Group 1 - The report maintains an investment rating of "Outperform the Market" for the non-bank financial sector [2][43]. - The core viewpoint emphasizes that the current market sentiment is recovering, supported by a series of policy implementations, which may lead to further increases in trading volume and index rebounds [43][44]. - The report highlights that the brokerage sector's price-to-book (PB) ratio is still at historical low levels, indicating potential for growth [44]. Group 2 - The report reviews the performance of the brokerage sector from 2003 to 2025, noting that the sector has experienced significant excess returns compared to the CSI 300 index during various market cycles [5][9]. - It identifies that the brokerage sector's beta has shown a slight recovery in 2025, currently at 1.66, indicating higher volatility compared to the market [5][6]. - The report points out that the current trading volume in the brokerage sector is positively correlated with the index performance, suggesting that increased trading activity often leads to higher index values [23][41]. Group 3 - The report indicates that the H-share brokerage stocks have shown stronger elasticity in the current market cycle, outperforming their A-share counterparts significantly [12][16]. - It notes that the average daily trading volume of the H-share brokerage sector has reached historical highs, with a peak of 15.3% of the total Hong Kong market trading volume [41]. - The report recommends specific brokerage firms such as GF Securities, East Money, CITIC Securities, Huatai Securities, and Hong Kong Exchanges for investment consideration [44].

城投债投资框架之一 城投债定价模型综述 2025 年 08 月 01 日 ➢ 城投债定价策略模型综述：城投债定价策略模型需要同时考虑城投平台自身 资质、平台所在区域基本面以及国家政策边际变化，结合短中长期的期限时间来 进行最终定价。 ➢ 城投债定价策略模型：主要包括五大部分，分别是城投平台层级、融资能力， 市场情绪及投资主体，政府信用及债务管控以及区域实力与竞争力，同时还需要 考虑一些特殊变量因素。 ➢ 不同视角看城投平台的考量因素：从短期视角（期限 0-1 年）来看，城投平 台层级、融资能力、市场情绪、投资主体是主要考量因素。从中期视角（期限 1- 3 年）来看，政府信用、区域经济基本面与债务管控机制是影响定价策略的重要 因素。而从长期视角（期限 3 年以上）来看，区域实力、省市竞争力是最大影响 因素，决定了城投债定价的底层逻辑。 ➢ 部分变量因素也会影响城投债定价策略：部分变量因素也正在时时影响城投 债的发行票面利率与二级市场估值，也能在中长期影响城投平台所在区域的产业 基本面与发展竞争力，因而需要给予一定程度重视。从短期来看，每年全国"两 会"正式公布的"赤字率"及"广义财政赤字"，当年的新增地方一般债 ...

- The report constructs a bottom-up framework for forecasting social financing (社融) by analyzing sub-items based on economic logic, high-frequency data, and seasonal characteristics[1][8][9] - The framework includes predictive models for various sub-items such as enterprise loans, resident short-term loans, government bonds, and corporate bonds, using specific economic indicators like PMI, housing sales data, and high-frequency issuance data[9] - For enterprise loans and resident short-term loans, the model employs rolling regression with PMI and Tangshan steel plant capacity utilization rate as independent variables[9] - Resident medium-to-long-term loans are forecasted based on housing mortgage data and three-stage characteristics of housing sales[9] - Enterprise bill financing is modeled using a rolling regression with a 5-year window, taking discount rates as exogenous variables[9] - Government bonds are tracked using high-frequency issuance and maturity data, with adjustments for discrepancies in reporting standards[9] - Corporate bonds are forecasted using a 5-year rolling regression to reallocate weights, effectively reducing reporting discrepancies[9] - Foreign currency loans are predicted using a 3-month average, considering correlations with RMB exchange rates and US-China bond yield spreads[9] - Trust loans and entrusted loans are forecasted by tracking issuance and maturity disclosures, with additional judgment for infrastructure-related increments[9] - Non-discounted bank acceptance bills are estimated using a 3-year average due to the cessation of high-frequency data publication[9] - Non-financial enterprise domestic stock financing is forecasted by deducting financial enterprise portions from monthly net equity financing data[9] - Loan write-offs are predicted using values from the same period last year, considering significant seasonal effects[9] - Asset-backed securities (ABS) issued by deposit-taking financial institutions are tracked using high-frequency ABS net financing data[9] - The July 2025 forecast predicts new social financing of approximately 1.53 trillion RMB, a year-on-year increase of 760 billion RMB, with a TTM month-on-month growth rate of 2.05% and a stock growth rate of 9.11%[8][9][18] - Structural predictions for July 2025 include government bonds net financing at 1.18 trillion RMB, corporate bonds net financing at 390 billion RMB, and resident medium-to-long-term loans at 5 billion RMB[9][18]

