证券研究报告 | 金融工程 gszqdatemark 2026 02 28 年 月 日 量化分析报告 择时雷达六面图：本周各项指标分数基本无变化 择时雷达六面图：基于多维视角的择时框架。权益市场的表现受到多维度 指标因素的共同影响，我们尝试从流动性、经济面、估值面、资金面、技 术面、拥挤度选取二十一个指标对市场进行刻画，并将其概括为"估值性 价比"、"宏观基本面"、"资金&趋势"、"拥挤度&反转"四大类，从而生成 [-1,1]之间的综合择时分数。 本周综合打分。本周市场的估值性价比分数略微下降，趋势&资金分数、 拥挤度&反转分数、宏观基本面分数均不变，综合打分位于[-1,1]之间，当 前的综合打分为-0.25 分，整体为中性观点。当前六面图各个维度的观点 如下： 流动性。本周货币方向发出看多信号，信用强度、货币强度信号中性， 信用方向发出看空信号，当前流动性得分为 0.00 分，综合来看信号中性。 经济面。本周增长强度、增长方向、通胀方向发出看空信号，通胀强度 信号中性，当前经济面综合得分为-0.75 分，综合来看信号偏空。 估值面。本周席勒 ERP 指标下降 0.05 分，PB 指标下降 0.10 分，AI ...

在当前快速变化的外汇市场中，提升决策的准确性和效率是参与者追求的核心目标。FXGT通过将先进的量化分析技术与市场洞察相结合，为个人及机构提 供了一套强大的决策支持体系。该体系旨在减少情绪干扰，提高策略的稳定性和执行效率，帮助用户更好地应对市场波动，捕捉潜在机会，实现更稳健的市 场参与。 FXGT深刻理解市场参与者对高效、客观决策工具的需求，致力于整合前沿的量化技术为用户赋能。其核心思路在于通过建立严谨的数理逻辑模型，对历史 及实时的市场数据进行深度挖掘和模式识别： 1. 信息处理与模式识别：系统能够高速处理新闻、经济指标、价格波动等结构化和非结构化数据，运用算法过滤噪音，识别可能被忽略的市场趋势和周期性 规律。 2. 策略回测与优化：在真实资金介入前，用户可以在安全的环境中，利用历史数据对预设的策略逻辑进行模拟验证。这有助于评估策略在不同市场环境下的 表现，发现潜在弱点，并进行迭代优化，从而提高策略的鲁棒性。 3. 风险管理的量化支撑：将风险管理原则量化，设置明确的、基于数据分析的参数（如动态调整的应对区间等）。这有助于用户更系统化地控制潜在波动带 来的影响，使操作过程更具纪律性。 外汇市场以规模庞大、流动性 ...

- The "Timing Radar Six-Facet Chart" is a multi-dimensional timing framework that evaluates equity market performance based on liquidity, economic conditions, valuation, capital flow, technical trends, and crowding metrics. It aggregates 21 indicators into four categories: "Valuation Cost-Effectiveness," "Macro Fundamentals," "Capital & Trend," and "Crowding & Reversal," generating a composite timing score ranging from [-1,1] [1][6][8] - Liquidity factors include "Monetary Direction Factor," which calculates the average change in monetary policy rates and short-term market rates over 90 days. A positive value indicates monetary easing, scoring 1 this week [10][11]. The "Monetary Strength Factor" uses the deviation of DR007 from the 7-year reverse repo rate, smoothed and standardized. This week, it scored 0, indicating a neutral signal [12][13]. The "Credit Direction Factor" evaluates the trend of medium-to-long-term loans over 12 months. A positive trend scores 1, which was observed this week [15][16]. The "Credit Strength Factor" measures the deviation of new RMB loans from expectations. This week, it scored -1, signaling a significant underperformance [18][19] - Economic factors include "Growth Direction Factor," derived from PMI data (manufacturing and non-manufacturing), calculating the 12-month average and year-over-year change. A downward trend scored -1 this week [20][21]. The "Growth Strength Factor" measures PMI deviations from expectations. This week, it scored -1, indicating significant underperformance [23][24]. The "Inflation Direction Factor" combines CPI and PPI data to assess inflation trends. An upward trend scored -1 this week [25][26]. The "Inflation Strength Factor" calculates CPI and PPI deviations from expectations. This week, it scored 0, indicating a neutral signal [27][31] - Valuation factors include "Shiller ERP," calculated as 1/Shiller PE minus the 10-year government bond yield, standardized over six years. This week, it scored -0.04, slightly improving [28][29]. The "PB Factor" uses the price-to-book ratio, standardized over six years, and scored -0.66 this week, showing a slight increase [32][33]. The "AIAE Factor" measures equity allocation relative to total market capitalization and debt, standardized over six years. This week, it remained at -1.00 [34][35] - Capital flow factors include "Margin Financing Increment," which compares the 120-day average increase in margin financing to the 240-day average. A positive trend scored 1 this week [37][39]. The "Trading Volume Trend" evaluates the logarithmic moving average distance of trading volume over 120 and 240 days. A positive trend scored 1 this week [40][42]. External capital flow indicators include "China Sovereign CDS Spread," which assesses the 20-day differential of CDS spreads. A positive differential scored -1 this week [44][45]. The "Overseas Risk Aversion Index" evaluates the 20-day differential of Citi RAI Index. A positive differential scored -1 this week [46][47] - Technical factors include "Price Trend," which uses moving average distances (120/240 days) to assess market trends and strength. This week, it scored 1, indicating a positive trend [49][50]. The "New Highs and Lows" metric evaluates the difference between new highs and lows among index constituents over the past year. A positive difference scored -1 this week [51][53] - Crowding metrics include "Option Implied Premium," derived from the implied premium in options markets. A positive signal scored -1 this week [55][59]. The "Option VIX Index" measures implied volatility expectations. A positive signal scored 1 this week [56][57]. The "Option SKEW Index" evaluates implied skewness in options markets. A positive signal scored -1 this week [60][62]. The "Convertible Bond Pricing Deviation" measures the deviation of convertible bond prices from model estimates, standardized over three years. A positive deviation scored -1 this week [63][64] - Composite timing score for this week is -0.30, indicating a neutral-to-bearish outlook [6][8][9]

