Investment Rating - The report maintains a "Buy" rating for Zhaoli Pharmaceutical (300181.SZ) with a current price of 17.18 yuan [1]. Core Views - The acquisition of the future pharmaceutical asset group is expected to enhance Zhaoli Pharmaceutical's performance by introducing new product lines and leveraging complementary advantages [5][8]. - The market for multi-trace element injection solutions is projected to grow, with significant demand in pediatric and adult nutrition support [6][7]. - The acquisition is valued at approximately 356 million yuan, corresponding to a PE ratio of about 6 times, indicating a favorable cost-benefit ratio [7][8]. Summary by Sections Company Overview - Zhaoli Pharmaceutical has a total share capital of 701 million shares and a market capitalization of 12.05 billion yuan [1]. - The stock has fluctuated between a low of 13.39 yuan and a high of 21.07 yuan over the past year [1]. Recent Developments - The company recently won a legal case against East China Pharmaceutical, which strengthens its market position [4]. - Zhaoli Pharmaceutical signed an agreement to acquire a multi-trace element injection asset group for 35.6 million yuan, which includes both marketed and research products [4][5]. Financial Performance - The asset group is expected to generate a net profit of approximately 60 million yuan in 2025, enhancing the company's profitability [7]. - The projected revenue for Zhaoli Pharmaceutical is expected to grow from 1.94 billion yuan in 2023 to 4.29 billion yuan by 2027, with a compound annual growth rate of 21.45% [9][13]. Market Potential - The overall market for multi-trace element injections is anticipated to reach around 1.8 billion yuan in 2024, with stable growth rates for existing products [6]. - The demand for these products is expected to continue rising, particularly in pediatric and adult critical care settings [6][11]. Valuation and Earnings Forecast - The report forecasts net profits of 655 million yuan in 2025, with a corresponding PE ratio of 18, indicating a positive outlook for the company's financial health [8][9]. - The company's return on equity (ROE) is projected to increase from 14.03% in 2023 to 28.18% by 2027, reflecting improved profitability [15].

分析师：赵格格 执业证书编号：S0930521010001 0755-23946159 zhaogege@ebscn.com 分析师：刘星辰 执业证书编号：S0930522030001 021-52523880 liuxc@ebscn.com 2025 年 12 月 15 日 总量研究 "猪周期"投资的新范式 ——《光大投资时钟》第二十七篇 作者 相关研报 黄金"狂欢"未歇，铜价能否共舞？—— 《光大投资时钟》系列报告第二十六篇 （2025-10-21） 黄金周：黄金上涨的三个新变量——《光大 投资时钟》系列报告第二十五篇（2025- 10-08） 美国政府停摆：可能性与市场影响——《大 国博弈》系列第八十九篇（2025-09-25） 稳定币：从数字美元到霸权上链 ——《大国 博弈》系列第八十八篇（2025-07-25） 特朗普为何加速推进 232 调查？——《大国 博弈》第八十七篇（2025-07-09） 关税大限将至，特朗普如何抉择？——《大 国博弈》系列第八十六篇（2025-07-03） 以斗争求合作，中方打到美方筹码底线—— 《大国博弈》系列第八十五篇（2025-05- 12） 中美会晤前哨观察：特朗 ...

