Investment Rating - The report maintains an "Increase" rating for the computer industry [4] Core Insights - The global AI market continues to thrive, with increasing adoption rates in B-end applications and maturing business models. Companies are expected to increase their AI budgets by an average of 75% over the next year, indicating that AI is now essential for business operations rather than experimental [11][13] - Sovereign AI construction is becoming a crucial part of national technology strategies, with significant investments from countries like Saudi Arabia and the UAE in collaboration with NVIDIA to enhance digital sovereignty [28][29] - The demand for computing power is surging due to a dramatic increase in the usage of large models, with IDC projecting that the volume of model calls in China's public cloud will reach 114.2 trillion tokens by 2024 [32][35] Summary by Sections AI Adoption and Business Impact - A survey by a16z indicates that the share of innovation budgets allocated to AI has dropped from 25% to 7%, reflecting a shift towards integrating AI into core IT and business budgets [13][15] - OpenAI remains the leading model provider, with 67% of its users deploying non-cutting-edge models in production, compared to 41% for Google and 27% for Anthropic [15][16] - Software development has emerged as a key use case for AI, with reports showing that nearly 90% of code in a high-growth SaaS company is now generated by AI, up from 10-15% a year ago [21] Sovereign AI Development - The report highlights the growing importance of sovereign AI, with countries investing in AI infrastructure to ensure control over their AI systems. For instance, Saudi Arabia's HUMAIN and the UAE's Khazna data center are collaborating with NVIDIA to build AI factories [28][29][31] - Chinese tech companies are also positioning themselves in the sovereign AI landscape, with initiatives aimed at establishing a "Digital Silk Road" strategy to provide AI infrastructure solutions in Southeast Asia, the Middle East, and Africa [31] Computing Power Demand - The report notes a significant increase in the daily token usage of large models, with the Doubao model reaching over 16.4 trillion tokens, a 137-fold increase from the previous year [38] - Google has reported a 50-fold increase in token processing, with its products and APIs handling over 480 trillion tokens monthly [41] - IDC's findings indicate that 80% of enterprises plan to adopt more generative AI in the next 18 months, with a focus on improving operational efficiency and customer satisfaction [32][35] Investment Opportunities - The report suggests focusing on companies involved in computing power, such as Cambrian, Alibaba, and Huagong, as well as those in the AI agent space like Alibaba and Tencent [7][46] - The automotive sector is also highlighted, with companies like Xpeng and Li Auto being noted for their advancements in autonomous driving technology [47]

- The report introduces a timing strategy for the Hang Seng Tech Index based on the CDS indicator. The strategy uses the rolling 4-year z-score of the CDS level and the 20-day difference to assign scores and portfolio allocations. For example, when the CDS is declining and at a low level, 100% of the portfolio is allocated to the Hang Seng Tech Index. Conversely, when the CDS is rising and at a high level, no allocation is made to the index[10][11] - The timing strategy achieves an annualized excess return of 9.8% relative to the Hang Seng Tech Index, with an annualized volatility of 20.4%, a maximum drawdown of 34.3%, and a Sharpe ratio of 0.80[10][16] - The report evaluates the macroeconomic environment using a six-cycle model based on monetary, credit, and growth factors. It identifies the current phase as "credit expansion (Stage 1)," which historically favors growth-oriented assets like the Hang Seng Tech Index. During this phase, the index demonstrates higher returns compared to other broad-based indices[14][17][19]

Employment Data Analysis - In June 2025, the U.S. added 147,000 non-farm jobs, exceeding the expected 110,000[2] - The unemployment rate fell to 4.1%, better than the anticipated 4.3%[2] - The labor participation rate was 62.3%, slightly below the expected 62.4%[2] Market Reactions - Following the non-farm data release, U.S. stock markets rose, with the S&P 500, Nasdaq, and Dow Jones increasing by 0.8%, 1.0%, and 0.8% respectively[3] - The 10-year U.S. Treasury yield rose by 6.3 basis points to 4.34%[3] - The U.S. dollar index increased by 0.4% to 97.1, while spot gold prices fell by 0.9% to $3,326.1 per ounce[3] Federal Reserve Outlook - The probability of a rate cut in July dropped from 25% to 0%, and the September cut probability decreased from 100% to approximately 73%[4] - The expected number of rate cuts for the year was revised down from 2.6 to 2.1[4] ADP Employment Data Discrepancy - The ADP report showed a loss of 33,000 jobs in June, significantly below the expected gain of 95,000[4] - The divergence between ADP and non-farm data is attributed to differences in statistical coverage and tariff impacts, with non-farm data considered more reliable[4] Economic Outlook - The report suggests that the U.S. economy remains resilient, supported by factors such as balance sheet recovery and accommodative monetary policy[4] - Risks include potential economic downturns, inflation pressures, and geopolitical conflicts that could exceed expectations[4]