- The "Timing Radar Six-Facet Chart" model is constructed based on multi-dimensional indicators including liquidity, economic fundamentals, valuation, capital flow, technical trends, and crowding sentiment. It aggregates 21 indicators into four categories: "Valuation Cost-Effectiveness," "Macroeconomic Fundamentals," "Capital & Trend," and "Crowding & Reversal," generating a comprehensive timing score ranging from [-1, 1] to assess market conditions [1][6][9] - **Liquidity Factors**: - **Monetary Direction Factor**: Calculated using the average change in central bank monetary policy tool rates and short-term market rates over the past 90 days. If the factor > 0, monetary policy is deemed expansionary; if < 0, it is deemed contractionary. Current score: 1 [12][14] - **Monetary Strength Factor**: Based on the "interest rate corridor" concept, calculated as deviation = DR007/7-year reverse repo rate - 1, smoothed and standardized using z-score. If the factor < -1.5 standard deviations, it indicates a 120-day future easing environment, scoring 1; if > 1.5 standard deviations, scoring -1. Current score: 0 [15][16] - **Credit Direction Factor**: Derived from monthly long-term loan data, calculated as the year-on-year change in the past 12 months' increment. If the factor rises compared to three months ago, it scores 1; otherwise, -1. Current score: 1 [18][20] - **Credit Strength Factor**: Captures whether credit metrics significantly exceed or fall short of expectations, calculated as (new RMB loans - median forecast)/forecast standard deviation. If the factor > 1.5 standard deviations, it scores 1; if < -1.5 standard deviations, it scores -1. Current score: -1 [22][23] - **Economic Factors**: - **Growth Direction Factor**: Based on PMI data (manufacturing, non-manufacturing, and Caixin manufacturing PMI), calculated as the year-on-year change in the 12-month average. If the factor rises compared to three months ago, it scores 1; otherwise, -1. Current score: -1 [26][28] - **Growth Strength Factor**: Captures whether growth metrics significantly exceed or fall short of expectations, calculated as (PMI - median forecast)/forecast standard deviation. If the factor > 1.5 standard deviations, it scores 1; if < -1.5 standard deviations, it scores -1. Current score: 1 [29][31] - **Inflation Direction Factor**: Calculated as 0.5 × smoothed CPI year-on-year + 0.5 × raw PPI year-on-year. If the factor decreases compared to three months ago, it scores 1; otherwise, -1. Current score: -1 [32][33] - **Inflation Strength Factor**: Captures whether inflation metrics significantly exceed or fall short of expectations, calculated as the average of CPI and PPI forecast deviations. If the factor < -1.5, it scores 1; if > 1.5 standard deviations, it scores -1. Current score: -1 [35][37] - **Valuation Factors**: - **Shiller ERP**: Calculated as 1/Shiller PE - 10-year government bond yield, standardized using z-score over the past six years. Current score: -0.10 [38][42] - **PB**: Processed as PB × (-1), standardized using z-score over the past six years, truncated at ±1.5 standard deviations. Current score: -0.75 [40][41] - **AIAE**: Represents aggregate investor allocation to equities, calculated as total market cap/(total market cap + total debt), standardized using z-score over the past six years. Current score: -1.00 [43][44] - **Capital Flow Factors**: - **Margin Financing Increment**: Calculated as the difference between 120-day average increment and 240-day average increment of margin financing balance. If the short-term increment exceeds the long-term increment, it scores 1; otherwise, -1. Current score: 1 [46][48] - **Turnover Trend**: Calculated as log turnover moving average distance = ma120/ma240 - 1. If the maximum distance of 10, 30, and 60 days is positive, it scores 1; if the minimum distance is negative, it scores -1. Current score: 1 [49][50] - **China Sovereign CDS Spread**: Represents overseas investors' pricing of China's economic and sovereign credit risk. If the smoothed 20-day difference of CDS spread < 0, it scores 1; otherwise, -1. Current score: 1 [52][54] - **Overseas Risk Aversion Index**: Captures overseas market risk preference using Citi RAI Index. If the smoothed 20-day