1. Report Industry Investment Rating No relevant content provided. 2. Core View of the Report The report systematically reviews the evolution of financial time - series forecasting models and constructs a prediction model for China's 10 - year treasury bond yield using a long - short - term memory (LSTM) neural network with historical time series as the single input variable, initially exploring the application of this deep - learning model in the fixed - income quantitative field [10]. 3. Summary by Relevant Catalog 3.1 Financial Time - Series Forecasting and Neural Network Models 3.1.1 Evolution of Financial Time - Series Forecasting Models Financial time - series forecasting has gone through three main development stages: traditional econometric models, traditional machine - learning models, and deep - learning models. Traditional econometric models have clear forms and strong interpretability but struggle to depict nonlinear and complex dynamic relationships. Traditional machine - learning models can perform nonlinear fitting and automatic feature screening but need manual feature extraction. Deep - learning models can automatically extract features from raw data and capture complex long - term time - series patterns, adapting well to the complex characteristics of financial time series [11][12]. 3.1.2 Neural Network Models and LSTM Models Neural network models are machine - learning models imitating the connection structure of human brain neurons. Recurrent neural networks (RNN) and their variants, such as LSTM, are designed for processing sequence data. LSTM solves the long - term dependence problem of traditional RNN through a "gating mechanism" and memory units, enhancing robustness to irregular data and being suitable for bond yield prediction [13][18]. 3.2 Treasury Bond Yield Prediction Based on Stacked LSTM Model 3.2.1 Stacked LSTM Model Stacked LSTM connects multiple LSTM layers in sequence, having advantages in long - sequence processing and multi - dimensional feature extraction, more suitable for complex time - series forecasting in financial scenarios [23]. 3.2.2 Construction of Treasury Bond Yield Prediction Model The report uses a classic and robust architecture of three - layer stacked LSTM + Dropout regularization to build a neural network model for predicting the 10 - year treasury bond yield. It only uses the historical time series of the 10 - year treasury bond yield as a single variable for prediction. The data is from the beginning of 2021 to December 12, 2025. After data processing and sample construction, a medium - complexity LSTM neural network model with about 130,000 adjustable parameters is built. The optimal model is obtained at the 27th training iteration, with an average absolute error of 1.43BP for the test set. The predicted yield on December 19, 2025, is 1.8330%, slightly lower than 1.8396% on December 12, 2025 [2][24][30]. 3.3 Follow - up Optimization Directions - Optimize model design: Adjust and optimize the design related to time windows, data processing, network architecture, and training strategies [3][36]. - Input multi - dimensional variables: Expand input variables from a single yield sequence to multi - dimensional variables such as macro, market, and sentiment to make the model more in line with economic logic and capture more comprehensive information [3][36]. - Build hybrid models: Combine the LSTM model with traditional econometric models or other machine - learning models to build hybrid models like ARIMAX - LSTM and CNN - LSTM - ATT, enhancing prediction accuracy [3][36]. - Introduce a rolling back - testing mechanism: Use a rolling time - window back - testing mechanism to update the model dynamically and make continuous predictions, improving the model's adaptability to market changes [3][36].

Report Industry Investment Rating No relevant information provided. Core Viewpoints of the Report - The main economic indicators in November 2025 further declined, with the year - on - year growth rate of industrial added value, the cumulative year - on - year growth rate of fixed - asset investment, and the year - on - year growth rate of total retail sales of consumer goods all showing a downward trend. However, the month - on - month growth rate of industrial added value increased, and the month - on - month decline of fixed - asset investment narrowed. [1][2] - In the bond market, investors should gradually become more optimistic about the bond market. The expected fluctuation center of the 10Y Treasury bond yield is 1.75%. In the long term, convertible bonds are still relatively high - quality assets, but attention should be paid to the structure. [3] Summary by Relevant Catalogs Event - On December 15, 2025, the National Bureau of Statistics released the economic data for November 2025, including the year - on - year growth rate of industrial added value above a designated size of 4.8%, the cumulative year - on - year decline of fixed - asset investment from January to November of 2.6%, and the year - on - year growth rate of total retail sales of consumer goods in November of 1.3%. [1][6][9] Comment Scale - above industrial production: year - on - year growth rate decreased but month - on - month growth rate increased - In November 2025, the year - on - year growth rate of industrial added value above a designated size was 4.8%, a 0.1 - percentage - point decrease from October. The month - on - month growth rate was + 0.44%, an increase from October. [2][6] - Among the three major categories, the year - on - year growth rate of the mining industry increased, while those of the manufacturing industry and the production and supply of electricity, heat, gas, and water decreased. [2][6] January - November fixed - asset investment: cumulative year - on - year decline widened, but the month - on - month decline in November narrowed - From January to November 2025, the cumulative year - on - year growth rate of fixed - asset investment was - 2.6%, with the decline widening. The month - on - month growth rate in November was - 1.03%, with the decline narrowing. [2][13] - The cumulative year - on - year growth rates of real estate, manufacturing, and general infrastructure investment from January to November all decreased, and the year - on - year growth rates of the three sub - items in November were all weak. [17] Total retail sales of consumer goods: year - on - year growth rate continued to decline, and the month - on - month growth rate was weaker than the seasonal average - In November 2025, the year - on - year growth rate of total retail sales of consumer goods was 1.3%, a decrease from the previous month. The month - on - month growth rate was - 0.42%, weaker than the seasonal average and lower than the same - period levels in 2023 and 2024. [2][18] - The year - on - year growth rates of different types of consumer goods all decreased in November compared with the previous month. [2][18] Bond Market Views Interest - rate bonds - Since August 2025, the yield of Treasury bonds has shown obvious differentiation. The short - end yield has fluctuated little and declined steadily, while the long - end yield, especially the 30 - year yield, has been on an upward trend, and the Treasury bond yield curve has steepened significantly. [3][22] - With the current loose capital situation and the weak fundamental trend, investors should gradually become more optimistic about the bond market, and the expected fluctuation center of the 10Y Treasury bond yield is 1.75%. [3][22] Convertible bonds - Since the beginning of 2025 (as of December 12), the change rate of the CSI Convertible Bond Index was + 16.5%, and the change rate of the CSI All - Share Index was + 21.8%. The performance of the convertible bond market was weaker than that of the equity market. [3][31] - Against the background of the slow - bull expectation of the equity market and the pattern where the demand in the convertible bond market is stronger than the supply and difficult to change, convertible bonds are still relatively high - quality assets in the long term, and more attention should be paid to the structure. [3][31]

