证券研究报告 | 2025年11月20日 股指分红点位监控周报 市场短期调整，四大主力合约均处于贴水状态 核心观点 金融工程周报 成分股分红进度 截至 2025 年 11 月 19 日： 上证 50 指数中，有 0 家公司处于预案阶段，0 家公司处于决案阶段，0 家公 司进入实施阶段，0 家公司已分红，3 家公司不分红； 沪深 300 指数中，有 1 家公司处于预案阶段，0 家公司处于决案阶段，0 家 公司进入实施阶段，0 家公司已分红，20 家公司不分红； 中证 500 指数中，有 0 家公司处于预案阶段，0 家公司处于决案阶段，0 家 公司进入实施阶段，0 家公司已分红，46 家公司不分红； 中证 1000 指数中，有 0 家公司处于预案阶段，0 家公司处于决案阶段，0 家公司进入实施阶段，0 家公司已分红，139 家公司不分红。 行业成分股股息率比较 我们对当前已披露分红预案的个股股息率进行了统计，其中，煤炭、银行和 钢铁行业的股息率排名前三。 已实现及剩余股息率 截至 2025 年 11 月 19 日： 上证 50 指数已实现股息率为 2.50%，剩余股息率 0.37%； 沪深 300 指数已实现股息 ...

Group 1: Dividend Progress of Constituent Stocks - As of November 19, 2025, there are no companies in the proposal, decision, or implementation stages for dividends in the Shanghai 50 Index, with 3 companies not distributing dividends [1] - In the CSI 300 Index, 1 company is in the proposal stage, while 20 companies are not distributing dividends [1] - The CSI 500 Index has no companies in any dividend stages, with 46 companies not distributing dividends [1] - The CSI 1000 Index also has no companies in any dividend stages, with 139 companies not distributing dividends [1] Group 2: Industry Dividend Yield Comparison - The current dividend yield statistics show that the coal, banking, and steel industries rank the highest in terms of dividend yield among disclosed dividend proposals [1][4] Group 3: Realized and Remaining Dividend Yields - As of November 19, 2025, the realized dividend yields are as follows: - Shanghai 50 Index: 2.50% realized, 0.37% remaining [1][7] - CSI 300 Index: 2.01% realized, 0.22% remaining [1][7] - CSI 500 Index: 1.24% realized, 0.05% remaining [1][7] - CSI 1000 Index: 0.94% realized, 0.04% remaining [1][7] Group 4: Tracking of Index Futures Premium/Discount - As of November 19, 2025, the annualized discounts for index futures are as follows: - IH main contract: 1.49% [1] - IF main contract: 4.27% [1] - IC main contract: 11.18% [1] - IM main contract: 14.26% [1] Group 5: Methodology for Index Dividend Estimation - The report discusses the importance of estimating dividend points for index futures, as the dividend distribution affects the index point drop [2][26] - The methodology includes calculating the dividend points based on the dividend amount, total market value, and weight of constituent stocks [28][30] - The report emphasizes the need for accurate daily weight data for constituent stocks to refine the dividend impact on the index [32]

Key Points - The article discusses the dividend progress of constituent stocks in major indices as of November 12, 2025, highlighting that no companies in the Shanghai 50, CSI 500, and CSI 1000 indices have declared or implemented dividends, while only one company in the CSI 300 is in the proposal stage [1][3] - The article provides a comparison of dividend yields across different industries, noting that coal, banking, and steel sectors have the highest yields [2][4] - The realized and remaining dividend yields for various indices are reported, with the Shanghai 50 index having a realized yield of 2.50% and a remaining yield of 0.36%, while the CSI 300 has 1.99% and 0.24%, respectively [8][7] - The article tracks the premium and discount of stock index futures, indicating that the IH main contract has an annualized premium of 3.92%, while the IC and IM contracts show significant discounts of 10.12% and 14.93% [4][12] - The methodology for calculating the impact of dividends on index futures is discussed, emphasizing the importance of considering dividend points in estimating futures premiums [20][34] - The accuracy of dividend point estimation is analyzed, showing that the model has a high prediction accuracy for the Shanghai 50 and CSI 300 indices, with deviations around 5 points, while the CSI 500 has a slightly larger deviation of about 10 points [36][34]