Investment Rating - The report maintains a "Buy" rating for the company [3][5]. Core Viewpoints - The company is positioned as a leader in energy planning and integrated services for both traditional and renewable energy, demonstrating resilient performance with steady growth in 2024 and Q1 2025 [1][13]. - The operational business segment is expected to continue growing, with significant contributions from renewable energy generation, energy storage, and hydrogen energy [2][10]. - The company has a strong market position in the construction of energy infrastructure, with a focus on optimizing its business structure and increasing the proportion of power engineering orders [19][23]. Summary by Sections Performance and Business Structure - The company achieved a total revenue of 436.7 billion yuan in 2024, a year-on-year increase of 7.6%, with traditional energy and renewable energy engineering growing by 12% and 13% respectively [13]. - The net profit attributable to shareholders for 2024 was 8.4 billion yuan, reflecting a 5% increase, outperforming the overall construction sector [13][19]. Operational Business Growth - In 2024, the investment and operational business generated revenue of 36.13 billion yuan, a 22.8% increase year-on-year, with a gross profit margin of 34.1% [2][10]. - The company has significantly increased its installed renewable energy capacity to 15.2 GW, a 60% year-on-year growth, contributing to 5.4 billion yuan in revenue from renewable energy operations [2][10]. Infrastructure and Market Position - The company is deeply involved in the construction of computing power infrastructure, leveraging its leadership in the energy sector to enhance its capabilities in this area [2][10]. - The company has been actively participating in the construction of national computing power hub centers, with plans for strategic acquisitions to strengthen its integrated investment and operation model [2][10]. Market Value Management - The company has implemented a detailed market value management plan, including a mid-term dividend policy with a payout ratio of 19.2% for 2024, indicating a commitment to enhancing shareholder value [3][10]. - The company has seen significant share buybacks and increased holdings by major shareholders, reflecting confidence in its future growth [3][10].

Employment Data Summary - In June, the U.S. added 147,000 non-farm jobs, exceeding the expected 110,000[2] - The unemployment rate fell to 4.1%, lower than the expected 4.3% and previous 4.2%[2] - Labor force participation rate was 62.3%, slightly below the expected and previous 62.4%[2] - Average hourly earnings increased by 0.2% month-on-month, below the expected 0.3% and previous 0.4%[2] Market Reactions - Following the non-farm data release, U.S. stock markets rose, with the S&P 500, Nasdaq, and Dow Jones increasing by 0.8%, 1.0%, and 0.8% respectively[2] - The 10-year U.S. Treasury yield rose by 6.3 basis points to 4.34%[2] - The U.S. dollar index increased by 0.4% to 97.1, while spot gold prices fell by 0.9% to $3326.1 per ounce[2] Fed Rate Expectations - The probability of a rate cut in July dropped from 25% to 0% after the non-farm data release[2] - The probability of a September rate cut decreased from 100% to approximately 73%[2] - The expected number of rate cuts for the year was revised down from 2.6 to 2.1[2] ADP vs Non-Farm Data - The ADP report showed a loss of 33,000 jobs in June, significantly below the expected gain of 95,000[3] - The divergence between ADP and non-farm data is attributed to differences in statistical coverage and the impact of tariffs[3] - Non-farm data is considered more reliable as it covers approximately 80% of employment positions compared to ADP's 17%[3] Economic Outlook - The strong non-farm data suggests resilience in the U.S. economy, supporting previous assessments[4] - The report indicates that if tariffs do not escalate further, a soft landing for the economy remains likely[4] - The Federal Reserve is expected to maintain a cautious stance given manageable economic downturn risks and rising inflation concerns[4]

Core Insights - The report highlights that in June, the top 100 real estate companies experienced a month-on-month sales increase, but the year-on-year decline has widened due to supply constraints and a high base from the previous year [6][7]. Real Estate Sector Summary - In the first half of 2025, the top 100 real estate companies achieved a total sales amount of 16,526.9 billion yuan, reflecting a year-on-year decrease of 10.8%, which is a 3.7 percentage point decline compared to the previous month [6]. - For June alone, the top 100 companies recorded a sales amount of 3,389.8 billion yuan, showing a year-on-year decrease of 22.8% but a month-on-month increase of 14.7% [7]. - The report maintains an "overweight" rating for the real estate sector, emphasizing the importance of policy-driven market dynamics and the potential for quality real estate companies to benefit from improved competitive conditions [8]. Investment Recommendations - The report suggests focusing on real estate stocks due to several factors, including the expectation of stronger policy support compared to previous years and the sector's role as an economic indicator [8]. - Specific investment directions include quality companies in first-tier and select second-tier cities, as well as local state-owned enterprises and property management firms [8].