1. Report Industry Investment Rating No relevant information provided. 2. Core View of the Report The report systematically reviews the evolution of financial time - series prediction models and constructs a prediction model for China's 10 - year Treasury bond yield using the Long Short - Term Memory (LSTM) neural network, with historical time - series as the single input variable, to explore the application of this deep - learning model in the fixed - income quantitative field [10]. 3. Summary by Directory 3.1 Financial Time - Series Prediction and Neural Network Models - **Evolution of Financial Time - Series Prediction Models**: Financial time - series prediction has gone through three main stages: traditional econometric models (e.g., ARIMA, GARCH), traditional machine - learning models (e.g., SVM, RF), and deep - learning models. Traditional econometric models have clear forms but struggle with nonlinear and complex dynamic relationships. Traditional machine - learning models can perform nonlinear fitting but need manual feature extraction. Deep - learning models can automatically extract features and capture long - term patterns, with RNN and its variants like LSTM being mainstream methods [11][12]. - **Neural Network Models and LSTM Models**: Neural network models mimic the human brain's neuron connection structure. RNN is designed for sequence data but has issues with long - term memory. LSTM solves the long - term dependency problem of RNN through a "gating mechanism" and memory units, enhancing robustness to irregular data. It is suitable for bond yield prediction due to its ability to handle long - term time series and filter noise [13][17][18]. 3.2 Treasury Bond Yield Prediction Based on Stacked LSTM Model - **Stacked LSTM Model**: Stacked LSTM connects multiple LSTM layers sequentially, offering advantages in long - sequence processing and multi - dimensional feature extraction, making it more suitable for complex financial time - series prediction [23]. - **Construction of Treasury Bond Yield Prediction Model**: - **Data Processing and Sample Construction**: The data is the yield of the 10 - year Treasury bond from the beginning of 2021 to December 12, 2025. First - order differences are calculated and standardized. Samples are constructed with the first - order differences of the past 60 trading days as input features and the first - order differences of the next week as the prediction target. The samples are divided into a training set (72%), a validation set (8%), and a test set (20%) [27]. - **Model Design and Evaluation**: The model architecture consists of LSTM, Dropout, and Dense layers. The training strategy involves 200 iterations with an early - stopping mechanism. Evaluation metrics include MSE, MAE, and RMSE [28][29]. - **Model Results**: A medium - complexity LSTM neural network model with about 130,000 adjustable parameters is built. The optimal model is obtained at the 27th iteration, and the early - stopping mechanism is triggered at the 77th iteration. The average absolute error for the test set is 1.43BP. The 10 - year Treasury bond yield is predicted to decline from December 15 - 19, 2025, with the predicted value on December 19, 2025, being 1.8330%, slightly lower than 1.8396% on December 12, 2025 [30]. 3.3 Follow - up Optimization Directions - **Model Design Optimization**: Adjust and optimize relevant designs such as time windows, data processing, network architecture, and training strategies [3][36]. - **Input Multi - Dimensional Variables**: Expand input variables from a single yield sequence to multi - dimensional variables such as macroeconomic, market, and sentiment variables to make the model more in line with economic logic and capture more comprehensive information [3][36]. - **Construct Hybrid Models**: Combine the LSTM model with traditional econometric models or other machine - learning models to build hybrid models like ARIMAX - LSTM and CNN - LSTM - ATT, leveraging different model advantages and improving prediction accuracy [3][36]. - **Introduce Rolling Back - testing Mechanism**: Use a rolling time - window back - testing mechanism to update the model dynamically and make continuous predictions, enhancing the model's adaptability to market changes [3][36].

