广发证券资深金工分析师 张钰东 SAC: S0260522070006 zhangyudong@gf.com.cn SAC: S0260512020003 广发证券首席金工分析师 安宁宁 anningning@gf.com.cn 广发证券 资深金工分析师 陈原文 SAC: S0260517080003 chenyuanwen@gf.com.cn 广发金工安宁宁陈原文团队 摘要 研究背景： 股票市场存在一定的隔夜相关性特征，日度收益可以拆解为隔夜收益和日间收益。结合近期学术成果，本报告从日间、隔夜等特征相关 度出发，刻画相似股票的关联特征。 隔夜涨跌幅相关性研究思路： 将多空信号和交易执行进行分离，通过信号与执行分离的机制，只捕捉跨股票的信息效应。基于夜收益和日间收益来 构建相关性矩阵，然后划分领先群组（Leader）和滞后群组（Lagger），在此基础上构建交易策略，仅从领先群组生成信号，仅在滞后群组内交 易。 实证研究： 基于相关文献方案，潜在的问题在于股票之间的特征值差异相对较小，进一步尝试直接基于特征进行KMEANS聚类分析，并基于平均数 值强度确认领先群组和滞后群组。测算显示，日度调仓下，A股的领先滞 ...

在好莱坞拍摄过的众多历史题材的电影中，有一部深入人心，那就是梅尔-吉布森主演的《勇敢的心》。在这部电影中，吉布森扮演的 威廉·华莱士 率领苏 格兰战士们反抗英格兰国王爱德华一世的军事征服。电影所展现的，正是苏格兰历史上一段波澜壮阔的篇章，那就是 13世纪末至14世纪初 苏格兰独立战 争 。 尽管华莱士本人最终被捕并遭处决，但他与后来的民族英雄罗伯特·布鲁斯共同点燃了苏格兰不屈的抗争之火。苏格兰人民凭借坚韧的民族意志和在班诺克 本战役等关键战场上的血战，最终成功地捍卫了自己的主权独立。在那个时代，苏格兰人向全世界证明了一件事：在面对民族大义时，他们宁愿流血，也不 会屈服。他们用剑和血肉，保住了自己的独立和骨气。 然而，军事的胜利并不意味着经济的昌盛。在随后的几个世纪里，苏格兰虽然保持了独立的主权，但其经济地位却日益尴尬。欧洲贸易的重心已经转移，相 较于财富滚滚而来的邻居英格兰，苏格兰在地理和气候上都不占优势，经济发展长期处于停滞状态。进入17世纪，尽管苏格兰国王詹姆士六世继承了英格 兰王位，实现了王室联合 （Union of the Crowns），但两国在议会和经济上仍是分离的。 英格兰将苏格兰视为经济上的竞 ...

