通常情况下，科技巨头股价大幅下跌对股票交易员来说是件糟糕的事。众所周知，标普500指数由科技股主导，尤其是“科技七巨头”，它们的估值对整个市场有着举足轻重的影响。
因此，甲骨文（Oracle）较9月高点已下跌40%以上；CoreWeave本周大跌，较7月历史高点已回调60%。这些消息本应引发市场剧烈震荡。两家公司都是AI“超大规模运营商”，专注于建设AI数据中心。至少在交易员看来，两者都为资助这些设施背负了过高债务。
例如，据《华尔街日报》报道，CoreWeave上周发行了22.5亿美元的可转换债券，这将稀释现有股东的权益。此前第三季度，CoreWeave报告了37亿美元的流动负债、103亿美元的非流动负债，以及391亿美元的数据中心未来租赁协议。该公司预计今年营收仅为50亿美元，但却声称未来拥有560亿美元的“收入储备订单”。
甲骨文与CoreWeave的市值崩盘，其跌幅规模堪比2000年或2008年的市场下跌。这难道不是AI泡沫破裂的证据吗？
某种程度上算是吧。然而，标普500指数本周仅小幅回落，该指数年内仍上涨16%。今早亚洲和欧洲市场普遍下跌——表明某种抛售正在进行中。但这看起来并不像甲骨文和CoreWeave所经历的那么惨烈。
现实情况是，投资者正在抛售那些似乎过度扩张的公司的个股。但他们对股市整体，尤其是某些科技股，仍普遍持乐观态度。例如，特斯拉（Tesla）本周走势强劲。
交易员仅逃离个别科技股，而且他们正在买入非科技板块的公司。标普500等权重指数在12月11日创下历史新高后，昨日微幅上涨。年初至今，等权重标普指数上涨10.2%，而常规的市值加权标普指数则上涨了16%。
“顾名思义，标普500等权重指数的每个成分股权重相等——这消除了超大市值成分股造成的扭曲，并显著改变了行业权重，包括科技板块的权重从市值加权标普500指数的33%降至等权重指数的仅13%，”总部位于夏洛特的LPL Financial首席技术策略师亚当・特恩奎斯特（Adam Turnquist）表示。
“近期对AI增长势头的担忧拖累了科技和通信服务板块，上周这两个板块分别下跌了2.3%和3.2%，”特恩奎斯特在《财富》看到的一份报告中写道。“资金轮动回更具经济周期性的板块已成为一个正在发展的主题，材料、金融和工业板块表现明显优于大盘。”
正如《金融时报》今早所言：“如果我们身处泡沫之中，它是否会以缓慢的‘嘶嘶’声，而非一声巨响结束？”（*）
译者：冯丰
审校：夏林
通常情况下，科技巨头股价大幅下跌对股票交易员来说是件糟糕的事。众所周知，标普500指数由科技股主导，尤其是“科技七巨头”，它们的估值对整个市场有着举足轻重的影响。
因此，甲骨文（Oracle）较9月高点已下跌40%以上；CoreWeave本周大跌，较7月历史高点已回调60%。这些消息本应引发市场剧烈震荡。两家公司都是AI“超大规模运营商”，专注于建设AI数据中心。至少在交易员看来，两者都为资助这些设施背负了过高债务。
例如，据《华尔街日报》报道，CoreWeave上周发行了22.5亿美元的可转换债券，这将稀释现有股东的权益。此前第三季度，CoreWeave报告了37亿美元的流动负债、103亿美元的非流动负债，以及391亿美元的数据中心未来租赁协议。该公司预计今年营收仅为50亿美元，但却声称未来拥有560亿美元的“收入储备订单”。
甲骨文与CoreWeave的市值崩盘，其跌幅规模堪比2000年或2008年的市场下跌。这难道不是AI泡沫破裂的证据吗？
某种程度上算是吧。然而，标普500指数本周仅小幅回落，该指数年内仍上涨16%。今早亚洲和欧洲市场普遍下跌——表明某种抛售正在进行中。但这看起来并不像甲骨文和CoreWeave所经历的那么惨烈。
现实情况是，投资者正在抛售那些似乎过度扩张的公司的个股。但他们对股市整体，尤其是某些科技股，仍普遍持乐观态度。例如，特斯拉（Tesla）本周走势强劲。
交易员仅逃离个别科技股，而且他们正在买入非科技板块的公司。标普500等权重指数在12月11日创下历史新高后，昨日微幅上涨。年初至今，等权重标普指数上涨10.2%，而常规的市值加权标普指数则上涨了16%。
“顾名思义，标普500等权重指数的每个成分股权重相等——这消除了超大市值成分股造成的扭曲，并显著改变了行业权重，包括科技板块的权重从市值加权标普500指数的33%降至等权重指数的仅13%，”总部位于夏洛特的LPL Financial首席技术策略师亚当・特恩奎斯特（Adam Turnquist）表示。
“近期对AI增长势头的担忧拖累了科技和通信服务板块，上周这两个板块分别下跌了2.3%和3.2%，”特恩奎斯特在《财富》看到的一份报告中写道。“资金轮动回更具经济周期性的板块已成为一个正在发展的主题，材料、金融和工业板块表现明显优于大盘。”
正如《金融时报》今早所言：“如果我们身处泡沫之中，它是否会以缓慢的‘嘶嘶’声，而非一声巨响结束？”（*）
译者：冯丰
审校：夏林
Normally it’s terrible for stock traders when giant tech companies’ shares fall dramatically. Everyone knows that the S&P 500 is dominated by tech stocks—the Magnificent Seven in particular—and their valuations have disproportionate influence over the market as a whole.
So the news that Oracle declined 2.66% yesterday and is now down 44% from its high in September, and that CoreWeave was down 8% yesterday and is down a staggering 60% since its all-time high in July, should have rocked markets to their core. Both companies are AI “hyperscalers” engaged in the business of building out AI data centers, and both—according to traders, at least—have taken on too much debt to fund those facilities.