difference < 0, it scores 1; otherwise, -1. Current score: -1 [55][56] - **Technical Factors**: - **Price Trend**: Calculated as moving average distance (ma120/ma240 - 1). Trend direction scores 1 if > 0, otherwise -1. Trend strength scores 1 if max(20) = max(60), otherwise -1. Comprehensive score = (direction score + strength score)/2. Current score: 1 [57][59] - **New Highs and Lows**: Calculated as the moving average of the difference between new lows and new highs among index constituents over the past year. If the smoothed difference > 0, it scores -1; otherwise, 1. Current score: -1 [60][62] - **Crowding Sentiment Factors**: - **Option Implied Premium**: Derived from the implied premium of options based on put-call parity. If 50ETF 5-day return > 0 and percentile > 70%, it scores -1; otherwise, 1. Current score: -1 [64][68] - **Option VIX**: Reflects expected volatility of options. If 50ETF 5-day return > 0 and percentile > 70%, it scores -1; otherwise, 1. Current score: -1 [65][67] - **Option SKEW**: Reflects expected skewness of options. If 50ETF 5-day return > 0 and percentile < 30%, it scores -1; otherwise, 1. Current score: -1 [69][70] - **Convertible Bond Pricing Deviation**: Calculated as (convertible bond price/model price - 1), standardized using z-score over the past three years. Higher deviation indicates higher crowding sentiment, scoring lower. Current score: -1 [71][72] - **Factor Testing Results**: - Liquidity factors: Monetary direction (1), Monetary strength (0), Credit direction (1), Credit strength (-1) [12][15][18][22] - Economic factors: Growth direction (-1), Growth strength (1), Inflation direction (-1), Inflation strength (-1) [26][29][32][35] - Valuation factors: Shiller ERP (-0.10), PB (-0.75), AIAE (-1.00) [38][40][43] - Capital flow factors: Margin financing increment (1), Turnover trend (1), China Sovereign CDS spread (1), Overseas risk aversion index (-1) [46][49][52][55] - Technical factors: Price trend (1), New highs and lows (-1) [57][60] - Crowding sentiment factors: Option implied premium (-1), Option VIX (-1), Option SKEW (-1), Convertible bond pricing deviation (-1) [64][65][69][71]

Core Insights - The article discusses the emergence of THPX signal source, which utilizes advanced quantitative analysis techniques to provide actionable predictive signals for the XAGBTC market, characterized by its complexity and high volatility [1][7]. Group 1: THPX Signal Source Overview - THPX is not merely a simple indicator but a comprehensive system that integrates data capture, intelligent processing, and forward-looking analysis to enhance users' understanding of XAGBTC's future trends [1][8]. - The system operates through three stages: multi-dimensional signal capture, intelligent processing and fusion, and forward-looking judgment output [3][5]. Group 2: Multi-Dimensional Signal Capture - THPX scans market data streams at a high frequency to extract potential value-changing signal factors, which may arise from subtle price movements or statistical anomalies within specific time windows [3]. Group 3: Intelligent Processing and Fusion - Captured signals undergo complex processing using advanced machine learning algorithms that automatically learn patterns, periodicities, and correlation rules from historical data, focusing on core information streams with predictive value [3][5]. Group 4: Forward-Looking Judgment Output - The processed information is transformed into clear, actionable market trend predictions, highlighting potential market directions, momentum changes, and key turning points [5][8]. - THPX enhances efficiency by automating data processing and complex calculations, significantly reducing the time required to convert information into insights [5][8]. Group 5: Value Proposition of THPX - THPX provides deeper insights by identifying subtle market patterns and early trend signals that traditional tools may overlook, allowing users to anticipate potential market movements [5][8]. - The clear and timely signal outputs serve as important reference points for decision-making, helping users align with market rhythms and maintain sensitivity during critical moments [5][8]. Group 6: Technological Empowerment - THPX represents a cutting-edge practice in market analysis empowered by modern quantitative technology, offering a scientific and systematic framework for understanding complex markets like XAGBTC [7][8]. - The development of THPX reflects a significant trend in financial technology, where market analysis is transitioning from empirical methods to highly intelligent and systematic approaches driven by big data and artificial intelligence [7][8].