Quantitative Models and Construction 1. Model Name: Odds Ratio + Win Rate Strategy - **Model Construction Idea**: This strategy combines the odds ratio and win rate metrics to allocate risk budgets across assets, aiming to optimize returns under historical data constraints [3][46] - **Model Construction Process**: - The odds ratio and win rate metrics are calculated for each asset based on historical data - The risk budgets derived from these two metrics are summed to form a composite score - Asset allocation is determined by the composite score, with higher scores receiving higher allocations - Current allocation recommendation: 11.5% equities, 2.2% gold, 86.3% bonds [3][46] - **Model Evaluation**: The model demonstrates stable performance with low drawdowns, making it suitable for risk-averse investors [3][46] 2. Model Name: Odds Ratio Enhanced Strategy - **Model Construction Idea**: Focuses on maximizing returns by overweighting high-odds assets and underweighting low-odds assets under a volatility constraint [40][41] - **Model Construction Process**: - Odds ratios are calculated for each asset - A fixed volatility constraint is applied to ensure risk control - Asset allocation is adjusted dynamically based on odds ratios - Current allocation recommendation: 15.6% equities, 2.9% gold, 81.5% bonds [40][41] - **Model Evaluation**: The strategy effectively balances risk and return, achieving consistent performance over time [40][41] 3. Model Name: Win Rate Enhanced Strategy - **Model Construction Idea**: Utilizes macroeconomic factors (e.g., monetary policy, credit, growth, inflation, and overseas conditions) to derive win rate scores for asset allocation [43][44] - **Model Construction Process**: - Win rate scores are calculated based on macroeconomic indicators - Asset allocation is determined by the win rate scores, favoring assets with higher scores - Current allocation recommendation: 6.6% equities, 1.7% gold, 91.7% bonds [43][44] - **Model Evaluation**: The strategy is robust in capturing macroeconomic trends, providing a defensive allocation approach [43][44] --- Model Backtesting Results 1. Odds Ratio + Win Rate Strategy - Annualized Return: 7.0% (2011–2025), 7.6% (2014–2025), 7.2% (2019–2025) - Maximum Drawdown: 2.8% (2011–2025), 2.7% (2014–2025), 2.8% (2019–2025) - Sharpe Ratio: 2.86 (2011–2025), 3.26 (2014–2025), 2.85 (2019–2025) [3][46][47] 2. Odds Ratio Enhanced Strategy - Annualized Return: 6.6% (2011–2025), 7.5% (2014–2025), 7.0% (2019–2025) - Maximum Drawdown: 3.0% (2011–2025), 2.4% (2014–2025), 2.4% (2019–2025) - Sharpe Ratio: 2.72 (2011–2025), 3.19 (2014–2025), 3.02 (2019–2025) [40][41][42] 3. Win Rate Enhanced Strategy - Annualized Return: 7.0% (2011–2025), 7.7% (2014–2025), 6.3% (2019–2025) - Maximum Drawdown: 2.8% (2011–2025), 2.3% (2014–2025), 2.3% (2019–2025) - Sharpe Ratio: 2.96 (2011–2025), 3.36 (2014–2025), 2.87 (2019–2025) [43][44][45] --- Quantitative Factors and Construction 1. Factor Name: Value Factor - **Factor Construction Idea**: Measures the relative attractiveness of value stocks based on odds, trends, and crowding metrics [18][20] - **Factor Construction Process**: - Odds: 0.2 standard deviations (higher indicates cheaper valuation) - Trend: -0.1 standard deviations (moderate level) - Crowding: -1.0 standard deviations (low crowding) - Composite Score: 1.0 (highest among all factors) [18][20] - **Factor Evaluation**: Strong trend and low crowding make it a top-performing factor [18][20] 2. Factor Name: Quality Factor - **Factor Construction Idea**: Focuses on high-quality stocks with favorable odds and low crowding, awaiting trend confirmation [20][21] - **Factor Construction Process**: - Odds: 1.4 standard deviations (high level) - Trend: -0.3 standard deviations (weak level) - Crowding: -0.8 standard deviations (low level) - Composite Score: 0.6 [20][21] - **Factor Evaluation**: Promising long-term potential but requires trend confirmation for stronger performance [20][21] 3. Factor Name: Growth Factor - **Factor Construction Idea**: Targets growth stocks with improving odds and moderate crowding [23][25] - **Factor Construction Process**: - Odds: 0.6 standard deviations (moderate level) - Trend: 0.02 standard deviations (neutral level) - Crowding: -0.1 standard deviations (moderate level) - Composite Score: 0.4 [23][25] - **Factor Evaluation**: Suitable for neutral allocation due to balanced metrics [23][25] 4. Factor Name: Small-Cap Factor - **Factor Construction Idea**: Captures small-cap stocks with strong trends but high crowding and low odds [26][28] - **Factor Construction Process**: - Odds: -0.5 standard deviations (low level) - Trend: 0.9 standard deviations (high level) - Crowding: 0.6 standard deviations (high level) - Composite Score: 0.0 [26][28] - **Factor Evaluation**: High uncertainty due to low odds and high crowding, requiring cautious approach [26][28] --- Factor Backtesting Results 1. Value Factor - Odds: 0.2 standard deviations - Trend: -0.1 standard deviations - Crowding: -1.0 standard deviations - Composite Score: 1.0 [18][20] 2. Quality Factor - Odds: 1.4 standard deviations - Trend: -0.3 standard deviations - Crowding: -0.8 standard deviations - Composite Score: 0.6 [20][21] 3. Growth Factor - Odds: 0.6 standard deviations - Trend: 0.02 standard deviations - Crowding: -0.1 standard deviations - Composite Score: 0.4 [23][25] 4. Small-Cap Factor - Odds: -0.5 standard deviations - Trend: 0.9 standard deviations - Crowding: 0.6 standard deviations - Composite Score: 0.0 [26][28]

