Core Viewpoint - The A-share market has seen a strong performance in brokerage stocks, with several companies experiencing significant price increases, indicating a positive trend in the sector [1]. Group 1: Brokerage Stock Performance - Huayin Securities (华林证券) reached a limit-up increase of 9.99%, with a total market capitalization of 48.2 billion and a year-to-date increase of 16.37% [2]. - Hua'an Securities (华安证券) increased by 9.70%, with a market cap of 35.5 billion and a year-to-date increase of 11.80% [2]. - Zhina Compass (指南针) rose by 8.02%, with a market cap of 89.2 billion and a year-to-date increase of 12.03% [2]. - Tonghuashun (同花顺) saw a rise of 7.79%, with a market cap of 192.5 billion and a year-to-date increase of 11.11% [2]. - Northeast Securities (东北证券) increased by 4.68%, with a market cap of 23.1 billion and a year-to-date increase of 5.69% [2]. - Dongfang Caifu (东方财富) rose by 4.00%, with a market cap of 390.4 billion and a year-to-date increase of 6.51% [2]. - Other notable increases include Guotai Haitong (国泰海通) at 3.87%, Huatai Securities (华泰证券) at 3.90%, and Tianfeng Securities (天风证券) at 3.80% [2]. Group 2: Market Indicators - The MACD golden cross signal has formed, indicating a positive momentum for these stocks [2].

（文章来源：每日经济新闻） 每经AI快讯，1月6日，券商股早盘走强，华林证券拉升封板，华安证券、国泰海通、华泰证券、东北 证券等拉升涨超3%。 ...

证券板块持续拉升，华林证券涨停，华安证券涨超6%，国泰海通、华泰证券、东北证券跟涨。 ...

利欧股份今日涨停，全天换手率31.32%，成交额110.90亿元，振幅6.71%。龙虎榜数据显示，机构净买 入8466.16万元，深股通净买入1.35亿元，营业部席位合计净卖出4.17亿元。 深交所公开信息显示，当日该股因日涨幅偏离值达8.00%上榜，机构专用席位净买入8466.16万元，深股 通净买入1.35亿元。 卖五 中信证券股份有限公司东阳吴宁西路证券营业部 2315.22 11167.32 （文章来源：证券时报网） 资金流向方面，今日该股主力资金净流出1534.80万元，其中，特大单净流入2.14亿元，大单资金净流出 2.30亿元。近5日主力资金净流入23.33亿元。 融资融券数据显示，该股最新（12月31日）两融余额为17.49亿元，其中，融资余额为17.43亿元，融券 余额为664.96万元。近5日融资余额合计增加1.75亿元，增幅为11.17%，融券余额合计增加103.23万元， 增幅18.38%。（数据宝） 利欧股份1月5日交易公开信息 | 买/卖 | 会员营业部名称 | 买入金额（万元） | 卖出金额（万元） | | --- | --- | --- | --- | | 买一 | 深股通专用 ...

Group 1 - The core viewpoint of the article emphasizes the performance of the "brokerage golden stocks" and their ability to track the performance of mixed equity funds, showcasing the analytical capabilities of brokerage firms [2][7][28] - In December 2025, the top-performing stocks in the brokerage golden stock pool included XW Communication, Maiwei Co., and Yaxiang Integration, with significant monthly increases [1][3][4] - The top three brokerages by monthly returns were Huachuang Securities, Guojin Securities, and Changcheng Securities, with returns of 17.26%, 12.74%, and 11.36% respectively, compared to 3.06% for the mixed equity fund index and 2.28% for the CSI 300 index [6][10] Group 2 - The brokerage golden stock pool showed a high allocation in the electronics (14.04%), non-ferrous metals (9.93%), and basic chemicals (8.96%) sectors, with notable increases in non-ferrous metals (+3.13%) and defense industry (+1.93%) [25][18] - The performance of the brokerage golden stock performance enhancement portfolio yielded an absolute return of 5.24% for December 2025 and 40.66% for the year, outperforming the mixed equity fund index by 2.18% and 7.47% respectively [33][27] - The article highlights the importance of analyst recommendations, noting that stocks with fewer prior recommendations tend to gain more market attention once included in the golden stock pool [22][20]