Core Insights - The article discusses the application value of the THPX signal source in analyzing the silver (XAG) and gold (XAU) markets in conjunction with Bitcoin, leveraging advanced AI technology and quantitative models to provide forward-looking strategy references for market participants [1][7] - The system integrates multi-dimensional indicators and dynamic learning capabilities to effectively identify market characteristics and cyclical patterns, significantly enhancing the adaptability and stability of strategies in complex market conditions [1][7] Group 1: Data Processing Architecture - The core of the solution lies in a multi-layered data processing architecture, with the first layer focusing on deep integration of cross-market data sources, including real-time volatility data of precious metals and Bitcoin, as well as global macroeconomic indicators and market sentiment indices [3] - The second layer deploys adaptive feature engineering modules that utilize recurrent neural networks and convolutional networks to automatically identify cross-market correlation features, capturing key driving factors such as risk sentiment transmission and capital flow trends [3] Group 2: Dynamic Modeling Strategies - The system implements dynamic modeling strategies tailored to the different market characteristics of XAG/BTC and XAU/BTC, with the silver combination focusing on industrial attributes and inflation resilience, while the gold combination emphasizes its financial attributes and risk-hedging consensus [3][5] - This differentiated modeling mechanism ensures that signal generation aligns closely with the intrinsic logic of the target assets [3] Group 3: Validation Mechanism - The dual-engine validation mechanism is a technical highlight, tracking market state transition patterns through LSTM and constructing a feature importance evaluation matrix using ensemble learning methods to optimize factor weights during each strategy iteration [5] - The system features a real-time feedback loop that automatically adjusts parameters when actual market movements deviate from predicted thresholds, maintaining sensitivity to market structure changes [5] Group 4: Practical Application and Advantages - Practical application validation shows that the solution demonstrates unique advantages in handling cross-market linkage trends, capturing signs of capital migration from Bitcoin to the silver market during periods of increased volatility in precious metals [7] - The THPX signal source solution establishes effective analytical pathways for special market combinations like XAG/BTC and XAU/BTC, overcoming the limitations of traditional analytical tools in cross-market correlation [7]

- Model Name: Timing Radar Six-Factor Model; Model Construction Idea: The model attempts to capture the performance of the equity market through multiple dimensions, including liquidity, economic conditions, valuation, capital flows, technical indicators, and crowding; Model Construction Process: The model selects 21 indicators from the aforementioned dimensions and categorizes them into four major categories: "Valuation Cost-Effectiveness," "Macro Fundamentals," "Capital & Trend," and "Crowding & Reversal." These indicators are then used to generate a comprehensive timing score ranging from [-1,1]; Model Evaluation: The model provides a comprehensive view of market conditions from multiple perspectives, making it a useful tool for market timing[1][6][9] - Factor Name: Liquidity Direction Factor; Factor Construction Idea: This factor aims to determine the direction of current monetary policy; Factor Construction Process: The factor is calculated using the average change in central bank policy rates and short-term market rates over the past 90 days. If the factor is greater than 0, it indicates a loose monetary policy, and if less than 0, it indicates a tight monetary policy; Current View: This week, the liquidity direction factor is greater than 0, signaling a bullish outlook, with a score of 1[11][13] - Factor Name: Liquidity Strength Factor; Factor Construction Idea: This factor is based on the "interest rate corridor" concept to measure the deviation of short-term market rates from policy rates; Factor Construction Process: The factor is calculated as the deviation = DR007/7-year reverse repo rate - 1, smoothed and z-scored. If the factor is less than -1.5 standard deviations, it indicates a loose environment for the next 120 trading days, with a score of 1. Conversely, if greater than 1.5 standard deviations, the score is -1; Current View: This week, the liquidity strength factor score is 0, indicating a neutral signal[14][15] - Factor Name: Credit Direction Factor; Factor Construction Idea: This factor measures