Core Viewpoint - The A-share market experienced a rebound this week, with a notable shift towards growth sectors, although overall capital withdrawal and reduced trading volume indicate a cautious market sentiment ahead of the holiday [2][3][5]. Market Overview - The A-share market rebounded significantly after a previous week's decline, with small-cap and growth sectors outperforming large-cap indices, marking a transition from a "value defense" phase to a "growth offensive" structure [3][7]. - The Sci-Tech 50 Index showed the most remarkable performance, rebounding strongly after a deep correction last week, leading all major indices in gains [3][7]. Industry Performance - The market displayed a clear rotation pattern, with the TMT (Technology, Media, and Telecommunications) sector leading the gains, while defensive sectors like consumer and financials continued to decline [10][24]. - High-growth and high-elasticity sectors attracted significant capital attention, while defensive sectors and previously strong sectors experienced capital outflows [10][24]. Trading Volume and Capital Flow - Despite the index rebound, there was a notable "price increase with volume decrease" phenomenon, with trading volume and turnover both declining by 12.3% and 13.6% respectively compared to the previous week [24][25]. - Main capital flows shifted from net inflows at the beginning of the week to a cumulative net outflow of 994.97 billion yuan by the end of the week, indicating rising risk aversion ahead of the holiday [34][36]. Individual Stock Performance - The top gainers were predominantly small-cap stocks driven by technology attributes and thematic factors, while the top losers included high-position stocks and traditional industry names [19][21]. - The top ten stocks by net inflow were mainly technology-related, reflecting a concentrated capital focus on this sector [40]. Market Sentiment - The number of stocks hitting the daily limit up showed a "stair-step decline," indicating a weakening buying momentum, while the number of stocks hitting the limit down increased, suggesting a shift towards caution in market sentiment [42][43]. - Margin trading data also indicated a slight decrease in financing balances, reflecting a trend of profit-taking and reduced positions ahead of the holiday [44]. Future Considerations - The upcoming "calendar effect" is expected to have a significant impact on market dynamics, particularly with a large number of shares set to be unlocked post-holiday, which may pressure stock liquidity [48][49].

Core Viewpoint - Soros Fund Management's latest 13F report reveals a significant shift in its investment strategy for Q1 2024, characterized by increased holdings in large-cap indices, utilities, and financial sectors, while substantially reducing exposure to technology and Chinese stocks, indicating a cautious outlook on the economic landscape [1][3]. Defensive Reallocation: Betting on Large-Cap Indices and Stable Income Assets - By the end of Q1, Soros Fund's top three new holdings were in defensive sectors, investing $96 million in American Electric Power (AEP), $93.6 million in Entergy Corp, and $56.66 million in JPMorgan Chase. The fund also significantly increased its position in the S&P 500 ETF while completely selling off the Russell 2000 small-cap ETF, highlighting a clear strategy of "selling small caps, buying large caps" [3]. - Analysts suggest that in the context of high interest rates and slowing economic growth, the utility sector's stable cash flow and strong anti-cyclical nature, along with large-cap stocks' better risk resilience compared to small-cap firms, reflect the fund's anticipation of increased market volatility [3]. Major Sell-Off: Technology Giants and Chinese Stocks Hit Hard - The fund completely exited 78 stocks and reduced holdings in 45 others, with technology and Chinese stocks being the hardest hit. Notable exits included Alibaba, TSMC, and Boeing, while Alphabet saw a 62.64% reduction and JD.com experienced a drastic 93.6% cut [4]. - The significant reduction in previously held semiconductor equipment firms and cloud computing companies is interpreted as a response to concerns over high valuations in the tech sector and a reassessment of exposure to Chinese stocks amid geopolitical risks [4]. Structural Increase: Betting on E-commerce Leaders and Energy Transition - Despite the sell-off, the fund increased its stake in Amazon by 30%, boosted FedEx holdings by 270%, and nearly 9-folded its position in First Solar. Even UnitedHealth Group, under investigation for insurance fraud, saw its holdings double to $1.12 billion [5]. - These moves align with the investment logic under the expectation of an "economic soft landing," where e-commerce logistics benefit from consumer resilience, renewable energy aligns with policy directions, and the healthcare sector maintains essential demand [5]. Institutional Interpretation: Preparing for Market Turbulence - Multiple Wall Street institutions interpret the fund's reallocation strategy as reflecting three strategic intentions: building a safety net through S&P 500 ETFs and utility stocks to hedge against potential recession risks, narrowing focus in the tech sector to prioritize artificial intelligence and other certain areas, and reducing exposure to emerging markets, particularly in geopolitically sensitive regions [6][7]. - Morgan Stanley strategists noted that the shift from aggressive growth to value defense indicates institutional investors are gearing up for a potentially prolonged high-interest rate environment and market volatility [7].

1. 情绪模型观点：市场情绪进一步下行 根据《从结构化视角全新打造市场情绪择时模型》文中提到的构建思路，目前我们用于构建市场情绪结构指标所用到 的细分指标如下表 在指标合成方法上，模型采用打分的方式，根据每个分项指标所提示的情绪方向和所处布林轨道位置计算各指标分 数，指标分数可分为（-1，0，1）三种情况，最终对各个指标分数等权求和。最终的情绪结构指标为求和后分数的20 日均线，如图1所示，指标整体围绕0轴在[-6,6]的范围内上下波动，近5年A股市场情绪波动较大，其中2023年大部分 时间指标都处于较低位置，直至2024年10月市场情绪得分突破2。 市场情绪自3月20日持续调整，当前已下降接近0轴，为0.1，数值较上周五(4/11)下降0.4，模型维持看空观点。 1.1 从分项指标出发：市场进一步缩量，资金不确定情绪增长 本周A股市场继续提示市场情绪下行，速度没有呈现减缓趋势。本周市场情绪不确定性增强，风险偏好程度下降是市 场情绪进一步调整的主要原因。 下表展示了4月以来的情绪结构各分项指标的分数情况，从分项指标出发，本周明显提示信号切换的指标为科创50成 交占比和300RSI指标，分别代表了市场风险偏好程 ...