证券研究报告 2025 年 11 月 21 日 湘财证券研究所 金融工程研究 策略双周报 期权波动率套利策略跟踪 ——波动率偏斜策略 相关研究： 核心要点： ❑ 波动率偏斜策略跟踪情况 波动率偏斜策略是通过价内合约与价外合约的隐含波动率差异进行套利交 易。正常情况下，VSI 指标会在一定范围内波动，但当不同期权合约的波 动率比值出现实质差异时，就存在相应的反向套利空间。 本年以来，认购子策略收益率为 8.49%，最大回撤为 2.93%；认沽子策略 收益率为-1.31%，最大回撤为 10.49%；组合策略收益率为 3.68%，最大回 撤为 5.57%。 近两周以来（2025 年 12 月 8 日至 2025 年 12 月 19 日），认购子策略的收 益率为 0.91%，最大回撤为 0.09%；认沽子策略收益率为 0.46%，最大回撤 为 0.29%；组合策略收益率为 0.68%，最大回撤为 0.13%。 ❑ 投资建议 近两周以来，标的资产以震荡走势为主，从 VSI 指标偏离情况来看，认购 合约和认沽合约都出现了轻微偏离但很快回归的现象，套利策略非常适用 于这类市场走势，从策略收益来看，认购和认沽子策略均获得了正 ...

- The high-frequency skewness factor had long-short returns of 0.67% last week, -1.18% in December, and 22.39% year-to-date 2025[5] - The intraday downside volatility factor had long-short returns of 0.87% last week, -1.33% in December, and 19.08% year-to-date 2025[5] - The post-open buying intention proportion factor had long-short returns of 0.66% last week, 0.61% in December, and 21.12% year-to-date 2025[5] - The post-open buying intention intensity factor had long-short returns of 0.46% last week, 0.94% in December, and 28.09% year-to-date 2025[5] - The post-open large order net buying proportion factor had long-short returns of -0.21% last week, 0.17% in December, and 22.11% year-to-date 2025[5] - The post-open large order net buying intensity factor had long-short returns of -0.25% last week, 0.38% in December, and 12.5% year-to-date 2025[5] - The intraday return factor had long-short returns of 0.35% last week, 0.91% in December, and 22.33% year-to-date 2025[5] - The end-of-day trading proportion factor had long-short returns of -0.94% last week, 1.04% in December, and 16.73% year-to-date 2025[5] - The average single transaction outflow proportion factor had long-short returns of -1.15% last week, -2.15% in December, and -8.11% year-to-date 2025[5] - The large order push-up factor had long-short returns of 0.41% last week, -0.93% in December, and 7.19% year-to-date 2025[5] - The GRU(10,2)+NN(10) factor had long-short returns of 1.13% last week, -0.47% in December, and 47.04% year-to-date 2025, with long-only excess returns of -0.2% last week, -0.26% in December, and 7.1% year-to-date 2025[5] - The GRU(50,2)+NN(10) factor had long-short returns of 1.66% last week, 0.19% in December, and 47.39% year-to-date 2025, with long-only excess returns of 0.15% last week, 0.06% in December, and 8.92% year-to-date 2025[5] - The multi-granularity model (5-day label) factor had long-short returns of 2.46% last week, 1.12% in December, and 68.13% year-to-date 2025, with long-only excess returns of 0.74% last week, -0.18% in December, and 24.48% year-to-date 2025[5] - The multi-granularity model (10-day label) factor had long-short returns of 2.26% last week, 1.11% in December, and 62.71% year-to-date 2025, with long-only excess returns of 0.76% last week, -0.5% in December, and 24.3% year-to-date 2025[5] - The weekly rebalanced CSI 500 AI-enhanced wide constraint portfolio had excess returns of 0.41% last week, -2.64% in December, and 5.46% year-to-date 2025[5] - The weekly rebalanced CSI 500 AI-enhanced strict constraint portfolio had excess returns of 0.92% last week, -1.62% in December, and 9.23% year-to-date 2025[5] - The weekly rebalanced CSI 1000 AI-enhanced wide constraint portfolio had excess returns of 1.55% last week, -2.69% in December, and 15.39% year-to-date 2025[5] - The weekly rebalanced CSI 1000 AI-enhanced strict constraint portfolio had excess returns of 1.48% last week, -1.45% in December, and 19.02% year-to-date 2025[5]

Quantitative Models and Factor Construction Quantitative Models and Construction Methods 1. **Model Name**: Dynamic Factor Adjustment Model **Model Construction Idea**: Combines factor momentum and reversal characteristics to dynamically adjust factor selection based on historical performance and failure probabilities[4][80][82] **Model Construction Process**: - Evaluate factor momentum using the average RankIC over the past 6 months and the average RankICIR over the past 3-12 months[4][82] - Calculate conditional failure probabilities by rolling one year of historical data to assess the likelihood of a factor transitioning from effective to ineffective[74][87] - Exclude factors with high failure probabilities and assign scores based on momentum and failure probabilities. Select the top N factors with the highest scores for equal-weighted scoring in each period[82][87][88] **Model Evaluation**: The model effectively balances momentum and reversal