Quantitative Models and Construction Methods Model 1: Core-Satellite Strategy - **Model Name**: Core-Satellite Strategy - **Model Construction Idea**: Increase the weight of benchmark constituent stocks to reduce tracking error[2] - **Model Construction Process**: - Allocate a portion of the portfolio directly to the benchmark index, while the remaining portion is actively managed[2] - Use grid search to find optimal parameters and construct the portfolio[2] - Example code provided by DeepSeek to achieve this[16] - For cases where the benchmark index has too many constituents, construct a substitute portfolio using methods like core leading stocks + industry representatives[20][23] - **Model Evaluation**: Effective in reducing tracking error by increasing the weight of benchmark constituent stocks[2][16] Model 2: Industry Neutralization - **Model Name**: Industry Neutralization - **Model Construction Idea**: Focus on stock selection to outperform the industry index by filling the weights of structurally underweighted sectors[3] - **Model Construction Process**: - Adjust individual stock weights to ensure the portfolio's industry exposure matches the benchmark index[31] - Use DeepSeek to generate the code for industry neutralization[31] - **Model Evaluation**: Significantly reduces tracking error by ensuring industry exposure consistency with the benchmark index[3][31] Model 3: Style Neutralization - **Model Name**: Style Neutralization - **Model Construction Idea**: Adjust stock weights to minimize style deviation from the benchmark index without changing the original holdings[4] - **Model Construction Process**: - Construct an optimization equation to solve for individual stock weights[4] - Use DeepSeek to generate the code for style neutralization, including multi-objective optimization[36][37] - **Model Evaluation**: Effective in reducing style deviation and improving portfolio performance relative to the benchmark[4][36] Model 4: Barbell Strategy - **Model Name**: Barbell Strategy - **Model Construction Idea**: Balance extreme growth and extreme value strategies to reduce tracking error[5] - **Model Construction Process**: - Implement a multi-strategy approach, such as equally weighting growth and value indices[5] - Use DeepSeek to generate the code for constructing and backtesting the barbell strategy[43] - **Model Evaluation**: Successfully reduces portfolio volatility and enhances performance by balancing different investment styles[5][46] Model Backtesting Results - **Core-Satellite Strategy**: - Average daily deviation reduced from 2.27% to 1.12% after adding 50% benchmark index weight[18] - Substitute portfolio tracking error relative to the benchmark index is 2.91%[28] - **Industry Neutralization**: - Maximum daily deviation reduced from 6.39% to 0.96% after industry neutralization[33] - **Style Neutralization**: - Average daily deviation reduced from 1.49% to 1.06% after style neutralization[38] - Relative to the benchmark index, the optimized portfolio's excess return improved from -9.55% to 3.55%[38] - **Barbell Strategy**: - Excess maximum drawdown reduced from over 30% to 19.88% after implementing the barbell strategy[46] - Annualized return and other performance metrics improved[50]

