传统宏观因子、宏观周期的高维度体系构建 2025 年 12 月 08 日 金 融 工 程 高维宏观周期驱动风格、行业月报(2025/11):经济 景气下行、通胀细分项下行看好大盘价值风格 投资要点： 金 融 工 程 定 期 报 告 宏观因子变量的构建：将宏观指数分别对宽基指数、代理宏观变 量做回归，选取 t 值显著的细分宏观变量，用过去一年标准差倒数加权 构建宏观因子变量。采用单边 HP 滤波器对宏观经济数据进行调整，消 除短期波动对长期趋势判断的影响。基于滤波变量，分别用因子动量 划分宏观趋势（上行、下行）和用时序百分位划分宏观状态（高、中、 低位）。 宏观因子升维的必要性：宏观因子 A 对宽基、风格和行业的价格 传导在 A 的不同边际变化不一致，且宏观因子 A 在宏观因子 B 的不同 状态下驱动宽基、风格和行业的收益方向也不同。同一状态及其边际 变化所对应的周期混乱，我们需要将宏观变量的边际与状态结合，综 合考虑宏观变量的变化趋势和所处的时序排位。 多信号驱动下的指数择时、风格轮动 小盘全指择时：在库存处于中等向上水平时预测值最高，因此推 荐配置中证全指。 2012 年 1 月末起至 2025 年 11 ...

Quantitative Models and Construction Methods 1. Model Name: Short-term Timing Strategy - **Model Construction Idea**: The model is based on historical data and statistical rules to identify short-term market timing signals, combining macroeconomic, valuation, sentiment, and liquidity indicators to generate a comprehensive timing signal[16][17][19] - **Model Construction Process**: 1. **Macroeconomic Indicators**: - Manufacturing PMI: If PMI > 50, it gives a positive signal; otherwise, a cautious signal. - Credit Pulse: The YoY growth rate of medium- and long-term RMB loans is used, with a higher percentile indicating a positive signal. - M1 YoY Growth Rate: Filtered using HP filter; higher percentiles indicate a positive signal. 2. **Valuation Indicators**: - PE Median Percentile: A higher percentile indicates a cautious signal due to mean reversion. - PB Median Percentile: A higher percentile also indicates a cautious signal due to mean reversion. 3. **Sentiment Indicators**: - Beta Dispersion: Neutral signal if within a certain range. - Volume Sentiment Score: Lower percentiles indicate a cautious signal. - Volatility: Neutral signal if within a certain range. 4. **Liquidity Indicators**: - Money Market Rate: Lower percentiles indicate a positive signal. - Exchange Rate Expectation: A stronger RMB against the USD gives a positive signal. - 5-day average net financing amount: Lower percentiles indicate a positive signal. 5. Combine all signals to generate a comprehensive timing signal[16][17][19] - **Model Evaluation**: The model demonstrates significant improvement over the benchmark strategy, with higher annualized returns, lower maximum drawdown, and better Sharpe ratio[18][21] 2. Model Name: Growth-Value Style Rotation Model - **Model Construction Idea**: The model uses a quantitative economic mid-cycle analysis framework, incorporating profitability, interest rate, and credit cycles to determine the relative attractiveness of growth versus value styles[26][27] - **Model Construction Process**: 1. **Macroeconomic Indicators**: - Profitability Cycle Slope: A steeper slope favors growth. - Interest Rate Cycle Level: Higher levels favor value. - Credit Cycle Strength: A stronger credit cycle favors growth. 2. **Valuation Indicators**: - Growth-Value PE Spread: A higher 5-year percentile indicates a preference for growth. - Growth-Value PB Spread: A higher 5-year percentile also indicates a preference for growth. 3. **Sentiment Indicators**: - Turnover Spread: A higher 5-year percentile indicates a preference for growth. - Volatility Spread: A higher 5-year percentile indicates a balanced preference for both growth and value. 4. Combine all signals to generate a comprehensive style rotation signal[26][27][28] - **Model Evaluation**: The strategy has shown significant improvement over the benchmark, with higher annualized returns, lower maximum drawdown, and better Sharpe ratio. However, in 2025, the strategy underperformed the benchmark slightly[27][29] 3. Model Name: Small-Cap vs. Large-Cap Style Rotation Model - **Model Construction Idea**: The model is based on 11 effective rotation indicators, including liquidity, sentiment, and valuation metrics, to determine the relative attractiveness of small-cap versus large-cap styles[30] - **Model Construction Process**: 1. **Indicators Used**: - Indicators such as R007, financing balance changes, trading volume, and sentiment metrics are analyzed. - For each indicator, a signal is generated to favor either small-cap or large-cap styles. 2. **Comprehensive Signal**: - Combine all individual signals into a comprehensive small-cap or large-cap rotation signal. - The model currently favors large-cap due to weak small-cap indicators such as low trading volume and negative sentiment[30][32] - **Model Evaluation**: The strategy has consistently generated positive annual excess returns since 2014, with a significant improvement over the benchmark in terms of annualized returns and maximum drawdown[31][32] --- Model Backtesting Results 1. Short-term Timing Strategy - Annualized Return: 16.41% - Annualized Volatility: 14.81% - Maximum Drawdown: 14.07% - Sharpe Ratio: 0.9655 - Return-to-Drawdown Ratio: 1.1667 - Monthly Win Rate: 66.24% - Quarterly Win Rate: 60.38% - Annual Win Rate: 78.57%[18][21] 2. Growth-Value Style Rotation Model - Annualized Return: 12.74% - Annualized Volatility: 20.80% - Maximum Drawdown: 43.07% - Sharpe Ratio: 0.5853 - Return-to-Drawdown Ratio: 0.2958 - Monthly Win Rate: 58.33% - Quarterly Win Rate: 59.62%[29] 3. Small-Cap vs. Large-Cap Style Rotation Model - Annualized Return: 19.73% - Annualized Excess Return: 12.67% - Maximum Drawdown: 40.70% - Average Turnover Interval: 20 trading days - Win Rate (per trade): 49.57%[32]

