Ai Bubble Watch: Why This Isn’t Dot-Com 2.0
“If everyone is thinking alike, then somebody isn’t thinking.” General George S. Patton
The market finally exhaled last week after a strong run since the April Liberation Day near-crash, as the government shutdown dragged on and tariff policy remains unclear. Stocks slipped. Bitcoin neared crash territory. None of it was surprising, but it was a useful wake-up call for investors to revisit their own diversification and risk appetite.
The S&P 500 is still up close to 15% this year despite higher tariffs, weak labor signals and consumer sentiment near historic lows. That kind of run, supported by skittish investors, leaves little room for error. The economy isn’t rolling over into a recession. But investors are rethinking what they are willing to pay for big tech and the AI buildout. The spending is massive. The timelines are long. The payoff is uncertain. Even strong narratives bend under that weight.
The warnings are getting louder as we focus on our Ai bubble watch. Don’t just take it from me. Jamie Dimon talked about a sharp correction. Jeff Bezos called AI an industrial bubble. Ray Dalio said tech is overpriced. The recent Bank of America fund manager survey shows more than half of professionals now tagging AI as a bubble. Michael Burry, known for his “Big Short” bet against the housing market in 2008, later made into the movie of the same name, is betting against the big-tech high flyers.
When builders, bankers and skeptics line up on the same side of a concern, it matters. This is not 1999. But it is stretched. A 10 to 20% reset is easy to picture if sentiment turns down further or AI timelines slip. At a minimum, markets and the S&P 500, led by big tech and the notorious Mag-7, may cool to more normal levels over the next few years as other undervalued “non-tech” stocks, sectors and assets gain traction if inflation continues to moderate.
While many investors are thinking alike, the AI momentum that looks bubble-like today while backed by trillion dollar companies (not start ups) may still fuel major transformation, growth, and profits over time — just not in a straight line. The same was true for the dot-com era. Companies like Amazon, Microsoft, and Google fell hard before becoming the backbone of the modern economy. History suggests the survivors of this boom will reshape the next decade, even if the road there is volatile.
Sector Watch
Including a mix of sizes, styles, sectors and global exposure helps reduce concentration risk, buffers against turbulence and improves overall valuation balance. No one knows which areas of the market will lead in any given period, so maintaining a thoughtful blend when building out your portfolio strategy is essential whether you are investing to- or through – retirement.
AI related companies sit mostly in tech, communication services and consumer discretionary. But leadership has begun to broaden. Financials, utilities, industrials and other value-oriented sectors now offer more attractive valuations and steadier fundamentals. With technology touching nearly every industry and almost 40% of the S&P 500 anchored in tech, according to Goldman Sachs research, the challenge for investors is getting meaningful tech exposure while still managing volatility and correlation if there is a slowdown, or worse.
We have not experienced a major recession in 17 years and are not predicting one. Still, investors nearing or in retirement cannot afford to be caught off guard if momentum shifts in tech heavy areas or an unexpected black swan event hits. A disciplined, balanced portfolio helps protect against the unexpected, whether the goal is growth, income or a durable retirement plan.
Dot-Com Redux
Jon here. The comparison between today’s generative AI boom and the dot-com era is imperfect but useful. Bubbles don’t burst because no one sees them. They burst when everyone sees them and keeps buying anyway out of greed and FOMO.
The fear today is that AI stocks look frothy. Bulls argue that if everyone keeps warning about a bubble, it can’t be one. History disagrees. The late 1990s boom kept inflating even after every major newspaper compared valuations to tulips. Investors chased quick gains, IPO pops and the greater-fool trade.
One of the clearest markers of that era was how easy it became to mint returns. Simply adding “.com” to a company name, like the notorious Pets.com, sparked an average 74% jump in less than two weeks, according to Jay Ritter at the University of Florida. Many of those firms had no revenue. Some had business plans that looked like napkin notes. The name alone became a catalyst.
The AI boom shares similar pressure points. Yet you can’t short the biggest name in the space because it’s private. OpenAI has traded at valuations that would make 1999 look tame. Fund managers know that sitting out means underperforming and maybe getting fired. That keeps the bubble alive. Maybe AI becomes indispensable and justifies the price. Maybe not. You only know it’s a bubble after it breaks.
Past Booms and Busts
Big tech’s AI buildout is starting to resemble past American overinvestment binges. Today’s spending dwarfs the dot-com era and rivals the railroad frenzy of the 1800s. According to estimates from Dell’Oro Group and Goldman Sachs, tech giants are on pace to pour nearly four hundred billion dollars into data centers and chips next year, and more than five trillion over the next five years.
Research from Kai Wu at Sparkline Capital finds that to break even on this level of investment, the industry would need roughly two trillion dollars in annual AI revenue by 2030. Current AI-related revenue is closer to twenty billion. That gap would have made 1990s telecom executives nervous.
History is blunt about who gets hurt when spending outruns reality. Railroad companies transformed the economy but bankrupted wave after wave of builders. Telecom companies buried the U.S. in fiber in the late 1990s, only to collapse 92% when demand lagged. Most of that fiber sat dark for a decade.
The pattern is consistent. Infrastructure arms races create enormous value for society, but the shareholders funding them often get crushed. Today’s AI race looks less like a clean growth story and more like a prisoner’s dilemma where no company can afford to slow down. At the same time we believe this is not an innovation head-fake but a new phase in tech innovation as we illustrate below going back to the inception of television in 1927.
