More than $3.2 trillion was wiped from the US stock market in a single session on Thursday — a number larger than Canada's annual GDP — after the tech-heavy Nasdaq fell 5.5% and the S&P 500 dropped 4.4%. The trigger was the Democrats' sweeping midterm victory, which investors read as a signal that tighter AI regulation and datacentre construction moratoriums are coming. Markets rebounded partially on Friday but failed to recover fully, with lingering fear of a second wave of selling. The sell-off didn't materialise out of nothing. The cyclically adjusted price-to-earnings (Cape) ratio had already climbed to its highest level since 2000 — the dot-com peak — signalling that US stock valuations were historically detached from underlying corporate profits. Alphabet lost more than $500 billion in value after reporting lower-than-expected earnings, a concrete demonstration that hype had outrun revenue. The magnificent seven tech stocks — Nvidia, Apple, Google, Microsoft, Meta, Amazon, and Tesla — account for more than a third of the S&P 500, meaning retirement savings worldwide are directly exposed to AI sentiment. The political trigger matters because it reveals a structural vulnerability: the AI investment thesis assumed unlimited runway for datacentre buildout and minimal regulatory friction. Democrats campaigned explicitly on challenging billionaire power and restraining big tech. Elon Musk poured nearly half a billion dollars into backing Republicans and relocated to Texas — the leading state for datacentre construction — and still lost. Shares in both SpaceX and Tesla fell sharply. The market priced in a policy environment that voters rejected. Opinion polls show most Americans oppose US datacentre construction. Governments worldwide have encouraged citizens to stockpile for emergencies including cyber-attacks. The public backlash against AI infrastructure was already a gathering force; the midterms gave it democratic legitimacy. Ben Barringer of Quilter Cheviot cautioned against oversimplifying: incoming Texas governor-elect Gina Hinojosa is expected to be pragmatic, ensuring resources are "more fairly allocated" rather than imposing blanket bans. The deeper issue is what the AI boom was actually producing. Central banks have been raising interest rates in response to Iran-war-driven inflation, and AI growth was softening the blow to the world economy. A sustained tech sell-off threatens stagflation — weak growth plus rising prices — by removing one of the few engines of economic expansion. Countries that pinned hopes on AI for jobs and productivity gains now face the possibility that neither the technology nor the investment cycle delivers on schedule. Mark Haefele of UBS Global Wealth Management argues the AI investment cycle remains intact and that fundamentals don't justify a prolonged downturn. This is the bull case: demand for semiconductors and AI tools won't vanish overnight. The bear case is that a virtuous circle has already become vicious, and investor confidence, once broken, is self-reinforcing. Both sides agree the market was overvalued; they disagree on how far the correction goes. The honest read: the correction was overdue and the political catalyst was predictable to anyone watching polls. What makes this dangerous is the concentration — one-third of the S&P 500 in seven stocks, all dependent on the same growth narrative. Whether this is a reset or a rout depends on whether AI revenues catch up to AI valuations before patience runs out.
