Investment Rating - The industry rating is "Maintain Overweight" [4] Core Views - The prices of aquatic products have generally increased in the first half of the year, and the growth rate of aquatic feed is expected to remain high in the second half. The production of aquatic feed is mainly related to fish prices and weather factors. The wholesale prices of the four major fish species remain high, with year-to-date price increases for carp, silver carp, grass carp, and crucian carp being 4.78%, 6.99%, 9.37%, and 7.00% respectively. Special fish prices have also seen significant increases, with year-to-date price increases for spotted catfish, live fish, California bass, and bullfrogs being 18.67%, 6.76%, 24.24%, and 22.81% respectively [1][11][12] Summary by Sections Aquaculture - The breeding volume of various special aquatic products has increased in recent years, particularly for frogs and catfish, with five-year compound growth rates of 17.5% and 11.3% respectively. Other species such as bass and mandarin fish have growth rates of 8.7% and 7.4%, while South American white shrimp has a growth rate of 4.8%. For 2024, the breeding volumes are projected to be 595,500 tons for live fish, 539,800 tons for mandarin fish, 938,500 tons for bass, and 840,100 tons for South American white shrimp [2][12] Swine Farming - The national price for lean meat pigs is 13.22 yuan/kg, down 3.4% from last week. Current valuations are still relatively low, and attention is drawn to leading companies with cost-effectiveness and high growth potential, including Muyuan Foods, Wens Foodstuff Group, Dekang Agriculture, and others [2][13] Poultry Farming - The price of broiler chickens is 7.02 yuan/kg, down 2.1% from last week, while the average price of chicken products is 8.65 yuan/kg, down 0.6%. The price of broiler chicks is 3.25 yuan each, down 4.4% from last week. There is a focus on potential price reversals following sentiment recovery, with recommendations for companies such as Yisheng Livestock, Shengnong Development, and others [3][13] Crop and Agricultural Products - The approval of genetically modified varieties is set to enter commercial sales after the public announcement period. The subsequent confirmation of actual transaction prices, profit distribution, and new season seed expansion will gradually clarify the growth potential of industry companies [3][13] Supporting Industries - The volatility of agricultural product prices has increased, and livestock prices have not yet entered a stable prosperity cycle. Leading feed companies are expected to replace smaller companies due to their advantages in procurement, scale, and capital. Attention is drawn to companies like Haida Group and Hewei Group [3][13]

Investment Rating - The report maintains a "Buy" rating for key companies in the optical module industry, specifically recommending Zhongji Xuchuang and Xinyi Sheng [4][10]. Core Insights - The optical module industry is experiencing unprecedented growth driven by the explosive demand for AI computing power, with a shift in focus from "where is the demand" to "how to deliver on time" [1][21]. - The core challenges for optical module manufacturers include capacity expansion, yield rates, and certification processes, which are critical for fulfilling orders [3][26]. - The report emphasizes the importance of delivery capability as a decisive factor in competition, highlighting leading companies like Zhongji Xuchuang and Xinyi Sheng for their technological innovation and rapid mass production capabilities [7][26]. Summary by Sections Investment Strategy - The report suggests focusing on the delivery capabilities of optical module manufacturers as the industry transitions into a phase driven by AI computing needs [1][21]. - Key companies to watch include Zhongji Xuchuang, Xinyi Sheng, and Tianfu Communication, among others in the optical communication sector [8][14]. Market Trends - The global AI server market is projected to reach $125.1 billion in 2024, with expectations to grow to $222.7 billion by 2028, indicating a significant increase in demand for optical modules [5][25]. - The demand for 800G optical modules is expected to reach approximately 7.5 million units in 2024 and 18 million units in 2025, with a supply gap predicted to be 40% to 60% by 2027 [5][25]. Production Challenges - The production of optical modules involves complex processes such as optical coupling, packaging, and reliability testing, which contribute to long lead times from order to delivery [3][23]. - The report identifies three main challenges: capacity expansion, yield improvement, and customer certification, which are critical for meeting the growing demand [11][24]. Recommendations - The report continues to favor the computing power sector and recommends investing in leading optical module companies while also considering smaller firms in the optical device space [7][26]. - It highlights the importance of domestic supply chains in the computing power industry, particularly in areas like liquid cooling [7][26].

