报告导读： 本报告对大小盘轮动月度策略、价值成长轮动月度策略以及风格因子表现进行 跟踪。根据量化模型信号， 1 月建议超配小盘风格，均衡配置价值成长风格。 大小盘风格轮动月度策略。 月度观点： 12 月底量化模型最新信号为 0.17 ，指向小盘。日历效应上，历史 1 月大盘相对占优；建议 1 月超配小盘风格。中 长期观点，当前市值因子估值价差为 0.89 ，近期有所下降，距离历史顶部区域 1.7~2.6 仍有距离，中长期并不拥挤，继续看好小盘。截止 12 月底，模型收 益 27.56% ，相对等权基准（ 26.84% ）的超额收益为 0.71% 。结合主观观点的策略收益 28.97% ，超额收益 2.13% 。策略构建详见报告《量化视角多 维度构建大小盘风格轮动策略 _20241102 》。 价值成长风格轮动月度策略。 月度量化模型信号为 0 ，建议 1 月等权配置成长和价值风格。截止 12 月底，模型收益 22.72% ，相对等权基准（ 20.4% ）的超额收益为 1.93% 。策略构建详见报告《量化视角多维度构建月度和周度价值成长风格轮动策略 _20250305 》。 风格因子表现跟踪。 8 个大类因子中 ...

固收- 2026 年度策略 - 时光倒流 20251230 摘要 中国经济不宜简单类比日本，中国企业投资和科技进步未停滞，海外投 资量级差异也导致通缩情况不同。美国投资者对贸易战关注减少，更关 注科技趋势和资本开支，中国则显现结构性业绩牛市迹象。 高质量发展强调可持续性，而非依赖基建和地产。房地产对 A 股的重要 性已大幅降低，不再是风险资产代理变量。应关注新兴产业和科技创新 带来的增长潜力。 A 股基本面已变，TMT 行业尤其是存储产业占据更重要位置，内存价格 与 A 股相关性高。房地产主要影响存量资产估值，对 GDP 贡献率下降。 房价与利率之间不存在稳定关系，消费领域受房地产市场波动影响也不 稳定，不同区域间消费与地产关系存在显著差异。 出口数据对资产定价重要，但过度依赖可能导致误判。全球通胀传导机 制显示中国长期通缩预期正在逆转，贸易流未断裂削弱了逆全球化解释 力。 预计 2026 年有降准和降息，但为配合反内卷，信贷量不太可能大幅增 加。央行可能扩大买债规模，以应对商业银行资产负债问题及社会风险 偏好提高，确保金融系统稳定。 2025 年量化模型遭遇挫折，因未能准确反映市场震荡特征。债券市场 面临利 ...

东证衍生品研究院金工高级分析师：徐凡（国债期货、基本面量化） 从业资格号： F03107676 投资咨询号： Z0022032 金工策略周报 东证衍生品研究院金工首席分析师：李晓辉（CTA） 从业资格号： F03120233 投资咨询号： Z0019676 东证衍生品研究院金工高级分析师：徐凡（国债期货、基本面量化） 从业资格号： F03107676 投资咨询号： Z0022032 ★国债期货行情简评： 上周四个期债品种均冲高回落，30年期主力合约报112.47元，10年期主力合约报107.985元，5年期主力合约报105.82元，2年期主力合约报 102.464元。期债基差方面，本周基差下行、IRR持续上行，跨期价差震荡偏强， ★国债期货择时策略： 十年期国债：今年表现来看，以夏普比排名分别为基差因子、风险资产和会员持仓，三者2025年的夏普比分别为1.68、1.93和0.59。 五年期国债：今年表现来看，以夏普比排名分别为高频资金流、日内量价、风险资产、会员持仓和基差因子，五者2025年的夏普比分别为2.51、 2.27、1.71、1.33和0.78。 两年期国债：今年表现来看，以夏普比排名分别为高频资金 ...

