By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel agreed that while the Shiller CAPE ratio suggests high valuations, the current environment's unique factors, such as AI-driven earnings and low real rates, may keep P/Es elevated. However, they also warned about potential risks like breadth risk, rising capex costs, and the fragility of earnings growth.
Risk: Breadth risk and the potential failure of AI capex to deliver broad margin gains across the S&P 500, leading to a sharper P/E compression.
Opportunity: The opportunity lies in monitoring the transition of hyperscalers from 'growth at any cost' to 'margin protection' as rates stay higher for longer, which could be a catalyst for multiple compression.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
In the history of the S&P 500's Shiller CAPE ratio, the market has rarely traded at this level.
The artificial intelligence boom has powered the market through many headwinds and bolstered S&P 500 earnings estimates.
History rarely repeats itself, but it often rhymes.
As of June 2, the benchmark S&P 500 (SNPINDEX: ^GSPC) index had hit 23 all-time highs in 2026. That's following 96 all-time highs in 2024 and 2025.
Now sitting just below its latest high, the market has absolutely crushed it despite facing several headwinds, including doubts about AI at the very beginning of the year; the Iran war, which drove up oil and gas prices; and now, concerns about elevated inflation.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
Exuberance over artificial intelligence and projected earnings growth of the S&P 500 has powered the market through these concerns. In fact, the stock market has traded at this high a valuation only one other time in its 69-year existence. History couldn't be any clearer about what happens next.
One way to assess the value of the S&P 500 is through the Shiller CAPE ratio, which compares the S&P 500's price to its 10-year average inflation-adjusted earnings per share (EPS).
The idea behind using a 10-year, inflation-adjusted average EPS is to smooth out the noise of the economic cycle or periods with differing inflation to get a more holistic view of how the market is trading. For a while now, the S&P 500's Shiller CAPE ratio has been creeping toward alarmingly high levels.
As you can see, dating back to the creation of the S&P 500 in 1957, the only other time the index had a higher Shiller CAPE ratio was in 2000, right before the dot-com bubble popped.
The tale bears similarities to the current situation. During the dot-com bubble, venture capital for start-ups soared amid the rise of the internet. Investors also poured into tech stocks, driven by fear of missing out (FOMO).
The fallout of the dot-com bubble was not pretty. The S&P 500 fell nearly 77% over the next two years.
The Shiller CAPE ratio also approached this level in late 2021 and early 2022, as people built up savings during the COVID-19 pandemic's height and pent-up demand emerged. However, inflation also got out of hand, and the Federal Reserve had to hike interest rates incredibly fast, leading to a 20% sell-off in 2022.
That's not nearly as dramatic as the dot-com bubble, but the point is that when the S&P 500 reaches these levels, a sharp sell-off has always followed.
Many would argue that the artificial intelligence boom is very different than the dot-com bubble.
The internet boom featured many start-ups without revenue receiving venture capital, and large tech and telecom companies funded much of the internet infrastructure build-out with debt. In 1999, there were 457 IPOs, most of which were related to the internet.
Today, much of the AI infrastructure build-out is being funded by massive, highly profitable companies that generate tremendous free cash flow (FCF). However, the hyperscalers have begun to see their FCF wells dry up and are taking on debt to fund infrastructure.
The IPO market has also been quite soft in recent years. In the U.S., there have been 63 IPOs year to date. There were 150 in 2024 and 250 last year, according to Renaissance Capital.
Of course, the other big difference is that many start-ups have been able to raise funding and stay in the private markets longer than they used to.
Some also believe that AI will be more transformational than the internet, and that the market is not yet at peak dot-com-era levels, so it may still have some runway.
Long-term investors with a five- or 10-year horizon don't need to do anything, as the market has bounced back from most of these steep sell-offs. But they should be aware of where the market is and check their portfolios to see if they hold AI stocks with ultra-high valuations that could be vulnerable to a sell-off.
I'm not saying to not own any of these, but be cognizant of your exposure to them.
If you are concerned about preserving capital over the next two to three years and highly valued AI stocks take up the majority of your portfolio, it may be time to take some gains, increase cash, or add exposure to sectors that are more resilient during market sell-offs.
Before you buy stock in S&P 500 Index, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and S&P 500 Index wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $449,393! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,366,006!
Now, it’s worth noting Stock Advisor’s total average return is 983% — a market-crushing outperformance compared to 212% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of June 3, 2026. *
Bram Berkowitz has no position in any of the stocks mentioned. The Motley Fool has no position in any of the stocks mentioned. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"Valuation extremes like the current CAPE level have historically foreshadowed meaningful drawdowns unless real yields stay suppressed and earnings breadth broadens."
While the headline grabs attention, CAPE is not a precise timing tool, and the article glosses over regime shifts that can sustain high valuations. The current environment features large-scale corporate buybacks, exceptionally resilient cash flows for mega-cap AI leaders, and a long stretch of low real rates that compress discount rates—conditions that can keep P/Es elevated longer than in the dot-com era. The piece downplays breadth risk and the possibility that AI-driven earnings are more concentrated than the index implies. If rates drift higher or if AI capex slows, multiple compression could arrive with little warning, even as a few stocks stay strong.
The strongest counterargument is that CAPE's predictive power has weakened in recent cycles. If real yields stay low or fall further, multiples can re-rate and breadth risks may be overstated.
