Bubble Trouble: AI rally shows cracks as investors question risks

NEW YORK: The biggest bout of volatility in U.S. stocks in months has revealed cracks in the artificial intelligence-related rally, raising questions about whether the market has been in the grips of a speculative bubble that may be popping.

Soaring valuations in AI stocks this year have stoked worries that Wall Street may be inflating another speculative bubble, with some of these fears coming to the fore after a stellar earnings report from AI bellwether Nvidia failed to rally the stock and broader market.

In recent days investors have been watchful of emerging signs of cracks in the AI investment narrative.

Several high-flying AI stocks have experienced sharp pullbacks, questions are mounting about when AI investments will translate into actual profits, and concerns are growing that enthusiasm may be outpacing the technology’s near-term capabilities.

Investor sentiment has been rattled recently by several developments, including a noticeable dip in retail traders’ enthusiasm for “buying the dip.”

This caution comes alongside a hit to Oracle bonds on news the company plans to increase its substantial debt to finance artificial intelligence infrastructure, and as lenders seek greater protection on loans to major tech companies, citing concerns over debt-financed AI investments.

The boom and recent retreat in AI and AI-adjacent stocks has drawn comparisons to some of history’s most notorious market manias, from the dot-com frenzy of the late 1990s to the more recent cryptocurrency boom.

At the heart of investors’ concerns about a market bubble are lofty valuations. Despite the recent pullback, valuations remain elevated and investors are wary of risks around customer capital spending and financing, plus challenges in expanding data center capacity amid energy constraints and memory chip shortages.

Alphabet Chief Executive Sundar Pichai recently said no company would be unscathed if the artificial intelligence boom collapses, but Nvidia CEO Jensen Huang on Wednesday shrugged off concerns.

Not all bubbles are created equal, and the warning signs that signal the buildup of excess in one bubble can differ significantly from those in another.

And so, anytime you’re going to have a proxy for the actual exposure, you might be sacrificing on something.

Historical data shows dramatic variations in how asset manias unfold – from the speed of their collapse to the years required for recovery.

The Japanese stock market crash of the early 1990s took decades to recover, while the 2021-2022 crypto crash played out in a matter of months.

Understanding these patterns matters for investors trying to gauge whether today’s AI enthusiasm represents rational exuberance over a transformative technology or the kind of speculative excess that ends in tears.

Some charts here help assess how the AI mania compares with historic bubbles and what stage it is in, if indeed this is a bubble.

OVER-VALUED

Its recent wobble notwithstanding, the U.S. stock market’s valuation has surged into territory that historically preceded major downturns, with the Buffett Indicator — a gauge favored by billionaire investor Warren Buffett — flashing warning signs.

The indicator, which compares total U.S. stock market capitalization to gross domestic product, recently rose above 200%, surpassing levels last seen at the pandemic-era market peak in 2021.

The metric currently stands near its highest level on record, surpassing even the dot-com bubble of 2000.

Named after the Berkshire Hathaway chairman, the ratio shows elevated readings before major market corrections since 1975.

Other stock valuation gauges also show elevated readings, though not at historical highs. The S&P 500’s price-to-earnings ratio has climbed to about 23 times, based on 12-month earnings estimates for its constituents, around its highest level in five years and well above its 10-year average of 18.7, according to LSEG Datastream.

A separate valuation measure – the CAPE ratio, or Shiller P/E ratio – which measures earnings averaged over 10 years to adjust for economic cycles, also shows elevated readings, though not yet at the heights touched during past bubbles.

EARLY DAYS

The Nasdaq’s current trajectory during the artificial intelligence boom bears a striking resemblance to its dot-com era path, albeit with less exuberance, backing the view that if this is a bubble it may still be early stage.

The tech-heavy index climbed roughly 100% in the three years following ChatGPT’s November 2022 launch, mirroring the early stages of excitement that followed Netscape’s August 1995 IPO.

NOT THERE YET

A key ingredient in past stock market bubbles has been runaway investor optimism, which appears to be absent now.

The American Association of Individual Investors survey, a weekly poll that measures investor sentiment among individual retail investors in the U.S. stock market, shows bullish sentiment at 38%, in line with its long-term average.

That’s a far cry from the 75% high hit in January 2000 or even the 57% high scaled during the meme-stock mania in 2021.While elevated bullishness is not a necessary precondition for a market reversal, heightened bullishness would have backed the view that investors have grown too complacent.

Historical data shows that prolonged periods of above-average optimism often precede market turbulence, as crowded trades and stretched valuations leave little room for disappointment.