Investment Rating - The report maintains a "Recommended" rating for the company [6] Core Insights - The company has launched an employee stock ownership plan aimed at enhancing long-term growth, covering up to 258 core technical and managerial personnel, with a total of no more than 1.8 million shares, representing 0.30% of the current total share capital [1][2] - The employee stock plan is priced at 24.97 CNY per share, which is a 47.03% discount compared to the current price, indicating a significant incentive to attract and retain talent [2] - The company is advancing in smart and electric vehicle technologies, becoming the first domestic supplier to mass-produce line control braking systems and EPB, with plans to expand production capacity significantly [3] - The company has made strategic acquisitions, including a 45% stake in Wanda, enhancing its product offerings and operational efficiency, with plans to establish a new suspension technology company [4] Financial Projections - Revenue projections for 2025-2027 are estimated at 130.75 billion CNY, 170.89 billion CNY, and 220.44 billion CNY respectively, with net profits of 15.94 billion CNY, 21.04 billion CNY, and 27.93 billion CNY [4][5] - The expected EPS for the same period is 2.63 CNY, 3.47 CNY, and 4.60 CNY, with corresponding PE ratios of 18, 14, and 10 [5][4] - The company anticipates a revenue growth rate of approximately 31.6% in 2025, with a net profit growth rate of 31.9% [5][9]

Investment Rating - The report initiates coverage with a "Buy" rating for the company [5]. Core Insights - The company is positioned as a leading player in the global printer market, with its brand BenQ continuously enhancing its competitive edge and solidifying its market leadership [3]. - The semiconductor segment, particularly in automotive and industrial applications, is expanding, creating new growth opportunities for the company [3]. Summary by Sections Printer Business - The company has established itself as a global leader in the printer industry, with a sales network spanning over 110 countries and regions. It has developed a comprehensive supply chain covering key components, printers, and printing management services [10][13]. - BenQ's market share in the global printer market reached 2.9% in Q4 2024, up from 2.6% in Q3 2024, and is projected to increase to 3.1% in Q1 2025 [13]. - The company has a strong R&D team of over 2,000 people and holds more than 6,200 patents, showcasing its technological capabilities [19]. - The A3 printer segment saw a remarkable sales increase of 131.44% year-on-year in 2024, with a quarterly growth of 401.41% in Q4 [23]. - The introduction of AI features and smart printing solutions has further strengthened its position in the consumer market [24]. Semiconductor Business - The semiconductor division has been active for over 20 years, with significant investments leading to accelerated growth. The company has established itself as a key player in the semiconductor industry, particularly in automotive electronics [2][49]. - The company has successfully launched several chips for automotive applications, including the GURC01 ultrasonic sensor, which has been adopted by major automotive manufacturers [2]. - The company has developed a comprehensive product matrix in the semiconductor field, including high-performance microcontrollers and automotive-grade chips [62]. Financial Forecast and Investment Recommendations - The projected net profits for the company from 2025 to 2027 are estimated at 3.93 billion, 13.56 billion, and 19.06 billion CNY, respectively, with corresponding PE ratios of 85X, 25X, and 18X [3]. - The report highlights the potential for significant profit growth, especially after the divestiture of Lexmark, which is expected to enhance the company's focus on its core printing business [48].

