Investment Rating - The report assigns an "Outperform" rating to multiple companies in the consumer staples sector, including Guizhou Moutai, Wuliangye, Luzhou Laojiao, and others, indicating a positive outlook for these stocks [1]. Core Insights - There is an intensified divergence among mid-to-high-end baijiu products, with no discount contraction observed in food and beverage categories [1]. - The report highlights that the wholesale prices of various baijiu brands have shown stability or slight changes, with some brands experiencing price increases while others remain flat or decrease [3][4][35]. - Discounts on representative products in condiments, liquid milk, and soft drinks have widened compared to the end of December, indicating a shift in pricing strategies [19][32]. Summary by Relevant Sections Baijiu Pricing - Guizhou Moutai's wholesale prices for Feitian (case and single bottle) and Moutai 1935 are reported at 1600, 1580, and 630 yuan respectively, with changes of 0, -10, and -10 yuan compared to last month [35]. - Wuliangye's eighth-generation price is 830 yuan, reflecting a +10 yuan increase from the previous month [35]. - Luzhou Laojiao's Guojiao 1573 remains at 850 yuan, unchanged from last month [35]. Discount Trends - The average discount rate for condiments decreased from 87.1% to 85.0%, while liquid milk discounts fell from 62.8% to 61.4% [19][22]. - In contrast, discounts for beer, infant formula, and convenience foods remained stable, with beer discounts slightly increasing from 80.3% to 80.7% [20][22][32]. Company Ratings - The report lists several companies with an "Outperform" rating, including: - Guizhou Moutai - Wuliangye - Luzhou Laojiao - Qingdao Beer - China Feihe - Others [1].

Investment Rating - The report indicates a mixed outlook for the Japanese consumer sector, with essential goods remaining resilient while discretionary spending shows signs of weakness [1]. Core Insights - The Japanese consumer sector in December exhibited three main characteristics: essential goods maintained resilience, discretionary spending weakened, and industry consolidation accelerated [3][15]. - Consumer confidence index in December recorded 36.2, slightly down from 36.4 in November, indicating a marginal decline in household consumption willingness [2][9]. - The core CPI in December rose by 2.4% year-on-year, showing a significant drop from November, while the CCCPI increased by 2.9%, indicating persistent high inflation levels [2][11]. Summary by Sections Macro - Inflation is slowing down, and the yen has appreciated significantly. The consumer confidence index in December was 36.2, slightly lower than November's 36.4 and below market expectations [2][9]. - The willingness to purchase durable goods decreased significantly, reflecting a marginal weakening in household consumption tendencies [9]. - Actual wages in November contracted by 2.8% year-on-year, marking an ongoing negative growth trend for 11 consecutive months [9][11]. Industry - The consumer sector is experiencing a general slowdown in growth rates across sub-industries, with duty-free consumption continuing to weaken. Essential goods like food and daily necessities show stable demand, while discretionary items like clothing and home goods are under pressure [3][15]. - External disturbances have notably impacted retail growth, with a warm winter suppressing demand for winter goods and a reduction in holiday shopping days negatively affecting retail sales growth by approximately 2 percentage points [3][15]. - The industry is witnessing an accelerated pace of consolidation, with significant mergers and acquisitions, such as the full acquisition of Welcia by Katsuya and subsequent TOB completion by Aeon [3][15]. Essential Companies - In December, same-store sales for Aeon decreased by 0.3%, while PPIH and 711 Japan reported increases of 3.0% and 1.9%, respectively [4][18]. - The drugstore sector saw Matsukiyo Cocokara's same-store sales decline by 4.6%, while Katsuya's increased by 0.3% [21][19]. - The beverage sector reported a 6% increase in sales for Suntory, while Asahi's sales dropped by 22% [4][22]. Discretionary Companies - In December, same-store sales for major restaurant chains like Saizeriya and Skylark increased by 18.7% and 11.0%, respectively, while clothing retailers like Uniqlo and ABC-MART saw declines of 6.6% and 0.4% [5][35]. - The overall sales for department stores in Japan fell by 1.1% year-on-year, marking the first negative growth in five years, primarily due to a warm winter and a significant drop in duty-free sales [40][41]. - The hotel sector reported a 5.1% increase in revenue, with a 86% occupancy rate, reflecting a recovery in tourism [43][44].

