There’s a lot of talk about an AI bubble these days. Could that, paradoxically, be a good thing?
“It appears to be conventional wisdom that we are in some form of a bubble”, said high-profile entrepreneur and marketing professor Scott Galloway in a recent podcast. He rattled off some recent headlines: Bill Gates likening AI to the dot-com bubble, short seller Michael Burry doubling down on AI bubble warnings, and even Vogue asking if fashion is “ready for the AI bubble to burst”.
Meanwhile, Bank of America’s latest fund manager survey shows more than half think AI stocks are already in bubble territory. As for retail investors, a recent American Association of Individual Investors (AAII) poll showed bearishness hitting its highest level since April’s tariff chaos.
However, a new study, Bubble Beliefs, suggests the ubiquity of bubble talk may actually be a sign that a crash isn’t coming. It found that during past boom-busts, analysts projected near-miraculous earnings and expected further gains, short interest stayed low and the media almost never uttered the B-word.
RM Block
In most cases, from Yahoo! in 1999 to any number of forgotten high-flyers, attention surged but few raised any doubts. Instead, “optimism portends crashes”, with the most bullish forecasts predicting the highest crash risk.
There were exceptions, such as the media clearly flagging 2021’s GameStop bubble. Generally, however, periods of heavier media coverage are actually linked to a slightly lower risk of a crash, highlighting how little awareness there usually is of an emerging bubble.
Today? Instead of euphoria, there is anxiety; instead of uncritical cheerleading, a chorus of warnings.
True, some AI valuations look very stretched. But historically, it seems, markets that are worried about being in a bubble tend not to be.
