市场形态周报（20250818-20250822）：本周指数普遍上涨-20250825

Quantitative Models and Construction Methods 1. Model Name: Heston Model - Model Construction Idea: The Heston model is used to calculate the implied volatility of near-month at-the-money options, serving as a market fear index. It reflects market participants' expectations of future volatility[8] - Model Construction Process: The Heston model is a stochastic volatility model where the variance of the asset price follows a mean-reverting square-root process. The model is defined by the following equations: $ dS_t = \mu S_t dt + \sqrt{v_t} S_t dW_t^S $ $ dv_t = \kappa (\theta - v_t) dt + \sigma \sqrt{v_t} dW_t^v $ where: - $S_t$: Asset price - $v_t$: Variance of the asset price - $\mu$: Drift term - $\kappa$: Rate of mean reversion - $\theta$: Long-term variance - $\sigma$: Volatility of variance - $W_t^S, W_t^v$: Two Wiener processes with correlation $\rho$[8] - Model Evaluation: The model effectively captures market fear and volatility expectations, providing a robust measure of implied volatility[8] --- Model Backtesting Results 1. Heston Model - Implied Volatility Results: - SSE 50: 20.3% (+2.93% WoW)[10] - SSE 500: 22.36% (+2.82% WoW)[10] - CSI 1000: 25.91% (+4.86% WoW)[10] - CSI 300: 19.21% (+1.12% WoW)[10] --- Quantitative Factors and Construction Methods 1. Factor Name: Multi-Signal Shape Timing Factor - Factor Construction Idea: This factor is based on the frequency and success rate of positive and negative signals derived from historical shape patterns. It aims to predict future highs and lows in the market[12] - Factor Construction Process: - Positive signals and negative signals are identified based on historical shape patterns - The success rate of these signals in predicting future highs and lows is calculated as: $ \text{Success Rate} = \frac{\text{Number of Correct Predictions}}{\text{Total Number of Predictions}} \times 100% $ - For the period from August 11 to August 15, 2025: - Positive signals: 3365 occurrences, average success rate of 70.33% - Negative signals: 3167 occurrences, average success rate of 27.82%[12] - Factor Evaluation: The factor demonstrates strong predictive power for positive signals, with a high success rate in identifying future market highs[12] 2. Factor Name: Industry Multi-Long-Short Shape Timing Factor - Factor Construction Idea: This factor is constructed by calculating the difference in the number of long and short signals within industry index constituent stocks. It aims to outperform respective industry indices through timing strategies[15] - Factor Construction Process: - For each industry index, the number of long and short signals is calculated daily - If no long or short signals are present, the respective count is set to zero - The difference between long and short signals (scissor difference) is calculated, and the ratio of this difference is used to construct the timing strategy[15] - Factor Evaluation: The factor outperforms all respective industry indices in backtesting, demonstrating excellent historical performance[15] --- Factor Backtesting Results 1. Multi-Signal Shape Timing Factor - Positive Signal Success Rate: 70.33%[12] - Negative Signal Success Rate: 27.82%[12] 2. Industry Multi-Long-Short Shape Timing Factor - Performance Metrics: - Outperformed respective industry indices in 100% of backtests[15] - Examples of Industry Results: - Machinery: Strategy annualized return 19.72%, maximum drawdown -42.41%; Index annualized return 4.63%, maximum drawdown -72.59%[16] - Retail: Strategy annualized return 19.75%, maximum drawdown -43.39%; Index annualized return -0.9%, maximum drawdown -77.37%[16] - Electronics: Strategy annualized return 22.54%, maximum drawdown -44.99%; Index annualized return 11.13%, maximum drawdown -58.54%[16] --- Additional Observations - Special Positive Shape Signals: Specific K-line patterns such as "Golden Needle Bottom," "Rocket Launch," and "Full Red" exhibit strong positive predictive effects[22][23] - Brokerage Golden Stock Shape Signals: Combining fundamental analysis with shape-based buy signals significantly improves portfolio returns and reduces maximum drawdowns[27]