量化周报：市场有望突破阻力-20250706

Quantitative Models and Construction Methods 1. Model Name: Three-Dimensional Timing Framework - Model Construction Idea: This model integrates three dimensions—liquidity, divergence, and prosperity—to determine market timing. It aims to identify full-position opportunities based on the upward trend of liquidity and prosperity and the downward trend of divergence[8][13][15] - Model Construction Process: 1. Liquidity Index: Measures the market's liquidity level 2. Divergence Index: Reflects the degree of market disagreement 3. Prosperity Index: Indicates the economic and market activity level 4. The framework combines these three indices to generate timing signals, with historical performance showing its effectiveness in identifying market opportunities[15][19][20] - Model Evaluation: The model is effective in identifying full-position opportunities when liquidity and prosperity are rising, and divergence is falling[8][15] --- Quantitative Factors and Construction Methods 1. Factor Name: Dividend Factor (dp_historical) - Factor Construction Idea: Measures the total cash dividends implemented over the past four quarters relative to the current market value[46][47] - Factor Construction Process: - Formula: $ dp_historical = \frac{\text{Total Cash Dividends (Last 4 Quarters)}}{\text{Current Market Value}} $ - The factor is market-cap and industry-neutralized to ensure robustness[45][46] - Factor Evaluation: Demonstrates strong performance, with a one-week excess return of 1.65% and a one-month excess return of 1.68%[47] 2. Factor Name: PEG Factor (peg) - Factor Construction Idea: Evaluates the price-to-earnings growth ratio to identify undervalued growth stocks[46][47] - Factor Construction Process: - Formula: $ peg = \frac{\text{PE Ratio}}{\text{Earnings Growth Rate}} $ - The factor is adjusted for market-cap and industry neutrality[45][46] - Factor Evaluation: Exhibits strong performance, with a one-week excess return of 1.42% and a one-month excess return of 5.30%[47] 3. Factor Name: Earnings Yield Factor (ep_fy3) - Factor Construction Idea: Uses the inverse of the forward three-year price-to-earnings ratio to assess valuation[46][47] - Factor Construction Process: - Formula: $ ep_fy3 = \frac{1}{\text{Forward PE (3-Year)}} $ - The factor is market-cap and industry-neutralized[45][46] - Factor Evaluation: Shows consistent performance, with a one-week excess return of 1.32% and a one-month excess return of 3.41%[47] 4. Factor Name: PE-Growth Ranking Factor (pe_g) - Factor Construction Idea: Ranks stocks based on the difference between PE rankings and earnings growth rankings[46][49] - Factor Construction Process: - Formula: $ pe_g = \text{PE Rank} - \text{Earnings Growth Rank} $ - The factor is adjusted for market-cap and industry neutrality[45][46] - Factor Evaluation: Performs well across different indices, with a one-week excess return of 4.84% in the CSI 300 index and 3.70% in the CSI 1000 index[49] --- Factor Backtesting Results 1. Dividend Factor (dp_historical) - One-Week Excess Return: 1.65%[47] - One-Month Excess Return: 1.68%[47] 2. PEG Factor (peg) - One-Week Excess Return: 1.42%[47] - One-Month Excess Return: 5.30%[47] 3. Earnings Yield Factor (ep_fy3) - One-Week Excess Return: 1.32%[47] - One-Month Excess Return: 3.41%[47] 4. PE-Growth Ranking Factor (pe_g) - One-Week Excess Return in CSI 300: 4.84%[49] - One-Week Excess Return in CSI 1000: 3.70%[49] --- Quantitative Portfolio Performance 1. CSI 300 Enhanced Portfolio - Absolute Return (Last Week): 2.67% - Excess Return (Last Week): 0.10% - Absolute Return (YTD): 5.01% - Excess Return (YTD): 4.98%[50][52] 2. CSI 500 Enhanced Portfolio - Absolute Return (Last Week): 4.59% - Excess Return (Last Week): 0.02% - Absolute Return (YTD): 2.97% - Excess Return (YTD): 2.76%[50][52] 3. CSI 1000 Enhanced Portfolio - Absolute Return (Last Week): 5.63% - Excess Return (Last Week): 0.24% - Absolute Return (YTD): 7.81% - Excess Return (YTD): 5.53%[50][52]