Report Industry Investment Rating No relevant content provided. Core Viewpoints of the Report The report comprehensively analyzes the secondary market and ETF product situation from July 21 to July 25, 2025. It includes the performance of major indexes and industries, the flow of funds in different types of ETFs, trading volume, and new fund launches and listings [1][2][3]. Summary by Related Catalogs 1. Secondary Market Overview - **Index Performance**: During the sample period, the top performers in weekly returns were the Science and Technology Innovation 50, Nikkei 225, and CSI 500, with returns of 4.63%, 4.11%, and 3.28% respectively. The PE valuation quantile of the CSI 500 was the highest at 99.59%, while that of the Nikkei 225 was the lowest at 72.13% [10]. - **Industry Performance**: In terms of returns, the top three industries were building materials, coal, and steel, with returns of 8.20%, 7.98%, and 7.67% respectively. The bottom three were banking, communication, and public utilities, with returns of -2.87%, -0.77%, and -0.27% respectively. In terms of valuation, the top three industries in valuation quantile were computer, national defense and military industry, and textile and apparel, all at 100.00%. The bottom three were agriculture, forestry, animal husbandry and fishery, comprehensive, and household appliances, at 23.14%, 39.05%, and 40.70% respectively [17]. 2. ETF Product Overview 2.1 ETF Market Performance - **By Product Type**: The average performance of industry - themed ETFs was the best, with a weighted average return of 2.96%, while that of bond ETFs was the worst, with a weighted average return of -0.27%. - **By Listing Plate**: The ETFs related to the Science and Technology Innovation Board and Japanese stocks performed well, with weighted average returns of 4.54% and 3.82% respectively. The ETFs related to US stocks and MSCI China A - share concept performed poorly, with weighted average returns of 0.40% and 1.65% respectively. - **By Industry Plate**: The cycle - sector ETFs had the best average performance, with a weighted average return of 7.21%, while the consumer - sector ETFs had the worst, with a weighted average return of 1.53%. - **By Theme**: Chip semiconductor and non - banking ETFs performed well, with weighted average returns of 4.98% and 4.93% respectively. Banking and innovative drug ETFs performed poorly, with weighted average returns of -2.87% and 0.88% respectively [21][24]. 2.2 ETF Fund Inflow and Outflow - **By ETF Category**: Industry - themed ETFs had the largest net inflow of funds at 231.53 billion yuan, while broad - based ETFs had the smallest at -148.46 billion yuan. - **By Listing Plate**: Hong Kong stock ETFs had the largest net inflow of funds at 106.88 billion yuan, while Science and Technology Innovation Board - related ETFs had the smallest at -39.97 billion yuan. - **By Industry Plate**: Cycle - sector ETFs had the largest net inflow of funds at 72.00 billion yuan, while biomedical - sector ETFs had the smallest at -4.03 billion yuan. - **By Theme**: Non - banking and robot ETFs had the largest net inflows of funds, at 59.10 billion yuan and 14.15 billion yuan respectively. Chip semiconductor and dividend ETFs had the smallest, at -10.89 billion yuan and -7.41 billion yuan respectively [2][29][32]. 2.3 ETF Trading Volume - **By ETF Category**: The daily average trading volume change rate of industry - themed ETFs increased the most, by 28.96%, while that of bond ETFs decreased the most, by -6.75%. - **By Listing Plate**: The daily average trading volume change rate of Japanese stock ETFs increased the most, by 72.66%, while that of US stock ETFs decreased the most, by -17.66%. - **By Industry Plate**: The daily average trading volume change rate of cycle - sector ETFs increased the most, by 98.87%, while that of biomedical - sector ETFs increased the least, by 17.86%. - **By Theme**: Non - banking and innovative drug ETFs had the largest 5 - day average daily trading volumes, at 290.98 billion yuan and 97.35 billion yuan respectively. Central state - owned enterprises and chip semiconductor ETFs had the largest increases in daily average trading volume change rate, at 39.26% and 37.77% respectively. Consumer electronics and military industry ETFs had the largest decreases, at -11.92% and -6.02% respectively [38][41][44][45]. 2.4 ETF Margin Trading During the sample period, the net margin purchase of all equity ETFs was -1.432 billion yuan, and the net short - selling was 0.47 billion yuan. The E Fund CSI Hong Kong Securities Investment Theme ETF had the largest net margin purchase, and the Huaxia CSI 1000 ETF had the largest net short - selling [2][51]. 2.5 ETF New Launches and Listings During the sample period, 10 funds were established and 8 funds were listed [3][53].