- The report introduces the methodology for calculating index dividend points, which is crucial for accurately estimating the premium or discount of stock index futures contracts. The calculation considers the dividend impact of index constituent stocks, using the formula: **Dividend Points = Σ (Dividend Amount / Total Market Value × Stock Weight × Index Closing Price)** This formula ensures that only stocks with ex-dividend dates between the current date and the futures contract expiration date are included[42][43][45] - The report details the process for estimating key components of the dividend points calculation, including: **1. Stock Weight:** Transitioning from approximate monthly estimates to precise daily weights using data disclosed by the China Securities Index Company[48][49] **2. Dividend Amount:** Estimation based on historical net profit distribution and dynamic prediction methods, considering whether companies have disclosed annual reports or profit forecasts[50][53] **3. Dividend Payout Ratio:** Using historical payout ratios for stable companies and applying truncation for ratios exceeding 100%[54][56] **4. Ex-Dividend Date:** Predicting dates based on historical intervals and linear extrapolation, with default dates applied for companies without prior dividend history[54][59] - The report evaluates the accuracy of the dividend points estimation model, showing that for 2023 and 2024, the model achieved high accuracy for the Shanghai 50 and CSI 300 indices, with errors around 5 points. However, the CSI 500 index had slightly larger errors, approximately 10 points. The model also demonstrated strong predictive accuracy for stock index futures contracts across these indices[60][64][65]

- The report introduces the methodology for calculating index dividend points, which is crucial for accurately estimating the premium or discount of stock index futures contracts. The calculation considers the dividend impact of index constituent stocks, including their dividend amount, market capitalization, weight, and index closing price[39][40][44] - The weight of index constituent stocks is refined from monthly data provided by mainstream data providers to daily data disclosed by the China Securities Index Company. This ensures higher accuracy in capturing the daily weight changes of individual stocks[45][46] - The estimation of dividend amounts involves predicting net profits and dividend payout ratios. Net profit prediction is based on historical profit distribution dynamics, categorizing companies into stable and unstable profit distribution groups. Stable companies use historical patterns for prediction, while unstable ones rely on the previous year's profit as a forecast[47][50] - Dividend payout ratio prediction uses historical averages. If a company paid dividends last year, the previous year's payout ratio is used. If not, the average of the past three years is applied. Companies with no dividend history are assumed not to pay dividends[51][53] - The prediction of ex-dividend dates employs a linear extrapolation method based on the stability of historical interval days between dividend announcement and ex-dividend dates. Default dates are set for companies with insufficient historical data or no dividend history[55][56] - The accuracy of the dividend point estimation model is validated by comparing predicted dividend points with actual dividend points for major indices like the SSE 50, CSI 300, and CSI 500. The model shows high accuracy for SSE 50 and CSI 300, with errors around 5 points, while errors for CSI 500 are slightly larger, around 10 points[57][61]

Group 1 - As of October 29, 2025, no companies in the Shanghai 50, CSI 300, CSI 500, and CSI 1000 indices are in the proposal, decision, or implementation stages for dividends, with a total of 3, 20, 46, and 138 companies respectively not distributing dividends [1][3] - The coal, banking, and steel industries have the highest median dividend yields among the disclosed dividend proposals [4] - The realized dividend yields as of October 29, 2025, are 2.45% for the Shanghai 50, 1.96% for the CSI 300, 1.20% for the CSI 500, and 0.92% for the CSI 1000, with remaining yields of 0.10%, 0.11%, 0.05%, and 0.04% respectively [7][6] Group 2 - The annualized premium and discount for the main futures contracts as of October 29, 2025, are 1.93% for IH, -0.96% for IF, -8.26% for IC, and -10.23% for IM [1] - The tracking of the premium and discount levels of stock index futures will consider the impact of dividend distributions on index points, which is crucial for accurately estimating the premium and discount of futures contracts [2][23] Group 3 - The dividend progress for the indices is expected to be concentrated in November and December, as shown in the forecasts for the Shanghai 50, CSI 300, CSI 500, and CSI 1000 indices [16] - The methodology for estimating dividend points involves calculating the dividend amounts based on the weights of constituent stocks, their market values, and the index closing prices [25][26]