Core Viewpoint - The article discusses the performance of various investment strategies in the A-share market, particularly focusing on value and dividend strategies, and their effectiveness compared to growth strategies over different time periods [7][15][22]. Group 1: Market Performance - The overall market showed slight fluctuations, with the CSI All Share Index slightly up, closing at 4.3 stars [1]. - Large and mid-cap stocks experienced minor gains, while small-cap stocks declined [2]. - Value style saw a slight decrease, whereas growth styles, including the ChiNext and STAR Market, experienced gains [3][4]. Group 2: Investment Strategies - The article references Siegel's book, "The Future for Investors," which indicates that high dividend and low P/E strategies can outperform the market index [7]. - In the A-share market, value strategies have shown more effectiveness in recent years compared to the U.S. market, with the CSI 300 Value Index and CSI Dividend Index being notable examples [15][22]. - The CSI 300 Value Index has increased from 1,000 points to 9,147 points (approximately 915% increase) from the end of 2004 to the end of 2024, while the CSI 300 Index rose from 1,000 points to 5,643 points [17][19]. Group 3: Fund Performance and Size - Despite the effectiveness of value strategies, the total scale of corresponding funds in the A-share market remains relatively small, with the CSI 300 Value Index fund only amounting to several billion, compared to over 1 trillion for the CSI 300 [24][25]. - The article notes that value strategies are not always effective, as seen during the growth style dominance from 2019 to 2021, which led to significant fund redemptions [26][27]. Group 4: Investor Behavior and Patience - A long-term effective strategy may not always yield results, as market styles rotate over time, with value styles being strong from 2016-2018, growth styles from 2019-2021, and value styles regaining strength from 2022 onwards [28]. - Many investors lack the patience to wait for a strategy to recover after a downturn, often selling during low periods instead of holding for potential future gains [29].

Core Viewpoint - The Q4 style rotation model indicates signals for small-cap and growth stocks, with a focus on sectors such as electric equipment and renewable energy for December [1][2]. Style Rotation Model - The Q4 style rotation model has issued signals favoring small-cap stocks, with a comprehensive score of -1 for the dual-driven rotation strategy as of September 30, 2025 [3]. - The value-growth style rotation model shows a comprehensive score of -3 for the dual-driven rotation strategy, indicating a preference for growth stocks [4]. Industry Rotation Insights - In November, the composite factor strategy yielded an excess return of -0.58%, while the single-factor long strategy had an excess return of -0.83% [4]. - For December, the single-factor long strategy recommends bullish sectors including banking, construction, non-bank financials, and electric equipment and renewable energy. The composite factor strategy suggests bullish sectors such as telecommunications, comprehensive finance, computer technology, electric equipment and renewable energy, and utilities [4].

