Furman: if the AI bubble pops, the market could fall 25%

Former White House economist Jason Furman says a bursting AI bubble would cut the broad US stock market by 25% — wiping out about $19 trillion of household wealth — though he does not believe AI is necessarily a bubble.

AI illustration for this news story: Furman: if the AI bubble pops, the market could fall 25%
AI-generated illustration, not a photo.

The question Furman tried to answer

If the artificial intelligence bubble bursts, the market is likely facing a much larger decline than investors currently realize, according to former White House economist Jason Furman. Furman, a previous chair of the US Council of Economic Advisers and now a professor at Harvard Kennedy School, wrote in an essay published on Sunday that the key to understanding the risks the economy faces is having an idea of how far stocks may slip if AI is a bubble set to pop.

The essay landed in the middle of a brutal stretch for AI stocks. Nvidia's shares have fallen for six straight weeks, the longest losing streak since 2012, wiping $1.7 trillion off the company's market value, and the market is bracing for Nvidia's earnings on Tuesday.

Five yardsticks, one baseline

Furman suggested a number of different methodologies to assess what the potential decline might be, measured from the November 13 close. If the seven biggest AI-related companies — Nvidia, Microsoft, Alphabet, Amazon, Meta, Broadcom and Oracle, about a fifth of the market — fell by half, the whole market would fall about 11%, not counting any consequences for other sectors.

Looking at valuation metrics combined with expectations of earnings declines, the fall looks to be in the region of 18% to 41%, he said. If the forward price-earnings ratio fell from 18.3 to its average since 1979 of 15.0, the market would fall 18%, or 34% if expected earnings also dropped 20%. If the Shiller price-earnings ratio returned to its 1979 average of 23.6, the market would fall 41%. And if the excess growth over nominal GDP since 2019 were reversed, the market would fall 34%.

If forced to pick, Furman would choose a 25% decline as the baseline for a bursting bubble, 35% as a bad case, and 50% as a severe and highly unlikely case. 'All of these declines are much larger than what options markets treat as likely,' Furman wrote.

Dark trading floor glowing red at night
AI-generated illustration, not a photo.

What the market is pricing

Options are currently pricing in an unusually low probability of a significant market decline. The odds of the S&P 500 falling 10% by December of next year are at 19%, while for a 50% decline they are 2%, based on Friday's closing prices on Cboe Global Markets.

Even the baseline case would be a big hit to household wealth: a 25% fall would wipe out about $19 trillion of the stock that households own directly and through mutual funds and retirement accounts, Furman said. His next posts will look at how a stock market decline of that size would feed into consumption and the financial system.

Furman's downside mapSource: MarketWatch
  • Base case if the bubble pops-25%
  • Bad case-35%
  • Worst case (highly unlikely)-50%
  • Seven biggest AI stocks down 50%Whole market -11%
  • Household wealth wiped out by a 25% fallAbout $19 trillion

Not necessarily a bubble

Furman noted that he does not necessarily believe AI is a bubble, pointing to technology developments like smartphones and jet planes, which proved not to be bubbles. The sober case, in his telling, is that today does not look like 2000: valuations are lower, profits are real, and demand for AI capacity still exceeds supply. Furman also said the environment is already different than it was in the lead-up to the dot-com crash, with today's largest market drivers — major tech companies including Alphabet and Nvidia — showing stellar revenue growth on the back of AI development.

The market was about 3% below its record on November 13. Whether the AI trade is a bubble or not, Furman's exercise puts a number on what is at stake if the optimists are wrong — and on how little protection the options market is currently pricing in.

Sources

Educational content, not investment advice. Facts as of the publish time above; figures come from the linked sources and markets change quickly. Genexus Alpha is not a registered investment adviser or broker.