Roughly three decades ago, the mainstream proliferation of the internet changed America forever. After a long wait, the next game-changing technology has arrived: artificial intelligence (AI).
Empowering software and systems with the tools to make split-second, autonomous decisions is a greater than $15 trillion global opportunity by 2030, according to PwC analysts. The rise of AI is also responsible for sending the Dow Jones Industrial Average (^DJI 0.56%), S&P 500 (^GSPC 0.11%), and Nasdaq Composite (^IXIC +0.35%) to record highs.
Nvidia (NVDA +2.59%) has been the face of the AI revolution, with its graphics processing units (GPUs) accounting for the lion’s share of chips deployed in enterprise data centers.
But AI application companies aren’t slouches, either. Data-mining specialist Palantir Technologies (PLTR 1.87%), which uses AI across both of its core platforms (Gotham and Foundry), has seen its shares skyrocket by more than 2,200% since the start of 2023.
Image source: Getty Images.
Although no trend offers a more sizable addressable opportunity than AI, this game-changing innovation isn’t without its risks. Based on what history has to say, the latest milestone for AI stocks should have Wall Street sounding the alarm.
AI concentration risk has hit its crescendo
On the one hand, the long-term future for AI hardware and applications appears bright. Businesses are aggressively spending on AI infrastructure and expect generative AI solutions and/or large language models to make various aspects of their operations more efficient over time.
On the other hand, investors have a terrible habit of overestimating the adoption and/or optimization of new technologies. While Nvidia’s parabolic sales growth makes it clear that AI adoption isn’t a concern, we’re likely years away from businesses optimizing AI solutions to boost sales and profits. In other words, we have a disconnect between AI stock valuations and near-term optimization/utility.
AI Bubble hits same concentration level that resulted in the bursting of previous bubbles, including the Dot Com 🚨🚨🚨 pic.twitter.com/Dlhof894h6
— Barchart (@Barchart) April 5, 2026
According to an analysis from Bank of America Global Research, Bloomberg, and Global Financial Data, there have been four concentration bubbles between the U.S. and Japanese stock markets since 1964:
- In the early 1970s, the “Nifty Fifty” (a group of roughly 50 time-tested companies traded on the New York Stock Exchange) hit a 40% concentration within the S&P 500.
- In the latter half of the 1980s, a relatively small percentage of Japanese stocks accounted for 44% of the MSCI ACWI.
- In the early 2000s, tech and telecom stocks peaked at a 41% concentration of the benchmark S&P 500.
- In 2026, the 10-largest AI stocks reached a 41% concentration of the S&P 500.
All four events share a common trait, beyond a 40% (or greater) concentration in their respective index: aggressive valuations. Several established Nifty Fifty stocks were sporting price-to-earnings (P/E) ratios of 50 to 100 in the early 1970s, which more than doubled the average P/E of the iconic S&P 500.
S&P 500 Shiller PE Ratio hits 2nd highest level in history 🚨 The highest was the Dot Com Bubble 🤯 pic.twitter.com/Lx634H7xKa
— Barchart (@Barchart) December 28, 2025
Meanwhile, stock valuations were, arguably, even more egregious in the lead-up to the dot-com bubble bursting. The S&P 500’s Shiller P/E Ratio hit its all-time high of 44.19 in December 1999, mere months before the S&P 500 and Nasdaq Composite would begin their respective peak-to-trough descents of 49% and 78%.
AI stocks are also historically pricey, with Palantir’s price-to-sales (P/S) ratio topping 100 earlier this year, and Nvidia’s P/S ratio exceeding 30 as recently as November.
All three previous historical concentration peaks above 40% were soon followed by bubble-bursting events. If history rhymes, once more, the AI revolution is running on borrowed time.
Image source: Getty Images.
Scarcity is both a catalyst and a crutch for AI companies
In addition to next-big-thing technologies needing time to mature and AI stock valuations being unsightly, competitive dynamics in the AI space threaten to remove a foundational catalyst: scarcity.
Make no mistake, companies with superior hardware and AI applications have been rewarded. Nvidia’s compute superiority in data centers and the lack of large-scale competition for Palantir’s software-as-a-service platforms have allowed these pillars of the AI revolution to thrive.
But a strong argument can be made that AI hardware demand swamping supply has been the biggest spark for AI stocks. When demand for a good or service outstrips its supply, its price rises until demand tapers off. Nvidia has been able to command significant pricing power for its GPUs thanks to ongoing GPU scarcity.
Today’s Change
(2.59%) $4.76
Current Price
$188.67
Key Data Points
Market Cap
$4.6T
Day’s Range
$184.32 – $190.00
52wk Range
$95.04 – $212.19
Volume
5.9M
Avg Vol
179M
Gross Margin
71.07%
Dividend Yield
0.02%
However, increasing competition from all angles can alter this dynamic. While most of Wall Street is focused on external rivals (e.g., Advanced Micro Devices), the biggest threat to the face of the AI revolution, Nvidia, comes from within.
For more than a year, several of Nvidia’s top customers by net sales have been internally developing GPUs for use in their data centers. Although these GPUs don’t pack the same punch as Nvidia’s AI hardware, they’re considerably cheaper and not backlogged.
Between hyperscalers utilizing their own AI chips and external rivals ramping up production, the GPU scarcity that’s fueled strong pricing power and sky-high gross margins should fade. This could represent the fatal blow for a historically AI-stock-concentrated S&P 500.