CoreWeave, for instance, offered a $2.25 billion convertible bond last week that will dilute existing shareholders, according to the Wall Street Journal. Previously, in Q3, CoreWeave reported $3.7 billion in current debt, $10.3 billion in noncurrent debt, and $39.1 billion in future lease agreements for data centers. The company expects to make only $5 billion in revenue this year but says it has $56 billion in “revenue backlog” coming in the future.
The collapse in market caps of Oracle and CoreWeave is on the scale of the declines we saw in 2000 or 2008. Surely that’s evidence of the AI bubble bursting?
Well, kinda. The S&P 500 stepped back only 0.16% yesterday. Futures are down 0.4% this morning. The index is still up 16% for the year. Markets in Asia and Europe were broadly down this morning—signaling that some kind of selloff is underway. But it doesn’t look as catastrophic as what’s happening to Oracle and CoreWeave.
The reality is that investors are selling off individual stocks of companies that seem to be overextended. But they are broadly bullish on stocks as a whole and certain tech stocks in particular. Tesla was up 3.56% yesterday, for instance, and Nvidia rose 0.73%.
As traders flee individual tech stocks, they are buying up non-tech-sector companies. The “equal-weight” S&P 500 was up marginally yesterday after hitting a record high on Dec. 11. The equal-weight S&P is up 10.2% year to date, compared to the regular, market-cap-weighted S&P at 16%.
“As the name implies, each index component [of the equal-weight S&P] carries an equal weighting—eliminating the distortion of the mega-cap components and significantly changing sector weightings, including technology, which drops from 33% on the [market-cap-weighted] S&P 500 to only 13% on the [equal-weight S&P],” according to Adam Turnquist, chief technical strategist for LPL Financial in Charlotte.
“Concerns over AI momentum have recently weighed on technology and communication services, which posted respective losses of 2.3% and 3.2% last week,” Turnquist wrote in a note seen by Fortune. “Rotational pressure back into more economically cyclical sectors has been a developing theme, with materials, financials, and industrials notably outperforming.”
As the Financial Times put it this morning, “If we’re in a bubble, might it end with a slow hiss, not a loud pop?”
•