Group 1 - The report indicates that domestic asset BL models 1 and 2 both recorded a weekly return of 0.1%, with December returns of 0.11% and year-to-date returns of 4.15% and 3.93% respectively [1][14]. - Global asset BL models 1 and 2 experienced a decline, with model 1 showing a weekly return of -0.14% and model 2 a slight decrease of -0.01%, while their year-to-date returns were 1.01% and 2.59% respectively [1][14]. - The domestic risk parity model achieved a weekly return of 0.04% and a year-to-date return of 3.68%, while the global risk parity model had a weekly return of 0.02% and a year-to-date return of 3.31% [20][21]. Group 2 - The report highlights that the macro factor-based asset allocation model yielded a weekly return of 0.07% and a year-to-date return of 4.48%, indicating its effectiveness in the current market environment [26][27]. - The performance of various asset classes was tracked, with SHFE gold showing a significant increase of 1.0%, while the South China commodity index and S&P 500 experienced declines of 1.21% and 0.78% respectively [7][9]. - The report emphasizes the importance of the Black-Litterman model, which integrates subjective views with quantitative models to optimize asset allocation, thus providing a more robust investment strategy [12][13].

微盘股，通常指市值较小、流动性相对偏低的上市公司股票。它们并非简单的"小盘股"，而是"小盘股 中的小盘股"。从过往实践来看，其投资标的往往具备以下鲜明特征： 近年来，微盘股投资策略备受市场瞩目，以其独特的波动性和高弹性吸引了众多目光。然而，对于大多 数投资者而言，微盘股投资仍笼罩着一层神秘面纱。它究竟是价值发现的沃土，还是一场高风险的博 弈？今天让我们来探讨下投资微盘股，到底投资的是什么？ 微盘股：一个怎样的"小众江湖"？ 策略如何落地：量化模型与严格风控 机构参与率低：大型机构资金鲜少涉足，使得股价不易被大额调仓所冲击，市场情绪相对独立。 筹码结构稳定：股价跌至低位后，现有股东抛售意愿降低，甚至核心股东有较强的做高股价动机，形成 天然的"安全垫"。 交投清淡，换手率低：日常成交金额小（多为千万级别），股价长期处于低估或无人问津的状态，便于 量化模型通过批量指标识别逆向投资标的。 估值修复弹性大：由于抛压轻，通常来说一旦获得市场关注，即使少量增量资金也能驱动股价快速、显 著地上涨。 投资的本质：一场"关注度"的博弈 与传统价值投资聚焦公司基本面（盈利能力、行业前景）不同，也区别于成长投资追逐高景气板块，微 盘 ...

Group 1: DeepSeek-V3.2 Series Release - The release of DeepSeek-V3.2 marks a significant advancement in open-source large models, achieving performance levels comparable to top closed-source models[3] - The Speciale version of DeepSeek-V3.2 has excelled in international competitions, ranking second in the ICPC and winning gold medals in the IMO, demonstrating its potential to reach human-level intelligence[4] - DeepSeek-V3.2 integrates thinking modes with tool invocation, enhancing the model's generalization and execution capabilities across complex scenarios[5] Group 2: Market Trends and Predictions - The 2025 Winter FORCE Conference is set to focus on Agentic AI, with significant updates expected for the Doubao model family and AI application capabilities[9] - Doubao model's daily token usage surged from 120 billion in May 2024 to over 30 trillion by September 2025, indicating a 253-fold increase in usage[10] - The report predicts that the 2026 monetary policy will emphasize "wide credit" rather than merely "wide loans," aligning with fiscal measures to support economic growth[35] Group 3: Company Coverage and Financial Projections - Faway Automobile Components (600742) is rated "Overweight" with a target price of RMB 14.10, based on stable automotive parts business and expansion into robotics and low-altitude economy[13] - Projected revenues for Faway are RMB 208.72 million, RMB 220.62 million, and RMB 231.65 million for 2025, 2026, and 2027 respectively, with net profits of RMB 6.30 million, RMB 6.99 million, and RMB 7.75 million[13] - The company is actively developing humanoid robots and EVTOL interior designs, leveraging its automotive parts manufacturing expertise[15]