Quantitative Models and Construction Methods Model Name: mHC-MLP - **Model Construction Idea**: The mHC-MLP model introduces manifold-constrained hyper-connections (mHC) into the traditional MLP framework to address issues such as low signal-to-noise ratio, non-stationarity, and extreme tail behavior in financial data. It achieves this by incorporating multi-stream residual channels, gated fan-in/fan-out mappings, and doubly stochastic manifold projections (via Sinkhorn-Knopp) to enhance numerical stability and extrapolation resistance[1][16][22]. - **Model Construction Process**: 1. **Multi-Stream Residual Channels**: The model expands the single residual channel in traditional ResNet to multiple parallel sub-streams, allowing independent feature representations and dynamic routing between streams[19][20]. 2. **Manifold Constraints**: - Residual mixing matrices are constrained to the Birkhoff polytope (doubly stochastic matrices), ensuring non-negativity, row sums of 1, and column sums of 1. This is achieved using the Sinkhorn-Knopp algorithm during training[22][23][54]. - Fan-in and fan-out mappings are constrained to non-negative values using sigmoid functions, ensuring that output features remain within the convex hull of input features[24]. 3. **Dynamic Routing Mechanism**: The model uses a combination of linear mixing (via residual matrices) and non-linear transformations (via MLP blocks) to balance feature interaction and noise suppression[49][50][51]. 4. **Deep Stacking**: The mHC-MLP extends the network depth to six layers, leveraging the numerical stability provided by manifold constraints to capture higher-order interactions[56][57]. 5. **Initialization and Regularization**: Parameters are initialized with minimal values (e.g., alpha = 0.01) to ensure stable gradient flow during early training stages. Regularization is achieved through manifold constraints rather than traditional dropout or L2 regularization[25][55]. - **Model Evaluation**: The mHC-MLP model demonstrates improved numerical stability, reduced overfitting, and enhanced robustness against noise. However, it may underperform in short-term, high-volatility scenarios due to its conservative nature[2][75][86]. --- Model Backtesting Results mHC-MLP Model - **Cumulative Return**: 49% (compared to 56% for the unconstrained MLP model)[75] - **t-Statistic**: Not explicitly mentioned for mHC-MLP - **IC_IR**: Not explicitly mentioned for mHC-MLP - **Turnover**: Lower than the unconstrained MLP model, indicating better stability[2][75] - **Maximum Drawdown**: Lower than the unconstrained MLP model, reflecting reduced risk exposure[2][75] --- Quantitative Factors and Construction Methods Factor Name: Barra MLP Factor - **Factor Construction Idea**: The Barra MLP factor leverages neural networks to capture non-linear interactions and complex relationships between Barra style factors and residual stock returns, overcoming the limitations of traditional linear factor models[30][31]. - **Factor Construction Process**: 1. **Baseline Risk Model**: A long-term risk model is constructed using the Barra CNE6 framework, incorporating one country factor, 31 industry factors, and 15 style factors (e.g., size, beta, momentum, value)[36][37][38]. 2. **Residual Return Extraction**: Stock returns are decomposed into common factor contributions and residual returns via cross-sectional regression. The residual returns serve as the prediction target for the MLP model[40]. 3. **Rolling Training**: The MLP model is trained using rolling windows of 24, 36, and 72 months to balance bias and variance. Features include the 15 style factors, and the target is the next-period residual return[41]. 4. **Multi-Period Signal Synthesis**: Predictions from the three training windows are standardized (Z-score) and combined using equal weighting or IC-based weighting to generate a composite factor[42][43]. 5. **Orthogonalization**: The composite factor is regressed against the 15 style factors to remove linear correlations, ensuring it provides incremental information[44]. 6. **Pure Factor Return Calculation**: The orthogonalized factor is incorporated into an enhanced Barra risk model, and its pure factor return is estimated via cross-sectional regression[45]. - **Factor Evaluation**: The Barra MLP factor effectively captures non-linear alpha signals and demonstrates significant cumulative returns and IC_IR values, validating its utility in quantitative strategies[46]. --- Factor Backtesting Results Barra MLP Factor - **Cumulative Return**: Over 15%[46] - **t-Statistic**: 2.8[46] - **IC_IR**: 0.45[46] - **Turnover**: Not explicitly mentioned - **Maximum Drawdown**: Not explicitly mentioned --- Composite Model: mHC-Enhanced Barra MLP Factor - **Model Construction Idea**: The mHC-enhanced Barra MLP factor integrates the mHC architecture into the Barra MLP framework to improve robustness and stability while retaining the ability to capture non-linear interactions[48]. - **Model Construction Process**: The MLP core in the Barra MLP factor is replaced with the mHC-MLP architecture, maintaining the same input features, target variables, and training framework. This modification introduces manifold constraints and dynamic routing to enhance numerical stability and reduce overfitting[48][49][50]. - **Model Evaluation**: While the mHC-enhanced factor demonstrates superior stability and robustness, it may lag in short-term, high-volatility markets due to its conservative design[75][86]. --- Composite Model Backtesting Results mHC-Enhanced Barra MLP Factor - **Cumulative Return**: Not explicitly mentioned - **t-Statistic**: Not explicitly mentioned - **IC_IR**: Not explicitly mentioned - **Turnover**: Lower than the original Barra MLP factor[2][75] - **Maximum Drawdown**: Lower than the original Barra MLP factor[2][75]