Market Overview - A-shares experienced a decline today, with the Shanghai Composite Index down 0.51% to 3878.00 points, the Shenzhen Component Index down 0.78%, the ChiNext Index down 1.12%, and the STAR Market Index down 0.95% [1][10] - The trading volume in the Shanghai and Shenzhen markets was approximately 16699.62 billion yuan, an increase of about 765.32 billion yuan compared to the previous trading day [1][10] - The market showed a low risk appetite, with 1443 stocks rising and 3876 stocks falling [1][10] Sector Performance - Dividend sectors performed well today, with transportation, non-ferrous metals, oil, mining, and coal showing positive results [1][10] - High-volatility sectors, including gaming, film and television, new energy vehicles, and computers, underperformed [1][10] - The market style showed that small-cap stocks lagged behind large-cap stocks, and growth stocks underperformed value stocks [1][10] Economic Outlook - The current macroeconomic state is characterized by a transition between old and new growth drivers, with a "K" shaped economic recovery [2][10] - Three sectors with growth potential identified are technology (AI revolution, policy support, overseas mapping), upstream anti-involution (solar, lithium batteries), and exports (global manufacturing recovery, positive overseas fiscal expectations) [2][10] - The technology and upstream sectors are still on an upward trend but carry risks due to previous significant gains [2][10] Investment Recommendations - Investors are advised to maintain a balanced allocation strategy, utilizing the "seesaw effect" to hedge daily volatility and optimize holding experiences [10] - Suggested ETFs for potential opportunities include non-ferrous metals 60 ETF (159881), mining ETF (561330), chemical leading ETF (516220), and industrial mother machine ETF (159667) [2][10] - As a hedging option, cash flow ETF (159399) is recommended [2][10] Bond Market Insights - The recent bond market environment shows a divergence between macro conditions and trading sentiment, with a weak nominal growth rate and a low interest rate environment supported by macro realities [7][16] - The People's Bank of China announced the purchase and sale of 50 billion yuan in government bonds, with the 30-year government bond yield rising by 2.40 basis points to 2.23% [14][16] - Financial institutions maintain a moderately optimistic outlook for the bond market in December, with a downward trend in funding rates observed since November [16][8]

