1️⃣ AI Related Information
Today, the overseas computing power sector experienced a significant decline, with the CPO index dropping by 3.37%, the PCB index by 5.72%, the communication index by 4%, and electronics by 2.3%. Our analysis is as follows: 1. One possible reason: the market is reacting to the electricity shortage in North America, raising concerns that data centers may not be completed on schedule, affecting computing power demand. However, we believe this does not change the mid-term industry trend, and North America’s energy resources, such as natural gas, are still relatively abundant, which can resolve this bottleneck, indicating a strong and sustained demand for computing power. 2. Another possible reason: concerns over OpenAI seeking government funding support. This should be understood by the market after subsequent documents are released. We believe that many CSP manufacturers have already started to shift from computing power to models to applications since Q2, and the future looks promising. 3. A third possible reason: changes in market style, with some institutions locking in profits towards the end of the year, leading to a decrease in risk appetite. From the market perspective, growth styles in overseas computing power, domestic computing power, and robotics are all adjusting today, with most rising sectors at low levels, indicating a high-low cut phenomenon, which confirms this reason. In summary, we believe the third reason is the main cause. Against the backdrop of unchanged fundamentals in the AI industry, and with a continuous high growth trend expected over the next five years, we remain optimistic about overseas computing power targets with good performance and low valuations, suggesting to buy on dips and welcome a bright spring.
【Investment Recommendation】 Recommended major companies related to data center electricity for long-term outlook: 20251110. Most favored: large space, good layout, and a high probability of becoming major companies—#Megmeet, currently has performance, and mid-term narrative: #Kstar, with UPS clients expanding and expected to manufacture HVDC products. High barriers in the segment: #Newray. Currently undervalued with certain market expectation gaps: #Shenghong Co., #Changbao Co.
【Why Models Should Be Emphasized More Going Forward, and Why Google Has a Late-Mover Advantage】 1. #Google is the pioneer of the large model Transformer architecture and the inventor of AlphaGo, with profound technical reserves. 2. If we compare the “computing power -> model -> application business” model to “computer host -> operating system -> desktop software,” most of us would choose the optimal ecological model/operating system (Windows) and then backtrack to match the computer host platform (X86 host) with Windows software ecosystem. Therefore, the role of the #model in leading traffic is severely underestimated at present. 3. #Google’s business has an impressive closed-loop ecosystem, from monopolizing search engine data, to self-designed/supplied TPU architecture, to the stunning performance of large models Gemini/Claude (with stakes, can be said to be in the same camp) in the Arena arena/user, and the absolute monopoly of the existing Android system in the mobile end, to the many mature businesses of the existing cloud platform…etc. All of #these existing businesses can achieve efficiency transformation through Google’s own AI empowerment. 4. #Google has remained resilient amidst multiple crises, including a temporary liquidity crisis due to government shutdowns, AI bubble doubts, and fluctuating Fed interest rates, with its healthy Capex spending and complete business ecosystem being significant reasons for short sellers to hesitate. 5. Additionally, overseas chain businesses may not necessarily land in Dongda (for reasons you understand), and in the future, there may be one strongest chain domestically and one strongest chain overseas. How to choose domestically (Deepseek/Qwen/Kimi/Hunyuan)? How to choose overseas (GPT/Gemini/Claude)? Competition remains fierce, and conclusions are difficult to draw, but undoubtedly, the #first mover with the optimal ecosystem has the strongest first-mover advantage. Recommended targets: Zhongji Xuchuang, Shenghong Technology, Saiwei Electronics, Dekoli, Tengjing Technology.
【Shenwan Hongyuan Chemical】Yake Technology: Beneficiary of the strong storage cycle, maintaining key recommendations. #From an industry perspective: 1) The global storage chip industry is accelerating into a “super cycle,” with product price increases exceeding expectations. Unlike previous “de-stocking recoveries,” this cycle’s driving force comes from real demand driven by AI, leading to a steeper and more sustained future storage demand curve; 2) SK Hynix, Samsung, and Micron continue to expand HBM capacity, while Changxin and Changcun are accelerating their IPOs, and domestic expansions are also rapidly advancing, bringing significant incremental demand for materials; 3) Changxin HBM3 is expected to achieve mass production by the end of the year, catalyzing the domestic HBM industry chain. #From a company perspective: 1) Precursor: Revenue growth rate exceeding 30% in 2025, with core customers like Changxin and SK Hynix continuing to ramp up, and zirconium and hafnium products achieving significant breakthroughs at Samsung, with a projected growth rate of 30% in 2026; 2) Photoresist: Domestic substitution by downstream panel manufacturers is accelerating, and the new capacity of Xianzi is expected to continue to improve; 3) Silicon micropowder: The Chengdu factory is operating smoothly, with a significant improvement in gross margin, and expansion in the copper-clad laminate field is progressing well; 4) Electronic gases: The Inner Mongolia project is advancing rapidly, expected to be put into production in 2026, driving profitability recovery in the gas sector; 5) LNG board materials: Sufficient orders, especially with continuous breakthroughs from Korean customers, and new orders expected to exceed expectations.