characteristics, reducing the impact of unstable factors and improving robustness in factor selection[82][87] 2. **Model Name**: "2+3" Dynamic Factor Model for Small-Cap Stocks **Model Construction Idea**: Combines two fixed factors (valuation and volatility) with three dynamically selected high-momentum factors to construct a robust small-cap stock selection model[98][99] **Model Construction Process**: - Fixed factors: Valuation (BTOP) and volatility (VOLATILITY) are always included due to their stable and significant performance in small-cap pools[98][99] - Dynamic factors: Exclude factors with conditional failure probabilities above 80% and select the top 3 factors based on medium- and long-term momentum scores[98][99] - Construct a portfolio of 50 equally weighted stocks based on the selected factors[98][103] **Model Evaluation**: The model demonstrates strong performance in small-cap pools, with high momentum and low reversal failure probabilities, making it robust against overfitting[98][103] 3. **Model Name**: "Exclusion + Scoring" Model for Large-Cap Stocks **Model Construction Idea**: Focuses on stricter exclusion of high-failure-probability factors and integrates failure information into the scoring process for large-cap stock selection[109][110] **Model Construction Process**: - Exclude factors with conditional failure probabilities above 70%[109][110] - Combine failure indicators into the momentum scoring model, selecting the top 5 factors with the highest comprehensive scores[109][110] - Construct a portfolio of 50 equally weighted stocks based on the selected factors[109][113] **Model Evaluation**: The model effectively addresses the high sensitivity and extreme reversals in large-cap pools, improving stability and performance[109][113] Model Backtesting Results 1. **Dynamic Factor Adjustment Model**: - Annualized return: 8.83% - Sharpe ratio: 0.42 - Excess annualized return: 11.47% - Maximum drawdown: 38.67%[103] 2. **"2+3" Dynamic Factor Model for Small-Cap Stocks**: - Annualized return: 8.83% - Sharpe ratio: 0.42 - Excess annualized return: 11.47% - Maximum drawdown: 38.67%[103] 3. **"Exclusion + Scoring" Model for Large-Cap Stocks**: - Annualized return: 8.40% - Sharpe ratio: 0.40 - Excess annualized return: 8.32% - Maximum drawdown: 36.40%[113] Quantitative Factors and Construction Methods 1. **Factor Name**: Valuation (BTOP) **Factor Construction Idea**: Measures the book-to-price ratio to capture undervalued stocks[8][39] **Factor Construction Process**: Calculate the ratio of book value to current market value for each stock[8][39] **Factor Evaluation**: Demonstrates stable and significant performance in small-cap pools, with strong selection ability in various market conditions[39][98] 2. **Factor Name**: Volatility (VOLATILITY) **Factor Construction Idea**: Measures the residual volatility of stock returns to identify low-risk stocks[8][50] **Factor Construction Process**: Calculate the standard deviation of residuals from a time-series regression of stock returns[8][50] **Factor Evaluation**: Performs well in both small-cap and large-cap pools, with low failure probabilities and consistent selection ability[50][98] 3. **Factor Name**: Earnings (EARNING) **Factor Construction Idea**: Measures earnings yield to capture profitability[8][39] **Factor Construction Process**: Calculate the ratio of earnings to market value for each stock[8][39] **Factor Evaluation**: Strong selection ability in large-cap pools, with stable performance across different market conditions[39][113] Factor Backtesting Results 1. **Valuation (BTOP)**: - RankICIR: Consistently ranks in the top 2 across small-cap pools[39][98] 2. **Volatility (VOLATILITY)**: - RankICIR: Demonstrates stable negative expression across all pools, with low failure probabilities[50][98] 3. **Earnings (EARNING)**: - RankICIR: Strong performance in large-cap pools, with high selection ability and stable expression[39][113]

高频选股因子周报（20251208- 20251212） 高频因子走势分化，多粒度因子显著回撤。AI 增强组合均 大幅度回撤。 本报告导读： 上周（特指 20251208-20251212，下同）高频因子走势分化，多粒度因子显著回撤。 AI 增强组合均大幅度回撤。 投资要点： | | | | [Table_Authors] | 郑雅斌(分析师) | | --- | --- | | | 021-23219395 | | | zhengyabin@gtht.com | | 登记编号 | S0880525040105 | | | 余浩淼(分析师) | | | 021-23185650 | | | yuhaomiao@gtht.com | | 登记编号 | S0880525040013 | [Table_Report] 相关报告 低频选股因子周报（2025.12.05-2025.12.12） 2025.12.13 绝对收益产品及策略周报（251201-251205） 2025.12.10 上周估值因子表现较好，本年中证 2000 指数增强 策略超额收益为 28.22% 2025.12.10 红利风格择时周报（1201 ...