1. Report Industry Investment Rating No industry investment rating information is provided in the report. 2. Report's Core View - The recovery signal of the fundamental situation still needs to be observed, the momentum factor remains bearish, and the market sentiment is rising from a low level. It is recommended to maintain an under - allocation of equity assets. In the short - term style, the small - cap style is expected to dominate in December, and the growth style will continue to dominate. It is recommended to focus on small - cap and growth styles. For fixed - income + funds, it is recommended to focus on relatively stable varieties, and for bond funds, focus on short - duration varieties [2][74]. 3. Summary According to the Catalog 11 - month Review Stock Market - A - shares declined, with the Shanghai Composite Index falling 1.67% and the Science and Technology Innovation 50 falling 6.24%. The Dow Jones Index rose 0.32%, and the Nasdaq Index fell 1.51%. Affected by concerns about the valuation of the US stock AI sector and the volatility of the Fed's interest - rate cut expectations, A - shares and Hong Kong stocks declined, and US stocks fluctuated [8][13]. Bond Market - US bond yields declined, with the 1 - year US bond yield dropping to 3.61% and the 10 - year US bond yield dropping to 4.02%. Domestic bond yields rose, with the 1 - year Treasury yield rising to 1.40% and the 10 - year Treasury yield rising to 1.84%, and the term spread widened. The decline in US bond yields was due to the volatility of the Fed's interest - rate cut expectations, while the rise in domestic bond yields was because the central bank's Treasury purchase volume was lower than expected [8][17]. Foreign Exchange Market - The US dollar index declined to 99.44, and the RMB appreciated. The on - shore exchange rate of the US dollar against the RMB rose to 7.08, and the off - shore exchange rate rose to 7.07. The decline in the US dollar index was due to the continuous volatility of the Fed's interest - rate cut expectations, and the RMB appreciation was supported by the weakening of the US dollar and the strong domestic economic fundamentals [20]. Commodity Market - Crude oil prices fell to $63.2 per barrel. Domestic commodity prices rose slightly after fluctuations, and overseas commodity prices fell overall after rising first and then falling. Among domestic commodities, precious metals and grains led the gains, while coal, coking, steel, minerals, agricultural and sideline products, non - metallic building materials, non - ferrous metals, energy, and chemicals declined [25]. Fund Market - The performance of the fund market in November was poor. The issuance scale increased to 94.6 billion yuan, a 31% increase from the previous month. Structurally, the issuance scale of equity funds was 45.3 billion yuan, a 30% increase from the previous month, accounting for 48% of the total issuance. ETF funds had a net inflow of 100.9 billion yuan (excluding money funds), and LOF funds had a net outflow of 320 million yuan. Among them, equity - type ETF products had a net inflow of 32.3 billion yuan, and equity - type LOF products had a net outflow of 440 million yuan. Active equity funds increased their positions in dividend, value - potential, and prosperous styles and reduced their positions in quality styles [30][36][37]. 12 - month Outlook Overseas Environment - The market's expectation of the Fed's interest - rate cut within the year fluctuated significantly, and the market expected the Fed to cut interest rates in December. The probability of the Fed cutting interest rates in December increased to over 80%. US bond yields first rose and then fell, with an overall decline [43]. Domestic Environment - The private - sector financing growth rate continued to decline, and the inflation factor rebounded from a low level. It is recommended to maintain an under - allocation of equity assets as the economic recovery signal still needs to be observed, and the momentum factor remains bearish [47]. Trading Perspective - The stock - market odds were close to the three - year average, and the A - share market sentiment was rising from a low level but had not yet returned to the optimistic range [48][52]. Market Style - The growth - value style rotation model shows that the market factor, US bond yield, and style momentum are all favorable for growth, and the growth style will continue to dominate. The small - large - cap style rotation model shows that the current credit environment, monetary environment, and long - and short - term style momentum all recommend the small - cap style [59][64]. Hong Kong Stock Market - The number of macro - indicators bullish on Hong Kong stocks decreased compared with last month. The model recommends an under - allocation of Hong Kong stocks as the private - sector financing growth rate, Hong Kong dollar M2 growth rate, and Chinese sovereign CDS spread are bearish, although the US dollar index and south - bound funds are bullish [68]. Domestic Bond Market - Short - term liquidity remained in a tight balance, and long - term interest rates rose. It is recommended to focus on short - duration bond funds as short - term bonds have better opportunities than long - term bonds [71]. Fund Allocation Strategy - It is recommended to maintain an under - allocation of equity assets and focus on small - cap and growth styles. For fixed - income + funds, focus on relatively stable varieties, and for bond funds, focus on short - duration varieties. Recommended funds include Dongwu Mobile Internet (001323.OF, medium - high risk), CITIC Prudential Multi - Strategy (165531.OF, medium - high risk), Harvest New Consumption (001044.OF, medium - high risk), BOC Steady Income (380009.OF, medium risk), and Penghua Stable Short - Term Bond (007515.OF) [2][74].

