expected.
• Adverse currency fluctuations may affect the company's performance.

What to watch

• Monitor news on incremental NAND capacity and capex, because Kokusai's equipment intensity in NAND WFE is higher than in DRAM.
• Monitor the ramp-up of new domestic Chinese memory fabs, order conversion, and the progress of DRAM and NAND capacity expansion.
• Monitor whether Chinese suppliers can simultaneously meet film uniformity, conformality, and throughput requirements in high-volume memory applications.
• Monitor mini-batch ALD penetration and product-mix changes in 200—300-layer and higher-layer NAND.