Investment Rating - The report maintains an "Outperform" rating for Zai Lab with a target price of HKD 35.25, indicating an expected return exceeding the market benchmark over the next 12-18 months [2][19]. Core Insights - Zai Lab is projected to achieve product net revenue of USD 465 million in FY25, representing a year-on-year growth of 16.5%. The core product, efgartigimod, is expected to generate USD 97.96 million, with a gross profit margin of 61.0% [3][16]. - The management emphasizes a dual-engine strategy focusing on introducing high-quality overseas products for commercialization in China and leveraging a cross-border R&D platform for efficient global market development [4][17]. - Despite a downward revision in revenue guidance for 2025 and a delay in profitability targets, the company has a robust portfolio with eight products approved in China and several differentiated global assets in its pipeline [4][17]. Financial Performance - Revenue forecasts for FY25-FY27 have been adjusted to USD 465 million, USD 501 million, and USD 717 million, respectively. Net loss estimates for the same period are revised to USD 180 million, USD 159 million, and USD 55 million [10][19]. - R&D expenses are expected to be USD 205 million, a decrease of 12.8% year-on-year, while SG&A expenses are projected at USD 284 million, down 5.0% [3][16]. Clinical Development Pipeline - Zocilurtatug (DLL3 ADC) is set to initiate three registrational clinical trials by the end of 2026, targeting small cell lung cancer and neuroendocrine carcinoma [5][21]. - Several early-stage products, including ZL-1503 for atopic dermatitis and ZL-6201 for various solid tumors, are anticipated to contribute to long-term global value [6][9][21].

wo[Table_Title] Research Report 29 Jan 2026 中国电子 China (Overseas) Technology 全球科技业绩快报：希捷 FY26Q2 Global Tech Earnings Snapshot: Seagate FY26Q2 姚书桥 Barney Yao 吕小潼 Xiaotong Lyu barney.sq.yao@htisec.com xt.lyu@htisec.com [Table_yemei1] Flash Analysis [Table_summary] (Please see APPENDIX 1 for English summary) 事件 Seagate FY26Q2 业绩表现：营业收入为 28.25 亿美元，同比增长 22%、环比增长 7%；non-GAAP 毛利率为 42.2%（环 比提升 210bp），经营利润率为 31.9%，EPS 为 3.11 美元。本季度 HDD 总容量出货达 190EB，同比增长 26%，其中 近线（Nearline）硬盘出货 165EB，同比增长 31%。按业务划分，数据中心收入为 22.24 亿美元（ ...

Investment Rating - The report does not explicitly provide an investment rating for the industry or specific companies involved in the DeepSeek-OCR 2 development. Core Insights - DeepSeek-OCR 2 represents a significant advancement in document understanding technology, particularly in handling complex layouts by utilizing a new visual encoder, DeepEncoder V2, which enhances the model's ability to parse text, tables, and formulas more accurately and efficiently [12][14]. - The model has achieved a score of 91.09% on the OmniDocBench v1.5 benchmark, indicating it has entered the top tier of document understanding models, with a notable improvement in reading order accuracy [14]. - The model's efficiency allows it to process complex documents with only 256 to 1120 visual tokens, significantly reducing computational load and latency for downstream applications [15]. Summary by Sections Model Upgrade and Features - The DeepSeek-OCR 2 model introduces a lightweight language model, Qwen2-500M, and a "causal flow query" mechanism that reorganizes visual tokens based on content logic, improving semantic continuity and recognition accuracy [13][14]. - The model's architecture allows for a more human-like understanding of document structure, which is crucial for processing complex documents like multi-column layouts and nested tables [12][13]. Performance Metrics - DeepSeek-OCR 2's edit-distance metric improved from 0.085 to 0.057, validating its structure-first reading approach [14]. - Compared to competitors, DeepSeek-OCR 2's performance is approaching industry leaders, with a document-parsing edit distance of 0.100, outperforming Gemini 3 Pro [14]. Real-World Applications - The model's open-source nature and moderate parameter size (3 billion) facilitate its integration into existing enterprise workflows, with potential applications in PDF-to-Markdown conversion and structured data extraction [15]. - Feedback from production environments indicates a significant reduction in text duplication rates, suggesting improved reliability in practical applications [15]. Long-Term Vision - The development of DeepSeek-OCR 2 is seen as an exploration of architectural innovation, aiming to enhance the capabilities of vision-language models and improve the generation of structured training data for large language models [16]. - The team has outlined clear iterative directions for future improvements, focusing on enhancing performance for text-dense documents [16].

