AI Infrastructure Debt Crisis Hammers Chip Stocks for Sixth Straight Session as CoreWeave Reaches Distress Levels

Concerns about unsustainable AI infrastructure spending drove chip stocks to their longest losing streak in over a decade, with the Philadelphia Semiconductor Index falling roughly 9% on the week. CoreWeave, a Nvidia-backed neocloud company that has taken on more than $51 billion in debt and obligations to finance its AI build-out, saw its 2031 bond yields top 15% and credit default swap spreads widen to levels generally viewed as distressed. Ark Invest moved contrarian, buying 192,566 CoreWeave shares as the stock slid, while Gartner flagged that most neocloud stocks carry Hold ratings or worse amid mounting AI infrastructure risks.

A Goldman Sachs Asia sales and trading note circulated widely, citing Anthropic's third-quarter revenue of $16.8 billion as a significant disappointment relative to expectations. The firm noted that a $5.4 billion implied monthly run-rate for July, reported in August, suggested the pace of growth stalled in August and September, amounting to close to zero growth over those two months and undermining a key pillar of the AI investment thesis: the accelerating annual recurring revenue trajectory of frontier models. Separately, Fed Governor Goolsbee flagged that if AI-driven demand were to overheat the economy and stoke inflation, the Fed would be compelled to respond, adding another layer of concern for heavily leveraged AI spending.

Despite the pressure in tech and semiconductors, broader markets showed resilience. The S&P 500 closed up 0.32% at 7,743.88, the Russell 1000 gained 0.37%, and the Dow surged more than 550 points as investors rotated away from AI and chip names into other sectors. The Nasdaq edged up just 0.07% to 26,711.24. After hours, Nasdaq 100 futures fell 0.5% as AI spending concerns persisted, though Asian stocks rose 0.6% the following session as chip shares stabilized and Japan's Topix gained 1.4%. Negative free cash flow and expected losses above $2.5 billion across the neocloud space intensified the scrutiny on whether AI capital expenditure can withstand a higher-rate environment.