Core Insights - The latest data reveals that the scale of public funds, particularly ETFs, has reached 4.31 trillion yuan by the end of Q2, marking a quarter-on-quarter growth of 13.42% [1][3] - The stock-type ETFs account for a significant 75.06% of the total ETF scale, indicating a strong preference for equity investments among fund managers [3] - The "national team," represented by Central Huijin Asset Management, has actively increased its holdings in at least 8 leading ETFs during Q2, with a total estimated investment exceeding 190 billion yuan [1][4][5] ETF Scale and Performance - As of the end of Q2, the total ETF scale is approximately 4.31 trillion yuan, up from 3.8 trillion yuan at the end of Q1, reflecting a growth of 13.42% [3] - Six stock-type ETFs have surpassed 100 billion yuan in scale, with the top three being: - Huatai-PB CSI 300 ETF: 3747.04 billion yuan (up 10.62%) - E Fund CSI 300 ETF: 2665.16 billion yuan (up 13.4%) - Huaxia CSI 300 ETF: 1967.01 billion yuan (up 23.82%) [3] - Other notable ETFs exceeding 100 billion yuan include: - Harvest CSI 300 ETF: 1695.71 billion yuan (up 14.43%) - Huaxia SSE 50 ETF: 1654.44 billion yuan (up 13.52%) - Southern CSI 500 ETF: 1134.38 billion yuan (up 19.57%) [3] National Team's Investment Activity - The "national team" has shown a clear trend of increasing their holdings in major ETFs, with significant purchases noted in Q2: - Huatai-PB CSI 300 ETF: 108.74 billion shares, approximately 420 billion yuan - E Fund CSI 300 ETF: 84.29 billion shares, approximately 310 billion yuan - Huaxia CSI 300 ETF: 92.88 billion shares, approximately 350 billion yuan - Harvest CSI 300 ETF: 55.4 billion shares, approximately 220 billion yuan [4][5] - Overall, the estimated total investment by the national team in these 8 ETFs during Q2 is around 1930 billion yuan [5] Market Outlook - Analysts suggest that the ongoing support from the national team signals a commitment to stabilizing the market, especially during periods of volatility [5][6] - The national team's future investment decisions will depend on the improvement of policy and external environments, as well as the stock market's valuation levels [7] - The A-share market is expected to continue presenting structural opportunities, with a focus on the cost-effectiveness of broad-based index investments [7]

- The report uses the Mahalanobis distance of weekly price and trading volume changes to measure industry crowding levels[3][17] - The construction process involves identifying industries in the first quadrant (price and volume both rising) and the third quadrant (price and volume both falling) and marking points outside the ellipse as industries with significant short-term deviations at a 99% confidence level[17] - The building materials industry experienced short-term trading overselling last week[3][18]

Core Viewpoint - The article introduces a comprehensive index map that includes various commonly used stock indices, their codes, selection rules, industry distribution, average and median market capitalization of constituent stocks, and the number of constituent stocks, which will be regularly updated for easy reference [1][2]. Group 1: Types of Indices - The index map includes several categories of stock indices: broad-based indices, strategy indices, industry indices, thematic indices, and overseas indices [4][8]. Group 2: Broad-based Indices - Examples of broad-based indices include: - CSI 300 (000300.SH): Comprises 300 large-cap stocks from the Shanghai and Shenzhen stock exchanges, with an average market cap of 1,888.48 billion and a median of 907.96 billion [5]. - CSI 500 (000905.SH): Includes 500 stocks ranked 301-800 by market cap, with an average market cap of 279.59 billion and a median of 259.32 billion [5]. - CSI 800 (000906.SH): Covers 800 stocks, with an average market cap of 882.92 billion and a median of 340.58 billion [5]. Group 3: Strategy Indices - Strategy indices focus on specific investment strategies, such as: - CSI Dividend (000922.CSI): Selects 100 stocks with high dividend yields and stable dividends, with an