The AI Bubble Might Be About to Burst, Experts Warn
When the global economy rallies around a new technology, the stock market tends to boom. In the past few years, that technology has been AI.
Recently, the S&P 500 has surged to record highs as it has ridden this wave, but the rally's foundations are looking increasingly precarious according to Stephen Wu, the Founder of Carthage Capital.
Earlier in his career, Stephen was an AI engineer at both Amazon and Microsoft, which has given him a unique insight into the meeting place between the technology and its performance on the market.
Right now, Stephen believes that the market may not be as strong as it first appears.
He believes that market conditions bear an unsettling resemblance to the dot-com bubble, which saw huge numbers of investments in web-based companies, many of which failed to turn a profit in the fullness of time.
"The S&P 500 is at record highs, but gains are concentrated only in a few tech companies – worse than the dot-com boom," he says.
His concern centres on market breadth, with equal-weight indices significantly underperforming their counterparts that are weighted on market cap.
Historically, this kind of pattern is followed by sharp corrections, Stephen suggests.
A two-tiered tech sector
Despite some concerns about valuation, hyperscale cloud providers are continuing to commit unprecedented amounts of capital to expanding AI infrastructure.
Amazon, for example, recently pledged US$30bn towards Nvidia's GPUs, while Microsoft has committed US$56bn and Meta US$40bn.
"Every major US cloud player committed billions to buying Nvidia's GPUs," Stephen says. “Nvidia’s new GB200 chips and the upcoming OpenAI release could generate fresh demand.”
The chipmaker, led by Jensen Huang, recently announced a partnership with OpenAI worth up to US$100bn, with Nvidia agreeing to supply data centre chips to the firm behind ChatGPT.
The elastic share prices of the tech sector’s major players does not tell the full story of the economy, though.
For Stephen, the industry feels more two-tiered than ever.
Beyond the shiny top layer, things are a little different.
“The economy right now feels like a two-story house,” he explains.
“Growth on the ground floor is solid – GDP bounced back 3.3% in Q2 after contracting in Q1 – but the upper floor is creaking.
“The stock market looks strong on the surface, but momentum is slowing and the gains are tied to only a few tech giants. With valuations above 22x forward earnings, the margin for error is thin.”
The risk of having a small set of market frontrunners
In recent months, a great deal of success stories in the market have focused on the sector’s giants, such as Nvidia, Google or Meta.
But the concentration of market gains in a handful of technology names has reached levels that exceed even the dot-com era.
“The risks are real,” Stephen warns. “Equal-weight underperformance is a tell. If AI leaders stumble, the rally in the broader market could unravel quickly.”
The S&P 500 trades above 22 times forward earnings, which leaves little room for disappointment.
Recent sessions have seen volatility return, with the index closing 0.6% lower on Tuesday despite hitting an intraday record of 6,699.52.
Nvidia shares retreated alongside other large-cap technology names including Apple.
At Carthage Capital, Stephen is positioning defensively.
"For investors, this is a time to take profits, trim risk and avoid leverage," he advises, adding that the fund is focusing on market-neutral strategies rather than chasing the final stages of what he characterises as a frothy rally.
The question facing investors is whether massive infrastructure spending can justify current valuations, or whether the market is repeating the mistakes of the dot-com era with AI cast in the starring role.