Core Insights - The report indicates that small-cap and value factors experienced a pullback, while high profitability and high growth factors performed relatively well [1] - The quant stock portfolio of top-performing funds achieved a weekly return of 4.43%, with a cumulative return of 52.54% for 2025 [1] Group 1: Multi-Factor Portfolio Performance - The aggressive and balanced portfolios had weekly returns of -4.10% and -3.85% respectively, underperforming the major indices [10] - For the year-to-date (YTD) 2025, the aggressive and balanced portfolios recorded cumulative returns of 69.47% and 55.27%, significantly outperforming the major indices [11] Group 2: Fund Holdings Performance - The exclusive holdings of top-performing funds yielded a weekly return of 4.43%, outperforming the total index of stock funds by 4.09% [26] - Since December 2025, these holdings have achieved a cumulative return of 7.58%, with an excess return of 6.65% [26] Group 3: Profitability, Growth, and Cash Flow Combination - The combination of profitability, growth, and cash flow achieved a weekly return of 1.12%, outperforming the CSI 300 index by 1.20% [28] - For 2025, this combination has a cumulative return of 88.82%, significantly higher than the CSI 300 index's return of 16.42% [28] Group 4: Low Valuation with Fundamental Support - The PB-profitability preferred portfolio had a weekly return of -2.64%, underperforming the CSI 300 index by 2.57% [30] - For the year-to-date 2025, this portfolio recorded a cumulative return of 19.82%, slightly outperforming the CSI 300 index [31] Group 5: Small-Cap Value and Growth Performance - The small-cap value preferred portfolio 1 had a weekly return of -2.84%, outperforming the micro-cap index by 1.85% [35] - The small-cap growth portfolio recorded a weekly return of -1.94%, outperforming the micro-cap index by 2.75% [39] Group 6: Single Factor Performance - In style factors, large-cap stocks outperformed small-cap stocks, and high-valuation stocks outperformed low-valuation stocks [42] - Technical factors showed negative excess returns across the board, with reversal and turnover factors contributing negatively [46] Group 7: Fundamental Factors - The ROE factor contributed positively, with a multi-factor return of 1.63% for the week [53] - The SUE factor also showed positive returns, indicating strong performance in fundamental analysis [53]

- The report introduces a method for calculating the dividend points of stock indices, which is crucial for accurately estimating the premium or discount of stock index futures contracts. The formula for dividend points is as follows: $ \text{Dividend Points} = \sum_{n=1}^{N} \left( \frac{\text{Dividend Amount of Component Stock}}{\text{Total Market Value of Component Stock}} \times \text{Weight of Component Stock} \times \text{Index Closing Price} \right) $ This calculation considers only the component stocks with ex-dividend dates between the current date (t) and the futures contract expiration date (T) [41] - The weight of component stocks is dynamically adjusted to reflect daily changes. The formula for calculating the weight is: $ W_{n,t} = \frac{w_{n0} \times (1 + r_{n})}{\sum_{i=1}^{N} w_{i0} \times (1 + r_{i})} $ Here, $w_{n0}$ is the weight of stock n on the last disclosed date, and $r_{n}$ is the non-adjusted return of stock n from the last disclosed date to the current date [45] - The estimation of dividend amounts involves predicting net profits and dividend payout ratios. The dividend amount is calculated as: $ \text{Dividend Amount} = \text{Net Profit} \times \text{Dividend Payout Ratio} $ For companies with stable profit distributions, historical patterns are used for prediction. For others, the previous year's profit is used as the estimate [47][50] - The dividend payout ratio is estimated using historical averages. If a company paid dividends in the previous year, the last year's ratio is used. If not, the average of the past three years is applied. If no historical data exists, the company is assumed not to pay dividends [51][53] - The ex-dividend date is predicted using a linear extrapolation method based on the stability of historical intervals between announcement dates and ex-dividend dates. If no reliable historical data is available, default dates are assigned based on typical dividend schedules [56] - The accuracy of the dividend point estimation model is evaluated. For the Shanghai 50 and CSI 300 indices, the annual prediction error is approximately 5 points, while for the CSI 500 index, the error is around 10 points. The model demonstrates high accuracy for predicting dividend points of stock index futures contracts [57][61]

