The artificial intelligence boom has divided Wall Street investors into two distinct groups:
1. Bulls who believe artificial intelligence is a transformational technology and think the massive CAPEX spending is logical, leading to years of profitability for AI-native companies.
2. Bears who believe that the massive CAPEX spending will lead to a poor return on investment and believe that much of the AI spending is “circular.”
In my view, rather than taking a one-sided stance, investors are best served by exercising some nuance when analyzing the AI market. From a bullish perspective, I have no doubt that AI is transformational. For instance, the other day, I plugged a 50-slide PowerPoint presentation into Meta Platform’s (META) “Muse” and gave it prompts. First, Muse took four years of data I provided and created a clean, accurate graphical representation. Then, I simply prompted the agentic AI platform to “Optimize” my slide show. Within seconds, Muse corrected any grammatical errors, perfected my layouts, and suggested some improvements. Before AI, optimizing this PowerPoint presentation would probably take two to three hours.
Conversely, history says that although the AI buildout is the largest in history, it will eventually end badly, because all bull markets do. Whether it was the railroad bull market of the late 1800s, the post-World War 2 industrial boom, or the internet boom of the late 1990s, the market is undefeated, and all bull markets end. In fact, most leading stocks in these eras fell 50% or more from their highs, never to recover.
Since the U.S. stock market began trading, one certainty has been that, in the long run, equity markets tend to rise. In other words, if you look at a 50-year chart of the S&P 500, it trends from the bottom left to the top right. Nevertheless, many industry-specific booms and busts have occurred during that time.
One interesting thing that I have noticed on social media is that “shock jocks” often garner more attention than bulls. Perhaps this phenomenon is due to the popularity of the Hollywood smash hit film “The Big Short,” or perhaps it is just human nature. Regardless, if you log into financial Twitter, chances are you will see bears beating their chest, using data points to connect the current AI bull market to that of the late 1990s or 1929.
“A bubble is any asset that goes up…that you don’t own.” ~ Art Hogan
But the bearishness isn’t limited to retail investors. This morning, I read a headline that famous investor Ray Dalio “Warns AI is a classic bubble nearing its bursting point.” Michael Burry, who became famous for his “Big Short” during the financial crisis, became bearish on AI leaders in late 2025. While these investors are well known and highly respected, history shows that their market calls are far from gospel. For example, Ray Dalio has warned of a structural debt crisis and market downturn for the past 12 years. Although the ballooning debt is certainly a concern, the S&P 500 Index was higher in 10 of the past 12 years on a total return basis, registering a 278% return. Meanwhile, AI-related stocks Michael Burry warned about in 2025, like NVIDIA (NVDA) and Palantir (PLTR), recently printed fresh all-time highs.
Bubble is Not a Four-Letter Word
In most Wall Street circles, the word bubble has a negative connotation. Because a bubble suggests that stock prices detach from fundamentals and ultimately come crashing down, such a reputation is partially earned. However, another way to view bubbles is as a period when stock prices rise rapidly over a short time.
“When I see a bubble forming, I rush in to buy, adding fuel to the fire. That is not irrational.” ~George Soros
I view bubbles as a rare time to generate a lot of portfolio alpha in a short time.
The question is, “Where in the cycle are we?” Five data points suggest that the market may just be entering its explosive phase, including:
1. Historical Precedent: The Nasdaq 100 precedent after the Netscape vs. ChatGPT launch suggests the meat of the equity move may be just starting. Thus far, the precedent is eerily similar.

2. Valuations: The trailing P/E for the Nasdaq 100 is currently ~29.5x. At the March 2000 Dot-com peak, the Nasdaq 100 trailing P/E was over 100x for many leaders. Cisco (CSCO), the market leader of that cycle, topped with a P/E of ~200x. NVIDIA, the current AI leader, has a P/E of just 34.67x.

3. Lack of IPO Animal Spirits: In 2026, there have been 251 IPOs. Conversely, in 2000, there were 446 IPOs. While future IPOs like Anthropic and OpenAI already generate billions in revenue, many dot-com IPOs were pre-revenue.
4. Price Action is Not Euphoric: In 1999, Qualcomm, the highest-flying internet stock, gained 2,600%. By early 2000, the stock continued its uptrend before forming a climax top pattern and topping. While AI stocks like SanDisk (SNDK) have seen similarly strong returns, key differences remain. Qualcomm (QCOM) shares were stretched more than 200% above the 200-day moving average, were near-vertical, and exhibited several classic overnight “exhaustion gaps.”

Conversely, SanDisk shares are just 47% above the 200-day moving average and have been basing since July – not exactly euphoric.

5. Retail Sentiment is Bearish: Historically, bubbles burst when retail investors work themselves into a frenzy. It happened with the real estate bust of 2008 (when many people began flipping real estate), and the internet bubble of 2000 (when many people became day traders). Today, retail sentiment is in the gutter. According to the American Association of Individual Investors, bearish sentiment is currently higher than bullish sentiment.

Bottom Line
The constant labeling of AI as a “bubble” shouldn’t deter investors from participating in one of history’s most transformative technological shifts. While market excesses and cyclical corrections are inevitable, today’s AI expansion rests on real revenue generation, reasonable valuations compared to previous manias, and a foundation of widespread institutional skepticism rather than euphoric retail mania.
Want the latest recommendations from Zacks Investment Research? Today, you can download 7 Best Stocks for the Next 30 Days. Click to get this free report
This article originally published on Zacks Investment Research (zacks.com).
| 45 min | |
| 1 hour | |
| 2 hours | |
| 2 hours | |
| 2 hours | |
| 2 hours | |
| 3 hours | |
| 3 hours | |
| 3 hours | |
| 3 hours | |
| 3 hours | |
| 3 hours | |
| 3 hours | |
| 3 hours | |
| 3 hours |
Join thousands of traders who make more informed decisions with our premium features. Real-time quotes, advanced visualizations, alerts, and much more.
Learn more about Finviz Elite