Investment Rating - The report maintains an "Overweight" rating for the electricity sector [2] Core Views - The new pricing policy for nearby consumption of renewable energy is expected to balance cost sharing and catalyze the development of green electricity direct connection projects. The policy will be implemented on October 1, 2025, requiring that renewable energy self-consumption accounts for at least 60% of total available generation and 30% of total consumption [5][10] - The recent surge in storage policies indicates a growing recognition of the value of regulatory power sources, with a target of adding over 100 million kilowatts of new storage capacity by 2027, leading to direct investments of approximately 250 billion yuan [5][10] - The implementation of AI in the energy sector is anticipated to create investment opportunities, with a focus on establishing a collaborative mechanism between computing power and electricity by 2030 [5][10] Summary by Sections Industry Views - The new policy for nearby consumption of renewable energy is expected to promote green electricity direct connection and address cost-sharing issues [5][10] - Storage policies are being introduced to enhance the value of regulatory power sources, with significant investment expected in new storage capacity [5][10] - The integration of AI in the energy sector is set to improve operational efficiency and create new investment avenues [5][10] Market Review - The Shanghai Composite Index closed at 3,870.60 points, up 1.52%, while the CSI 300 Index closed at 4,522.00 points, up 1.38%. The CITIC Power and Utilities Index closed at 3,077.52 points, up 0.90%, underperforming the CSI 300 Index by 0.49 percentage points [55][56] Key Company Recommendations - Recommended stocks include New天绿色能源 (H), 龙源电力, 中闽能源, and 吉电股份 in the green electricity sector, and 华能国际, 华电国际, and 宝新能源 in the thermal power sector [6][7]

Investment Rating - The report maintains an "Increase" rating for the food and beverage industry, indicating a positive outlook for the sector [5]. Core Insights - The white liquor sector is showing signs of bottoming out, with the upcoming peak season expected to improve the fundamentals. The report highlights the long-term value of leading brands such as Kweichow Moutai and Wuliangye, while also identifying short-term elastic stocks like Luzhou Laojiao and Jiangsu Yanghe [1][2]. - In the beverage segment, Budweiser plans to expand its investment in Xiamen, while Baijiu's major shareholder is transferring shares, indicating confidence in the future of the company. The report emphasizes the importance of focusing on high-growth stocks and market leaders in the beverage sector [3]. - The snack industry is experiencing rapid growth, with the number of stores for "Mingming Hen Mang" surpassing 20,000. The report notes that the milk price is stabilizing, which may lead to opportunities in the dairy sector post-harvest season [4]. Summary by Sections White Liquor - Kweichow Moutai is leading the industry with a focus on product and channel transformation, aiming for market recovery as the peak season approaches. The report anticipates a gradual improvement in the fundamentals of the white liquor sector [2]. Beer and Beverage - Budweiser's strategic upgrade includes consolidating its operations in the Asia-Pacific region. The report suggests monitoring the recovery of the restaurant sector and the performance of leading beer brands like Yanjing Beer and Zhujiang Beer [3]. Snacks and Dairy - The snack market is characterized by intense competition, with leading brands leveraging scale and supply chain advantages. The report also notes a slight recovery in raw milk prices, indicating potential growth in the dairy sector [4].