Quantitative Models and Construction Methods - **Model Name**: Minute Herding Effect Factor Cluster **Construction Idea**: Focus on the trading behavior of followers after significant actions by "trend funds" using minute-level data [13][14][18] **Construction Process**: 1. **Event Identification**: Detect actions of trend funds through anomalies in volume, price changes, volatility, and price-volume correlation [13][14] 2. **Factor Definition**: Measure herding strength by analyzing post-event price, volume, price-volume correlation, and other metrics [14][18] 3. **Data Frequency**: Use minute-level data to identify events and define factors [14][18] **Evaluation**: Effective in capturing herding behavior at the minute level [18] - **Model Name**: Tick-by-Tick Herding Effect Factor Cluster **Construction Idea**: Apply discrete factor definitions directly to tick-by-tick data to capture herding effects [1][11][20] **Construction Process**: 1. **Event Identification**: Identify liquidity shock events using tick-by-tick order and trade data, introducing the concept of "aggressiveness" for orders [21][22][25] 2. **Factor Definition**: Analyze post-event metrics such as order volume, trade volume, imbalance indicators, and price-volume correlation [30][31][61] 3. **Factor Production**: Generate approximately 20,000 factors, retaining the top 50 based on performance and low correlation [63][84] **Evaluation**: Demonstrates strong predictive power with annual ICIR values exceeding 2 [63][84] - **Model Name**: Tick-by-Tick Herding Effect Composite Factor **Construction Idea**: Combine the top 10 factors with the highest information ratio into a composite factor [67][85] **Construction Process**: 1. Select the top 10 factors based on information ratio from the tick-by-tick factor cluster [67][85] 2. Equally weight these factors to create the composite factor [67][85] **Evaluation**: Highly effective with robust performance metrics, even after neutralizing common style and industry factors [67][71][85] Model Backtesting Results - **Minute Herding Effect Composite Factor**: - Monthly IC Mean: 0.085 - Annual ICIR: 3.18 - Monthly RankIC Mean: 0.116 - Annual RankICIR: 4.10 - Annual Return: 41.59% - Annual Volatility: 12.56% - Information Ratio: 3.31 - Monthly Win Rate: 82.91% - Maximum Drawdown: 10.06% [18] - **Tick-by-Tick Herding Effect Factor Cluster**: - Annual ICIR Absolute Value: >2 for all 50 factors [63][65] - Example Factor (Factor 16): - Monthly IC Mean: 0.057 - Annual ICIR: 2.82 - Monthly RankIC Mean: 0.072 - Annual RankICIR: 3.01 - Annual Return: 25.86% - Annual Volatility: 9.11% - Information Ratio: 2.84 - Monthly Win Rate: 76.92% - Maximum Drawdown: 6.38% [64][65][66] - **Tick-by-Tick Herding Effect Composite Factor**: - Monthly IC Mean: 0.080 - Annual ICIR: 3.49 - Monthly RankIC Mean: 0.101 - Annual RankICIR: 3.74 - Annual Return: 44.26% - Annual Volatility: 10.90% - Information Ratio: 4.06 - Monthly Win Rate: 89.74% - Maximum Drawdown: 10.66% [67][85] - **Pure Tick-by-Tick Herding Effect Composite Factor** (Neutralized for Style and Industry): - Monthly IC Mean: 0.044 - Annual ICIR: 3.33 - Monthly RankIC Mean: 0.046 - Annual RankICIR: 3.03 - Annual Return: 19.53% - Annual Volatility: 6.36% - Information Ratio: 3.07 - Monthly Win Rate: 78.63% - Maximum Drawdown: 5.13% [71][85] Index Enhancement Portfolio Performance - **CSI 300 Index Enhancement Portfolio**: - Excess Annual Return: 8.89% - Tracking Error: 3.50% - Information Ratio: 2.54 - Monthly Win Rate: 77.78% - Maximum Drawdown: 2.96% [75][86] - **CSI 500 Index Enhancement Portfolio**: - Excess Annual Return: 13.46% - Tracking Error: 5.31% - Information Ratio: 2.54 - Monthly Win Rate: 79.49% - Maximum Drawdown: 5.15% [78][86] - **CSI 1000 Index Enhancement Portfolio**: - Excess Annual Return: 17.23% - Tracking Error: 4.78% - Information Ratio: 3.61 - Monthly Win Rate: 84.62% - Maximum Drawdown: 4.14% [80][86]

Core Viewpoint - The article discusses the tendency of retail investors in the A-share market to engage in short-term trading despite evidence suggesting that long-term holding of quality stocks yields higher returns. It highlights the psychological factors driving this behavior and the resulting financial consequences. Group 1: Retail Investor Behavior - Retail investors in the A-share market have an average holding period of only 32 days, with an annual turnover rate exceeding 600% [1] - Investors who hold quality stocks for over five years have a threefold higher probability of making a profit compared to short-term traders [1] - The allure of immediate financial gratification leads many investors to prefer short-term trading over long-term strategies [2] Group 2: Psychological Factors - The human tendency for instant feedback drives retail investors to engage in short-term trading, as they can see daily price fluctuations and realize profits quickly [2] - Retail investors often perceive themselves as "prophets," relying on market rumors and trends rather than fundamental analysis, which leads to poor investment decisions [4][5] - Behavioral finance concepts such as greed and fear significantly impact retail investors, causing them to make irrational decisions during market fluctuations [6][8] Group 3: Market Dynamics - The A-share market is characterized by a high proportion of retail trading, with nearly 80% of transactions coming from retail investors, leading to a high turnover rate and a tendency for "bulls to be short-lived" [8] - The prevalence of short-term trading creates a market environment where retail investors frequently chase trends, often resulting in losses when market conditions change rapidly [4][10] - Stories of short-term trading success are often amplified, overshadowing the more common experience of long-term investors who quietly accumulate wealth [10]