【Changjiang Electronics】Dongshan Precision Update: The company’s 1.6t optical module products are currently being sampled for overseas customers. The rumors about low-price share grabbing are unfounded; next year, optical chips will be in short supply, and overseas customers do not determine shares based on price. We remain optimistic about the AI chain PCB, assembly, and storage!
【Changjiang Electric New】North American AI Electricity Shortage Recommends—Some Thoughts on Market Concerns. Our previous calculations regarding the North American AI electricity shortage have garnered significant attention. After a week of roadshows, the market has common concerns about some details, and we have made some reflections. #Is the US currently facing an electricity shortage? 【Yes】 In 2024, the average power outage duration for US electricity users is 662.6 minutes, up 81% year-on-year, marking a significant 10-year high; at the same time, from January to August 2025, the average sales electricity price across the US is 13.54 cents/kWh, up 5.2% year-on-year. #Can unit retirements and load transfers solve the US electricity shortage? 【No】 From 2025 to 2030, approximately 54G of thermal power units in the US are planned to be retired; as of January 2024, US cryptocurrency mining facilities account for about 10.3GW. Assuming all these power sources are released, they can only solve about 30% of the electricity gap for 2025-2030, and there will still be an electricity gap each year. #Are US data centers implementing energy storage? 【Yes】 Currently, there are increasingly more projects for data centers to build energy storage; typically, the GW Ranch off-grid power project in Texas is equipped with 5GW of gas power generation and 1.8GW of energy storage. #What is the policy for US data centers connecting to the grid? 【Non-discriminatory】 On October 23, the US Department of Energy proposed a rule to FERC, requesting FERC to initiate a rule-making process to ensure that large electricity loads (such as AI data centers) can connect to interstate grid systems in a timely, orderly, and non-discriminatory manner, and requiring the completion of rule-making by April 30, 2026. Therefore, we emphasize that the degree and sustainability of the North American AI electricity shortage may exceed expectations, and it is expected to drive the industry chain beyond expectations. #Continue to recommend: 1) Energy storage – Sunshine, Artas, Haibo, Trina, Tongrun, Xidian, Shuneng, etc.; 2) Transformers – Siyuan, Jinpan, Igor, TBEA, Xidian, Mingyang Electric, Samsung, Guangxin, Shenma, Huaming, Jinbei, etc.; 3) SST/HVDC – Sifang, Kstar, Megmeet, Chint, Liangxin, Hongfa, Shenghong, Zhongheng, etc.; 4) Large storage battery cells – CATL, Yiwei, Xinwangda, Penghui, Zhongchuang, Guoxuan, etc.; 5) Lithium battery materials – Tianci, Dofluor, Tianji, Fulian, Yuneng, Shangtai, Putailai, Zhongke, Enjie, Xingyuan, etc.
2️⃣ Robotics Related Information
【Changjiang Robotics】Emotional disturbances are the main factor, with many expectations for industry progress in November! Today, the robotics sector adjusted significantly, mainly influenced by two factors: 1) False news: It was rumored that Tesla would adjust its cumulative delivery target for robots to 1 million units by 2035. This was a market misinterpretation; Musk’s compensation plan spans ten years, with one goal being the cumulative delivery of 1 million robots, but the market misread it as Tesla planning to deliver 1 million robots only by 2035. 2) After the Tesla shareholder meeting, the market believes there is a window period, with no catalysts for short-term industry advancement, but there are still many expectations both domestically and internationally: 1) Tesla: Many suppliers have provided feedback, and November is expected to receive fixed-point orders, and T is very optimistic about the production capacity guidance for 2026. The actual fixed-point orders in the industry chain are expected to create significant catalysts. 2) Domestic chain: Xiaopeng Robotics is gaining attention, and Yushu and Leju are preparing for IPOs, which are expected to positively impact the related industry chain after landing.
