金工定期报告：基于技术指标的指数仓位调整月报202412

Quantitative Models and Construction Methods 1. Model Name: Single Signal Model - Model Construction Idea: This model is based on technical indicators derived from price-volume data, aiming to generate buy/sell signals for position adjustments to achieve excess returns [3][6] - Model Construction Process: - A total of 27 technical indicators were constructed, each generating signals based on price-volume "divergence" concepts [3][6] - These indicators were tested under specified backtesting conditions across three broad-based indices (CSI 300, CSI 500, CSI 1000) and 31 SW Level-1 industry indices [3][6] - The average annualized excess return across 34 indices was 3.75% [3] 2. Model Name: Multi-Signal Model - Model Construction Idea: This model combines multiple technical indicators to enhance performance through correlation analysis and signal integration [3][6] - Model Construction Process: - Two strategies were developed: - 5-Signal Strategy: Combines five technical indicators - 7-Signal Strategy: Combines seven technical indicators with buy/sell differentiation [3][6] - Signals were processed using two methods: - Post-Merge Processing (Rolling Stable Strategy): Signals are merged first, then processed, suitable for low-risk investors [3][6] - Pre-Merge Processing (Rolling Momentum Strategy): Signals are processed first, then merged, suitable for high-risk, bullish investors [3][6] 3. Model Name: Rolling Signal Combination - Model Construction Idea: Rolling signal combinations aim to optimize signal stability and performance by adjusting rebalancing frequency [3][6] - Model Construction Process: - Backtesting showed that reducing rebalancing frequency to T+10 improved stability, with an average annualized excess return of 3.99% [3][6] - Two strategies were defined: - Rolling Stable Strategy: Focuses on stability and lower risk - Rolling Momentum Strategy: Focuses on momentum and reducing missed opportunities [3][6] --- Model Backtesting Results 1. Single Signal Model - Annualized Excess Return: 3.75% on average across 34 indices [3] 2. Multi-Signal Model - 5-Signal Strategy: - CSI 300: 0.56% - CSI 500: 0.00% - CSI 1000: 0.00% [7][8] - 7-Signal Strategy: - CSI 300: -0.87% - CSI 500: 0.71% - CSI 1000: -1.43% [8] 3. Rolling Signal Combination - Rolling Stable Strategy: - CSI 300: -0.87% - CSI 500: -0.70% - CSI 1000: -2.36% [10] - Rolling Momentum Strategy: - CSI 300: -0.78% - CSI 500: -0.12% - CSI 1000: -0.77% [10]