Core Viewpoint - The article discusses the impact of "narrative trading" on asset pricing, emphasizing that asset pricing is influenced not only by fundamentals but also by popular narratives such as the restructuring of the dollar credit system and the new technological revolution [1][12]. Group 1: Narrative Economics - The influence of narratives on economic phenomena consists of a series of elements: a popular, easily spread story, public behavior, and an epidemiological model for macro-level dissemination [2][16]. - The concept of "herding behavior" is used to illustrate how narratives affect micro-level decision-making, with varying strengths across different phases of narrative development [2][18]. Group 2: Herding Effect in Asset Allocation - Traditional studies of herding behavior focus on individual stocks and short-term market sentiment, but the current narrative-driven environment poses challenges for asset allocation due to the breakdown of continuity in global fiscal, monetary, and trade environments [3][20]. - The article suggests that the herding effect can be quantified and applied to investment portfolio optimization and asset timing strategies [3][20]. Group 3: Measurement of Herding Effect - Four common indicators of herding behavior are identified: Cross-Sectional Absolute Deviation (CSAD), the quadratic coefficient of return dispersion, standard deviation of beta coefficients, and cross-correlation [4][23]. - The CSAD index, which measures the deviation of asset returns from the average, indicates the presence of herding behavior when returns cluster around a certain average level [4][23]. Group 4: Current State of Herding Effect - The CSAD index for major asset classes shows a right-skewed distribution, indicating a tendency for extreme herding behavior, with a mean-reverting characteristic suggesting that extreme trends are difficult to maintain [5][28]. - Since May 2025, the CSAD has decreased significantly, indicating a rapid herding effect, but has started to rebound slightly, suggesting a potential shift towards more balanced asset performance [5][28]. Group 5: Strategy Integration - The article proposes integrating the herding factor into a macro risk parity framework, which has shown superior annualized returns compared to traditional models [6][34]. - The new framework suggests increasing allocations to equities and commodities while reducing bond exposure, indicating a shift in investment strategy based on herding behavior [6][34]. Group 6: Domestic Equity Market Analysis - The herding effect in the domestic equity market, as measured by the CSAD, has shown a decline in right-skewness, indicating lower dispersion compared to historical levels [7][40]. - The herding effect has gone through phases of fermentation, intensification, and now a slight loosening, suggesting a gradual return to individual rationality among investors [7][40].

Core Insights - The People's Bank of China (PBOC) has established a macro-prudential policy framework post the 2008 financial crisis, leading to a unique management practice in China [1] Summary by Categories Governance Mechanism - Strengthened the centralized leadership of the Communist Party and enhanced the PBOC's macro-prudential management functions [1] Policy Framework - Released the "Macro-Prudential Policy Guidelines" in 2021, clarifying the management approach and policy framework [1] - Established a differentiated reserve requirement system in 2003, introduced a dynamic adjustment mechanism in 2010, and upgraded to Macro-Prudential Assessment (MPA) in 2016 to promote stable growth in monetary credit [1] Regulatory Framework - Developed a comprehensive regulatory framework for systemically important financial institutions, including guidelines and assessment methods for systemically important banks and insurance companies [1] Cross-Border Financing - Set up macro-prudential adjustment parameters for cross-border financing to implement counter-cyclical adjustments on capital flows [1] Financial Market Management - Conducted dynamic observation and assessment of bond market operations, enhancing risk alerts for financial institutions to mitigate risk accumulation [1] - Collaborated with the China Securities Regulatory Commission to establish two monetary policy tools to support the capital market [1] Currency Stability - Maintained the decisive role of the market in exchange rate formation, ensuring the stability of the RMB at a reasonable and balanced level to prevent significant volatility risks [1] Real Estate Financial Management - Dynamically adjusted mortgage down payment ratios and interest rates as part of the macro-prudential management of real estate finance [1] Financial Holding Companies - Established a regulatory framework for financial holding companies, which is now under the purview of the Financial Regulatory Bureau [1] Market Expectation Management - Actively managed market expectations during significant market fluctuations to correct "herd behavior" and maintain stability in stock, bond, and foreign exchange markets [1]

Core Insights - The Chinese AI industry is at a critical juncture, with the financial sector leading in technology adoption while recognizing associated risks [1][2] - Financial institutions are heavily investing in AI, with state-owned banks like Bank of Communications allocating significant resources to digital transformation [1] - AI applications in finance have shown substantial efficiency improvements, particularly in retail lending and risk management [1] Investment in AI - Bank of Communications has invested approximately 12 billion RMB annually in technology, representing about 5.4% of total revenue since 2021 [1] - The bank's workforce includes over 10,000 technology personnel, accounting for more than 10% of total employees [1] AI Applications and Efficiency - AI has improved service efficiency in retail lending by 3.5 times through end-to-end applications in credit access, marketing, and review processes [1] - AI technology has achieved over 80% accuracy in fraud prevention [1] - Operational management has seen a release of over 60% of manual productivity through AI authorization processes [1] Risks Associated with AI - Potential risks in AI applications include cybersecurity, data security, and model safety, necessitating clearer boundaries between public and private data rights [2] - The need for enhanced personal privacy protection is emphasized as data collection increases [2] Value Judgment and Market Stability Risks - The risk of "value deviation" arises from the public's limited financial knowledge, which may lead to skewed perceptions due to information silos [3] - The "herding effect" could pose risks to market stability if financial institutions utilize homogeneous large models for investment advice and risk assessment, potentially leading to market and liquidity risks [3] Human Oversight in Financial Decisions - It is crucial for humans to remain in control of financial decision-making, as AI lacks the emotional intelligence necessary for responsible financial management [3]

