Impactive Capital's Wolfe: hard to tell when the AI bubble might burst

what happened

Lauren Taylor Wolfe, managing partner at Impactive Capital, told CNBC, as reported by Seeking Alpha, that it is very hard to tell when an AI bubble might burst. She said the technology is an incredible tool that has boosted productivity, but questioned whether it has an effective business model and will generate returns for investors.

Goldman Sachs estimates hyperscalers will spend about $3 trillion on AI capital expenditure in 2026-2028, while earning competitive returns would require $5 trillion to $10 trillion in revenue. Wolfe singled out neocloud providers that borrow money, saying their cost of capital is rising and compute costs are inflating while AI token prices keep falling. She said CoreWeave's borrowing costs have risen 15%.

Wolfe also said corporate finance departments are scrutinizing AI budgets more closely. She noted that Google, for the first time in her view, had no free cash flow, and urged investors to watch these developments closely.

why it matters

For markets, the growing skepticism about AI returns matters because it puts focus on debt levels and the cost of AI infrastructure. This could weigh on valuations of tech companies and on cloud providers that rely on borrowing. One fund manager's view is not a forecast, but it reflects sentiment among some investors.

tags

instruments