Core Viewpoint - Northeast Securities has initiated coverage on Sanhua Intelligent Control (02050), projecting revenue growth from 2025 to 2027, with corresponding net profits and PE ratios, and has given a "Buy" rating [1] Group 1: Financial Projections - The company is expected to achieve revenue of 322.6 billion, 371.3 billion, and 441.2 billion yuan for the years 2025, 2026, and 2027 respectively, with net profits of 42.4 billion, 51.2 billion, and 60.3 billion yuan [1] - For 2025, the company forecasts a net profit between 38.7 billion and 46.5 billion yuan, representing a year-on-year increase of 25% to 50%, and a non-recurring net profit between 36.8 billion and 46.1 billion yuan, reflecting an 18% to 48% increase [2] Group 2: Business Performance - The traditional refrigeration business has shown strong performance with a revenue of 103.9 billion yuan in the first half of 2025, marking a year-on-year growth of 25.5% and a gross margin of 28.2%, up by 0.65 percentage points [3] - The automotive parts business has seen a revenue of 58.7 billion yuan in the first half of 2025, with a modest year-on-year growth of 8.8%, influenced by Tesla's production and sales improvements [4] Group 3: Strategic Developments - The company has established a dedicated division for humanoid robots and set up a production base in Thailand to support Tesla's mass production needs, with expectations of significant revenue contributions starting in 2026 [5]

Group 1 - The core viewpoint of Northeast Securities is that Sanhua Intelligent Controls (002050) is expected to see significant revenue growth from 2025 to 2027, with projected revenues of 32.26 billion, 37.13 billion, and 44.12 billion yuan, and net profits of 4.24 billion, 5.12 billion, and 6.03 billion yuan respectively, leading to a PE ratio of 35, 29, and 25 times [1] - The company has released a performance forecast for 2025, expecting a net profit of 3.87 to 4.65 billion yuan, representing a year-on-year increase of 25% to 50%, and a non-net profit of 3.68 to 4.61 billion yuan, with a year-on-year increase of 18% to 48% [1] Group 2 - The traditional refrigeration business has shown strong performance, with a revenue of 10.39 billion yuan in the first half of 2025, reflecting a year-on-year growth of 25.5%, and a gross margin of 28.2%, up by 0.65 percentage points [2] - The automotive parts business is recovering, with a revenue of 5.87 billion yuan in the first half of 2025, showing a year-on-year growth of 8.8%, driven by improved production and sales from major client Tesla, which delivered 497,000 vehicles globally in Q3, a year-on-year increase of 7.4% [3] - The company is reducing its reliance on Tesla, as it begins to see contributions from other major clients such as General Motors and domestic new energy vehicle manufacturers like Xiaomi, Li Auto, and XPeng, while also making progress with European clients like Mercedes-Benz and Volkswagen [3] Group 3 - The company has established a clear strategic position in humanoid robotics, being a core supplier of electromechanical actuators for Tesla's humanoid robots, and has set up a dedicated humanoid robotics division along with a production base in Thailand to support customer demand [4] - Significant revenue contributions from the humanoid robotics segment are expected to begin in 2026 [4]