Investment Rating - The report does not explicitly state an investment rating for LVMH, but it implies a cautious optimism regarding the company's performance and strategies moving forward [7]. Core Insights - LVMH's revenue in 2025 was resilient at €80.8 billion, with an organic growth rate of -1% year-on-year, but profits were significantly impacted by foreign exchange headwinds and pressures in high-margin sectors, leading to a 9% decline in operating profit to €17.8 billion [8][9]. - The company experienced a notable divergence in performance across its segments, with Watches & Jewelry and Selective Retailing showing strong growth, while Fashion & Leather Goods and Wines & Spirits faced challenges [9][10]. - Management emphasized the importance of monitoring foreign exchange impacts on operational performance and profit sensitivity, indicating a strategic focus on cost discipline and selective investments to maintain cash flow and improve operational quality [8][12]. Summary by Sections Revenue Performance - LVMH achieved €80.8 billion in revenue for 2025, with a reported decline of 5% year-on-year, primarily due to foreign exchange effects [8][11]. - The organic growth rate improved in the second half of 2025, reaching +1% in Q4, after a -3% decline in the first half [8][9]. Segment Analysis - In Q4 2025, organic revenue growth rates were as follows: Watches & Jewelry +8%, Selective Retailing +7%, Perfumes & Cosmetics -1%, Fashion & Leather Goods -3%, and Wines & Spirits -9% [9][10]. - For the full year, Selective Retailing grew by 4%, driven by Sephora's strong performance, while Watches & Jewelry saw a 3% increase, largely supported by the Japanese market [9][10]. Brand Performance - Core brands such as LV, Dior, and Tiffany showed varying strategies: LV maintained stability through new product launches and limited editions, Dior focused on aggressive marketing and product expansion, while Tiffany is in a transformation phase to enhance its brand and product mix [10][12]. - The report highlights the importance of high-end demand, noting a structural split in customer sensitivity to economic conditions, with ultra-high-net-worth individuals being more stable compared to the broader high-income segment [12][13]. Geographic Insights - LVMH's revenue mix remained balanced across regions, with the U.S., Europe, and Asia (excluding Japan) each contributing 26% to total revenue, while Japan's share decreased to 8% [11]. - The report notes that the U.S. and Europe showed stable growth patterns, while Japan experienced a significant decline of 12% in organic growth for FY25 [11]. Future Outlook - Management expressed a medium-to-long-term optimistic outlook, citing rising global living standards as a support for high-quality product demand, but acknowledged short-term uncertainties due to geopolitical and macroeconomic factors [7][13]. - The controlling Arnault family plans to increase its stake in the company, reflecting confidence in LVMH's long-term value creation capabilities [13].

研究报告 Research Report 28 Jan 2026 西锐 Cirrus Aircraft (2507 HK) 调研反馈：产品叠加服务铸就的护城河 US site tour feedback: great product paired with excellent service build a durable moat 观点聚焦 Investment Focus [Table_Info] 维持优于大市 Maintain OUTPERFORM 评级 优于大市 OUTPERFORM 现价 HK$61.00 目标价 HK$70.20 HTI ESG 2.6-3.3-3.5 E-S-G: 0-5, (Please refer to the Appendix for ESG comments) 市值 HK$22.33bn / US$2.86bn 日交易额 (3 个月均值) US$4.35mn 发行股票数目 365.99mn 自由流通股 (%) 20% 1 年股价最高最低值 HK$73.40-HK$20.55 注：现价 HK$61.00 为 2026 年 01 月 28 日收盘价 资料来源： Factset ...