Investment Rating - The report maintains an "Overweight" rating for the C-REITs sector, indicating a positive outlook for investment opportunities in this area [6]. Core Insights - The C-REITs market is expected to benefit from a low interest rate environment in 2025, presenting favorable allocation opportunities. The report suggests three main investment strategies: focusing on policy-driven themes with resilient and undervalued projects, recognizing the market's acknowledgment of weak-cycle assets, and paying attention to original equity holders with ample asset reserves for future growth [4][6]. Summary by Sections REITs Index Performance - The CSI REITs total return index decreased by 0.78% this week, closing at 1070 points, while the CSI REITs index fell by 0.84% to 840 points. In comparison, the Shanghai Composite Index rose by 1.38% and the real estate sector (Shenwan) saw a significant increase of 5.98% [10][11]. Year-to-Date Performance - As of September 12, the CSI REITs total return index has increased by 10.56% year-to-date, while the CSI REITs index has risen by 6.39%. The Hang Seng Index has shown the highest growth at 31.55% this year [2][10]. Secondary Market Performance - The C-REITs secondary market experienced a general pullback this week, with a total market capitalization of approximately 221.46 billion yuan and an average market value of about 3 billion yuan per REIT. Among the listed REITs, 12 increased in value while 61 decreased, resulting in an average weekly decline of 0.95% [3][12]. REITs Valuation Performance - The internal rate of return (IRR) for listed REITs has shown significant differentiation, with the top three being Huaxia China Communications REIT at 9.7%, Ping An Guangzhou Guanghe REIT at 8.7%, and CICC Hubei Keti Guanggu REIT at 7.7%. The price-to-net asset value (P/NAV) ratio ranges from 0.7 to 1.8, with Huaxia China Communications REIT having a lower P/NAV of 0.7 [3][12]. Investment Recommendations - The report emphasizes the importance of focusing on resilient and high-quality undervalued projects, particularly in high-energy cities, and suggests that the logistics and factory leasing demand improvements present significant opportunities. It also notes that while weak-cycle assets have been recognized by the market, current prices reflect market expectations, necessitating a strategic approach to timing and asset selection [4][6].

Quantitative Models and Construction Methods 1. Model Name: CSI Artificial Intelligence Theme Index - **Model Construction Idea**: The index selects stocks from companies involved in providing foundational resources, technologies, and application support for artificial intelligence, aiming to reflect the overall performance of AI-themed listed companies[16][20] - **Model Construction Process**: - **Sample Space**: The sample space is consistent with the CSI All Share Index, including A-shares and depositary receipts issued by red-chip companies. Exclusions include ST/*ST securities, and securities listed on the STAR Market or Beijing Stock Exchange for less than one or two years, respectively. Other securities must have been listed for over one quarter unless their average daily market capitalization ranks in the top 30 since listing[17] - **Investability Screening**: Securities with average daily trading volume in the top 80% over the past year are selected[18] - **Selection Method**: 1. From the investable sample space, companies involved in fields such as big data, cloud computing, machine learning, intelligent chips, etc., are shortlisted 2. The top 50 securities by average daily market capitalization over the past year are selected as index components[19] - **Index Adjustment**: The index is adjusted semi-annually, with changes implemented on the next trading day after the second Friday of June and December[20] 2. Model Name: AI Industry Profitability Forecast - **Model Construction Idea**: The model forecasts the revenue and net profit growth of the CSI Artificial Intelligence Theme Index to evaluate its profitability and growth potential[27] - **Model Construction Process**: - **Revenue Forecast**: - 2025E: 9,723 billion RMB (+22.62% YoY) - 2026E: 11,198 billion RMB (+15.17% YoY) - 2027E: 12,954 billion RMB (+15.68% YoY) - **Net Profit Forecast**: - 2025E: 816 billion RMB (+89.59% YoY) - 2026E: 1,076 billion RMB (+31.86% YoY) - 2027E: 1,334 billion RMB (+23.95% YoY)[27] Model Backtesting Results 1. CSI Artificial Intelligence Theme Index - **Revenue Growth**: - 2025E: 22.62% - 2026E: 15.17% - 2027E: 15.68%[27] - **Net Profit Growth**: - 2025E: 89.59% - 2026E: 31.86% - 2027E: 23.95%[27] Quantitative Factors and Construction Methods 1. Factor Name: Market Capitalization Distribution - **Factor Construction Idea**: The factor evaluates the distribution of market capitalization among index constituents to assess concentration and risk[24] - **Factor Construction Process**: - Stocks with market capitalization >1,000 billion RMB account for 58.43% of the index weight - Stocks with market capitalization between 500–1,000 billion RMB account for 18.77% - Stocks with market capitalization between 200–500 billion RMB account for 22.37% - Stocks with market capitalization <200 billion RMB account for only 0.43%, with no stocks below 100 billion RMB[25] 2. Factor Name: Industry Exposure - **Factor Construction Idea**: The factor identifies the industry distribution of index constituents to highlight thematic focus[22][23] - **Factor Construction Process**: - **Primary Industry Exposure**: - Electronics: 35.88% - Computers: 30.96% - Communications: 24.72% - **Tertiary Industry Exposure**: - Integrated Circuits: 29.53% - Network Equipment: 24.30% - General Computing Equipment: 9.90%[23] 3. Factor Name: Conceptual Exposure - **Factor Construction Idea**: The factor evaluates the thematic relevance of index constituents based on their exposure to AI-related concepts[26] - **Factor Construction Process**: - TMT: 94.63% - Technology Leaders: 92.36% - AI+: 66.91% - AI Applications: 44.71% - AI Computing Power: 51.66%[26] Factor Backtesting Results 1. Market Capitalization Distribution - **Weight Distribution**: - >1,000 billion RMB: 58.43% - 500–1,000 billion RMB: 18.77% - 200–500 billion RMB: 22.37% - <200 billion RMB: 0.43%[25] 2. Industry Exposure - **Primary Industry Weights**: - Electronics: 35.88% - Computers: 30.96% - Communications: 24.72%[23] - **Tertiary Industry Weights**: - Integrated Circuits: 29.53% - Network Equipment: 24.30% - General Computing Equipment: 9.90%[23] 3. Conceptual Exposure - **Conceptual Weights**: - TMT: 94.63% - Technology Leaders: 92.36% - AI+: 66.91% - AI Applications: 44.71% - AI Computing Power: 51.66%[26]