Core Insights - The A-share market in 2025 closed positively, with major indices showing significant gains, and brokerage "golden stock" portfolios delivering impressive results [1] - The top ten brokerage golden stock portfolios achieved returns exceeding 50%, with the highest being Guoyuan Securities at 84.08% [1][5] Brokerage Performance - Guoyuan Securities led with a return of 84.08%, with its best stock being Zhongtung Gaoxin [2][5] - Northeast Securities ranked second with a return of 68.38%, highlighting its best stock as Wukuang Resources [2][6] - Other notable performers include Kaiyuan Securities (67.1%), Huaxin Securities (62.23%), and Dongxing Securities (61.98%) [2] Stock Selection Strategies - Guoyuan Securities focuses on selecting stocks based on China's new economic growth drivers and emerging industries, aiming to identify companies with strong fundamentals and market expectations [5] - Northeast Securities employs a collaborative approach with nearly 30 teams contributing to stock recommendations, emphasizing technology and cyclical sectors [6][7] - Dongguan Securities utilizes a systematic research framework that integrates macroeconomic analysis with industry insights to select high-quality stocks [8] Research Methodologies - Huazhong Securities emphasizes a dual approach, combining top-down macroeconomic analysis with bottom-up stock selection to identify high-potential sectors and individual stocks [9] - The overall performance of brokerage golden stock portfolios in 2025 significantly outperformed previous years, with the lowest return among the top 20 brokerages nearing 40% [5]

Core Insights - The wealth management industry is undergoing a significant transformation driven by stringent regulations, which are reshaping the industry order and increasing the cost of violations [2][16]. Regulatory Trends - The regulatory landscape is characterized by three major trends: penetrating accountability, multi-faceted penalties, and full-cycle supervision [16]. - There is a clear signal of "zero tolerance" towards violations, with a notable increase in the number of penalties issued [2][16]. Violations and Penalties - As of December 26, 2025, at least 166 penalties have been issued against 57 brokerage firms for violations related to wealth management business, highlighting issues such as mismanagement of personnel and inadequate compliance [2][3]. - Over one-third of brokerages have faced administrative measures due to violations in wealth management since 2025, primarily involving branch offices [3]. Common Violations - Six prevalent types of violations have been identified: 1. Inadequate compliance management of personnel, with examples including unauthorized trading and improper account handling [3][4]. 2. Failure to effectively implement investor suitability management, with instances of providing incorrect answers to knowledge assessments [4]. 3. Unauthorized promises of returns during financial product sales, indicating a focus on quantity over quality in brokerage practices [5]. 4. Illegal solicitation of clients, with several firms found to be assigning marketing tasks to non-marketing personnel [6][7]. 5. Failure to report significant events that could impact management and client rights in a timely manner [8]. 6. Multiple issues often exist within the same brokerage, leading to severe operational impacts [9]. Impact of Violations - The consequences of violations extend beyond warnings, with some branches facing business suspensions for serious infractions [9][10]. - Increased internal compliance checks and regulatory discussions have been mandated for firms with identified issues [10][11]. Employee Accountability - Nearly 97 penalties have been issued to individual employees for violations related to wealth management, with a concentration on sales promotion and internal controls [13][14]. - The regulatory focus on employee misconduct reflects the ongoing challenges in transitioning to a "buy-side advisory" model [14][15]. Compliance Management Risks - The rise of online channels for business has introduced new compliance risks, with several penalties issued for violations related to online marketing practices [15].