当地时间2月14日，ARK Invest CEO兼CIO凯茜·伍德在其视频栏目《ITK》2月节目中表示，近期市场大幅波动更多是程序化交易驱动，而不是基 本面发生了同等幅度的变化。她在节目开头直言："这轮波动大多是算法交易'制造'的。算法不会像我们一样做研究。" 伍德说，这种波动会"把人吓坏"，但也会制造定价错误。"去年4月关税风波时，很多人慌了。那时卖出的人，后来一年都在后悔。"她把当前行情 形容为"爬上忧虑之墙"，并称这类市场往往更强。 算法为何会把波动"做出来" 伍德口中的"算法"，核心不是对企业现金流和竞争格局做判断，而是按规则对风险敞口做机械调整。她用一句话概括近期盘面特征："先卖再说、 回头再问。" 从交易机制看，程序化策略常以价格趋势、波动率、相关性、仓位风险预算等为触发条件： 她还点到另一层放大器：技术面主导的交易心态在上升。"现在很多人只做技术分析。"在她看来，越多人盯着同一条均线、同一个"关键点位"， 越容易出现踩踏式的同向交易。 算法无法理解的"结构性转型" 在AI资本开支争议升温之际，木头姐把美股市场的"急涨急跌"归因于算法卖盘的连锁反应。 针对近期科技股尤其是软件板块的剧烈波动，伍德认 ...

Core Insights - Alibaba has launched a new generation of open-source model, Qwen3.5-Plus, which is claimed to outperform Gemini 3 Pro, making it the strongest open-source model globally [1] Model Performance - Qwen3.5-Plus features a total of 397 billion parameters, with only 17 billion activated, achieving superior performance compared to the trillion-parameter Qwen3-Max model [1] - The deployment memory usage has been reduced by 60%, and inference efficiency has significantly improved, with maximum inference throughput increased by up to 19 times [1] Cost Efficiency - The API pricing for Qwen3.5-Plus is set at 0.8 yuan per million tokens, which is only 1/18th of the cost of Gemini 3 Pro [1] Technological Advancements - Qwen3.5 has achieved breakthroughs in native multimodal capabilities through pre-training on mixed text and visual data, demonstrating excellent performance across various benchmarks in reasoning, programming, and agent intelligence [1] - The model has also excelled in authoritative evaluations of visual understanding, winning several performance accolades [1] Accessibility - The Qwen3.5-Plus model is now integrated into the Qwen APP and PC platform, with developers able to download it from the Modao community and HuggingFace, or access API services directly through Alibaba Cloud [1]

在此背景下，有两个关键特征值得关注。其一，投资空间依然广阔。预计到2028年，AI相关投资将累计增长20%，但目前实际投入尚不足这一规 模的20%。这意味着，绝大部分投资机会仍在前方。 其二，融资结构正在生变。与2025年及之前的支出阶段不同，下一阶段的建设将更加依赖多元化的信用市场，包括有担保与无担保融资、证券化 与结构化产品，以及合资企业模式。即将到来的资本支出规模过于庞大，仅靠股权融资已难以支撑，信用将在系统性融资中扮演核心角色。 当前，股市疲软已向信用市场传导。年初至今，标普软件指数已下跌23%。这一压力在信用领域表现尤为明显，特别是在杠杆贷款和商业开发公 司（BDC）等对软件行业敞口较大的板块。 摩根士丹利警告，相关板块情绪可能持续低迷，信用投资者或需等待更长时间或更大幅度的价格修正，才会重新入场。尽管目前违约率仍处低 位，但随着AI应用加速推进及不确定性持续，信用市场的价格下跌或将扩大和加深。 摩根士丹利指出，人工智能正从单纯的资本市场热点，转变为引发信用市场结构性压力的真实变量。该行北美固定收益研究主管Vishwanath Tirupattur认为，市场对AI的认知正出现显著分化：一方面是科技巨 ...

尽管增长远逊预期，但分析师认为，这尚不足以动摇日本央行今年晚些时候加息的计划。经济重返增长区间本身为央行继续收紧货币政策留出了 空间，尽管复苏力度温和。 消费与投资增长乏力 从细项数据来看，占GDP比重过半的私人消费在第四季度仅微增0.1%，较三季度0.4%的增速明显放缓。尽管企业近年来持续加薪，但食品价格 持续上涨令家庭支出承压，2025年全年实际工资月均录得同比下降。 资本支出同样表现乏力，仅增长0.2%，远低于路透调查预期的0.8%。三菱日联研究咨询首席经济学家Shinichiro Kobayashi指出： "利率上升与工资增长可能对中小企业构成双重风险，进而制约其资本支出能力。" 出口方面则呈现分化。第四季度出口下降0.3%，降幅较上季度的1.4%有所收窄，显示关税的初步冲击已逐步消化，但入境旅游遭受打击。 受高市早苗去年11月言论影响，我国发布赴日旅游警告，来自中国的游客数量在第四季度明显下滑。据新华社，中国驻东京旅游办事处主任欧阳 安表示，高市早苗近期错误言论已对中日文旅交流产生实质性影响。随着中国政府发布旅游、留学提醒和预警，航空公司也陆续推出免费退改政 策，中国游客赴日意愿降低。 央行加息路径 ...