3️⃣ Consumer Sector Related Information
【Zhejiang Merchants Food and Beverage】White liquor sector surges: Pay attention to PPI turning point signals and the left-side configuration window (20251110). Today, the food and beverage sector rose by 2.43%, with the white liquor sector increasing by 2.48%, led by Shede Liquor, Luzhou Laojiao, Jiu Gui Jiu, and Gujing Gong Jiu. We believe the surge in the sector today is mainly due to: 1) Macroeconomic indicators catalyzing: On November 9, CPI and PPI data were released, with October CPI rising by 0.2% year-on-year and 0.2% month-on-month; PPI decreased by 2.1% year-on-year, with the decline narrowing by 0.2 percentage points compared to last month, and month-on-month turning from flat to an increase of 0.1%. Previous research has shown that the #white liquor index is strongly correlated with PPI, so we suggest paying attention to PPI turning point signals. 2) Sector fund rotation: The consumer sector performed strongly today, possibly influenced by fund rotation from high to low. 3) Fundamental bottoming improvement: ① Rhythm judgment: In Q3 2025, leading liquor companies’ reports are actively clearing, and they are still bottoming out, with sales improvement not yet significant, but the demand side has reached a low point. Based on recent liquor company performance communication meetings, #we expect Q4 2025 to still be in the clearing stage, Q1 2026 to remain cautious under a high base, and Q2 2026 to see performance turning positive under a low base in H2 2026. ② Moutai’s batch price has recently stabilized in the range of 1600-1700 yuan, with the current scattered Moutai/box price at 1640/1660 yuan. #Considering that this range is basically the cost line for distributors, and the annual quota for general Moutai has been basically completed (some distributors still have about 8% quota remaining), and general Moutai inventory has dropped to less than one month, we expect the recent general Moutai batch price may be the low point of this adjustment period, but non-standard products may still be under pressure. Previous research on the adjustment period of white liquor has shown that #liquor company performance expectations hitting bottom + Moutai batch price hitting bottom are two major signals for stock price bottoming. Additionally, expectations lead, and stock prices and valuations may bottom out before fundamentals and rebound before fundamentals, so we believe attention should be paid to the current left-side configuration window for white liquor, with mid-term focus on inventory turning points. Investment recommendations: We believe that the current liquor company performance and Moutai batch price have reached the bottom range, and we should also pay attention to PPI turning point signals and white liquor sector configuration opportunities, as well as new consumption investment opportunities represented by “Da Zhen.” Based on this, we recommend: ① Prioritize leading brands such as Guizhou Moutai/Shanxi Fenjiu/Wuliangye; ② Recommend mid-range priced brands with share increase logic & low valuations, and high elasticity such as Zhenjiu Lidu/Luzhou Laojiao/Gu Jing Gong Jiu/Jinshi Yuan/Yingjia Gong Jiu/Shede Liquor, etc.
【Huachuang Food and Beverage】Food and beverage viewpoint update: White liquor expectations bottoming out to welcome buying points, and mass-market products focus on yellow wine catalysis and restaurant reversal potential. 20251110 #White liquor: Reports accelerating clearance, expectations for bottoming out are clear. 1⃣ Moutai supply easing, entering a low single-digit phase; Wuliangye determines the industry’s clearing speed, with the worst report stage confirmed this year; regional leader Gujing’s report has significantly declined beyond the sales decline, clearly starting to de-stock, and Yanghe Jinshiyuan’s report reflects real sales; 2⃣ Report rhythm outlook: Q4 2025 will continue the clearance extent of Q3 2025 reports (Lao Jiao may expand the decline), and Q1 2026 report decline will narrow. If demand improves, Q2 2026 may turn positive under a low base in H2 2026. 3⃣ We are currently welcoming a turning point where supply is fully released, and the supply-demand gap is converging. August was the first buying point at the bottom of the cycle (the extreme suppression of the liquor ban was lifted, and reports began to release pressure), #currently is the second buying point at the bottom of the cycle (reports are fully released, supply-demand gap is converging, and medium to long-term funds can further layout), and next year’s Q1 report may be the third buying point (demand bottom recovery, reports with low base starting to improve); 🚩Target selection: One is not killing performance【Moutai Fenjiu】; the other is fully cleared with potential catalysts for mass-market liquor during the Spring Festival【Gujing】; focus on【Zhenjiu】. #Mass-market products: Key recommendations include Kuaijishan, focusing on restaurant reversal catalysts, and continue to layout yeast beer. 1⃣ Kuaijishan Lanting’s sales are significantly accelerating, and this year’s goal of doubling 200 million is expected to be achieved without issue. The factory price has already increased in October, and channel feedback is positive, making it worth recommending as the peak season approaches; 2⃣ The third-quarter report continues to show differentiation, and the upward cycle of yeast operations has not yet reached its endpoint, with recent declines being a good time to increase positions; restaurant bottoms attract funds to layout, and companies with operational highlights have high elasticity, prioritizing Babi and Baoli; 3⃣ Layout leading beer and dairy companies with improving ROE. 🚩Selected targets include【Kuaijishan, Anqi】, restaurant preferences【Babi, Baoli】, bottom layout【Runpi, Yili, Yanjin】, and closely hold【Dongpeng, Nongfu】.