the tightness of credit transmission from commercial banks to the real economy; Factor Construction Process: The factor is calculated using the monthly value of medium- and long-term loans, the incremental value over the past 12 months, and the year-over-year change. If the factor is rising compared to three months ago, it is bullish, with a score of 1. Conversely, if falling, the score is -1; Current View: This week, the credit direction factor is identified as an upward trend, signaling a bullish outlook, with a score of 1[17][19] - Factor Name: Credit Strength Factor; Factor Construction Idea: This factor captures whether credit indicators significantly exceed or fall short of expectations; Factor Construction Process: The factor is calculated as (new RMB loans for the month - median forecast) / forecast standard deviation. If the factor is greater than 1.5 standard deviations, it indicates a significantly above-expectation credit environment for the next 60 trading days, with a score of 1. Conversely, if less than -1.5 standard deviations, the score is -1; Current View: This week, the credit strength factor score is -1[20][22] - Factor Name: Growth Direction Factor; Factor Construction Idea: This factor is based on PMI data to measure economic growth direction; Factor Construction Process: The factor is calculated using the 12-month average of PMI values and their year-over-year change. If the factor is rising compared to three months ago, it is bullish, with a score of 1. Conversely, if falling, the score is -1; Current View: This week, the growth direction factor is identified as an upward trend, signaling a bullish outlook, with a score of 1[22][24] - Factor Name: Growth Strength Factor; Factor Construction Idea: This factor captures whether economic growth indicators significantly exceed or fall short of expectations; Factor Construction Process: The factor is calculated as (PMI - median forecast) / forecast standard deviation. If the factor is greater than 1.5 standard deviations, it indicates a significantly above-expectation growth environment for the next 60 trading days, with a score of 1. Conversely, if less than -1.5 standard deviations, the score is -1; Current View: This week, the growth strength factor score is -1, signaling a bearish outlook[25][26] - Factor Name: Inflation Direction Factor; Factor Construction Idea: This factor measures the current inflation level's impact on monetary policy; Factor Construction Process: The factor is calculated as 0.5 × smoothed CPI year-over-year value + 0.5 × raw PPI year-over-year value. If the factor is lower compared to three months ago, it indicates a downward inflation environment, with a score of 1. Conversely, if higher, the score is -1; Current View: This week, the inflation direction factor is identified as an upward trend, signaling a bearish outlook, with a score of -1[27][28] - Factor Name: Inflation Strength Factor; Factor Construction Idea: This factor captures whether inflation indicators significantly exceed or fall short of expectations; Factor Construction Process: The factor is calculated as the average of CPI and PPI forecast deviations, where forecast deviation = (reported value - median forecast) / forecast standard deviation. If the factor is less than -1.5, it indicates a significantly below-expectation inflation environment for the next 60 trading days, with a score of 1. Conversely, if greater than 1.5 standard deviations, the score is -1; Current View: This week, the inflation strength factor score is -1, signaling a bearish outlook[29][33] - Factor Name: Shiller ERP; Factor Construction Idea: This factor adjusts for economic cycle fluctuations in corporate earnings to assess current market valuation levels; Factor Construction Process: The factor is calculated as Shiller ERP = 1 / Shiller PE - 10-year government bond yield, with the past 6 years' z-score used as the score; Current View: This week, the Shiller ERP has risen, with the score increasing to 0.06[30][31] - Factor Name: PB; Factor Construction Idea: This factor adjusts the PB indicator similarly to ERP; Factor Construction Process: The factor is calculated as PB × (-1) and z-scored over the past 6 years, standardized to ±1 after 1.5 standard deviation truncation; Current View: This week, the PB has declined, with the score rising to -0.38[34][35] - Factor Name: AIAE; Factor Construction Idea: This factor reflects the overall market risk preference based on the aggregate investor allocation to equities; Factor Construction Process: The factor is calculated as AIAE = total market cap of CSI All Share Index / (total market cap of CSI All Share Index + total debt of the real economy), with the past 6 years' z-score used as the score; Current View: This week, the AIAE has declined, with the score rising to -0.72[36][37] - Factor Name: Margin Trading Increment; Factor Construction Idea: This factor measures market sentiment based on the source of leveraged funds; Factor Construction Process: The factor is calculated as the difference between the 120-day average increment and the 240-day average increment of margin trading balance. If the 120-day average increment is greater than the 240-day average increment, it is bullish, with a score of 1. Conversely, if less, the score is -1; Current View: This week, the short-term increment of margin trading is greater than the long-term increment, signaling a bullish outlook, with a score of 1[38][39] - Factor Name: Trading Volume Trend; Factor Construction Idea: This factor measures market activity and sentiment based on trading volume; Factor Construction Process: The factor is calculated as the moving average distance of log trading volume = ma120 / ma240 - 1. If the maximum of the moving average distance over 10, 30, and 60 days is positive, it is bullish, with a score of 1. Conversely, if the minimum is negative, the score is -1; Current View: This week, the trading volume trend signals a bullish outlook, with a score of 1[41][42] - Factor Name: China Sovereign CDS Spread; Factor Construction Idea: This factor represents the pricing level of China's economic and sovereign credit risk by overseas investors; Factor Construction Process: The factor is calculated as the 20-day difference of the smoothed CDS spread. If the difference is less than 0, it indicates a bullish outlook, with a score of 1. Conversely, if greater than 0, the score is -1; Current View: This week, the 20-day difference of the CDS spread is greater than 0, signaling a bearish outlook, with a score of -1[45][46] - Factor Name: Overseas Risk Aversion Index; Factor Construction Idea: This factor captures overseas market risk preference based on the Citi RAI Index; Factor Construction Process: The factor is calculated as the 20-day difference of the smoothed risk aversion index. If the difference is less than 0, it indicates a bullish outlook, with a score of 1. Conversely, if greater than 0, the score is -1; Current View: This week, the 20-day difference of the risk aversion index is less than 0, signaling a bullish outlook, with a score of 1[47][48] - Factor Name: Price Trend; Factor Construction Idea: This factor measures the market trend and strength based on moving average distance; Factor Construction Process: The factor is calculated as the moving average distance = ma120 / ma240 - 1. If the distance is greater than 0, the trend direction score is 1, otherwise -1. If the maximum of the distance over

- The report discusses the construction of the A-share prosperity index, which is based on the Nowcasting target of the year-on-year net profit of the Shanghai Composite Index's parent company. The index reflects the high-frequency prosperity trend of the A-share market. The current prosperity index is at 21.37, showing an increase of 15.95 compared to the end of 2023, indicating an upward cycle[30][33][34] - The A-share sentiment index is constructed by dividing the market into four quadrants based on the direction of changes in volatility and trading volume. Among these, the quadrant with rising volatility and declining trading volume shows significant negative returns, while the others show significant positive returns. The sentiment index includes bottom-warning and top-warning signals, with the current comprehensive signal being bullish[34][40][45] - The report applies the BARRA factor model to construct ten major style factors for the A-share market, including size (SIZE), beta (BETA), momentum (MOM), residual volatility (RESVOL), non-linear size (NLSIZE), valuation (BTOP), liquidity (LIQUIDITY), earnings yield (EARNINGS_YIELD), growth (GROWTH), and leverage (LVRG). Recent factor performance indicates that high-profitability stocks performed well, while leverage factors underperformed[60][61][62] - The pure factor returns for the past week show that industry factors such as non-ferrous metals, national defense, and telecommunications outperformed the market-cap-weighted portfolio, while banking and media factors experienced significant pullbacks. Among style factors, the earnings factor achieved high excess returns, while residual volatility showed significant negative excess returns[61][65][70] - The report evaluates the performance of enhanced index portfolios. The CSI 500 enhanced portfolio achieved a weekly return of 0.60%, underperforming the benchmark by 0.34%. Since 2020, the portfolio has generated an excess return of 49.21% relative to the CSI 500 index, with a maximum drawdown of -5.73%[48][50][52] - The CSI 300 enhanced portfolio achieved a weekly return of 1.40%, outperforming the benchmark by 0.12%. Since 2020, the portfolio has generated an excess return of 38.20% relative to the CSI 300 index, with a maximum drawdown of -5.86%[55][56][58]