average market cap of 1,970.77 billion [6]. - Shanghai Dividend (000015.SH): Comprises 50 stocks with high dividend yields from the Shanghai Stock Exchange, with an average market cap of 2,828.36 billion [6]. - Shenzhen Dividend (399324.SZ): Includes 40 stocks with stable dividend histories and high dividend ratios, with an average market cap of 984.44 billion [6]. Group 4: Industry Indices - Industry indices represent specific sectors, such as: - CSI Medical (930641.CSI): Selects stocks involved in traditional Chinese medicine production and sales, with an average market cap of 151.90 billion [7]. - CSI Real Estate (399393.SZ): Comprises 50 stocks from the real estate sector with significant market capitalization and liquidity, with an average market cap of 189.85 billion [7]. Group 5: Thematic Indices - Thematic indices focus on specific investment themes, such as: - CSI Consumer (000932.SH): Selects major consumer industry stocks from the CSI 800 index, with an average market cap of 1,155.60 billion [7]. - CSI Innovation (931152.CSI): Includes up to 50 representative companies involved in innovative drug research, with an average market cap of 485.93 billion [7].

- The report introduces the **RSI (Relative Strength Index)** as a quantitative factor, constructed to measure market conditions by comparing average gains and losses over a 12-day period. The formula is: $ RSI = 100 - 100 / (1 + RS) $ where RS represents the ratio of average gains to average losses over the specified period. RSI values above 70 indicate overbought conditions, while values below 30 suggest oversold conditions [2] - Another quantitative factor mentioned is **Net Purchase Amount (NETBUY)**, which calculates the net inflow or outflow of funds for ETFs. The formula is: $ NETBUY(T) = NAV(T) - NAV(T-1) * (1 + R(T)) $ where NAV(T) is the net asset value of the ETF on day T, NAV(T-1) is the net asset value on the previous day, and R(T) is the return on day T [2] - The report also tracks **Institutional Holdings Ratio**, which estimates the proportion of ETF shares held by institutions based on the latest annual or semi-annual reports. This excludes holdings by linked funds and may involve approximations due to data limitations [3] - The report provides **daily intra-day price trends** for ETFs using 5-minute interval data, highlighting the highest and lowest prices of the day with red markers. However, some data gaps may exist due to limitations in the source [2] - The report categorizes ETFs into "Broad-based" and "Thematic" groups based on the indices they track, such as major indices like CSI 300, CSI 500, and industry-specific indices like Non-bank Financials and Red Chips [2] - The report includes a detailed table of ETF performance metrics, such as RSI values, net purchase amounts, trading volumes, management fees, institutional holdings ratios, and T+0 trading availability. For example, the RSI for CSI 300 ETFs ranges from 68.14 to 75.77, while institutional holdings ratios vary significantly across ETFs [4] - The report highlights thematic ETFs such as **Consumption Electronics**, **Non-bank Financials**, **Renewable Energy**, **Semiconductors**, and **Artificial Intelligence**, providing metrics like RSI, net purchase amounts, and institutional holdings ratios. For instance, the RSI for Consumption Electronics ETFs ranges from 59.68 to 60.40, while institutional holdings ratios range from 23.10% to 58.47% [6] - The report also includes performance metrics for international ETFs tracking indices like the **Hang Seng**, **Nikkei 225**, **Nasdaq 100**, and **S&P 500**, with RSI values ranging from 46.50 to 72.83 and institutional holdings ratios varying widely [4][6]

- The report uses Mahalanobis distance to measure industry crowding based on weekly price and transaction volume changes[3][17][18] - The construction process involves identifying industries where price and transaction volume deviate significantly, with industries outside the ellipse in quadrant 1 indicating short-term significant crowding[17] - Last week, the building materials industry showed significant trading crowding[18]