Quantitative Models and Construction Methods 1. Model Name: Black-Litterman (BL) Model - **Model Construction Idea**: The BL model is an improvement over the traditional mean-variance optimization (MVO) model. It integrates subjective views with quantitative models using Bayesian theory to optimize asset allocation weights. This approach addresses the sensitivity of MVO to expected returns and provides a more robust asset allocation solution[12][13]. - **Model Construction Process**: - The BL model combines subjective views of investors with market equilibrium returns to derive optimized portfolio weights. - The model uses the following formula to calculate the posterior expected returns: $ \mu = [( \tau \Sigma )^{-1} + P^T \Omega^{-1} P]^{-1} [( \tau \Sigma )^{-1} \Pi + P^T \Omega^{-1} Q] $ - $\mu$: Posterior expected returns - $\tau$: Scalar representing the uncertainty in the prior estimate of returns - $\Sigma$: Covariance matrix of asset returns - $\Pi$: Equilibrium returns derived from market capitalization weights - $P$: Matrix representing the views on assets - $\Omega$: Covariance matrix of the views - $Q$: Vector of expected returns based on the views - The optimized portfolio weights are then derived using the posterior expected returns and the covariance matrix[12][13]. - **Model Evaluation**: The BL model effectively addresses the sensitivity of MVO to expected returns and provides a more robust and efficient asset allocation framework. It also allows for the incorporation of subjective views, making it more flexible and practical for real-world applications[12]. 2. Model Name: Risk Parity Model - **Model Construction Idea**: The risk parity model aims to equalize the risk contribution of each asset in a portfolio. It is an improvement over the traditional mean-variance optimization model and focuses on diversifying risk rather than capital allocation[17][18]. - **Model Construction Process**: - Step 1: Select appropriate underlying assets. - Step 2: Calculate the risk contribution of each asset to the portfolio using the formula: $ RC_i = w_i \cdot \sigma_i \cdot \rho_{i,portfolio} $ - $RC_i$: Risk contribution of asset $i$ - $w_i$: Weight of asset $i$ - $\sigma_i$: Volatility of asset $i$ - $\rho_{i,portfolio}$: Correlation of asset $i$ with the portfolio - Step 3: Solve the optimization problem to minimize the deviation between actual and target risk contributions, subject to the constraint that the sum of weights equals 1[18][19]. - **Model Evaluation**: The risk parity model provides a balanced risk allocation across assets, making it suitable for achieving stable returns across different economic cycles. It is particularly effective in reducing portfolio volatility and drawdowns[18]. 3. Model Name: Macro Factor-Based Asset Allocation Model - **Model Construction Idea**: This model constructs a macro factor system covering six key risks: growth, inflation, interest rates, credit, exchange rates, and liquidity. It bridges macroeconomic research with asset allocation by translating macroeconomic views into actionable portfolio strategies[21][22]. - **Model Construction Process**: - Step 1: Calculate the factor exposure levels of assets at the end of each month. - Step 2: Use a risk parity portfolio as the benchmark and calculate the benchmark factor exposure. - Step 3: Based on macroeconomic forecasts for the next month, assign subjective factor deviation values. For example, if inflation is expected to rise, assign a positive deviation to the inflation factor. - Step 4: Combine the benchmark factor exposure with the subjective factor deviations to derive the target factor exposure for the portfolio. - Step 5: Solve the optimization problem to determine the asset allocation weights for the next month[22][25]. - **Model Evaluation**: This model effectively incorporates macroeconomic views into asset allocation, providing a systematic framework for translating macroeconomic insights into portfolio decisions. It is particularly useful for capturing macroeconomic trends and their impact on asset performance[21]. --- Model Backtesting Results 1. Black-Litterman (BL) Model - **Domestic Asset BL Model 1**: Weekly return: -0.32%, November return: 0.05%, 2025 YTD return: 4.0%, annualized volatility: 2.18%, maximum drawdown: 1.31%[14][16][17] - **Domestic Asset BL Model 2**: Weekly return: -0.15%, November return: 0.08%, 2025 YTD return: 3.77%, annualized volatility: 1.95%, maximum drawdown: 1.06%[14][16][17] - **Global Asset BL Model 1**: Weekly return: -0.17%, November return: -0.26%, 2025 YTD return: 0.78%, annualized volatility: 2.0%, maximum drawdown: 1.64%[14][16][17] - **Global Asset BL Model 2**: Weekly return: 0.01%, November return: 0.08%, 2025 YTD return: 2.7%, annualized volatility: 1.59%, maximum drawdown: 1.28%[14][16][17] 2. Risk Parity Model - **Domestic Asset Risk Parity Model**: Weekly return: -0.27%, November return: -0.09%, 2025 YTD return: 3.6%, annualized volatility: 1.32%, maximum drawdown: 0.76%[20][28] - **Global Asset Risk Parity Model**: Weekly return: -0.2%, November return: -0.07%, 2025 YTD return: 3.04%, annualized volatility: 1.42%, maximum drawdown: 1.2%[20][28] 3. Macro Factor-Based Asset Allocation Model - **Macro Factor-Based Asset Allocation Model**: Weekly return: -0.31%, November return: -0.01%, 2025 YTD return: 4.43%, annualized volatility: 1.55%, maximum drawdown: 0.64%[27][28]

