6 月中旬：边际乐观，逢低建仓——主动量化周报

Quantitative Models and Construction Methods 1. Model Name: Annualized Discount Model for CSI 500 Futures - Model Construction Idea: The model identifies optimal entry points for building positions based on historical performance when the annualized discount of CSI 500 futures exceeds a certain threshold, indicating market pessimism. [1][11] - Model Construction Process: - The model uses the annualized discount rate of the next-month contract of CSI 500 index futures as the key metric. - Historical data from 2017 onwards is analyzed to determine the relationship between the discount rate and subsequent returns. - Key findings: - When the annualized discount exceeds 15%, holding the index for more than 12 trading days results in average cumulative returns trending upward. - Holding for over 33 trading days yields a probability of positive cumulative returns exceeding 50%. - Holding for over 50 trading days increases the probability of positive returns to approximately 60%. - Formula: $ \text{Annualized Discount} = \frac{\text{Spot Price} - \text{Futures Price}}{\text{Futures Price}} \times \frac{365}{\text{Days to Maturity}} $ - Spot Price: Current index level - Futures Price: Price of the futures contract - Days to Maturity: Remaining days until the futures contract expires [11] - Model Evaluation: The model effectively captures market pessimism and identifies potential rebound opportunities, making it a useful tool for timing market entry. [11] --- Model Backtesting Results 1. Annualized Discount Model for CSI 500 Futures - Key Metrics: - Holding for 12 trading days: Average cumulative returns trend upward. - Holding for 33 trading days: Positive return probability > 50%. - Holding for 50 trading days: Positive return probability ~60%. [1][11] --- Quantitative Factors and Construction Methods 1. Factor Name: Proprietary Active Trader Activity Indicator - Factor Construction Idea: This factor measures the activity level of speculative funds (e.g., proprietary traders) to gauge market sentiment and risk appetite. [3][13] - Factor Construction Process: - Data Source: Derived from "Dragon and Tiger List" (龙虎榜) data. - The indicator tracks the marginal changes in active trader participation over time. - Observations: - From late April, the indicator showed a consistent decline, reflecting reduced risk appetite and cautious market sentiment. - Recently, the indicator has shown marginal improvement, suggesting a potential rebound in risk appetite. [3][13] - Factor Evaluation: The factor provides timely insights into the behavior of speculative funds, which can serve as a leading indicator for shifts in market sentiment. [3][13] 2. Factor Name: BARRA Style Factors - Factor Construction Idea: These factors assess the performance of various style attributes (e.g., momentum, volatility, size) to understand market preferences. [23][24] - Factor Construction Process: - Data Source: BARRA factor model. - Key Observations for the Week: - Fundamental factors (e.g., profitability) showed significant positive excess returns. - Stocks with high short-term momentum and high volatility outperformed. - Size-related factors (e.g., market capitalization) continued to underperform, indicating a preference for mid- to small-cap stocks. - Formula: Factor returns are calculated as the weighted average of stock returns within each style category. [23][24] - Factor Evaluation: The factors effectively capture shifts in market preferences, providing actionable insights for portfolio adjustments. [23][24] --- Factor Backtesting Results 1. Proprietary Active Trader Activity Indicator - Key Metrics: - Indicator showed consistent decline from late April, reflecting reduced risk appetite. - Recent marginal improvement suggests a potential rebound in speculative activity. [3][13] 2. BARRA Style Factors - Key Metrics: - Momentum: +0.2% weekly return. - Volatility: +0.2% weekly return. - Profitability: +0.3% weekly return. - Size: -0.5% weekly return. - Nonlinear Size: -0.3% weekly return. [23][24]