- The report discusses the methodology for calculating index dividend points, which is crucial for accurately estimating the premium or discount of stock index futures contracts. The formula for dividend points is: $ Dividend\ Points = \sum_{n=1}^{N} \frac{Dividend\ Amount}{Total\ Market\ Value} \times Component\ Weight \times Index\ Closing\ Price $ This calculation considers the dividend amount, total market value, component weight, and index closing price, ensuring precision in estimating the impact of dividends on index points[39][44][57] - Component weights are refined from rough estimates to precise daily data. The formula for adjusting weights is: $ W_{n,t} = \frac{w_{n0} \times (1 + r_{n})}{\sum_{i=1}^{N} w_{i0} \times (1 + r_{i})} $ Here, $ w_{n0} $ represents the weight of component $ n $ at the last disclosed date, and $ r_{n} $ is the non-adjusted return of the component stock. This ensures daily accuracy in weight calculations[45][46] - Net profit prediction is based on historical profit distribution dynamics. Companies are categorized into stable and unstable profit distribution groups. Stable companies use historical patterns for prediction, while unstable ones rely on the previous year's profit as the forecast value[47][50] - Dividend payout ratio prediction uses historical averages. For companies with stable operations, the previous year's payout ratio is used. If no dividends were issued in the past, the average of the last three years is applied. If the payout ratio exceeds 100%, it is capped[51][53] - Ex-dividend date prediction employs a linear extrapolation method based on historical intervals. If a company has not disclosed its ex-dividend date, historical data is used, or default dates are applied based on the calendar year[51][56] - The accuracy of the dividend point estimation model is evaluated. For the Shanghai 50 and CSI 300 indices, the annual prediction error is approximately 5 points, while for the CSI 500 index, the error is around 10 points. The model demonstrates high accuracy for stock index futures contracts, particularly for Shanghai 50 and CSI 300 futures[57][61][66]

- The report introduces the dividend point estimation method for stock indices, emphasizing its importance in accurately calculating the premium or discount of stock index futures contracts, which track price indices rather than total return indices[39][12][44] - The dividend point estimation formula is presented as follows: $$Dividend\ Points = \sum_{n=1}^{N} \frac{Dividend\ Amount_{n}}{Total\ Market\ Value_{n}} \times Weight_{n} \times Index\ Closing\ Price$$ Here, \(n\) represents the index constituent stock, \(Weight_{n}\) is the stock's weight in the index, and \(Dividend\ Amount_{n}\) is the dividend amount of the stock[39][44][40] - The estimation process involves several steps: 1. Determine whether the company has disclosed dividend amounts and ex-dividend dates. If not, estimate these values[42] 2. Estimate dividend amounts using the formula \(Dividend\ Amount = Net\ Profit \times Dividend\ Payout\ Ratio\), where net profit is dynamically predicted based on historical profit distributions[47][50] 3. Predict dividend payout ratios using historical averages, applying truncation if the ratio exceeds 100%[51][53] 4. Forecast ex-dividend dates using historical intervals and linear extrapolation methods[56][55][53] - The report highlights the accuracy of the dividend point estimation model, noting that for indices like the SSE 50 and CSI 300, the prediction error is within 5 points, while for CSI 500, the error is around 10 points[57][61][62] - The backtesting results for stock index futures contracts show varying annualized premium or discount levels: - IH main contract: annualized premium of 0.55%[4][13] - IF main contract: annualized discount of 2.85%[4][13] - IC main contract: annualized discount of 10.08%[4][13] - IM main contract: annualized discount of 12.15%[4][13] - The historical percentile levels of the main contracts are also analyzed: - IH contract is at the 49th percentile[27][29] - IF contract is at the 24th percentile[27][29] - IC contract is at the 12th percentile[27][32] - IM contract is at the 8th percentile[27][32] - The report concludes that the deeper the market capitalization coverage of the index, the greater the discount observed in the futures contracts[24][27][12]