Amara’s Law
Amara’s Law explains a lot about what we’re seeing today with generative ai. People tend to overestimate new technology in the short run and underestimate it in the long run. That’s the loop investors are stuck in. The AI story is real, but the timelines are off. Markets are trying to price a decade of transformation into a few quarters of earnings. That creates tension between what the tech can deliver today and what investors hope it will deliver tomorrow.
Every major innovation cycle follows this same arc. The hype peaks before the infrastructure is ready. Expectations outrun adoption. Then the air comes out, not because the technology fails but because reality needs time to catch up. AI fits the pattern. The spending is massive, the use cases are early, and the payoff will take longer than Wall Street’s patience. That’s how bubbles form even when the underlying trend is legitimate.
Innovation History since 1927
The idea of a “bubble” is often overused, yet the definition is not always clear. Markets move in cycles, and investor perceptions shift over time. For every true bubble, like late 1990s dot-com stocks or mid-2000s housing, there are many periods when feared bubbles never formed. After the 2008 crisis, investors expected repeated bursts, yet the following decade became the longest bull market on record.
So the question of whether we are in a bubble is different from whether the market will face a setback. As the eras below show, from the machine age to personal computers, the internet and today’s AI wave, market pullbacks are normal. Transformational change takes time, and the eventual winners are rarely the first movers.
Ai Bubble Watch and High Stock Valuations
To separate routine pullbacks from real bubble risk, you have to focus on value. What matters in investing isn’t just the price you pay but what you receive in future earnings and cash flows. Valuation metrics like price to sales and price to earnings help measure this. They show not only what a share costs, but what you are getting for that cost.
The accompanying chart highlights the Shiller price to earnings ratio, which uses inflation adjusted earnings over the past ten years to smooth out economic cycles. Today’s level of about 38x means investors are paying $38 for every dollar of historical earnings. That is well above the long-term average of roughly 27x. These elevated levels reflect the market’s struggle with inflation, policy uncertainty and sharp swings in technology stocks.
With many measures showing the market is expensive by historical standards, there are a few points investors should keep in mind. First, valuations don’t predict short term returns. They simply show what investors are willing to pay based on expectations about the future. Even when stocks look pricey, markets can rise for long stretches if business fundamentals hold up. This is why timing the market is often counterproductive.
Next, while today’s environment echoes the 1990s tech boom, there are important differences. Unlike many dot com era firms, today’s market leaders are profitable, established and supported by strong balance sheets. As with past technology waves, AI could ultimately support broader economic growth.
Finally, high valuations don’t always “pop.” They can fall if prices decline, but they can also normalize if earnings grow. Some of today’s enthusiasm reflects expectations for stronger future profits. Recent earnings results have supported part of that optimism.
Invest According to Your Plan, Not Exuberance
AI is a real technological shift and will reshape parts of the economy, but the path will be uneven. Spending may overshoot. Valuations may need time to settle. Markets will move faster than fundamentals. Instead of trying to time a bubble, we advise investors in times of growth and uncertainty to anchor to their long-term plan and the following basics that keep portfolios resilient:
-Your portfolio is designed to benefit from long-term innovation while still meeting income and risk goals.
-Diversification matters. Heavy concentration in any single sector raises risk even when the story sounds compelling.
-A balanced mix of growth, value, bonds and other assets helps stabilize returns across cycles. It won’t capture every surge, but it keeps you on track.
The bottom line: AI is real and the innovation is meaningful, but the spending, expectations and valuations are moving faster than the fundamentals. The Ai bubble watch is real but not dotcom redux. Markets can stay optimistic for long stretches, yet they can also reset quickly when liquidity tightens or growth timelines slip. Your best move is to stay diversified, stay disciplined and avoid chasing the hottest stories at the hottest prices. Innovation creates opportunity, but your plan is what protects your wealth through every cycle.
For more information on our firm or to request a complimentary investment and retirement check-up, call (561) 210-7887 or email jon.ulin@ulinwealth.com.
Author: Jon Ulin, CFP® is the founder and Managing Principal of Ulin & Co. Wealth Management, an independent advisory firm based in South Florida for over 20 years. As a fiduciary wealth advisor, Jon helps successful individuals, families, and business owners nationwide with multi-generational planning, investment management, and retirement strategies. Learn more about Jon and our team at About/CV.
Note: Diversification does not ensure a profit or guarantee against loss. You cannot invest directly in an index.
Information provided on tax and estate planning is not intended to be a substitute for specific individualized tax or legal advice. We suggest that you discuss your specific situation with a qualified tax or legal advisor.
You cannot invest directly in an index. Past performance is no guarantee of future returns. Diversification does not ensure a profit or guarantee against loss. All examples and charts shown are hypothetical used for illustrative purposes only and do not represent any actual investment. The information given herein is taken from sources that are believed to be reliable, but it is not guaranteed by us as to accuracy or completeness. This is for informational purposes only and in no event should be construed as an offer to sell or solicitation of an offer to buy any securities or products. Please consult your tax and/or legal advisor before implementing any tax and/or legal related strategies mentioned in this publication as NewEdge Advisors, LLC does not provide tax and/or legal advice. Opinions expressed are subject to change without notice and do not take into account the particular investment objectives, financial situation, or needs of individual investors.