Investment Rating - The report maintains an "Accumulate" rating for the space computing industry [6] Core Insights - Space computing represents a new paradigm for AI computing infrastructure, extending capabilities from terrestrial to orbital environments, leveraging unique conditions in space for enhanced efficiency and autonomy [12][17] - The demand for computing power is expected to surge, with projections indicating that global AI data center (AIDC) electricity demand could reach 347 GW by 2030, highlighting the urgent need for innovative solutions like space computing to address energy and cooling challenges [2][67] - Several companies, including Starcloud and ADA Space, are actively developing space computing projects, indicating a strong market potential and commercial viability for this technology [3][15] Summary by Sections What is Space Computing? - Space computing involves deploying data centers in low or medium Earth orbit, utilizing modular server nodes capable of AI inference and training, thus forming "orbital data centers" [17][20] - It is distinct from edge computing, as it integrates distributed AI architecture in space, enhancing real-time processing and collaboration capabilities [20] Energy Constraints and Cooling Challenges - The report emphasizes that the explosive growth in computing demand is straining existing energy supplies, particularly in the U.S., where electricity infrastructure has not kept pace with demand [13][69] - Space computing offers significant advantages in energy efficiency and cooling, as it can harness solar energy continuously and utilize the vacuum of space for effective heat dissipation [14][14] Global Deployment Progress - Numerous companies are initiating space computing projects, with notable examples including Starcloud's plans for a gigawatt-scale data center in space and ADA Space's launch of AI satellites [3][15] - The report suggests that space computing is transitioning from conceptual validation to engineering implementation, with strong commercial prospects [3][15] Investment Recommendations - The report advises investors to focus on companies that are strategically positioning themselves within the space computing supply chain, such as SOCE, RKLB in the U.S., and Shunhao Co., Hangang Co., Putian Technology, and China Satellite in China [4][15]