一周观点 总量研究 本周观点 | 总量研究 2 | | --- | | 本周观点 2 | | 重点报告 2 | | 行业研究 4 | | 本周观点 4 | | 重点报告 5 | | 公司研究 6 | | 重点报告 6 | | 重点报告摘要 7 | | 总量研究 7 | | 行业研究 10 | | 公司研究 12 | | 领域 | 一周观点 | 分析师 | | --- | --- | --- | | | 新一轮政策部署护航，A 股跨年行情可期。一方面，未来国内经济政策有望持续发力，经济增 | | | | 长有望保持在合理区间，进一步夯实资本市场繁荣发展的基础；另一方面，政策红利释放，有 | | | 策略 | 望提振市场信心，进一步吸引各类资金积极流入；此外，历史来看，"十三五"和"十四五" | 张宇生 | | | 开局之年 A 股市场均有不错的表现，历史上开局之年的积极表现有望在 2026 年得到延续。 | | | | 本周国内权益市场指数普遍上涨，创业板指上涨 1.86%,周期主题基金表现占优，消费、医药 | | | 金工 | 主题基金净值回调。国内市场新成立基金 39 只，合计发行份额为 365.89 亿份 ...

Investment Rating - The report maintains an "Overweight" rating for the steel and non-ferrous metals sectors [5] Core Insights - The blast furnace capacity utilization rate is expected to remain below last year's levels, indicating potential challenges in production efficiency [1] - The liquidity indicators show a decline in the growth rate difference between M1 and M2, which may impact market dynamics [10] - The construction and real estate sectors are experiencing a downturn, with significant declines in new construction and sales areas [22][79] Summary by Relevant Sections Liquidity - The M1 and M2 growth rate difference was -3.1 percentage points in November 2025, down by 1.10 percentage points month-on-month [10][18] - The BCI small and medium enterprise financing environment index was 52.50 in November 2025, reflecting a slight increase of 0.17% from the previous month [10][18] Construction and Real Estate Chain - The national blast furnace capacity utilization rate was 86% this week, down by 1.16 percentage points [9][42] - The cumulative year-on-year decline in new construction area for the first ten months of 2025 was -19.80% [22] - The cumulative year-on-year decline in commodity housing sales area for the same period was -6.80% [22] Industrial Products Chain - The operating rate of semi-steel tires was 71.57%, up by 0.65 percentage points [2] - The price of electrolytic aluminum was 22,070 yuan/ton, down by 0.36% [2] - The price of tungsten concentrate reached 374,000 yuan/ton, reflecting a week-on-week increase of 5.65% [2] Price Relationships - The price ratio of London spot gold to silver reached a new low since July 2021 [3] - The price of rebar was 3,250 yuan/ton, down by 0.61% [9][42] - The price of iron ore was 785 yuan/ton, down by 0.6% [9] Export Chain - The CCFI comprehensive index for container shipping rates was 1,118.07 points, up by 0.29% [3] - The U.S. crude steel capacity utilization rate was 75.70%, down by 0.10 percentage points [3] - New export orders in China's PMI for November 2025 were 47.60%, up by 1.7 percentage points [3] Valuation Metrics - The PB ratio for the steel sector relative to the Shanghai and Shenzhen markets is currently at 0.51, with historical highs reaching 0.82 [4] - The overall steel industry gross profit was 152 yuan/ton, reflecting a week-on-week increase of 3.6% [9]

2025 年 12 月 15 日 总量研究 交易信心有所提振，后市仍将震荡上行 ——金融工程市场跟踪周报 20251215 要点 本周市场核心观点与市场复盘： 本周（2025.12.08-2025.12.12，下同）A 股震荡上行，市场量能有所提振。截 至周五（2025.12.12，下同），主要宽基指数量能择时指标均转为看多信号。沪 深 300、中证 500 指数时序波动率、横截面波动率均有提升，Alpha 环境好转。 资金面方面，融资增加额环比提升，资金交易情绪积极。 12 月中央经济工作会议进一步提振市场信心，资金面叠加情绪指标改善下，市 场或进一步震荡上行。风格方面，本周市场大市值风格显著、基本面因子表现相 对占优，市场或逐步实现资金面驱动向基本面驱动的过渡。中长线持续看好"红 利+科技"配置主线。 本周上证综指下跌 0.34%，上证 50 下跌 0.25%，沪深 300 下跌 0.08%，中证 500 上涨 1.01%，中证 1000 上涨 0.39%，创业板指上涨 2.74%，北证 50 指数 上涨 2.79%。 截至 2025 年 12 月 12 日，宽基指数来看，上证指数和上证 50 指数处于估 ...

