Core Viewpoint - The small-cap stocks, represented by the CSI 2000 and National CSI 2000 indices, have shown a significant recovery after a decline of over 6% since reaching their peak on August 27, with various ETFs tracking these indices experiencing notable gains in recent trading sessions [1][3]. Group 1: Market Performance - The CSI 2000 index and National CSI 2000 index saw a cumulative decline of over 6% since their peak on August 27, but have since continued to rise [1]. - As of September 4, several ETFs tracking the CSI 2000 index reported daily gains ranging from 1.76% to 2.01%, with year-to-date performance showing increases between 28.94% and 53.11% [3][4][8]. - The Wind Micro-cap Index has surged by 68.72% year-to-date, while the CSI 2000 and CSI 1000 indices have increased by 32.47% and 24.47%, respectively [6]. Group 2: Fund Performance and Flows - The largest ETF tracking the CSI 2000 index is the Huatai-PB CSI 2000 ETF, with a latest scale of 2.329 billion [7][10]. - The year-to-date performance of various ETFs shows that the China Merchants CSI 2000 Enhanced ETF leads with a gain of 53.11%, followed by the Silver Hua CSI 2000 Enhanced ETF at 45.21% [8][10]. - In terms of net fund flows, the Huatai-PB CSI 2000 ETF experienced a net outflow of 1.467 billion, while the China Merchants CSI 2000 Enhanced ETF saw a net inflow of 585 million [8]. Group 3: Investment Trends - The current market environment favors small-cap stocks due to a focus on marginal changes in industry structure during periods of rapid technological iteration and policy encouragement for innovation [6]. - Historical patterns suggest that micro-cap stocks may face a weakening trend, although structural opportunities may still exist [7]. - The investment landscape is characterized by institutional investors holding significant pricing power, which typically benefits large-cap stocks, while individual investors tend to favor small-cap stocks [6].

Quantitative Models and Construction Methods 1. Model Name: Wantushi AI Model - **Model Construction Idea**: The Wantushi AI model evaluates indices based on their potential upward probability and selects ETFs with favorable characteristics for investment[6] - **Model Construction Process**: 1. The model assigns a score to indices based on their upward probability. Indices with scores above 0.8 are selected[6] 2. For each selected index, the corresponding ETFs are identified[6] 3. Among these ETFs, those with a 30-day average daily trading volume exceeding 30 million RMB are retained[6] 4. Finally, the ETF with the lowest IOPV premium rate for each index is chosen[6] - **Model Evaluation**: The model systematically filters ETFs based on liquidity and valuation metrics, ensuring a focus on high-quality investment options[6] --- Backtesting Results of Models 1. Wantushi AI Model - **Index Code: 932000 (CSI)** - Upward Probability: 93.52% - Selected ETF: 159552 (CSI 2000 Enhanced ETF)[7] - **Index Code: 399006 (SZ)** - Upward Probability: 93.37% - Selected ETF: 159977 (Tianhong ChiNext ETF)[7] - **Index Code: 399673 (SZ)** - Upward Probability: 92.49% - Selected ETF: 159682 (ChiNext 50 ETF)[7] - **Index Code: 399850 (SZ)** - Upward Probability: 90.15% - Selected ETF: 159350 (Shenzhen 50 ETF by Fuguo)[7] - **Index Code: 000852 (SH)** - Upward Probability: 85.86% - Selected ETF: 159680 (1000 Enhanced ETF)[7] - **Index Code: 399303 (SZ)** - Upward Probability: 85.82% - Selected ETF: 159628 (Guozheng 2000 ETF)[7] - **Index Code: 399293 (SZ)** - Upward Probability: 85.03% - Selected ETF: 159814 (ChiNext Large Cap ETF)[7] - **Index Code: 399330 (SZ)** - Upward Probability: 84.81% - Selected ETF: 159901 (Shenzhen 100 ETF)[7] - **Index Code: 399296 (SZ)** - Upward Probability: 84.46% - Selected ETF: 159967 (ChiNext Growth ETF)[7] - **Index Code: 000905 (SH)** - Upward Probability: 83.23% - Selected ETF: 510580 (CSI 500 ETF by E Fund)[7] - **Index Code: 000906 (SH)** - Upward Probability: 81.05% - Selected ETF: 515800 (800 ETF)[7]

Group 1 - The strategy for low to medium risk asset allocation includes risk parity and risk budgeting models, where risk parity allocates equal risk weights across different assets, while risk budgeting allows investors to set asset risk weights based on their risk preferences [1] - The correlation between major asset classes such as equities (A-shares, Hong Kong stocks, US stocks), bonds, and commodities (precious metals, energy, chemicals) is relatively low, making it suitable to construct portfolios using corresponding ETFs [1] - The long-term correlation between bonds and equities or commodities ranges from 0 to -30%, indicating a "stock-bond seesaw" effect due to the counter-cyclical nature of interest rates affecting bond yields, while equities and commodities reflect the health or expectations of the real economy [1] Group 2 - A simple construction method for the model involves selecting broad-based indices for equities such as CSI 300 ETF, CSI 500 ETF, ChiNext ETF, and National 2000 ETF, while the bond portion can include government bond ETFs, policy financial bond ETFs, and local government bond ETFs [2] - For the commodity portion, gold ETFs and commodity futures ETFs can be included, with advanced construction methods allowing for a core-satellite approach or sector rotation strategy for equities [2]

深圳证券交易所 SHENZHEN SHENZHEN STOCK EXCHANGE 深交所ETF投资问答(39) kes i m g T t x 154 B d UN - 编者按 - 」 风险平价／风险预算模型 ▪ 风险预算放型 放松对风险均衡配 置的严格要求,投 资者可以根据自身 风险偏好去设定各 资产风险权重。 风险平分旗型 属于风险预算模型 的一个特例,是对 投资组合中不同资 产分配相同的风险 权重的一种资产配 置理念。 选择大类资产 E3 商品 les ... 风险预算模型 风险模型 2% 2% 2 000 资产组合构建 举列 大类资产之间的相关性较低，适合选 ....... 择相应的ETF资产构建组合。 例如,债券和权益类、商品类ETF的长 期相关性在0至-30%之间，"股债晓晓 板"模式长期存在,短期也会有"股债 双杀"的局面。 近年来我国指数型基金迅速发展,交易型开 放式指数基金(ETF) 备受关注。为帮助广 大投资者系统全面认识ETF，了解相关投资 方法,特摘编由深圳证券交易所基金管理部 编著的《深交所ETF投资问答》（中国财政 经济出版社2024年版）形成图文解读。本 篇是第39期，继续为大家 ...