随着信用风险的增加，人工智能（AI）相关股票的吸引力可能已不再像过去那样牢不可破。TS Lombard的经济学家达里奥·帕金斯（Dario Perkins）写道：“直到最近，投资者还无需担心AI产业建设过程中的杠杆率或信用问题，因为大型科技公司（超大规模企业）都是用自身的现金流来进行投资。但情况已经发生了变化——如今债务已成为主要的融资来源。”
Artificial intelligence stocks may not be the sure-fire bet they once were as credit risks mount. "Until recently, investors didn't need to worry about leverage or credit in the AI buildout because hyperscalers were investing out of their cashflow. Things have changed, and debt is now the marginal source of funding," wrote Dario Perkins, economist at TS Lombard.
事实上，上周美国股市遭遇重创，原因是投资者担心CoreWeave公司为推动AI业务发展而背负了过高的债务负担：该公司的股价上周暴跌了25%，本周又进一步下跌了2%。周四，当英伟达（Nvidia）首席执行官黄仁勋（Jensen Huang）向CNBC表示AI领域的资金极为稀缺、这正在阻碍公司的发展时，相关股票再次下跌。黄仁勋对CNBC的阿琼·卡尔帕尔（Arjun Kharpal）说：“从基本面来看，对AI相关服务的需求非常高；但问题在于缺乏足够的营运资金来支撑这种增长——由于AI领域涉及众多行业，资金供应严重不足。”
Indeed, U.S. equity markets took a hit last week as investors feared CoreWeave may be taking on too much debt in order to finance its AI build-out. The stock plunged 25% last week and is down another 2% this week. Stocks fell again on Thursday after Nvidia CEO Jensen Huang told CNBC that capital for AI is scarce and holding back growth. .SPX 5D mountain SPX 5-day chart "If you look at the fundamentals, the demand for the neo clouds is off the charts," Huang told CNBC's Arjun Kharpal . "What is missing in the middle is the working capital to fuel that growth … because there's just so much demand across so many sectors of AI, there's a scarcity of capital."
越来越多的AI公司开始依赖债务来推动业务扩张，但帕金斯警告称，借贷速度可能正在过快上升。“仅2027年一年，AI公司预计将发行约7000亿美元的新债券，其中超大规模企业将发行4000亿美元，其余部分则由AI生态系统中的其他企业承担。相比之下，这个数字超过了整个欧元区政府的债券发行总额（尤其是那些面临债务危机的国家）。”
AI companies are increasingly turning to debt to fund expansion, and Perkins at TS Lombard warns that borrowing may be rising too quickly. "In 2027 alone, AI companies are expected to issue around $700 billion in new bonds, split between the hyperscalers ($400 billion) and the rest of the AI ecosystem. As a comparison, that is more than the total issuance of all euro-area governments combined (you know, the ones facing a 'debt crisis')," he said.
帕金斯强调：“如果信用市场真的成为主要的融资渠道，那么AI生态系统之外的机构很可能会停止提供信贷支持——即使科技行业内部的企业仍愿意继续大规模投资。”此外，信贷成本也很有可能上升。美联储预计在下周的会议上维持其基准利率在3.75%至4%的区间内不变；此前该机构已在9月份将基准利率上调了0.25个百分点。
This is important because, "if credit markets are now the marginal source of funding, it is possible that someone outside the AI ecosystem will turn off the credit taps – even if tech insiders are happy to keep investing on such a gigantic scale," Perkins said. On top of that, credit costs are likely to rise. The Federal Reserve is expected to keep its overnight benchmark rate unchanged in a 3.75%-4% range when it meets next week, after lifting rates a quarter percentage point in September.
但根据芝加哥商品交易所集团的“美联储观察”工具，美联储在12月会议上再加息25个基点的概率为75%。如果油价持续高企并推高消费品价格，可能会迫使美联储进一步收紧政策。更积极的一面是，许多AI公司的业务基本面看起来相当稳健。Meta平台公司公布，9月当季营收同比增长26%。FactSet分析师约翰·巴特斯上周在一份报告中表示，“七巨头”公司群体最近一个季度的盈利预计同比增长33%。投资者现在面临的问题是：AI公司能否为其迅速膨胀的债务规模提供合理依据？
But there's a 75% chance that the central bank hikes rates by another quarter point at its December meeting, according to the CME Group's FedWatch tool. If oil prices stay elevated and drive up consumer goods prices, it could force the Fed to tighten policy further. More positively, business fundamentals for many AI companies look solid. Meta Platforms reported 26% revenue growth in the September quarter versus the year-earlier period . FactSet analyst John Butters said in a note last week that earnings for the " Magnificent Seven " group of companies are expected to have grown 33% on a year-over-year basis in the latest quarter. The question for investors now is: Can AI companies justify their rapidly growing mountain of debt?