入冬后的申城，寒气袭人，却挡不住股民们的热情。三五成群的投资者们裹着厚外套，手里拿着热茶， 聚集在广东路西藏中路，热烈讨论着行情走势。这个自发形成的"马路股市沙龙"，每逢周末便成为沪上 股民交流学习的重要场所。 在人群中，一位戴着眼镜的中年股民正在草稿纸上画着技术图形，向周围人讲解均线系统的实战应 用。"突破关键阻力位需要成交量的确认，这是技术分析的基本功。"他的讲解引来阵阵讨论，有人补充 着各类指标的适用条件，还有人分享着择时、择股、波段操作的方法论。 股民们正在街头交流炒股心得。（复原当时情景，AI辅助生成） 退休工人王师傅是这里的常客，他操着一口地道的上海话感慨："现在的小年轻炒股，不像我们当年光 靠听消息了，都讲究要学真东西。这波行情来了，我们每周都找个主题好好研究。上周刚聊完经济大势 怎么影响股市，这周就在琢磨各个板块是怎么轮着涨的。" 旁边几位年轻股民正凑在一起，盯着平板电脑上跳动的曲线。"你看，我们正在测试这个指标策略在不 同行情下灵不灵光。"其中一个戴眼镜的男生推了推眼镜说："炒股不能凭感觉，得看数据说话，就像做 实验一样，得反复验证。我就在研究当下最流行的量化模型！" 从经验分享到工具革新 ...

Core Insights - THPX Signal Source has launched the WTIBTC Intelligent Trend Capture System to assist investors in identifying market trends and trading signals related to the unique correlation between WTI crude oil and Bitcoin [1][3][8] Group 1: Product Overview - The WTIBTC system integrates advanced algorithms and big data analysis to provide real-time market trend identification and trading signals for investors focusing on the WTI-BTC asset combination [1][3] - The system aims to alleviate "information anxiety" by presenting clear trend paths amidst the complex interactions between WTI crude oil prices and Bitcoin market performance [3][5] - It employs a comprehensive processing capability, utilizing a scientific framework that tracks various indicators, including price change rates and market sentiment [3][5] Group 2: Adaptive Learning and Risk Management - The system incorporates an adaptive learning core that continuously optimizes its analysis models based on historical signal validation and current market data [5][8] - It emphasizes the importance of risk control, providing signals and analysis reports that assist in decision-making while integrating risk considerations [5][8] - The system is designed to be an "intelligent advisor" rather than a replacement for user decision-making, maintaining user control over trading decisions [5][8] Group 3: User Applications - The WTIBTC system caters to various user types, from long-term investors optimizing their WTI-cryptocurrency asset allocation to short-term traders capturing WTI-BTC trading opportunities [7][8] - It effectively compensates for the information processing limitations of time-constrained investors while serving as a supplementary tool for experienced traders [7][8] - The system aims to enhance overall decision-making efficiency and judgment dimensions for investors [7][8]

Group 1 - The core viewpoint emphasizes that trading success is a collective effort of the team, highlighting the importance of discipline and adherence to a predetermined trading plan [1] - The team achieved success through a combination of strategic determination and tactical flexibility, focusing on a "defensive first, offensive second" approach and timely execution based on volatility cycles [1][2] - The trading strategy during the competition was primarily based on "trend following and sector rotation," utilizing options for hedging, which allowed for enhanced returns during market upswings and protection during downturns [1][2] Group 2 - The market characteristics this year include uncertainty in direction, increased event-driven trading, and fluctuating volatility cycles, leading the team to focus on volatility pricing and risk exposure management rather than directional predictions [2][3] - The team concentrated on specific sectors such as the Sci-Tech 50, non-ferrous metals, gold, crude oil, agricultural products, and the Hang Seng Tech Index, selecting these based on long-term fundamentals and liquidity [2] - The entry timing strategy is based on fundamental analysis for direction and technical analysis for timing, with a focus on market sentiment and volatility levels [3] Group 3 - Risk control is implemented through a dual system of "hard stop-loss" and "logical stop-loss," with options serving as both a stop-loss tool and a means of risk transfer [3] - The company emphasizes the importance of withdrawing principal after significant gains to maintain a healthy trading mindset, advocating for diversified positions and gradual building of positions [3] - The futures market is viewed as a platform for self-improvement and understanding, where successful investing relies on decisive actions at critical moments rather than frequent trading [4]