Group 1 - The core message of the report indicates a significant shift in AI capital expenditure from a "money printing machine" to a "money shredding machine," suggesting potential liquidity and asset pricing turmoil [1] - The expected capital expenditure for hyperscalers has surged to $740 billion by 2026, up from a previous estimate of $670 billion, raising concerns about extreme financial consequences for major tech companies [2] - The narrative in the market is shifting from "awe of AI" to "being impoverished by AI," with a clear signal for investors being the announcement of capital expenditure cuts by a major AI player [4] Group 2 - The report highlights a rapid spread of AI disruption effects into traditional service sectors, described as "wildfire AI disruption," impacting various industries sequentially [5][6] - The first sector affected by AI disruption was Indian tech stocks, which have not seen any buying support since being labeled as "AI victims," indicating a prolonged recovery period for such stocks [6] Group 3 - Political factors are intensifying the asset rotation, with a notable divergence in support for Trump between Wall Street and Main Street, leading to potential aggressive affordability policies ahead of the midterm elections [10][11] - The report suggests a strategy of "long Main Street, short Wall Street" is becoming effective, with significant asset performance divergence since the interest rate cut on October 29 [11] Group 4 - A historical shift in the correlation between the Japanese yen and the Nikkei index has been noted, indicating a potential long-term bull market for Japan, although a strong yen could negatively impact exporters [12][13] - Despite a $463 billion inflow into global equities, the Bull & Bear Indicator remains in the "sell" zone, suggesting that the adjustment in risk assets is not yet over [16] Group 5 - The report outlines a "great rotation" in asset leadership over the past 50 years, with significant political and financial events altering the flow of investments, indicating a potential shift towards emerging markets and small-cap stocks as the next structural leaders [17][20] - Emerging markets and small-cap stocks are expected to benefit from a transition in service-oriented to manufacturing-oriented investments, alongside rising costs in the AI arms race [20]

Core Insights - The storage industry is emerging from a downturn and entering a "super cycle" driven by AI demand, as indicated by the recent earnings reports from major storage companies [1] Group 1: Key Signals from Earnings Reports - Inventory levels have dropped below the "warning line," with SK Hynix's inventory turnover days decreasing from a peak of 233 days in Q1 2023 to just 127 days, indicating potential supply chain shortages [2] - Average Selling Prices (ASP) are experiencing significant recovery, with Samsung's DRAM ASP soaring by 40% quarter-over-quarter and SK Hynix's NAND prices increasing by 32% [2] - Major companies are aggressively increasing capital expenditures in response to surging demand for High Bandwidth Memory (HBM), with SK Hynix's capital spending projected to rise from 7 trillion KRW in Q4 2024 to 12 trillion KRW by Q3 2025 [2] Group 2: Market Trends and Price Movements - HBM4 production is exceeding expectations, which will influence future market share in AI computing [3] - The industry outlook is extremely optimistic, with projections for a long-term "storage super cycle," as noted by Nanya Tech's president, who claims the current cycle is better than the 2017-2018 cloud server boom [3] - The Memory Indicator from Bank of America has risen to 124, indicating an upward cycle, compared to an average of 103 in the first half of 2025 [4] Group 3: Current Market Conditions - DRAM spot prices remain high, with 16Gb DDR5 prices at $38, a year-on-year increase of 709%, and 16Gb DDR4 prices reaching $78, up 2445% year-on-year [6] - Despite some OEMs indicating temporary assembly line halts due to rising storage costs, overall sales data from Taiwan shows significant month-over-month growth exceeding 20% and year-on-year doubling for several manufacturers [7] - SSD prices have surged, with a weekly increase of 40% and a monthly increase of 60%, reflecting market concerns about potential shortages in the second half of the year [9] Group 4: Insights from Semicon Korea - The recent Semicon Korea event showcased a booming industry, with equipment suppliers benefiting from record capital expenditures by major players like Samsung and SK Hynix [10] - HBM production is time-consuming, requiring extensive equipment to maintain high output levels, which may impact supply timelines [12] - The adoption of Hybrid Bonding technology may face delays, as manufacturers prefer existing methods for HBM production [12]