Group 1: Report Overview - The report focuses on the performance comparison benchmarks of US active funds, aiming to provide insights for China's public fund market after the release of the "Action Plan for Promoting the High - quality Development of Public Funds" [2] Group 2: Investment Rating - Not provided in the report Group 3: Core Views - The US has established requirements for performance comparison benchmarks with broad - based indices as the main and narrow - based indices as supplementary. The CFA Institute also offers benchmark - setting guidelines [4][9] - US active funds mainly use single - index benchmarks. Stock - type funds use S&P 500 as a single - benchmark index; multi - benchmark funds prefer broad - based and narrow - based index combinations. Hybrid funds often use composite benchmarks, and bond - type funds have concentrated single - benchmarks and diverse multi - benchmarks [4][22] - US stock - type funds with S&P 500 as the benchmark have higher correlation, lower tracking error, and a lower proportion of significantly underperforming the benchmark compared to Chinese ordinary stock - type funds with CSI 300 as the main benchmark [5][53] - Capital Group and Fidelity, two leading active equity fund companies, have different benchmark - setting characteristics. Capital Group mainly uses single - benchmarks, while Fidelity has a more balanced distribution of single - and multi - benchmarks [61] Group 4: Summary by Directory 1. US Active Fund Performance Comparison Benchmark Overview - **Performance Comparison Benchmark Policy**: Since 1993, the SEC has required funds to compare their total returns with the total returns of appropriate broad - based indices, and also encourages the use of narrow - based indices. In 2022, the definition of broad - based indices was revised. The CFA Institute also provides benchmark - setting guidelines [9][10][13] - **US Active Fund Classification**: According to SEC naming rules, 80% of a fund's assets should be invested in line with its name. The ICI classifies mutual funds into major asset categories. As of April 2025, the US mutual fund market was worth $27.97 trillion, with stock - type funds being the largest in scale [15][16] - **US Active Fund Performance Comparison Benchmark Type Distribution**: Among 4938 US active mutual funds, 56.3% are stock - type funds and 32.4% are bond - type funds as of March 17, 2025. 63.4% of funds use single - benchmarks, 31.6% use multi - benchmarks, and 5.0% use composite benchmarks [19][22] 2. Stock - Type Fund Benchmark Analysis - **Single Benchmark**: Single - benchmark stock - type funds have high index concentration and diverse index selection, mainly using S&P 500. Among 1848 single - benchmark stock - type funds, S&P 500 is used 320 times [26] - **Multi - Benchmark**: Multi - benchmark stock - type funds often use broad - based and narrow - based index combinations. 846 out of 913 multi - benchmark stock - type funds use 2 indices as benchmarks. Large - scale multi - benchmark stock - type funds mainly use broad - based and style indices [30][35] 3. Hybrid Fund Benchmark Analysis - Among 239 hybrid funds, 122 use composite benchmarks, mostly composed of 2 indices. The equity index weight in composite benchmarks ranges from 5% to 85%. The most commonly used combination is S&P 500*60% + Bloomberg US Aggregate*40% [37][40] 4. Bond - Type Fund Benchmark Analysis - **Single Benchmark**: Bloomberg US Aggregate and Bloomberg Municipal are the most commonly used single - benchmarks for bond - type funds, with high benchmark concentration [46] - **Multi - Benchmark**: Multi - benchmark bond - type funds have diverse benchmark combinations, reflecting investment characteristics in regions, bond types, durations, and credit ratings. Large - scale multi - benchmark bond funds use diverse benchmark combinations [48][50] 5. US Active Fund Return vs Benchmark Comparison - **Correlation and Tracking Error Analysis**: The average correlation coefficient between US stock - type funds with S&P 500 as the benchmark and