Group 1 - The core viewpoint of the article highlights the misinterpretation of market adjustments as opportunities, emphasizing that many retail investors are still trapped in the illusion of a bull market despite significant underperformance among stocks [1][2][6] - A report from CITIC Securities indicates that only 62% of stocks outperformed the index in the first eleven months of 2025, suggesting that 4 out of 10 stocks are quietly lagging behind [2] - The article discusses the phenomenon of retail investors mistakenly believing that all adjustments in a bull market are buying opportunities, leading to significant losses during market downturns [2][4] Group 2 - The article describes a "cat-and-mouse game" in the market, where the ability to predict policy changes is crucial for investment success, contrasting the current environment with past market behaviors [6][10] - Data shows that stocks favored by insurance capital have outperformed the market by an average of 17 percentage points over the past five years during year-end rallies, indicating a disconnect between retail investors and institutional movements [6][10] - The article emphasizes the importance of quantitative analysis in understanding market dynamics, suggesting that modern investors need to focus on data-driven insights rather than speculative trends [10][12] Group 3 - The article identifies three key traits of successful investors who can navigate through market fluctuations: sensitivity to policy shifts, precision in tracking capital flows, and a deep understanding of their cognitive limitations [13] - It suggests that the ability to interpret market language through quantitative data is more valuable than traditional research methods, highlighting the evolution of information asymmetry in financial markets [10][12] - The author shares a personal insight about the importance of a long-term quantitative system in understanding real market transactions, reinforcing the need for data literacy among investors [12][13]

Quantitative Models and Construction Methods 1. Model Name: Timing Radar Six-Facet Diagram - **Model Construction Idea**: The model evaluates equity market performance through a multi-dimensional perspective, selecting 21 indicators across liquidity, economic fundamentals, valuation, capital flows, technical trends, and crowding. These are categorized into four major dimensions: "Valuation Cost-Effectiveness," "Macroeconomic Fundamentals," "Capital & Trend," and "Crowding & Reversal," generating a composite timing score within the range of [-1,1][1][6][8] - **Model Construction Process**: - The model aggregates 21 indicators into four categories, each representing a specific market dimension - The composite score is calculated as a weighted average of these dimensions, normalized to fall between [-1,1] - Detailed construction methods for individual indicators are referenced in the report "Timing Radar Six-Facet Diagram: A Multi-Dimensional Timing Framework"[6][8] - **Model Evaluation**: The model provides a comprehensive and systematic framework for market timing, integrating multiple dimensions of market dynamics[1][6] --- Model Backtesting Results 1. Timing Radar Six-Facet Diagram - **Liquidity Score**: 0.25 (Neutral to slightly positive signal)[8] - **Economic Fundamentals Score**: -0.25 (Neutral to slightly negative signal)[8] - **Valuation Score**: -0.33 (Neutral to slightly negative signal)[8] - **Capital Flows Score**: 0.00 (Neutral signal)[8] - **Technical Trends Score**: 0.00 (Neutral signal)[8] - **Crowding Score**: 0.50 (Neutral to slightly positive signal)[8] - **Overall Composite Score**: 0.02 (Neutral signal, down from 0.04 last week)[6][8] --- Quantitative Factors and Construction Methods 1. Factor Name: Liquidity Factors - **Factor Construction Idea**: Liquidity factors assess monetary and credit conditions, capturing signals from central bank policies and market interest rates[11][13][16][19] - **Factor Construction Process**: - **Monetary Direction Factor**: Measures the average change in central bank policy rates and short-term market rates over 90 days. Formula: If >0, monetary policy is deemed expansionary; if <0, contractionary[11] - **Monetary Strength Factor**: Based on the "interest rate corridor" concept, calculated as: $ \text{Deviation} = \frac{\text{DR007}}{\text{7-Year Reverse Repo Rate}} - 1 $ Smoothed and z-scored to form the factor. Thresholds: <-1.5 SD (expansionary, score=1), >1.5 SD (contractionary, score=-1)[13] - **Credit Direction