Core Viewpoint - Recent fluctuations in gold prices have led to a significant drop, causing uncertainty among buyers regarding whether to purchase or wait for further price changes [1][4][7]. Group 1: Gold Market Dynamics - On October 24, gold prices experienced a sharp decline, with spot gold dropping by 1.92% to $4054.44 per ounce, while COMEX futures fell by 1.91% to $4066.4 per ounce [1]. - The decline in gold prices has not resulted in a surge of buying activity; many consumers remain hesitant to purchase gold despite the price drop [7][9]. - In Beijing, there has been an increase in customers looking to buy investment gold bars, with reports of shortages in 10-gram investment gold bars due to heightened demand [9][10]. Group 2: Consumer Behavior and Sentiment - Many consumers are experiencing a "wait and see" approach, with some expressing a desire to wait for further price drops before making purchases [9][10]. - A notable trend is the "herd effect," where consumers tend to buy more as prices rise and hesitate when prices fall, leading to emotional decision-making [13]. - Some consumers have successfully capitalized on price fluctuations, with individuals reporting profits from selling gold bars at higher prices than their purchase costs [10][11]. Group 3: Investment Strategies - Experts suggest that consumers looking to invest in gold should consider purchasing investment-grade gold products, such as gold bars or ETFs, rather than jewelry, which incurs higher processing fees [17]. - A long-term investment strategy, such as regular purchases of small amounts of gold, is recommended for those who view gold as a savings tool rather than a speculative investment [15].

Market Overview - On October 24, gold prices experienced a significant drop, with spot gold falling by 1.92% at one point, and later narrowing the decline to 1.75%, priced at $4054.44 per ounce. COMEX futures gold also dropped by 1.91%, reaching $4066.4 per ounce. Silver prices followed suit, with spot silver down 1.94% to $47.9 per ounce [2]. Consumer Behavior - Despite the recent drop in gold prices, there has not been a surge in gold purchases. Reports from various gold shops in Shanghai indicate a lack of customer activity, with many potential buyers hesitant due to uncertainty about future price movements [3]. - In a specific Chinese gold store, only one customer was browsing, expressing reluctance to buy at current prices, which she still considered high. She indicated that she would only consider purchasing if prices dropped further [5]. - In contrast, there has been an increase in customers looking to sell gold, particularly at Beijing Cai Bai, where the gold buyback counter saw long queues. The sales staff noted a rise in customers buying investment gold bars, with some items currently out of stock [5][6]. Investment Insights - Some investors are taking advantage of the current market conditions to realize profits. For instance, one investor sold 100 grams of gold bars at approximately 931 yuan per gram, having initially purchased them at under 300 yuan per gram, resulting in a profit of over 60,000 yuan [6]. - Another investor reported selling 200 grams of gold bars, making a profit of about 18,000 yuan, and is now looking to reinvest when prices drop again [6]. - The current market sentiment reflects a typical herd mentality, where consumers are more inclined to buy when prices are rising and hesitant to purchase during declines, leading to emotional decision-making [7]. Investment Strategies - Experts suggest that consumers looking to invest in gold should avoid purchasing gold jewelry due to high processing fees that diminish resale value. Instead, they recommend investment-grade gold products such as gold bars, panda coins, or ETFs for better returns [11]. - Some consumers have adopted a systematic investment approach, treating gold purchases as a long-term savings strategy, which helps mitigate the impact of short-term price fluctuations [9].

Core Viewpoint - The recent sharp decline in gold prices, with a 6% drop on October 22 marking the largest single-day decrease in twelve years, has left many potential buyers in a dilemma about whether to purchase gold now or wait for further price changes [1][3][16]. Group 1: Price Movement - Gold prices fell from 1292 CNY per gram on October 21 to 1235 CNY on October 22, and further down to 1223 CNY on October 23, totaling a decline of 69 CNY per gram over two days [3][5]. - The significant drop in gold prices has not led to an expected surge in customer traffic at gold stores, as many consumers still perceive current prices as high [3][5]. Group 2: Consumer Behavior - Despite the price drop, consumers like Ms. Huang express reluctance to buy, stating that gold remains expensive and indicating a desire for further price reductions before making a purchase [5][12]. - Some consumers are opting to buy gold in Hong Kong due to lower prices and promotional offers, despite the additional costs of travel [12][18]. Group 3: Market Sentiment - The volatility in gold prices has created a challenging environment for buyers, with many feeling anxious about making the right decision [14][16]. - The phenomenon of "buying high and selling low" reflects a typical herd mentality among consumers, leading to fluctuating confidence in gold investments [16][18]. Group 4: Investment Insights - Experts suggest that consumers looking to invest in gold should consider purchasing investment-grade gold products, such as gold bars or ETFs, rather than gold jewelry, which carries high processing fees [18].