- The report introduces the dividend point estimation method for stock indices, emphasizing its importance in accurately calculating the premium or discount levels of stock index futures contracts, which track price indices rather than total return indices[39][12][44] - The dividend point estimation formula is provided as follows: $$ \text{Dividend Points} = \sum_{n=1}^{N} \frac{\text{Dividend Amount of Component Stock}}{\text{Total Market Value of Component Stock}} \times \text{Weight of Component Stock} \times \text{Index Closing Price} $$ This formula requires the dividend amount, total market value, weight of component stocks, and index closing price, with the condition that the ex-dividend date of the stock must fall between the current date and the futures contract expiration date[39][44][45] - The report details the process of estimating component stock weights, transitioning from imprecise monthly data to precise daily data provided by the China Securities Index Company, ensuring accuracy in daily weight calculations[45][46] - The estimation of dividend amounts involves predicting net profits and dividend payout ratios. Net profit prediction uses historical profit distribution dynamics, categorizing companies into stable and unstable profit distribution groups, and applying respective methods for estimation[47][50][52] - Dividend payout ratio prediction employs historical averages, with adjustments based on past dividend behavior. Companies with no dividend history are assumed not to pay dividends, and ratios exceeding 100% are capped[51][53][56] - Ex-dividend date prediction uses historical intervals for linear extrapolation, considering factors like past dividend dates, shareholder meeting schedules, and default dates for companies with no dividend history[51][55][56] - The accuracy of the dividend point estimation model is validated through comparisons of predicted and actual dividend points for indices like the SSE 50, CSI 300, and CSI 500. The model demonstrates high accuracy, with errors generally within 5 points for SSE 50 and CSI 300, and slightly larger errors for CSI 500, around 10 points[57][61][66]

Group 1 - The core viewpoint of the article emphasizes the tracking of dividend progress and yield across major stock indices, highlighting the lack of dividend announcements from many companies within these indices as of October 22, 2025 [1][3][5] - The article provides a detailed comparison of dividend yields across different industries, with coal, banking, and steel sectors ranking the highest [5][9] - The realized and remaining dividend yields for major indices are reported, with the Shanghai 50 Index showing a realized yield of 2.26% and a remaining yield of 0.29%, while the CSI 1000 Index shows a realized yield of 0.91% and a remaining yield of 0.03% [9][18] Group 2 - The article tracks the premium and discount of stock index futures, noting that as of October 22, 2025, the IH main contract has an annualized premium of 0.55%, while the IF, IC, and IM contracts show annualized discounts of 2.85%, 10.08%, and 12.15% respectively [1][4][12] - It discusses the importance of considering dividend impacts when calculating the premium and discount of stock index futures, as dividends can lead to a natural decline in index points [2][24] - The article outlines a systematic approach to estimating dividend points for stock indices, emphasizing the need for accurate data on constituent stock weights and dividend amounts [26][29][30] Group 3 - The article includes a detailed analysis of the dividend progress for various indices, indicating that no companies within the Shanghai 50, CSI 300, CSI 500, and CSI 1000 indices have announced dividends as of the specified date [1][3][18] - It presents a visual representation of the dividend yield trends and remaining yields for the indices since 2025, illustrating the overall dividend landscape [8][9] - The article also highlights the historical context of premium and discount levels for stock index futures, providing insights into market sentiment and investor behavior [10][13][40]