1. Report Industry Investment Rating No information provided regarding the industry investment rating in the given report. 2. Core Viewpoints of the Report - With the expansion of global AI demand and the capital expenditure on computing power by North American cloud - computing giants, the domestic computing hardware supply chain is expected to continue its high - growth trend. Domestic large - models are predicted to iterate rapidly in the second half of the year, and the AIDC scale is expected to further expand. The pan - AI sector is a relatively scarce high - growth area, and investment opportunities in this sector are recommended to be focused on in Q3 [1][10]. - Currently, convertible bond valuations are at a relatively high historical level due to the continuous inflow of fixed - income funds and the recovery of stock market expectations. The idea of achieving excess returns in the high - valuation range is to bet on the elasticity of the underlying stocks of convertible bonds. When the capital situation of convertible bonds is stable and the stock market expectations do not change significantly, the valuation of the convertible bond market is unlikely to shrink actively. Buying convertible bonds corresponding to high - elasticity underlying stocks (such as those in the AI sector) at a high - risk preference position can easily generate excess returns during an upward wave [1][10]. 3. Summary by Relevant Catalog 3.1 Overall Logic and Layout Ideas - The domestic computing hardware supply chain is expected to maintain high growth due to global AI demand expansion and North American cloud - computing giants' capital expenditure on computing power. Domestic large - models will iterate quickly in the second half, and AIDC will expand. The pan - AI sector is a high - growth area, and Q3 investment opportunities are recommended [1][10]. - For convertible bond investment, with high valuations, the strategy is to invest in convertible bonds of high - elasticity underlying stocks to gain excess returns in an upward market [1][10]. 3.2 Individual Bond Selection 3.2.1 Unex Electronics/Unex Convertible Bond - Unex is a global leader in electronic design and manufacturing services, leading in the SiP module field. It has 30 manufacturing service sites across four continents, providing comprehensive services to global brand customers [16]. - In 2024, its revenue was 60.691 billion yuan, almost flat year - on - year. Cloud and storage product revenue increased by 13.35% due to AI - driven server demand [16]. - In 2025, it aims to accelerate business in AI accelerator cards. It is also developing power modules and motherboards for AI servers. It participates in providing Wi - Fi SiP modules for a North American AI glasses customer's third - generation product and has obtained an order for N - in - one motherboard modules, expected to bring significant revenue in 2026 [2][22][24]. 3.2.2 Huamao Technology/Huamao Convertible Bond - Huamao is a leader in the automotive passive safety field, with products covering airbags, seat belts, etc. In 2024, it released an action plan, strengthening its automotive parts business and entering the semiconductor and computing manufacturing fields [25]. - In 2024, its revenue was 2.213 billion yuan, up 7.67% year - on - year. Net profit was 277 million yuan, up 14.64%. It plans to expand in the semiconductor and computing manufacturing sectors by increasing investment and integrating the supply chain of Fuchuang Youyue [25][30]. - Fuchuang Youyue provides one - stop electronic manufacturing services, especially in high - speed optical module manufacturing for global computing industry chains. It has shipped to 7 of the top 20 global optical module manufacturers in 2024, with over 3.5 million 800G optical module PCBA shipments [31][32]. 3.2.3 Bowei Alloy/Bo 23 Convertible Bond - Bowei's main businesses are new materials and international new energy. Its new materials are high - performance non - ferrous alloy products, widely used in AI, 6G, etc. In 2024, its alloy strip business sales increased by 42.23% and net profit increased by 171.12% [33]. - Its high - speed connector, shielding, and lead - frame materials are crucial for computing servers and data centers. Products like boway19920 and boway70318 meet the requirements of high - computing servers [35]. 3.2.4 Sangfor Technologies/Sangfor Convertible Bond - Sangfor focuses on enterprise - level network security, cloud computing, and IT infrastructure. Its network security business uses cloud security and AI for active monitoring and protection [36]. - It has been developing cloud computing since 2012, launching multiple products. In 2024, it released the AICP platform for large - model development, aiming to lower the threshold of using AI technology [37]. 3.2.5 Minglida/Mingli Convertible Bond - Minglida's products are mainly used in photovoltaic, energy storage, new - energy vehicles, and security. In 2024, its sales declined due to the inventory reduction in the photovoltaic and energy - storage industries. However, demand is recovering in 2025 [39]. - It has made breakthroughs in the new - energy vehicle business with leading global customers. It plans to expand in the robot and liquid - cooling industries and expects increased revenue from server and automotive liquid - cooling [39][43][44].

