The US stock market suffered one of its worst weeks in recent years, technology shares plunged sharply, and economist Oleg Ustenko noted a shift in investor sentiment.
Falls in indices, some heavy losses and volatility
The Nasdaq 100 lost 10.4% over the week, and intraday declines from the October peak exceeded 20%, although the index did not close below that threshold. Technology stocks were hit hardest: Nvidia shares fell 14% in a single trading session, while Anthropic’s stock dropped roughly 8.5% after its market debut.
The implied volatility gauge VIX rose over a few weeks from about 20 to 35.5 points, reflecting increased market nervousness. Investors began trimming risky positions, and losses affected more than just individual tech firms.
Scale of AI investment and companies’ financial obligations
Estimates suggest aggregate investment in artificial intelligence infrastructure could exceed $4 trillion by 2029, while the largest cloud providers alone spent around $500 billion on capital expenditures in 2025. Roughly 75% of projected investment will be attributed to the largest technology companies, which already account for about 35% of total US market capitalisation.
Balance-sheet debt of the seven largest companies in the AI sector is estimated to exceed $800 billion by the end of 2026, and the potential maximum nominal value of their off-balance-sheet obligations tops $3 trillion. For most major cloud providers, free cash flow turned negative in 2026 as infrastructure development costs outstripped operating receipts.

Anthropic IPO, amount raised and financial results
On 19 November, Anthropic carried out an initial public offering (IPO), raising $81.3 billion through the IPO and a further $26 billion via a private placement, a total of $107.3 billion. The company’s valuation reached approximately $1.5 trillion, and demand during the placement was nearly seven times the offering.
Over the nine months to 2026, Anthropic reported $33.24 billion in revenue but an operating loss of $1.26 billion. A large share of expenses relates to building out computing infrastructure, underlining the high cost of deploying the technology.
Assessments of investor sentiment and possible consequences
Ustenko observes that over the past two years investors were positioned for rapid profit growth from technology companies, but are now more fearful of losses than of missing out on gains. He highlights the question of whether the hundreds of billions of dollars invested in data centres, chips and compute capacity will pay off.
The economist suggests doubts about returns on AI investment could have broader consequences: declines in the value of large companies could affect pension funds, banks, bond markets and consumer demand across countries. At the same time, he does not characterise the current situation as a crisis, but rather as a change in investor behaviour.