Inflation and Employment Trends - The U.S. August CPI increased by 2.9% year-on-year, matching expectations and marking the highest level in the past seven months[1] - Core CPI remained stable at 3.1% year-on-year, consistent with previous values[1] - The month-on-month CPI adjusted for seasonal factors rose by 0.4%, exceeding the expected 0.3%[1] Employment Data and Market Reactions - Initial jobless claims unexpectedly surged to 263,000, the highest level since June 2023, indicating significant employment risks[3] - Following the CPI release, the probability of interest rate cuts by the Federal Reserve increased to 90% for September, October, and December[3] - Major U.S. stock indices rose post-CPI announcement, with the S&P 500, Nasdaq, and Dow Jones increasing by 0.9%, 0.7%, and 1.4% respectively[3] Core Inflation Components - Food prices rose by 0.5% month-on-month, while energy prices increased by 0.7%, driven by a notable rise in gasoline prices[2] - Core goods prices increased by 0.3% month-on-month, with clothing, new cars, and used cars showing rebounds[2] - "Super core inflation," excluding food, energy, and housing, was reported at 0.12% month-on-month, consistent with previous months[2] Federal Reserve Outlook - The current economic environment suggests that employment risks outweigh inflation risks, leading to expectations of rate cuts in the near term[4] - Future rate decisions will depend on whether employment data stabilizes and the nomination of the next Federal Reserve chair[4]

Investment Rating - The report maintains a "Buy" rating for the company [6] Core Views - The acquisition of Chiplet Technology will enhance the company's CPU matrix and strengthen its position in the RISC-V sector, which is expected to drive significant growth in AI ASIC and RISC-V technology [3][4] - The company has achieved record-high new orders, totaling 1.205 billion yuan, representing an 85.88% year-on-year increase, with AI-related orders accounting for approximately 64% [4] - The company is positioned as a leading player in the AI ASIC and IP market, benefiting from the ongoing AI industry wave [4] Financial Performance - The total revenue for Chiplet Technology in 2024 is projected to be 77.947 million yuan, with a near break-even point achieved in the reporting period [2] - Revenue forecasts for the company are 3.179 billion yuan in 2025, 4.057 billion yuan in 2026, and 5.883 billion yuan in 2027, with year-on-year growth rates of 36.9%, 27.6%, and 45.0% respectively [5] - The company is expected to achieve a net profit of 0.1 million yuan in 2025, 0.6 million yuan in 2026, and 1.4 million yuan in 2027, with significant year-on-year growth rates [5] Market Position - Chiplet Technology is recognized as one of the leading RISC-V CPU IP providers in China, with over 300 global authorized clients and applications across various sectors including AI, automotive electronics, and IoT [2] - The company has established partnerships with multiple leading RISC-V firms and has provided customized chip solutions for over 20 clients [3]

Investment Rating - The report maintains a "Buy" rating for the company [6] Core Views - The company has shown impressive performance with a significant increase in revenue and net profit for the first half of 2025, achieving a revenue of 1.319 billion yuan, up 67.55% year-on-year, and a net profit of 72 million yuan, up 85.60% year-on-year [1] - The self-owned brand revenue has surged, accounting for 46% of total revenue, with a remarkable growth rate of 242.42% year-on-year [2] - The overall profitability has improved, with a gross margin of 56.92%, an increase of 11.39 percentage points year-on-year, primarily driven by the higher contribution from the self-owned brand business [3] Business Segments Summary - **Brand Management**: Revenue reached 335 million yuan, up 52.53% year-on-year, contributing 25% to total revenue. Notable performance includes a 165% increase in GMV for Bayer's products during the 618 shopping festival [2] - **Agency Operations**: Revenue was 380 million yuan, down 2.79% year-on-year, accounting for 29% of total revenue [2] - **Self-Owned Brands**: Revenue was 603 million yuan, up 242.42% year-on-year, with the brand "Zhanjia" generating 444 million yuan, a 157.11% increase year-on-year [2] Financial Projections - The company is projected to achieve revenues of 2.789 billion yuan, 3.825 billion yuan, and 4.798 billion yuan for 2025, 2026, and 2027 respectively, with year-on-year growth rates of 57.9%, 37.2%, and 25.4% [3][5] - Net profit is expected to reach 180 million yuan, 249 million yuan, and 337 million yuan for the same years, with growth rates of 70.4%, 38.5%, and 35.0% respectively [3][5]