Quantitative Models and Construction Barra Risk Model - **Model Name**: Barra Risk Model (Barra CNE6) - **Model Construction Idea**: The model aims to reduce the dimensionality of asset returns, enabling the calculation of covariance matrices between assets, which are essential for portfolio optimization. It uses constrained weighted least squares (WLS) to address multicollinearity and heteroscedasticity issues, constructing pure factor portfolios that isolate exposure to individual factors [3][9][11] - **Model Construction Process**: 1. The cross-sectional asset returns are modeled using a multi-factor linear regression: $R_{t}=\alpha+\beta\lambda_{t}+\varepsilon_{t}$ Here, $\beta$ represents factor exposures, $\lambda_{t}$ denotes factor returns, and $\varepsilon_{t}$ is the residual [9][10] 2. The covariance matrix of asset returns is derived as: $\Sigma_{R}=\beta\Sigma_{A}\beta^{T}+\Sigma_{E}$ $\Sigma_{A}$ is the covariance matrix of factors, and $\Sigma_{E}$ is the covariance matrix of residuals [11][12] 3. Factor exposures are standardized using market capitalization-weighted normalization: $$\widehat{\boldsymbol{\beta}_{t-1}^{j}}=\frac{{\boldsymbol{\beta}_{t-1}^{j}}-\frac{\sum_{i}^{N}s_{i,t-1}\beta_{i,t-1}^{j}}{\sum_{i}^{N}s_{i,t-1}}}{s t d({\boldsymbol{\beta}_{t-1}^{j}})}$$ Here, $s_{i,t-1}$ represents the market capitalization of stock $i$ at time $t-1$ [18][32] 4. Industry factor returns are constrained to ensure neutrality: $\sum_{i=1}^{P}s_{I_{i}}\lambda_{i}^{I_{i}}=0$ [18][22] 5. Factor returns are estimated using constrained WLS: $$\lambda_{t}=C_{t}(C_{t}\beta_{t-1}W^{-1}\beta_{t-1}C_{t})^{-1}C_{t}\beta_{t-1}W^{-1}R_{t}$$ Here, $W$ is the weight matrix, and $C_{t}$ represents constraints [20][25] - **Model Evaluation**: The model effectively isolates factor exposures, enabling better evaluation of factor returns. However, pure factor portfolios have low investability due to constraints like short-selling limitations [19][21] --- Quantitative Factors and Construction Style Factors - **Factor Names**: Size, Volatility, Liquidity, Momentum, Quality, Value, Growth, Dividend Yield - **Factor Construction Idea**: These factors represent different market characteristics, such as size, volatility, and growth, and are used to explain asset returns and identify systematic risks [3][15][26] - **Factor Construction Process**: 1. **Size**: Logarithm of market capitalization (LNCAP) [114] 2. **Volatility**: Includes Beta, historical sigma, daily standard deviation, and cumulative range [114] 3. **Liquidity**: Calculated using turnover ratios (monthly, quarterly, annual) and annualized traded value ratio [114] 4. **Momentum**: Includes short-term reversal, seasonality, industry momentum, and relative strength [114][115] 5. **Quality**: Includes earnings variability, accruals, profitability metrics, and investment quality [114][116] 6. **Value**: Includes book-to-price ratio, earnings-to-price ratio, and enterprise multiple [114][116] 7. **Growth**: Historical growth rates for earnings per share and sales per share [114][116] 8. **Dividend Yield**: Dividend-to-price ratio [114][116] - **Factor Evaluation**: Single-factor tests show limited stock selection ability, with low significance and effectiveness. However, after constructing pure factor models, the significance of factors improves, especially for Volatility and Momentum [66][78] Residual Factor - **Factor Name**: Residual Factor - **Factor Construction Idea**: Residuals represent the unexplained portion of stock returns after accounting for industry, style, and country factors. They are tested for nonlinear relationships with stock returns [79][82] - **Factor Construction Process**: 1. Residuals are derived from the regression model: $R_{t}=\beta_{t-1}C_{t}\gamma_{t}+\delta_{t}$ Here, $\delta_{t}$ represents residuals [23][79] 2. Residuals are used as stock selection factors and tested using layered backtesting [79][82] - **Factor Evaluation**: Residual factors exhibit strong nonlinear relationships with stock returns, showing robust stock selection ability. Middle-layer