Factor**: Uses 12-month growth in medium-to-long-term loans. If growth is higher than three months ago, score=1; otherwise, score=-1[16] - **Credit Strength Factor**: Measures deviation of new RMB loans from expectations: $ \text{Credit Strength Factor} = \frac{\text{New Loans} - \text{Median Expectation}}{\text{Standard Deviation of Expectation}} $ Thresholds: >1.5 SD (positive, score=1), <-1.5 SD (negative, score=-1)[19] - **Factor Evaluation**: These factors effectively capture liquidity conditions and their implications for market performance[11][13][16][19] 2. Factor Name: Economic Factors - **Factor Construction Idea**: Economic factors reflect growth and inflation dynamics, using PMI and inflation data to assess economic trends and surprises[22][25][27][28] - **Factor Construction Process**: - **Growth Direction Factor**: Based on PMI data (manufacturing and non-manufacturing), calculated as: $ \text{Growth Direction Factor} = \text{PMI 12-Month Average} \to \text{YoY Change} $ If higher than three months ago, score=1; otherwise, score=-1[22] - **Growth Strength Factor**: Measures PMI surprises: $ \text{Growth Strength Factor} = \frac{\text{PMI} - \text{Median Expectation}}{\text{Standard Deviation of Expectation}} $ Thresholds: >1.5 SD (positive, score=1), <-1.5 SD (negative, score=-1)[25] - **Inflation Direction Factor**: Combines CPI and PPI data: $ \text{Inflation Direction Factor} = 0.5 \times \text{Smoothed CPI YoY} + 0.5 \times \text{Raw PPI YoY} $ If lower than three months ago, score=1; otherwise, score=-1[27] - **Inflation Strength Factor**: Measures CPI and PPI surprises: $ \text{Inflation Strength Factor} = \text{Mean of CPI and PPI Surprises} $ Thresholds: <-1.5 SD (positive, score=1), >1.5 SD (negative, score=-1)[28] - **Factor Evaluation**: These factors provide a robust framework for assessing economic conditions and their market implications[22][25][27][28] 3. Factor Name: Valuation Factors - **Factor Construction Idea**: Valuation factors assess equity cost-effectiveness using metrics like Shiller ERP, PB, and AIAE[32][35][38] - **Factor Construction Process**: - **Shiller ERP**: $ \text{Shiller ERP} = \frac{1}{\text{Shiller PE}} - \text{10-Year Treasury Yield} $ Z-scored over six years to derive the factor score[32] - **PB**: $ \text{PB Factor} = \text{PB} \times (-1) $ Z-scored over six years, with 1.5 SD truncation, normalized to ±1[35] - **AIAE**: $ \text{AIAE} = \frac{\text{Total Market Cap of CSI All-Share}}{\text{Total Market Cap + Total Debt}} $ Multiplied by -1 and z-scored over six years to derive the factor score[38] - **Factor Evaluation**: These factors effectively capture valuation dynamics and investor sentiment[32][35][38] 4. Factor Name: Crowding Factors - **Factor Construction Idea**: Crowding factors measure market sentiment and overreaction using derivatives and convertible bond pricing[59][65][67] - **Factor Construction Process**: - **Option Implied Basis**: Derived from put-call parity, reflecting market sentiment. Thresholds: If 50ETF 5-day return <0 and percentile <30%, score=1; if return >0 and percentile >70%, score=-1[59] - **Option VIX**: Measures implied volatility. Thresholds: If 50ETF 5-day return <0 and percentile >70%, score=1; if return >0 and percentile >70%, score=-1[60] - **Option SKEW**: Reflects skewness expectations. Thresholds: If 50ETF 5-day return <0 and percentile >70%, score=1; if return >0 and percentile <30%, score=-1[65] - **Convertible Bond Mispricing**: $ \text{Mispricing} = \frac{\text{Convertible Bond Price}}{\text{Model Price}} - 1 $ Multiplied by -1 and z-scored over three years to derive the factor score[67] - **Factor Evaluation**: These factors provide valuable insights into market sentiment and potential reversals[59][65][67] --- Factor Backtesting Results 1. Liquidity Factors - **Monetary Direction Factor**: Score = 1 (Positive signal)[11] - **Monetary Strength Factor**: Score = 0 (Neutral signal)[13] - **Credit Direction Factor**: Score = 1 (Positive signal)[16] - **Credit Strength Factor**: Score = -1 (Negative signal)[19] 2. Economic Factors - **Growth Direction Factor**: Score = 1 (Positive signal)[22][24] - **Growth Strength Factor**: Score = 0 (Neutral signal)[25] - **Inflation Direction Factor**: Score = -1 (Negative signal)[27] - **Inflation Strength Factor**: Score = -1 (Negative signal)[28] 3. Valuation Factors - **Shiller ERP**: Score = 0.07 (Slightly positive signal)[34] - **PB**: Score = -0.37 (Negative signal)[35] - **AIAE**: Score = -0.69 (Negative signal)[38] 4. Crowding Factors - **Option Implied Basis**: Score = 1 (Positive signal)[59] - **Option VIX**: Score = 1 (Positive signal)[60] - **Option SKEW**: Score = 1 (Positive signal)[65] - **Convertible Bond Mispricing**: Score = -1 (