研究报告 Research Report 28 Jan 2026 安踏体育 ANTA SPORTS (2020 HK) [Table_Info] 维持优于大市 Maintain OUTPERFORM 评级 优于大市 OUTPERFORM 现价 HK$77.90 目标价 HK$91.40 HTI ESG 0.7-1.6-3.5 E-S-G: 0-5, (Please refer to the Appendix for ESG comments) 市值 HK$217.86bn / US$27.94bn 日交易额 (3 个月均值) US$93.59mn 发行股票数目 2,797mn 自由流通股 (%) 47% 1 年股价最高最低值 HK$104.40-HK$75.50 注：现价 HK$77.90 为 2026 年 01 月 27 日收盘价 资料来源： Factset 1mth 3mth 12mth 绝对值 -7.2% -13.3% -0.1% 绝对值（美元） -7.4% -13.6% -0.2% 相对 MSCI China -10.3% -11.6% -33.7% [Table_Profit] Rmb mn Dec ...

Financial Performance - Halliburton reported a net profit of $589 million in Q4 2025, exceeding consensus expectations due to strict cost control and high-margin international projects[1] - Operating cash flow reached $1.2 billion, with free cash flow totaling $875 million for the quarter[1] - Total free cash flow for 2025 amounted to $1.9 billion, with 85% allocated to a $1 billion share buyback, up from $525 million in 2024[1] Business Segments - Drilling and Evaluation segment revenue was $2.389 billion, surpassing consensus estimates of $2.361 billion, with a profit margin of 15% compared to the expected 14.2%[2] - Completion and Production segment revenue reached $3.268 billion, exceeding expectations of $3.074 billion, with a profit margin of 17% against a forecast of 16.3%[2] Risks - Potential risks include declining oil and gas capital expenditures, reduced demand for oil services, cost overruns, and falling oil prices[3]

Investment Rating - The report does not explicitly state an investment rating for the industry or specific companies involved in the research Core Insights - The Engram model proposed by DeepSeek and Peking University introduces a "Conditional Memory" mechanism that separates static knowledge recall from complex computations, significantly improving computational efficiency and task performance [1][2] - Engram-27B demonstrates systematic improvements over MoE-27B across multiple benchmarks, particularly excelling in long-context tasks [1][3] - The architecture allows for the offloading of large parameter tables to host memory, maintaining controllable inference throughput impact, thus validating the feasibility of "separation of storage and computation" [1][6] Summary by Sections Event - In January 2026, DeepSeek and Peking University released a paper on the Engram model, achieving significant performance improvements in various benchmarks while maintaining computational efficiency [1][17] Commentary - Engram innovatively decouples the recall of fixed knowledge from complex model computations, allowing models to focus on deeper reasoning tasks, thus enhancing overall efficiency [2][18] Performance Optimization - The study reveals an optimization path for resource allocation, indicating that transferring some model capacity to a conditional memory module can lead to a "U-shaped" performance trend, with a clear optimal performance range [3][19] - Replacing approximately 20% of traditional parameter capacity with conditional memory can yield significant improvements in knowledge-intensive tasks [3][19] Long Context Processing - Engram effectively offloads local repetitive details to memory lookup, allowing the backbone network to focus on global information integration, which is crucial for long-text processing [4][20] - In experiments, Engram-27B consumed only about 82% of the baseline pre-training computation while achieving higher accuracy in long-text retrieval tasks [4][20] System-Level Design - Engram's deterministic addressing mechanism allows for data pre-fetching from host memory, alleviating pressure on high-bandwidth memory (HBM) and controlling inference overhead to within 3% even with large memory tables [6][22] - The innovation shifts the focus from GPU memory constraints to CPU memory capacity and interconnect technologies, potentially redefining the critical constraints of AI systems [6][23] Impact on Chinese Large Models - Engram's ability to transfer memory-type parameters to scalable system memory enhances model capabilities while reducing reliance on high-end HBM, providing a clearer path for efficiency-driven technological advancement in China's large model industry [7][24] - The open-sourcing of related papers and code lowers barriers for industry validation and development, facilitating faster deployment and commercialization of large models in cost-sensitive environments [7][26]