Core Insights - The rapid evolution of AI technology is leading to significant changes in various industries, surpassing the impact of the COVID-19 pandemic [1][12] - AI has transitioned from being an "assistive tool" to an "independent executor," capable of completing complex tasks autonomously [2][6][21] - The speed of AI advancements is accelerating, with each breakthrough occurring in shorter intervals, leading to a potential restructuring of the labor market [2][6][11] Industry Impact - AI's capabilities now extend beyond parameter growth to independently completing complex tasks, reducing workflows that previously took weeks to mere minutes [2][6] - The ability of AI to self-iterate and improve its own systems marks a critical turning point, breaking the limitations imposed by human researchers [7][33] - Industries such as law, finance, healthcare, and customer service are experiencing profound changes, with AI expected to replace many cognitive jobs within the next few years [21][34] Employment Landscape - The nature of work is being fundamentally reshaped, with repetitive and standardized tasks being automated, while roles requiring complex judgment and emotional interaction will evolve [8][34] - Professionals are encouraged to embrace AI tools to enhance their capabilities, as those who resist change may find themselves marginalized [8][12] - The article outlines a three-step framework for professionals to build resilience in the AI era: scene-based learning, focusing on irreplaceable core skills, and maintaining dynamic adaptability [9][12] Societal and Economic Repercussions - AI is redefining wealth distribution, educational foundations, and occupational structures, leading to a "winner-takes-all" effect in various industries [11][12] - The philosophical implications of AI's capabilities challenge traditional notions of work and human value, as machines increasingly perform tasks previously thought to require human intelligence [11][12] - The urgency for individuals and organizations to adapt to these changes is emphasized, as the window for adjustment is rapidly closing [12][46]

Core Viewpoint - Investors are retreating from risk assets and turning to U.S. Treasury bonds as concerns grow over the sustainability of AI investment returns and high market valuations [1][4] Group 1: Market Trends - Morgan Stanley's report indicates a shift in market sentiment, with investors becoming increasingly sensitive to the negative externalities of the AI investment cycle [1] - The S&P 500 index has reached new highs, but stocks directly impacted by AI disruptions have begun to decline [2][4] - A basket of 108 AI-impacted stocks has shown a significant divergence from the broader market, indicating a potential peak in AI optimism [2][4] Group 2: Economic Indicators - Morgan Stanley has raised its 2026 U.S. GDP growth forecast from 2.4% to 2.6%, driven by capital expenditures from hyperscalers [1] - However, the firm warns that this growth comes with risks, particularly if investment returns do not materialize, leading to potential overinvestment issues [1][4] Group 3: Investor Sentiment - High-income groups (earning over $100,000) have shown a notable shift in their perception of the economy since the beginning of the year, reflecting concerns over asset price volatility [5][6][7] - The decline in confidence among affluent consumers is often a precursor to economic recession, suggesting a cautious outlook for the economy [7][9] Group 4: Inflation and Monetary Policy - Recent inflation data has surprised to the downside, with January CPI growth at 0.17%, below expectations, and core CPI at 0.30% [12][13] - This unexpected decline in inflation is reshaping Federal Reserve policy expectations, with markets pricing in potential interest rate cuts by mid-2026 [13] - The Fed's recent bond purchasing operations have provided liquidity support, further benefiting short-term U.S. Treasuries [13]

风险提示及免责条款 市场有风险，投资需谨慎。本文不构成个人投资建议，也未考虑到个别用户特殊的投资目标、财务状况或需要。用户应考虑本文中的任何 意见、观点或结论是否符合其特定状况。据此投资，责任自负。 恒生科技指数跌0.19%。洛阳钼业、宁德时代将纳入恒生指数成份股数，均涨超1%。阿里巴巴跌1.8%。 ...

2月15日，OpenAI首席执行官奥特曼在《印度时报》发表的文章中表示，印度已成为其全球关键战略市 场。目前，该国拥有1亿周活跃ChatGPT用户，贡献了全球总活跃用户数的约八分之一，用户规模仅次 于美国。该表态在AI影响峰会开幕前夕作出，峰会于周一在新德里启动，旨在探讨AI技术的应用与治 理。 这一数据凸显了印度在全球人工智能版图中的战略地位。在全球周活跃用户总数已突破8亿并向9亿迈进 的背景下，印度超过10亿的庞大互联网用户基数与年轻的人口结构，成为各大科技公司争夺新兴市场增 量的战略支点。 在产品侧，OpenAI推出了定价低于5美元的ChatGPT Go套餐，并面向印度用户提供为期一年的免费使用 权，旨在通过降低准入门槛抢占增量市场。 奥特曼对此表示： "凭借对可及性、实用AI素养以及支持广泛采用的基础设施的关注，印度有能力扩大技术受 益者范围，并帮助塑造民主AI如何大规模采用。" 政府峰会汇聚全球AI领袖 尽管印度市场的规模为OpenAI带来了显著的流量红利，但其商业化进程仍面临挑战。印度市场高度价 格敏感，加之基础设施方面的限制，使得AI技术从大规模采用转向深度变现与持续经济影响的过程， 较发达经济 ...