Core Insights - The report suggests focusing on the financial, non-ferrous metals, electronics, and machinery sectors for December [1] - The value-growth rotation strategy has a composite score of 5, indicating a higher score for growth style as of November 30, 2025 [3][6] - The small-cap style has a higher score in the size rotation strategy, with a composite score of 4 [8] Style Rotation Insights - The large-cap stocks are more sensitive to economic prosperity, while growth stocks benefit more from liquidity easing [3][6] - The value-growth rotation strategy yielded a growth index return of -2.85% and a value index return of 0.35% in November 2025 [6] - The size rotation strategy showed a return of -2.46% for the CSI 300 and -2.30% for the CSI 1000 in November 2025 [8] Industry Rotation Insights - The report constructs a four-dimensional engine with macro, fundamental, technical, and crowding indicators for industry index rotation [11] - The top five industries for December based on the industry rotation composite score are banking, electronics, machinery, non-ferrous metals, and non-bank financials [3][23] - The bottom five industries are coal, real estate, construction, oil and petrochemicals, and textiles and apparel [3][23] Macro Indicators - The macroeconomic growth dimension is in the "expansion strengthening/recession alleviation" phase, while the liquidity dimension is in the "easing intensification/tightening slowdown" phase as of November 30, 2025 [13] - The report recommends allocating to the large financial and midstream manufacturing sectors based on these macro indicators [13] Fundamental Indicators - The top five industries based on fundamental indicators are non-bank financials, non-ferrous metals, electronics, telecommunications, and electric equipment and new energy [17] - The bottom five industries based on fundamental indicators are real estate, coal, construction, agriculture, forestry, animal husbandry, and textiles and apparel [17] Technical Indicators - The top five industries based on technical indicators are electronics, banking, telecommunications, non-ferrous metals, and machinery [18] - The bottom five industries based on technical indicators are coal, construction, food and beverage, oil and petrochemicals, and real estate [18] Crowding Indicators - The industries with high crowding indicators include basic chemicals, electric equipment and new energy, agriculture, real estate, and textiles and apparel [22] - The industries with low crowding indicators are machinery, non-bank financials, automobiles, computers, and food and beverage [22]

Core Viewpoint - The AI sector currently shows a relatively low level of crowding risk, indicating potential investment opportunities in the long term, while short-term strategies may favor value styles [4][42][43]. Group 1: AI Sector Crowding Analysis - The crowding score for the AI sector has decreased from a high point in September 2025, reflecting a return to a safer range, with no current crowding signals triggered [2][29][43]. - The crowding monitoring model includes five observational indicators: returns, trading volume, trading structure, valuation, and market sentiment, which collectively indicate a low risk of crowding in the AI sector [2][6][43]. - The AI-related indices, such as the STAR 50 Index and the CSI AI Theme Index, have shown a significant drop in crowding scores, suggesting that the previous high trading enthusiasm has subsided [8][22][29]. Group 2: Style Rotation Model - The style rotation model for 2025 has a high success rate, achieving a monthly win rate of 73% as of November 27, 2025, effectively capturing important style momentum and switching points [3][37]. - Short-term recommendations lean towards value styles due to current market dynamics, while long-term growth styles still hold potential for development [3][41]. - The model indicates that macroeconomic factors, such as a narrowing year-on-year decline in PPI, favor value styles in the short term, while growth styles may benefit from optimistic long-term economic expectations [41]. Group 3: Future Outlook - The AI sector's crowding risk has been alleviated, and there are still long-term growth opportunities, although value styles may have a relative advantage in the short term [4][42][43]. - Institutional investors are likely to adopt defensive strategies as year-end approaches, which may further support the preference for value styles [44].

钛媒体App 11月30日消息，银河证券表示，11月市场行情呈现出高低切换特征，资金从高估值成长股流 向低估值周期股及红利资产，防御板块吸引力上升。12月来看，市场仍处于风格频繁切换阶段，或以结 构性行情为主。一方面，作为年末最关键的政策窗口，重要政策会议在科技创新、扩内需、反内卷、稳 地产等方面或将作出重点部署。另一方面，一系列产业会议可能成为主题行情的重要催化剂，关注即将 召开的2025"人工智能+"产业生态大会、脑机接口大会等。而外部环境依然存在不确定性，12月美联储 议息会议表态对全球市场流动性的潜在影响值得关注。（广角观察） ...

银河 证券表示，11月市场行情呈现出高低切换特征，资金从高估值成长股流向低估值周期股及红利资 产，防御板块吸引力上升。12月来看，市场仍处于风格频繁切换阶段，或以结构性行情为主。一方面， 作为年末最关键的政策窗口，重要政策会议在科技创新、扩内需、反内卷、稳地产等方面或将作出重点 部署。另一方面，一系列产业会议可能成为主题行情的重要 催化剂，关注即将召开的2025" 人工智能 +"产业生态大会、 脑机接口大会等。而外部环境依然存在不确定性，12月美联储议息会议表态对全球 市场流动性的潜在影响值得关注。 ...