股指分红点位监控周报：IH及IF主力合约升水，IC及IM主力合约贴水-20250722

Quantitative Models and Construction Methods Dividend Points Estimation Model - Model Name: Dividend Points Estimation Model - Model Construction Idea: This model estimates the dividend points of stock indices to account for the impact of constituent stock dividends on index futures' premium/discount levels. It is essential for accurately evaluating the fair value of index futures contracts[11][39][44]. - Model Construction Process: 1. Formula: The dividend points during the period from time $t$ (current date) to $T$ (futures contract expiration date) are calculated as: $\text{Dividend Points} = \sum_{n=1}^{N} \left( \frac{\text{Dividend Amount of Stock } n}{\text{Total Market Value of Stock } n} \times \text{Weight of Stock } n \times \text{Index Closing Price} \right)$ Here: - $N$: Number of constituent stocks - $t < \text{Ex-Dividend Date} \leq T$: Only stocks with ex-dividend dates within the specified period are considered[39]. 2. Steps: - Obtain constituent stock weights and market values. - Estimate dividend amounts for stocks without disclosed data using: - Net Profit Estimation: Based on historical profit distribution patterns or prior-year profits for unstable companies[47][50]. - Dividend Payout Ratio Estimation: Use historical averages or prior-year ratios, with adjustments for outliers[51][53]. - Predict ex-dividend dates using historical intervals or default dates if no data is available[56]. 3. Adjustments: - Use daily weight data from the China Securities Index Company for accuracy[46]. - Handle special cases like stock adjustments, unlocking, or rights issues to minimize bias[46]. - Model Evaluation: The model demonstrates high accuracy for indices like the SSE 50 and CSI 300, with prediction errors typically within 5 points. For the CSI 500, the error margin is slightly larger, around 10 points[57][61]. --- Model Backtesting Results Dividend Points Estimation Model - SSE 50 Index: Prediction error within 5 points for annual dividend points in 2023 and 2024[61][65]. - CSI 300 Index: Prediction error within 5 points for annual dividend points in 2023 and 2024[61][66]. - CSI 500 Index: Prediction error within 10 points for annual dividend points in 2023 and 2024[61][70]. --- Quantitative Factors and Construction Methods Realized and Remaining Dividend Yield Factors - Factor Name: Realized Dividend Yield and Remaining Dividend Yield - Factor Construction Idea: These factors measure the dividend yield already distributed and the expected remaining yield for the year, providing insights into the dividend contribution to index returns[16]. - Factor Construction Process: 1. Formulas: - Realized Dividend Yield: $\text{Realized Dividend Yield} = \sum_{i=1}^{N_1} \left( \frac{\text{Dividend Amount of Stock } i}{\text{Market Value of Stock } i} \times \text{Weight of Stock } i \right)$ - Remaining Dividend Yield: $\text{Remaining Dividend Yield} = \sum_{j=1}^{N_2} \left( \frac{\text{Expected Dividend Amount of Stock } j}{\text{Market Value of Stock } j} \times \text{Weight of Stock } j \right)$ Here: - $N_1$: Number of stocks that have already paid dividends - $N_2$: Number of stocks yet to pay dividends[16]. 2. Data Inputs: - Realized dividend amounts for stocks that have already distributed dividends. - Estimated dividend amounts for stocks yet to distribute dividends, based on historical payout ratios and profit estimates[47][51]. - Factor Evaluation: These factors provide a clear breakdown of dividend contributions to index returns, aiding in the analysis of dividend-related investment strategies[16]. --- Factor Backtesting Results Realized and Remaining Dividend Yield Factors - SSE 50 Index: - Realized Dividend Yield: 2.04% - Remaining Dividend Yield: 0.17%[3][16]. - CSI 300 Index: - Realized Dividend Yield: 1.56% - Remaining Dividend Yield: 0.27%[3][16]. - CSI 500 Index: - Realized Dividend Yield: 1.09% - Remaining Dividend Yield: 0.15%[3][16]. - CSI 1000 Index: - Realized Dividend Yield: 0.86% - Remaining Dividend Yield: 0.07%[3][16].