groups outperform top and bottom groups significantly [79][82] --- Backtesting Results Pure Factor Model - **Size**: Annualized return -2.75%, annualized volatility 0.026, maximum drawdown 35.53%, Sharpe ratio -1.08 [76][77] - **Volatility**: Annualized return 1.93%, annualized volatility 0.049, maximum drawdown 12.43%, Sharpe ratio 0.39 [76][77] - **Liquidity**: Annualized return -5.90%, annualized volatility 0.033, maximum drawdown 60.88%, Sharpe ratio -1.81 [76][77] - **Momentum**: Annualized return -5.57%, annualized volatility 0.042, maximum drawdown 58.64%, Sharpe ratio -1.32 [76][77] - **Growth**: Annualized return -0.21%, annualized volatility 0.015, maximum drawdown 9.24%, Sharpe ratio -0.15 [76][77] - **Dividend Yield**: Annualized return -0.85%, annualized volatility 0.016, maximum drawdown 17.09%, Sharpe ratio -0.52 [76][77] - **Quality**: Annualized return 0.35%, annualized volatility 0.016, maximum drawdown 8.45%, Sharpe ratio 0.23 [76][77] - **Value**: Annualized return 1.38%, annualized volatility 0.028, maximum drawdown 13.83%, Sharpe ratio 0.49 [76][77] Residual Factor - **Middle Layer (Group 5)**: Annualized return 17.98%, annualized volatility 26.94%, Sharpe ratio 0.68, maximum drawdown 52.50% [82] - **Top vs Bottom Layer (Group 5 vs Group 10)**: Excess annualized return 13.58%, excess Sharpe ratio 1.50 [82] --- Index Attribution Results Positive Excess Return Indices - **Indices**: CSI 500 (3.41%), ChiNext Index (18.23%) - **Key Drivers**: Small-cap, high volatility, low liquidity, high growth, low dividend yield styles; leading sectors include non-ferrous metals, electronics, communication, and new energy [101][110] Negative Excess Return Indices - **Indices**: CSI 1000 (-0.22%), CSI A500 (-1.60%), CSI 300 (-4.30%), SSE 50 (-10.27%) - **Key Drivers**: Large-cap, low volatility, high liquidity, low growth, high dividend yield styles; underperforming sectors include banking, non-bank finance, and food & beverage [101][110]

- The report highlights the performance of the A-share market, with the CSI A50 leading the gains and the ChiNext Index showing strong performance[1][6] - The Dividend Enhanced Portfolio outperformed the CSI Dividend Total Return Index, with the Central SOE High Dividend 30 Portfolio and the Balanced Dividend 50 Portfolio achieving weekly excess returns of approximately 0.41% and 0.75%, respectively[6][21] - The Electronic Enhanced Portfolio also outperformed the Electronic Total Return Index, with the Electronic Balanced Allocation Enhanced Portfolio and the Electronic Sector Preferred Enhanced Portfolio achieving weekly excess returns of approximately 1.78% and 1.53%, respectively[6][29]

Group 1 - The report emphasizes the importance of identifying trend strength through moving average strategies, which can help in recognizing market trends effectively [3][4][9] - It introduces three key moving average indicators: moving average arrangement score, moving average dispersion distance, and moving average time series change, which collectively help in assessing market trends [8][10][13] - The report suggests that a higher composite score indicates a stronger upward trend, while a lower score suggests a stronger downward trend [13][21] Group 2 - The report outlines a funding flow strategy that aggregates various funding flow indicators to identify market consensus, focusing on both institutional and retail investor behaviors [26][28] - It highlights the significance of funding flow volatility over mere funding direction, suggesting that stable funding behavior can indicate potential market reversals [31][41] - The report proposes a combined approach of trend strength and funding consensus to select indices with clear trends and stable funding, enhancing investment decision-making [42][39] Group 3 - The report presents historical performance data for industry rotation and ETF rotation strategies, showing annual returns and excess returns compared to equal-weighted benchmarks [50][52] - It indicates that the industry rotation strategy has shown significant excess returns in certain years, particularly in 2020 with a return of 76.84% [50] - The ETF rotation strategy also demonstrated strong performance in 2019 and 2020, with excess returns of 30.33% and 26.62% respectively [52]