Market Overview - The market continued to rise in July, supported by both policy and liquidity[5] - Policies aimed at "anti-involution" have led to higher elasticity in commodity prices, driving cyclical industries to lead the market[5] - Future focus will remain on liquidity support and the gradual increase in risk appetite, although the likelihood of market adjustments is rising[5] Investment Recommendations - The "Golden Stock Portfolio" has achieved a year-to-date return of 30.37%[15] - Key recommended stocks include: - China Pacific Insurance (601601.SH) - Innovent Biologics (1801.HK) - Bairun Food (002568.SZ) - Muyuan Foods (002714.SZ) - Filihua (300395.SZ) - CATL (300750.SZ) - Hubei Yihua (000422.SZ) - Luoyang Molybdenum (603993.SH) - Conch Cement (600585.SH) - North Huachuang (002371.SZ)[12] Risk Factors - Risks include macroeconomic performance falling short of expectations, policy implementation delays, and overseas expansion not meeting projections[12]

Investment Rating - The report maintains a "Recommended" rating for the company, considering its stable profitability and global technological leadership [4]. Core Insights - The company reported a revenue of 178.886 billion yuan for the first half of 2025, a year-on-year increase of 7.27%, and a net profit attributable to shareholders of 30.485 billion yuan, reflecting a growth of 33.33% [1]. - The overseas business has shown strong performance, generating 61.208 billion yuan in revenue, a 21.14% increase year-on-year, with a gross margin of 29.02%, significantly higher than the domestic business [2]. - The company has accelerated its development in the battery-swapping ecosystem, planning to build at least 500 battery swap stations by 2025, with a long-term goal of expanding to 10,000 stations [3]. Summary by Sections Financial Performance - The company achieved a total battery system production of 310 GWh in the first half of 2025, with Q2 output expected to be close to 150 GWh, showing a continuous increase [1]. - The gross margin for the power battery system was 22.41%, while the energy storage battery system gross margin was 25.52%, indicating strong profitability [1]. Overseas Business - The company's global market share for power batteries reached 38.1% from January to May 2025, maintaining a leading position [2]. - The company has secured large-scale energy storage project orders in emerging markets such as the Middle East and Australia, particularly in high-growth scenarios like AI data centers [2]. Technological Innovation - The company has launched several innovative products, including the second-generation supercharging battery and a large-capacity energy storage system solution, reinforcing its industry leadership [2]. Future Projections - Revenue projections for 2025-2027 are estimated at 406.5 billion, 497.9 billion, and 600.1 billion yuan, with year-on-year growth rates of 12.3%, 22.5%, and 20.5% respectively [3]. - The net profit attributable to shareholders is projected to be 68.2 billion, 80.4 billion, and 98.6 billion yuan for the same period, with corresponding growth rates of 34.4%, 17.9%, and 22.6% [3].