"The Shiller CAPE is an imperfect tool for measuring modern, software-driven index profitability, making historical comparisons to 2000 fundamentally flawed."
The article leans on the Shiller CAPE ratio to signal a 2000-style collapse, but it ignores the fundamental shift in S&P 500 composition. Today’s index is dominated by high-margin, asset-light tech giants with massive moats and recurring revenue, unlike the debt-laden telecom and speculative hardware firms of the dot-com era. While a CAPE ratio near 35x is undeniably expensive, it reflects a structural increase in profitability and lower capital intensity. The real risk isn't a valuation reset based on historical 'rhymes,' but rather a failure of AI to move from experimental Capex to tangible operational efficiency gains in the bottom line for the broader S&P 500.
If interest rates remain 'higher for longer,' the discount rate applied to these future AI earnings will compress multiples regardless of how high-quality the underlying cash flows are.
"Shiller CAPE alone is a poor timing tool; what matters is whether S&P 500 earnings growth of ~19% can justify 30x multiples or if AI capex becomes value-destructive."
The article conflates valuation extremity with inevitable drawdown, but ignores a critical difference: earnings growth justification. Yes, Shiller CAPE is near 2000 levels (~30x), but S&P 500 forward earnings growth is ~19% YoY versus dot-com's 0% real earnings growth. The 2022 correction happened because rates rose 425bps in 12 months—a shock, not a valuation reversion. Today's hyperscalers fund AI capex from FCF, not debt-fueled speculation. The article's 77% and 20% historical precedents are real, but they conflate different causal mechanisms. The actual risk isn't valuation alone; it's whether AI capex delivers ROI or becomes stranded.
If AI ROI disappoints and capex doesn't translate to earnings growth, we could see multiple compression from 30x to 18-20x even without a rate shock—a 35-40% drawdown. The article's point that 'history rhymes' has teeth if fundamentals don't hold.
"CAPE extremes plus rising debt-funded capex point to subpar forward returns over the next 24 months even if recession is avoided."
The article correctly flags the Shiller CAPE near dot-com peaks, but underplays how today's earnings base is already far higher and more concentrated. With S&P 500 EPS growth still projected above 15% for 2026-27, a re-rating lower could occur without a 2000-style collapse if AI capex sustains margins. The overlooked risk is not the initial multiple compression but the debt-funded hyperscaler buildout now visible in rising interest expense at Microsoft and Meta, which could cap FCF and trigger faster de-rating once growth disappoints. Long-term holders should monitor Q3 capex trends closely rather than assume mean reversion is inevitable.
AI-driven productivity gains could justify CAPE staying above 30 for a decade, as occurred in the 1990s before the final leg higher, rendering historical comparisons misleading.
"Breadth risk can blunt CAPE-driven bulls, causing sharper P/E compression if AI ROI remains concentrated in a few megacaps."
Gently push back on Gemini: even if CAPE reflects higher profitability, its Achilles’ heel is breadth risk. A few AI-enabled megacaps can sustain index EPS, but if AI capex fails to deliver broad-margin gains across the S&P 500, the broad stock universe can disappoint while the mega-caps stay elevated. That dynamic creates a skewed cushion for multiples today, meaning a surprise earnings miss outside the AI cluster could trigger a sharper, more abrupt P/E compression than implied.
"The rising cost of capital will force a shift toward margin discipline, which is the most likely trigger for P/E compression."
Grok, your point on rising interest expense is the missing link. While Claude notes hyperscalers fund capex via FCF, you correctly identify that the cost of capital is no longer zero. If rates stay higher for longer, the hurdle rate for AI projects rises, forcing a pivot from 'growth at any cost' to 'margin protection.' This transition is the real catalyst for multiple compression, not just a simple valuation mean-reversion as the original article suggests.
"The valuation risk isn't CAPE itself but the cliff-edge collapse in forward earnings growth if capex ROI disappoints or rates stay elevated."
Gemini and Grok both identified rising capex costs, but neither quantified the margin trap: if hyperscalers' FCF growth slows from 25% to 8% due to higher rates and capex intensity, S&P 500 forward earnings growth collapses from 19% to single digits. That's the inflection point. CAPE at 30x on 19% growth is defensible; CAPE at 30x on 5% growth triggers a 40% re-rating. The article's real insight isn't valuation extremity—it's that the earnings growth cushion is fragile and rate-dependent.
"Earnings concentration creates a narrower, more abrupt re-rating path than Claude's uniform margin-trap math implies."
Claude's 19-to-5% earnings collapse scenario assumes uniform margin pressure across the index, yet ChatGPT's breadth point shows the opposite risk: AI leaders can protect their own FCF while non-tech names absorb the capex drag. That asymmetry could produce a stealthier de-rating, where the S&P 500 multiple compresses mainly because the denominator stops growing outside the mega-caps rather than from a broad rate shock.
The panel agreed that while the Shiller CAPE ratio suggests high valuations, the current environment's unique factors, such as AI-driven earnings and low real rates, may keep P/Es elevated. However, they also warned about potential risks like breadth risk, rising capex costs, and the fragility of earnings growth.
The opportunity lies in monitoring the transition of hyperscalers from 'growth at any cost' to 'margin protection' as rates stay higher for longer, which could be a catalyst for multiple compression.
Breadth risk and the potential failure of AI capex to deliver broad margin gains across the S&P 500, leading to a sharper P/E compression.