Quantitative Models and Construction Methods 1. Model Name: Multi-Cycle Timing Model for Domestic Interest Rates - **Model Construction Idea**: The model uses kernel regression algorithms to identify support and resistance lines of interest rate trends. It evaluates the breakthrough patterns of interest rate movements across different investment cycles to form multi-cycle composite timing signals[10][24]. - **Model Construction Process**: - **Data Input**: Yield-to-Maturity (YTM) data for 5-year, 10-year, and 30-year government bonds[6][10]. - **Cycle Classification**: - Long cycle: Monthly frequency - Medium cycle: Bi-weekly frequency - Short cycle: Weekly frequency[10][21]. - **Signal Generation**: - A signal is generated when at least two cycles show consistent directional breakthroughs (upward or downward). - For example, for the 5-year YTM, the current signal is "bearish" as both the long and medium cycles show upward breakthroughs, while the short cycle shows no signal[10]. - **Scoring Mechanism**: - Each cycle contributes one "vote" for upward or downward breakthroughs. - A composite score is calculated based on the total votes, and the final signal is determined[10][13][17]. - **Model Evaluation**: The model effectively captures interest rate trends and provides actionable timing signals for different bond maturities[10][24]. 2. Model Name: Multi-Cycle Timing Model for US Interest Rates - **Model Construction Idea**: The domestic timing model is applied to the US Treasury market to generate timing signals for 10-year US Treasury YTM[21]. - **Model Construction Process**: - **Data Input**: 10-year US Treasury YTM data[21]. - **Cycle Classification**: Same as the domestic model (long, medium, and short cycles)[21]. - **Signal Generation**: - The current signal is "neutral" as only the short cycle shows an upward breakthrough, while the long and medium cycles show no signal[21]. - **Model Evaluation**: The model demonstrates adaptability to international markets, providing consistent timing signals for US Treasuries[21]. --- Model Backtesting Results 1. Multi-Cycle Timing Model for Domestic Interest Rates - **5-Year YTM**: - Long-term annualized return: 5.48% - Maximum drawdown: 2.88% - Return-to-drawdown ratio: 1.91 - Short-term annualized return (since end-2024): 2.11% - Maximum drawdown: 0.59% - Return-to-drawdown ratio: 3.57 - Long-term excess return: 1.07% - Short-term excess return: 0.87%[6][28][29]. - **10-Year YTM**: - Long-term annualized return: 6.06% - Maximum drawdown: 2.74% - Return-to-drawdown ratio: 2.21 - Short-term annualized return (since end-2024): 2.39% - Maximum drawdown: 0.58% - Return-to-drawdown ratio: 4.14 - Long-term excess return: 1.65% - Short-term excess return: 1.36%[28][29]. - **30-Year YTM**: - Long-term annualized return: 7.34% - Maximum drawdown: 4.27% - Return-to-drawdown ratio: 1.72 - Short-term annualized return (since end-2024): 3.03% - Maximum drawdown: 0.92% - Return-to-drawdown ratio: 3.31 - Long-term excess return: 2.43% - Short-term excess return: 2.97%[28][29][33]. 2. Multi-Cycle Timing Model for US Interest Rates - The report does not provide specific backtesting results for the US model, but the current signal is "neutral" based on the latest data[21]. --- Quantitative Factors and Construction Methods 1. Factor Name: Interest Rate Structure Indicators (Level, Term, Convexity) - **Factor Construction Idea**: Transform YTM data into structural indicators (level, term, and convexity) to analyze the interest rate market from a mean-reversion perspective[7]. - **Factor Construction Process**: - **Level Structure**: Represents the average interest rate level. - Current value: 1.63% - Historical percentiles: 25% (3 years), 15% (5 years), 7% (10 years)[7]. - **Term Structure**: Represents the slope of the yield curve. - Current value: 0.45% - Historical percentiles: 34% (3 years), 21% (5 years), 23% (10 years)[7]. - **Convexity Structure**: Represents the curvature of the yield curve. - Current value: 0.01% - Historical percentiles: 26% (3 years), 16% (5 years), 13% (10 years)[7]. - **Factor Evaluation**: These indicators provide a comprehensive view of the interest rate market's structural characteristics, aiding in timing and allocation decisions[7]. --- Factor Backtesting Results - **Interest Rate Structure Indicators**: - The report does not provide specific backtesting results for these factors, but their historical percentiles indicate their relative positioning in the market[7].