Quantitative Models and Construction Methods 1. Model Name: Deep Learning Stock Return Prediction Model - **Model Construction Idea**: This model is based on a deep learning framework tailored to the current market environment. It integrates daily and minute-frequency inputs to predict stock returns and generate trading signals based on historical rolling thresholds[1][10][22] - **Model Construction Process**: - **Input Layer**: Combines 51 technical/sentiment daily features, 7 basic daily price-volume indicators, 10 enhanced style factors, and 52 minute-frequency features aggregated to daily frequency[22] - **Training Layer**: Utilizes meta-learning to adapt to new market data dynamically, avoiding overfitting to historical data[14] - **Output Layer**: Employs LinSAT neural networks to impose constraints on the output, ensuring specific objectives like controlling style and industry exposures[18] - **Loss Function**: Multi-period mean squared error (MSE) is used to stabilize predictions for timing strategies[22] - **Formula**: Multi-period return prediction as \( y = (n, 1) \), where \( n \) represents the number of stocks[22] - **Model Evaluation**: Demonstrates robustness in adapting to market changes and controlling exposures, with significant predictive power for timing strategies[10][22] 2. Model Name: SimStock - **Model Construction Idea**: SimStock uses self-supervised learning to predict stock similarities, incorporating both static and dynamic correlations. It leverages contrastive learning to dynamically capture time-series information beyond traditional industry and style classifications[2][47][48] - **Model Construction Process**: - **Input**: Past 40-day price-volume data, Barra style factors, and capital flow indicators[52] - **Positive and Negative Sample Construction**: Positive samples are generated as \( X_{pos} = X + (1-\alpha)X_{rand} \), where \( \alpha = 0.75 \) and \( X_{rand} \) is a random feature sample[52] - **Embedding**: LSTM initializes dynamic attention weights, and CLS tokens aggregate sequence information into stock attribute vectors[52] - **Similarity Calculation**: Stock similarity is measured using cosine similarity between attribute vectors[52] - **Model Evaluation**: Effectively identifies stocks with high similarity, primarily within the same industry, but without clear patterns in market capitalization or sub-industry[56] 3. Model Name: Improved GRU Model with SimStock Integration - **Model Construction Idea**: Enhances the GRU-based stock return prediction model by initializing hidden states with SimStock-generated stock attribute vectors, improving stability across different stock types[57][59] - **Model Construction Process**: - **Initialization**: SimStock attribute vectors replace the GRU model's initial hidden state[57] - **Training**: Retains the same training setup as the baseline GRU model, with adjustments to incorporate the new initialization[59] - **Model Evaluation**: Demonstrates improved predictive performance and stability, particularly in timing strategies across diverse stocks[60][63] 4. Model Name: Index Timing Model - **Model Construction Idea**: Aggregates individual stock signals into index signals using weighted predictions based on market capitalization, followed by threshold-based signal generation[77] - **Model Construction Process**: - **Aggregation**: Combines stock return predictions into index return predictions using market-cap weights[77] - **Signal Generation**: Uses the 60th percentile of past-year predictions as the buy threshold and the 40th percentile as the sell threshold[77] - **Holding Period**: Maintains positions for at least 5 trading days to reduce turnover[77] - **Model Evaluation**: Effective in generating excess returns, particularly in high-volatility sectors[79][82][84] --- Model Backtest Results 1. Deep Learning Stock Return Prediction Model - **Cumulative Excess Return**: 77% over 5 years[33] - **Annualized Return**: 27%[33] - **Excess Return vs. Stocks**: 11.3% (pre-cost)[33] 2. SimStock - **Cumulative Excess Return**: 109% over 5 years[60] - **Annualized Return**: 30%[60] - **Excess Return vs. Stocks**: 14.8% (pre-cost)[60] - **Daily Win Rate**: 57.4%[60] - **Holding Probability**: 45.7%[60] 3. Index Timing Model - **HS300**: Annualized Return 5.1%, Excess Return 5.6%, Max Drawdown 7.7%[79] - **CSI500**: Annualized Return 12.4%, Excess Return 12.2%, Max Drawdown 7.1%[82] - **CSI1000**: Annualized Return 15.1%, Excess Return 14.9%, Max Drawdown 11.3%[84] 4. Sector Timing - **Best Sector**: Electric Power Equipment & New Energy, Annualized Return 36%, Excess Return 31.1%[101] --- Quantitative Factors and Construction Methods 1. Factor Name: Reinforced Style Factor (PPO Model) - **Factor Construction Idea**: Uses PPO reinforcement learning to predict market style preferences, generating more interpretable and robust risk factors compared to traditional deep learning[12] - **Factor Construction Process**: - **Input**: Traditional style factors and recent stock price-volume data[12] - **Reward Function**: Stability-penalized market return goodness-of-fit[12] - **Output**: Enhanced style factor representing AI market preferences[12] - **Factor Evaluation**: Provides a stable and interpretable representation of market style dynamics[12] --- Factor Backtest Results 1. Reinforced Style Factor - **RankIC**: Weekly average of 4.5% since 2019[36] - **Annualized Return**: 23.2% for long-only portfolios, Excess Return 18.3% vs. CSI800[36]

Group 1: PMI Overview - The manufacturing PMI for July 2025 is reported at 49.3%, a decrease of 0.4 percentage points from the previous month[3] - The non-manufacturing business activity index and composite PMI output index are at 50.1% and 50.2%, respectively, both down by 0.4 and 0.5 percentage points from last month[3] - The decline in July's PMI is attributed to both temporary and structural factors, including extreme weather and natural disasters[3] Group 2: Factors Influencing PMI - Extreme weather conditions, such as high temperatures and natural disasters, have negatively impacted PMI readings, contributing to a seasonal decline in manufacturing[4] - The "anti-involution" trend may cause a temporary slowdown in production, but it is expected to improve price expectations significantly[5] - The new export orders index fell to 47.1%, indicating a slowdown in export demand, reflecting the effects of previous strong export activities[6] Group 3: Price Indicators - The PMI raw material purchase price index increased by 3.1 percentage points to 51.5%, while the PMI factory price index rose by 2.1 percentage points to 48.3%[5] - The marginal improvement in price expectations does not indicate a substantial recovery in prices, as the transition from negative to positive PPI growth is still pending[5] Group 4: Non-Manufacturing Sector - The construction PMI and service PMI recorded values of 50.6% and 50.0%, with respective declines of 2.2 and 0.1 percentage points[6] - The construction sector is expected to face less downward pressure due to ongoing policy support for infrastructure projects[8]