Investment Rating - Maintain "Buy" rating for the industry [5] Core Insights - The report highlights a rapid explosion in overseas industrial and commercial energy storage demand starting from 2025, driven by electricity pricing dynamics and the need for backup power solutions [12][15] - The demand for industrial energy storage is catalyzed by high electricity prices in Europe, rising generation costs in Southeast Asia, and frequent power outages in Africa [19][59] - The report estimates significant energy storage demand potential in Europe (approximately 66 GWh), Southeast Asia (approximately 50 GWh), and Africa (approximately 30 GWh) [3][19] Summary by Sections Supply and Demand Analysis - The underlying demand for industrial energy storage is driven by electricity pricing policies and the need for emergency backup power, with system costs decreasing further stimulating demand [15] - In Europe, high industrial electricity prices and dynamic pricing mechanisms are expected to boost energy storage demand [25][39] - In Southeast Asia, rising generation costs are pushing up local electricity prices, leading to a strong demand for cost-effective off-grid energy storage solutions [19][44] - In Africa, frequent power outages and rising electricity prices are catalyzing the shift towards off-grid energy storage solutions [59] Overseas Electricity Prices - European industrial electricity prices are significantly higher than those in the US, with small commercial users facing the highest costs [25][33] - In Southeast Asia, average electricity prices are relatively low but represent a high percentage of GDP, which is expected to drive energy storage demand [53][44] - African electricity prices are increasing rapidly, with significant implications for energy storage solutions [61] Demand Space - The report estimates that the industrial energy storage demand in Europe is around 66 GWh, while Southeast Asia and Africa have potential demands of approximately 50 GWh and 30 GWh, respectively [3][19] - The profitability of energy storage systems in Europe is supported by peak and off-peak pricing, with a payback period of around four years [3][19] - The report emphasizes the strong growth potential for energy storage systems in emerging markets, particularly in Africa and Southeast Asia [59][19] Investment Recommendations - The report recommends focusing on inverter manufacturers with strong positions in Africa, Latin America, and Europe, such as DeYue Co., Airo Energy, and Jinlang Technology [4] - For system integration, leading companies like Sungrow Power Supply and Haibo Technology are highlighted [4] - In the battery cell segment, companies with vertical integration capabilities, such as Pylon Technologies, are recommended [4]

Investment Rating - The report maintains an "Overweight" rating for the real estate industry [5][35] Core Viewpoints - The report highlights that the sales amount of the top 100 real estate companies increased month-on-month in June, but the year-on-year decline has widened due to supply constraints and high base effects from the previous year [2][14] - The report emphasizes that the real estate sector serves as an early-cycle indicator, making it a key economic barometer [5][35] - It notes that the competitive landscape is improving, with leading state-owned enterprises and a few mixed-ownership and private companies performing well in land acquisition and sales [5][35] Summary by Sections Sales Performance - In the first half of 2025, the top 100 real estate companies achieved a total sales amount of CNY 16,526.9 billion, a year-on-year decrease of 10.8% [2][14] - In June alone, the top 100 companies recorded a sales amount of CNY 3,389.8 billion, down 22.8% year-on-year but up 14.7% month-on-month [2][14] - The report indicates that different tiers of companies experienced varying degrees of sales decline, with the top 10 companies seeing the largest drop of 14.0% year-on-year [3][18] Leading Companies - The report identifies several leading companies with notable sales performance in June, including China Overseas Land & Investment with CNY 28.26 billion, followed by Poly Developments and Green Town China [4][31] - It mentions that 29 out of the top 40 companies achieved month-on-month sales growth, with China Construction East winning the highest growth rate of 503.8% [4][31] Investment Recommendations - The report suggests focusing on real estate-related stocks due to several reasons, including anticipated policy support and the sector's role as an economic indicator [5][35] - Recommended stocks include Green Town China, China Overseas Development, and Poly Developments among others [5][35]