此主题在企业行为与部位配置上的迹象愈来愈多：根据 Bloomberg 对电话会议逐字稿的分析，管理层在电话会议中提及「AI disruption」的次数相较前一季几乎翻倍。美国银行最新的基金经理人调查指出，对企业支出过度的担忧升至纪录高点，约 35% 警告企业投资过头；25% 将「AI 泡沫」列为最大尾端风险，30% 表示大型科技公司 AI 资本支出最可能引发信用危机。JPMorgan 策略师预估，超大规模云端业者的 AI 相关资本支出在未来 12 个月将成长 53%；同时 DTCC 数据显示，1 年前仍相当稀少的高评级大型科技发行人单一名称信用衍生品，如今已成为美国金融类股以外最活跃交易的合约之一。
策略师将下一阶段定位为「选股」，指出滚动式修正正从软体扩散至金融服务、运输、保险与商用不动产等领域；即便如此，仍有人认为若超大型科技股趋于稳定，这种波动可能带来逢低布局的机会。AI 之外的总体背景同样牵动市场：英国失业率升至 5 年新高，联准会重申通膨目标为 2%，官员表示在更多证据出现前利率可能需要维持不变，同时也有人认为若通膨持续降温仍有降息空间。企业消息则凸显当前资金流与风险预算的规模：一项 $141 billion 的美国 ICBM 计划重整目标在年底前完成、一项 $1 billion 封闭式基金募资行动、近 $170 million 的比特币买入且约一半以永续优先股融资、一笔 $9.9 billion 的医疗科技交易、约 $2.5 billion 的住宅服务收购、到 2035 年的 $100 billion AI 资料中心计划，以及一项超过 $7 billion 的和解推动。
Stocks swung in a volatile session as investors weighed 2 opposing AI narratives: rapid disruption risk versus doubts that heavy AI spending will pay off soon. The S&P 500 finished up 0.1% after erasing a nearly 1% intraday drop, while a major software ETF fell 2.2%, highlighting dispersion beneath near-record index levels. Other cross-asset moves underscored the risk-off tone: Bitcoin hovered around $67,500 after extending a February slide, spot gold fell below $4,900/oz, and the US 10-year yield ticked up 1 basis point to 4.06%.
Evidence of the theme showing up in corporate behavior and positioning is mounting: mentions of “AI disruption” on management calls almost doubled versus the prior quarter, according to a Bloomberg transcript analysis. Bank of America’s latest fund manager survey flagged a record level of concern about excessive corporate spending, with about 35% warning companies are overinvesting; 25% named an “AI bubble” as the top tail risk, and 30% said big-tech AI capex was the most likely trigger for a credit crisis. JPMorgan strategists projected AI-related capex at hyperscalers to rise 53% over the next 12 months, while DTCC data indicated single-name credit derivatives on high-grade Big Tech issuers that were scarce 1 year ago are now among the most actively traded US contracts outside financials.
Strategists framed the next phase as “selectivity,” with rolling corrections spreading from software into areas like financial services, transportation, insurance, and commercial real estate, even as some view the volatility as a potential buying opportunity if mega-cap tech stabilizes. The macro backdrop remained in play alongside AI: UK unemployment hit a 5-year high, the Fed’s inflation goal was reiterated at 2%, and officials signaled rates may need to stay steady until more evidence emerges, while others still saw scope for cuts if inflation keeps cooling. Corporate headlines reinforced the scale of current money flows and risk budgeting: a $141 billion US ICBM program restructuring target by year-end, a $1 billion closed-end fund raise effort, nearly $170 million of Bitcoin bought with about half financed via perpetual preferred stock, a $9.9 billion medical-tech deal, a roughly $2.5 billion residential-services acquisition, a $100 billion AI data-center plan by 2035, and a settlement push exceeding $7 billion.