For nearly two years, the artificial intelligence boom has been the undisputed driver of Wall Street’s gains. No name has embodied this era of seemingly unstoppable growth more than Nvidia (NVDA). But beneath the surface of the relentless rally, fault lines are emerging.
Since November, major indices have largely treaded water. A growing list of high-profile AI players have stumbled after reporting lackluster quarterly results. While this doesn’t yet confirm a market top, it is the kind of hesitation that often precedes a pullback. If the market is finally at a tipping point—or worse, if the AI bubble is poised to burst—most stocks will feel the pain.
However, three names stand out as particularly vulnerable. Two are the obvious pillars of the AI trade. The third is a bank that most investors would never connect to silicon chips.
No company has benefited more from the AI frenzy than Nvidia. Since early 2023, its stock has surged over 1,100%, briefly making it the world’s most valuable public company. But that meteoric rise is precisely what makes it so exposed.
Nvidia’s share price has gone nowhere since last August, a sign that investors are beginning to scrutinize the sustainability of its AI-driven growth. The central concern is simple: Are companies pouring too much capital into AI hardware with too little to show for it? If the AI industry hits headwinds, demand for Nvidia’s data center processors—which now account for three-quarters of its revenue—could evaporate quickly. A correction in AI sentiment would almost certainly begin with Nvidia.
Most investors don’t realize that the world’s most advanced chips are manufactured using a complex light-based process called extreme ultraviolet (EUV) lithography. ASML (ASML) dominates this niche, producing the multi-million-dollar machines that chipmakers use to “print” semiconductors.
This monopoly has made ASML a darling of the AI trade, with its stock up another 35% since the start of the year. However, its very success is also its Achilles’ heel. Each EUV system costs roughly $400 million, making customers highly sensitive to any economic slowdown. Last year, ASML sold 300 new EUV systems, down from 380 in 2024—a dip that surfaced even before any broad AI slowdown, hinting at softening demand. If a market correction takes hold, would-be buyers may delay or cancel orders, delivering a heavy blow to ASML’s top line. Trading at nearly 50 times this year’s projected earnings, the stock has little room for disappointment.
Now for the twist: a bank appearing on a list of AI-vulnerable stocks. JPMorgan Chase (JPM) doesn’t sell chips or build lithography machines. Yet if an AI bubble bursts, the resulting market weakness would inevitably spill into the broader economy—and that is precisely where JPMorgan’s risks lie.
First, strategic acquisitions often dry up during periods of uncertainty, directly impacting the bank’s M&A and advisory business, which accounts for about 10% of its revenue. Second, if investors flee equities, JPMorgan’s stock trading revenue could plunge—another roughly one-fifth of its total business. But the biggest threat is net interest income, which makes up nearly half of the company’s top line. Economic malaise typically leads to lower interest rates, squeezing profit margins on loans and sapping demand for new borrowing. A severe downturn would pressure all three pillars simultaneously.
Despite these vulnerabilities, the market had largely priced JPMorgan’s stock as if no such risk existed—until recently. After peaking in January, the shares have formed a pattern of lower lows and lower highs, a classic technical sign that a peak may be in place. It is a subtle signal that doubts about its near-term future are brewing. And history suggests that major sell-offs often begin just like this: quietly, and one step at a time.
