美国经济分析局口径的最新数据显示,2026年第二季度美国企业在“计算机和外设”上的支出增加205亿美元,达到4,201亿美元的纪录高位。这一项目常被市场视为AI相关硬件投资的关键代理指标。
背景上,BofA策略师Savita Subramanian最新表示,押注AI资本开支、同时做空消费类主题的“简单交易”正在变得更难,因为仓位已经相当拥挤,AI支出与消费疲弱都已被部分定价。此前媒体引述其观点称,标普500消费必需品指数过去12个月涨4.6%,而消费可选指数跌3.3%。
对市场而言,这一趋势继续利好与AI基础设施相关的公司,尤其是微软、亚马逊、Alphabet和Meta等超大规模云厂商,以及服务器、电子设备和零部件供应链。与此同时,若资金持续向资本开支集中,消费、零售和部分非AI制造环节可能面临估值与需求预期的再平衡。
需要注意的是,BofA的表述强调的是“交易更难做”,并非判断AI投入放缓;相反,其核心意思是AI支出仍强,但市场对这条主线的拥挤度在上升。
The latest BEA-backed data show U.S. business spending on computers and peripheral equipment rose by $20.5 billion in Q2 2026 to a record $420.1 billion, reinforcing the scale of AI-related hardware investment. Bank of America, meanwhile, says the once-simple trade of buying AI capex beneficiaries and fading consumer themes is getting harder as positioning becomes crowded.
That backdrop matters because the spending surge remains one of the clearest macro tailwinds for AI infrastructure names. But it also means much of the AI capex story may already be embedded in prices, while consumer-related sectors could see renewed scrutiny if portfolio rotation persists.
For investors, the main beneficiaries are still the hyperscalers and their supply chain — especially Amazon, Microsoft, Alphabet, Meta, and makers of servers, components, and networking gear. The flip side is that consumer discretionary and other non-AI areas may struggle to attract incremental capital if the market keeps favoring capex over consumption.
BofA’s message was not that AI investment is slowing; it was that the market trade around it is becoming more crowded. In other words, the spending boom is still intact, but the easy relative-value expression around it may be fading.