S&P 500 in the past three years is 0.91, higher than that of Chinese ordinary stock - type funds with CSI 300 as the main benchmark. The tracking error of US funds is also lower [53][54] - **Excess Return Analysis**: Less than 10% of US single - benchmark stock - type funds with S&P 500 as the benchmark significantly underperformed the benchmark in the past three years, a lower proportion compared to Chinese stock - type funds with CSI 300 as the main benchmark [59] 6. Benchmark Setting of Leading Active Equity Fund Companies - **Capital Group**: As of October 3, 2024, it had 94 products with a total management scale of $2.4 trillion. Stock - type funds accounted for 67% of the scale. The company mainly uses S&P 500 or MSCI ACWI as single - benchmarks [64][68] - **Fidelity**: As of October 4, 2024, its management scale was $2.95 trillion, with similar active and passive product scales. Stock - type funds accounted for 79% of the scale. Single - and multi - benchmark funds are evenly distributed, with single - benchmark funds mainly using S&P 500 and multi - benchmark funds using broad - based and industry/style index combinations [73][76] 7. Summary - The report introduces US active fund performance comparison benchmark policies and industry guidelines, and analyzes current benchmark - selection characteristics. US active funds mainly use single - index benchmarks, and different types of funds have different benchmark - selection preferences [84][85] - US stock - type funds with S&P 500 as the benchmark have better performance in terms of correlation, tracking error, and excess return compared to Chinese stock - type funds with CSI 300 as the main benchmark [86] - Capital Group and Fidelity have different benchmark - setting characteristics, and both show certain abilities to obtain excess returns [87]

Quantitative Factors and Construction Methods 1. Factor Name: RSI (Relative Strength Index) - **Factor Construction Idea**: RSI measures the relative strength of price movements over a specific period to identify overbought or oversold market conditions[2] - **Factor Construction Process**: - The formula for RSI is: $ RSI = 100 - \frac{100}{1 + RS} $ where $ RS $ is the ratio of the average gain to the average loss over a 12-day period[2] - RSI > 70 indicates an overbought market, while RSI < 30 indicates an oversold market[2] - **Factor Evaluation**: RSI is a widely used technical indicator for short-term market sentiment analysis[2] 2. Factor Name: Net Subscription (NETBUY) - **Factor Construction Idea**: This factor calculates the net subscription amount of ETFs to gauge investor demand[2] - **Factor Construction Process**: - The formula for NETBUY is: $ NETBUY(T) = NAV(T) - NAV(T-1) \times (1 + R(T)) $ where $ NAV(T) $ is the net asset value on day T, $ NAV(T-1) $ is the net asset value on the previous day, and $ R(T) $ is the return on day T[2] - **Factor Evaluation**: This factor provides insights into the capital flow dynamics of ETFs, reflecting investor sentiment and market positioning[2] --- Factor Backtesting Results 1. RSI Factor - **HS300 ETFs**: RSI values range from 60.24 to 70.30, with most ETFs hovering around the overbought threshold of 70[4] - **CSI500 ETFs**: RSI values range from 66.04 to 66.37, indicating moderate strength[4] - **CSI800 ETFs**: RSI values range from 69.54 to 71.77, with some ETFs entering overbought territory[4] - **CSI1000 ETFs**: RSI values range from 63.76 to 64.97, suggesting neutral to slightly strong market conditions[4] - **Thematic ETFs**: RSI values vary significantly, with some themes like "New Energy" reaching as high as 74.71, while others like "Semiconductors" remain around 50[6] 2. Net Subscription Factor - **HS300 ETFs**: Net subscription values range from -6.69 billion to 1.51 billion, indicating mixed investor sentiment[4] - **CSI500 ETFs**: Net subscription values range from -2.65 billion to 0.54 billion, reflecting weak demand[4] - **CSI800 ETFs**: Net subscription values range from 0.00 billion to 0.45 billion, showing stable capital flows[4] - **CSI1000 ETFs**: Net subscription values range from -2.69 billion to 0.55 billion, indicating varied investor interest[4] - **Thematic ETFs**: Net subscription values vary widely, with some themes like "Semiconductors" showing strong inflows (up to 6.28 billion), while others like "Consumer Electronics" exhibit moderate inflows (up to 1.15 billion)[6]

