See all articles

AI, high valuations, and market risks: What investors should know

AI opportunity: Market insights for Q3 of 2025.

By Ben Bakkum, CFA, CFP®

Sr. Investment Strategist, Betterment |

Published Oct. 06, 2025

AI, high valuations, and market risks: What investors should know Oct 6, 2025 9:30:00 AM AI opportunity: Market insights for Q3 of 2025. Key takeaways U.S. stocks are expensive but not at dotcom levels. AI spending is boosting valuations, though expectations are tamer. An AI slowdown could weigh on markets, making diversification key. Shifting some stock exposure to bonds can reduce risk. Valuations don’t predict near-term moves—stay balanced. Despite signs of a slowing US economy, client portfolios have rallied, thanks to a historic bull market. Yet a sense of unease permeates much of investors’ discourse. Inevitable comparisons to past bull runs that ended in crashes are top of mind, leaving investors to ponder the age-old question: Will this time be different?” And specifically: Is the hype around AI just another speculative market bubble? The Shiller Cyclically Adjusted Price-to-Earnings (CAPE) ratio serves as a popular gauge of frothiness in the valuations of stocks. Developed by Yale economist Robert Shiller, the indicator divides the S&P 500’s price by its companies’ inflation-adjusted earnings averaged over the past 10 years, showing how cheap or expensive the market looks while smoothing out short-term fluctuations. U.S. large-cap stocks are about as expensive as they were at peak pandemic recovery levels in 2021, though still below dotcom heights. Bullishness, largely on the back of the AI investment boom, has pushed valuations into expensive territory, so investors should temper their expected returns. However, it doesn’t necessarily mean that the market will not rally further, or that a selloff is imminent. Goldman Sachs Global Investment Research notes that long-term earnings growth expectations, implied by market pricing, are still below 2021 and dotcom-era levels, which might indicate that investor expectations have not over-extended themselves as much as before. Goldman also highlights an important difference: The average P/E ratio of the 10 largest tech, media, and telecom stocks today is 31x—compared with 41x, at the height of the dotcom bubble, in December 1999. Still, a slowdown in AI investment seems inevitable, likely placing downward pressure on stock prices. This risk underscores the importance of portfolio diversification, both across asset classes and geographies. As we often note, stock valuations outside the U.S. look less elevated, even after year-to-date rallies in international developed and emerging markets that have outpaced the U.S. The market is also contending with fresh tariff announcements, along with the lingering effects of those imposed earlier in the summer. Higher import costs have added pressure to goods prices, but inflation has remained low enough for the Federal Reserve to cut rates in September in a bid to support the labor market. More accommodative monetary policy should support economic growth and portfolio performance, though it carries risks of higher inflation and financial instability. These developments ultimately outweigh the significance of yet another short government shutdown in the eyes of the market. For investors with shorter time horizons and lower risk tolerances, shifting slightly from stocks to bonds can align a portfolio with a waning risk appetite. Lower volatility in inflation and interest rates can now potentially support risk-adjusted returns in fixed income as well. However, we are loath to suggest ever completely giving up on stocks. Price-to-earnings ratios have climbed, but these metrics have a poor track record of predicting near-term market moves, and investors should balance their portfolios for both downside and upside risks. AI is a powerful force reshaping markets, but it shouldn’t dominate portfolio decisions on its own. The task is to keep portfolios anchored to long-term goals, weighing transformative innovations like AI alongside broader risks and opportunities—not reacting to headlines.

Key takeaways

U.S. stocks are expensive but not at dotcom levels.
AI spending is boosting valuations, though expectations are tamer.
An AI slowdown could weigh on markets, making diversification key.
Shifting some stock exposure to bonds can reduce risk.
Valuations don’t predict near-term moves—stay balanced.

Despite signs of a slowing US economy, client portfolios have rallied, thanks to a historic bull market. Yet a sense of unease permeates much of investors’ discourse. Inevitable comparisons to past bull runs that ended in crashes are top of mind, leaving investors to ponder the age-old question: Will this time be different?” And specifically: Is the hype around AI just another speculative market bubble?

The Shiller Cyclically Adjusted Price-to-Earnings (CAPE) ratio serves as a popular gauge of frothiness in the valuations of stocks. Developed by Yale economist Robert Shiller, the indicator divides the S&P 500’s price by its companies’ inflation-adjusted earnings averaged over the past 10 years, showing how cheap or expensive the market looks while smoothing out short-term fluctuations. U.S. large-cap stocks are about as expensive as they were at peak pandemic recovery levels in 2021, though still below dotcom heights.

Bullishness, largely on the back of the AI investment boom, has pushed valuations into expensive territory, so investors should temper their expected returns. However, it doesn’t necessarily mean that the market will not rally further, or that a selloff is imminent. Goldman Sachs Global Investment Research notes that long-term earnings growth expectations, implied by market pricing, are still below 2021 and dotcom-era levels, which might indicate that investor expectations have not over-extended themselves as much as before. Goldman also highlights an important difference: The average P/E ratio of the 10 largest tech, media, and telecom stocks today is 31x—compared with 41x, at the height of the dotcom bubble, in December 1999.

Still, a slowdown in AI investment seems inevitable, likely placing downward pressure on stock prices. This risk underscores the importance of portfolio diversification, both across asset classes and geographies. As we often note, stock valuations outside the U.S. look less elevated, even after year-to-date rallies in international developed and emerging markets that have outpaced the U.S.

The market is also contending with fresh tariff announcements, along with the lingering effects of those imposed earlier in the summer. Higher import costs have added pressure to goods prices, but inflation has remained low enough for the Federal Reserve to cut rates in September in a bid to support the labor market. More accommodative monetary policy should support economic growth and portfolio performance, though it carries risks of higher inflation and financial instability. These developments ultimately outweigh the significance of yet another short government shutdown in the eyes of the market.

For investors with shorter time horizons and lower risk tolerances, shifting slightly from stocks to bonds can align a portfolio with a waning risk appetite. Lower volatility in inflation and interest rates can now potentially support risk-adjusted returns in fixed income as well. However, we are loath to suggest ever completely giving up on stocks. Price-to-earnings ratios have climbed, but these metrics have a poor track record of predicting near-term market moves, and investors should balance their portfolios for both downside and upside risks.

AI is a powerful force reshaping markets, but it shouldn’t dominate portfolio decisions on its own. The task is to keep portfolios anchored to long-term goals, weighing transformative innovations like AI alongside broader risks and opportunities—not reacting to headlines.