Group 1: Macro and Market Overview - The financial data for November shows a recovery due to increased fiscal efforts, with social financing growth supported by accelerated government bond issuance and faster conversion of fiscal spending into general deposits [2] - A favorable liquidity environment is highlighted, with significant growth in corporate bond financing contributing positively to social financing [2] - The A-share market is expected to perform well in the upcoming year-end, supported by ongoing domestic economic policy efforts and historical trends indicating strong performance in the first year of the 13th and 14th Five-Year Plans [3] Group 2: Bond Market Insights - The secondary market for REITs has seen a decline in prices, with the weighted REITs index closing at 180.06 and a weekly return of -0.23% [5] - Credit bond issuance increased significantly, with 369 bonds issued totaling 459.51 billion yuan, a 35.34% increase from the previous period [6] - Investors are advised to adopt a comprehensive view when analyzing financial aggregate data, focusing on a balanced understanding of the market [4] Group 3: Industry-Specific Research - In the float glass industry, the trend of increasing concentration among leading companies is expected to continue, with recommendations to focus on Xinyi Glass and Qingdao Huadong Glass [10] - The photovoltaic glass sector is anticipated to see a clearing out of smaller companies at the industry cycle's bottom, leading to increased concentration among leading firms, with a focus on Xinyi Solar and Flat Glass Group [10] - The banking sector is experiencing a slowdown in credit expansion, with social financing in November at 2.5 trillion yuan, maintaining an 8.5% growth rate [11] Group 4: Company-Specific Analysis - Zhongyou Engineering has successfully launched a new material project, with projected net profits of 738 million yuan, 825 million yuan, and 929 million yuan for 2025-2027, respectively [20] - The company is rated as "buy" due to its strategic expansion into emerging business areas [20] - Hualan Biological is increasing its investment in innovative products and has a high dividend payout ratio, enhancing its long-term investment value [22]

Investment Ratings - The report maintains a "Buy" rating for both the power equipment and environmental protection sectors [1]. Core Views - The Central Economic Work Conference emphasizes the application of green electricity and promotes the development of hydrogen, ammonia, methanol, and energy storage, indicating a positive outlook for investment opportunities in green energy sectors in 2026 [3]. - Domestic energy storage saw significant growth in November, with newly installed capacity reaching 4.51GW/13.03GWh, reflecting a month-on-month increase of 57.14% in power and 74.66% in capacity [3][7]. - The report highlights the importance of hydrogen and green fuels as new growth points, with expectations for increased investment in these areas due to supportive policies and market conditions [4]. Summary by Sections Energy Storage - Domestic energy storage is experiencing a boom, with November's new installations showing a 45.95% year-on-year increase in power and a 49.6% increase in capacity [3][7]. - The report anticipates that independent energy storage tenders will maintain a good level in 2026, supported by a complete revenue model through energy markets and auxiliary services [3]. Hydrogen and Green Fuels - The report suggests that hydrogen and methanol will play a crucial role in the non-electric applications of green electricity, with significant investment expected in these areas [4]. - The development of zero-carbon parks and factories is also highlighted as a key initiative for 2026 [3]. Wind Power - The report notes that in 2024, onshore wind power installations are expected to reach 75.8GW, a year-on-year increase of 9.68%, while offshore wind installations are projected to be 4.0GW, a decrease of 40.85% [8]. - The bidding capacity for wind power equipment in 2024 is expected to be 164.1GW, a 90% increase year-on-year [13]. Lithium Battery - The report indicates that the demand for lithium batteries remains strong, with December's retail sales of new energy vehicles expected to show a bright performance despite a year-on-year decline of 17% [19]. - The supply chain for lithium batteries is expected to stabilize, with ongoing negotiations for long-term contracts and price adjustments [22][23].