Group 1: Equity Market Performance - The domestic equity market showed strong performance last week, with major indices rising over 3% and daily trading volume increasing to around 1.5 trillion yuan [1] - The Shanghai Composite Index reached a new high for the year, surpassing the 3400-point mark, which is a significant resistance level [1] - Financial, computer, and military industries performed well, while the oil and petrochemical sectors saw declines due to falling oil prices [1] Group 2: Macroeconomic Observations - Despite significant tariff impacts and a slowdown in the domestic real estate market since April, the overall macroeconomic environment remains stable, supported by a strong industrial chain and manufacturing capabilities [1] - The policy of replacing old consumer goods has also contributed positively to the macroeconomic performance [1] Group 3: Bond Market Analysis - The bond market exhibited a fluctuating pattern last week, influenced by seasonal factors and the stock-bond relationship [2] - The strong performance of the stock market initially suppressed bond market performance, but increased liquidity from the central bank and insurance demand supported the bond market [2] - By the end of the week, the yields on 10-year government bonds and 30-year government bonds rose by 1 basis point, reaching 1.65% and 1.85% respectively [2]

Quantitative Models and Construction Methods Model 1: RSI (Relative Strength Index) - **Model Name**: RSI (Relative Strength Index) - **Model Construction Idea**: The RSI is used to measure the speed and change of price movements. It is primarily used to identify overbought or oversold conditions in a trading instrument. - **Model Construction Process**: - The RSI is calculated using the following formula: $$ RSI = 100 - \frac{100}{1 + RS} $$ where RS (Relative Strength) is the average of 'n' days' up closes divided by the average of 'n' days' down closes. Typically, a 14-day period is used. - The RSI value ranges from 0 to 100. An RSI above 70 indicates that the market is overbought, while an RSI below 30 indicates that the market is oversold.[2] Model 2: Net Purchase (NETBUY) - **Model Name**: Net Purchase (NETBUY) - **Model Construction Idea**: This model calculates the net purchase amount of an ETF to understand the inflow and outflow of funds. - **Model Construction Process**: - The net purchase amount is calculated using the following formula: $$ NETBUY(T) = NAV(T) - NAV(T-1) \times (1 + R(T)) $$ where NETBUY(T) is the net purchase amount on day T, NAV(T) is the net asset value on day T, NAV(T-1) is the net asset value on the previous trading day, and R(T) is the return on day T.[2] Model Backtesting Results - **RSI Model**: - RSI values for various ETFs range from 37.69 to 80.86, indicating different levels of market conditions from oversold to overbought.[4][6] - **Net Purchase Model**: - Net purchase values for various ETFs range from -6.38 billion to 99.72 billion, indicating significant variations in fund inflows and outflows.[4][6] Quantitative Factors and Construction Methods Factor 1: Institutional Holdings - **Factor Name**: Institutional Holdings - **Factor Construction Idea**: This factor measures the percentage of an ETF's holdings that are owned by institutional investors. - **Factor Construction Process**: - The percentage of institutional holdings is derived from the latest annual or semi-annual reports of the ETF, excluding the holdings of corresponding linked funds. The data is an estimate and may have some deviations.[3] Factor Backtesting Results - **Institutional Holdings Factor**: - Institutional holdings percentages for various ETFs range from 2.79% to 96.29%, indicating varying levels of institutional interest and confidence in these ETFs.[4][6]