中国半导体设备：2026年一季度模型更新，订单增长加速

China Semicap: 1Q26 model update, order growth reaccelerating in this call(link to NAURA, AMEC, Piotech models). The stocks performed fairly well YTD(Piotech +45%, AMEC +41%, NAURA+20%),but we think more upside awaits this yearwith the surging WFE demand from local fabs. We suggest investors should stay in themarket and wait for the catalysts on memory IPO, which could come anytime in the next fewmonths.Re-iterateOutperform on all three names,rankedAMEC>Piotech>NAURAdue to higherexposure to memory. +852 21232654qingyuan.lin@bernsteinsg.com +85221232694zheng.cui@bernsteinsg.com +852 2123 2626francis.ma@bernsteinsg.com the companies post earnings, consistentfeedback is China WFE spending in 2026 is2026,and localization ratiorampupquicklydue to theexport control (US vendors are nolongersellingtolocal DRAM players).Japanese vendors seem to stillbenefit a lotfromtheDRAMcapacityexpansion as the current baseline isfocusing onde-AmericanizationandJapanese vendors have maintained a high share for now.NAND has less visibility on 2H26but 1H26 order almost as strong as last year full year, indicating that the capacity expansioncould also be fasterthan expected.As NAND already de-Americanized in 2024,looks likeJapanese vendors are further losing share since local tools are cheaper and has caught upthe companies admit that it's a conservative assumption), but also will need to wait till 2H26won't grow as fast with a higher base last year, but the demand from memory outsourcingis pushing local mature logic fabs to continue capacity expansion,and WFE localizationalso going up steadily. Note that the localization ratio in logic is lower than memory,andthe US vendors mainly will benefit from the logic capacity expansion as most fabs can stillreacceleratingfrom2025at~30% NAURAdeliveredweak earningsfor4Q25and1Q26,whichhasbeen thecorereasonforits underperformance vs AMEC/Piotech. Tone from the company on profitability outlookwas still not positive for 2026 due to continuation on aggressive hiring, but expect the newemployeeaddition toslowdown in2H26andthus expectNPM improvement in2027.more cautious price reaction into the print. On the other hand, both AMEC and Piotechare expected to achieve NPM improvement with normalizing R&D ratio and reductionon material consumption for new tool development. We highlight Piotech who has notbeen very aggressive in hiring and is starting to capitalize part of their R&D spending coulddeliver very impressive NPM improvement for the next few quarters, which should givemarket confidenceto trade up the stockfurther.We don't expect competition to add morepressure on GPM as companieshave been reducing costof components to offset somepricing pressure, and improvement in product mix should also help. Continue on next page... wait forthe next catalyst.Although the market has alreadybeen positive onmemorycapacity expansion,our channelchecks showthat theupside is probablygoing tobe evenbigger than the bull buy side expectation. On top of that, beyond the new fabs ramping upin 2026 that is well expected (new DRAM fab in Shanghai and fab 3 for NAND in Wuhan),now we expect two new fabs for DRAM (one in Beijing and the other in Hefei) and two newfabs for NAND (Fab 4&5) will be ramping up in 2027, thus the strong growth momentum inmemory order will continue into 2027, bring further upside to AMEC/Piotech. In addition,the CXMT IPO and YMTC prospectus could come any time in the next few months, so it'salso worthwhileto hold forthat new catalyst to come.In terms of valuation, we raise 12-monthPTofNAURAtoCNY680 (was 650),AMECtoCNY500 (was460),Piotechto CNY580 (was 450).Welargely revise up the revenue projection given the positive signals frombothMemory&AdvanceLogicbutrevisedownGPMassumptionforallthreecompanies.WevalueNAURA/AMEC/Piotechat40x/7Ox/48xP/E(was48x/76x/48x)basedon2027EEPS of RMB 16.41/7.18/12.40 was(13.31/6.05/9.43), respectively. (PVD, CVD), Dry Etch (ICP), Thermo Processes, and Cleaning,as well as a more diverse client base covering leading logic, DRAM, AMEC (Outperform, CNY 5oo): Primarily focus on Dry Etch (CCP,ICP) with rapid expansion in Deposition (ALD, LPCVD, EPI),commonly perceived as the domestic WFE company with the best technology and widest global recognition, continue to benefitfrom the WFE domestic substitution in China with acceleration share gain. Piotech (Outperform,CNY58o):Rising domestic WFEvendor primarilyfocus on Deposition (PECVD,HDPCVD,SACVD,ALD)with expansion in W2W and c2w hybrid bonding equipment for advanced packaging. Piotech has a strong track record ofproduct innovation, which will allow it to benefit from the WFE domestic substitution in China with acceleration share gain halfoftotal4,oo0+newhireswillbeR&Demployees. expenditurecapitalized,differentfromNAURA/AMEC References to"Bernstein" or theFirm" in these disclosures relate to the following entities: Bernstein Institutional Services LLC (April 1, 2024 onwards), Sanford C. Bernstein & Co., LLC (pre April 1