Service prices have shifted from decline to increase! Pay attention to hotel elasticity! 【Shenwan Retail and Service】
In October, policies such as expanding domestic demand continued to take effect, coupled with the National Day and Mid-Autumn Festival holidays, the Consumer Price Index (CPI) rose by 0.2% month-on-month and 0.2% year-on-year, while the core CPI, excluding food and energy prices, rose by 1.2% year-on-year, with the increase expanding for the sixth consecutive month. Service prices shifted from decline to increase. Service prices rose by 0.2% from a 0.3% decline last month, exceeding seasonal levels by 0.2 percentage points, impacting CPI’s month-on-month increase by about 0.07 percentage points. In services, the overlap of the National Day and Mid-Autumn Festival holidays led to strong travel demand, with hotel accommodation, plane tickets, and tourism prices rising by 8.6%, 4.5%, and 2.5%, respectively, all exceeding seasonal levels!! Hotel market: In October, the weekly RevPAR data for four weeks was +12%/-2%/+1%/6%; price indicators were 12%/2%/2%/5%. Recommendations include: Jinjiang Hotels: (Management expenses in Q3 decreased significantly, with remarkable results from the discontinuation of private connections, and CRS proportion significantly increased. Overseas, Q3 currently shows a slight increase in losses, but having passed the investment period, coupled with expectations of European interest rate cuts and RP recovery, is expected to gradually reduce losses.) Shouqi Hotels, Atour, Huazhu. Calm encounters: Hotel stocks surged today due to catalysts: 1) Data beta continues to improve: industry declines are narrowing, Jinjiang remained flat in September, and Huazhu’s Q3 expectations show a slight decline almost returning to positive. 2) Changes in Jinjiang Group’s chairman previously raised expectations for performance improvement. 3) The weekend CPI returning to positive catalyzed the hotel sector, which itself has two eyes on ADR, and CPI improvement is conducive to further price increases and profit margin enhancement. 4) Transportation + hotel data indicate that business travel demand has improved from September to October. Core conclusions: 1. It is expected that the hotel sector will see significant valuation recovery in Q4 2025 and Q1 2026. 2. Key catalysts: 1) Continuous improvement in operating data and narrowing declines; 2) Expectations for company-specific transformations boosting sentiment; 3) Sector fund preferences. 3. Recommended targets: Huazhu, Atour, Jinjiang, Shouqi. Recent industry communication indicates: 1. Next year, data and performance improvement is likely to become a consensus compared to this year. 2. Recent business travel improvements, transportation, and hotel data have verification, but OTA perceptions are not obvious, so sustainability is debatable. 3. The bottom is a consensus, but some leading companies’ performance and valuation space have not formed a consensus. 4. A-shares are more recognized for beta, and the alpha based on actual conditions remains highly debatable. Reiterating changes in the industry fundamentals (as explained last week): Change 1: Continuous month-on-month improvement in operating data, narrowing declines. Change 2: Overall supply is still accelerating, but structural differentiation. Change 3: Leading companies are changing, generally having elastic valuations and space that exceed industry beta: 1. Hotel valuations below 18x PE are historical bottoms, with a central tendency of 25x, and prosperity cycles at 30-40x, and ROE improvement hedges against slowing growth, so the range is basically stable. 2. Current hotel valuations: 2025/2026 net profit attributable to the parent and PE valuations for Huazhu: 4.1/4.5 billion, 21/18x PE; Atour: 1.77/2.4 billion, 21/15x PE; Shouqi: 0.9/1.05 billion, 18/15x PE; Jinjiang: 0.92/1.1 billion, 25/21x PE. 3. It is expected that valuations will first recover towards historical average valuations (25x PE).