The S&P 500 Is Flashing a Warning Not Seen Since the Dot-Com Bubble; History Suggests This Could Happen Next

Historically, the S&P 500 has been one of the world’s greatest wealth-generating machines, returning an average annual return of over 10% since its launch in 1957. But it has been far from smooth sailing. And if you bought shares at the wrong times (such as the peak of the dot-com bubble in 2000 or before the great financial crisis of 2007-2008), it would take several years to recover the value of your original investment.

Is 2026 another bad time to buy? While it’s impossible to know for sure, several historical parallels offer clues about what might happen next.

Image source: Getty Images.

Stocks are historically pricey

The cyclically adjusted price-to-earnings (CAPE) ratio is a stock market valuation metric that compares the S&P 500’s current price with its inflation-adjusted earnings over the past decade. The long duration of the comparison helps smooth out the impacts of the business cycle, allowing investors to identify periods when shares are unusually pricey.

Right now, the market has a CAPE ratio of 42.5, a level not seen since it peaked at 44.2 in 1999 during the dot-com bubble. And there are some sharp parallels between the two time periods.

Just as in the late 1990s (when the internet was becoming mainstream), the world is experiencing a technology megatrend: Generative artificial intelligence (AI), which promises to revolutionize the way we live and do business. In both scenarios, the “pick and shovel” providers that supply hardware and physical infrastructure capture the lion’s share of early profits while frontier software remains speculative.

Will this time be different?

Stock market crashes occur because people assume the current bubble will be different from the last bubble. That said, the current generative AI boom differs starkly from the dot-com craze over 25 years ago. Unlike the late-1990s rally, which was driven by unprofitable companies with shaky business models, today’s boom has been led by large and successful technology companies like Nvidia.

The chipmaker earned an eye-popping net income of $58.3 billion in the first quarter alone. And other AI infrastructure leaders, such as Micron Technology and Sandisk, have also enjoyed explosive operational improvements that help justify the recent growth in their stock prices.

That said, demand for AI infrastructure depends on its end users believing they can use it to create profitable consumer-facing services in the future. And an elephant in the room could bring the party to a screeching halt: China.

The world’s second-largest economy is rapidly gaining ground on American frontier models while also offering lower prices. If this trend continues, we could see gross margins begin to erode as the industry races to the bottom, similar to what happened when Chinese competition hit other emerging technologies, such as electric vehicles.

Chinese companies are also entering the infrastructure market, with the Hefei-based chipmaker CXMT rising to become the country’s largest company, with a market capitalization of 3.3 trillion yuan, or $487.7 billion. CXMT plans to pour its resources into mass-producing advanced memory for AI data centers. And this could eventually lead to a glut in the market that hurts the prospects for its booming U.S. rivals.

What should investors do?

While the signs are increasingly pointing to market overvaluation, this doesn’t necessarily mean investors should sell all their stocks or short the S&P 500. Timing the market is difficult. And even when you get things right, government and monetary policy actions (such as lowering interest rates or passing stimulus packages) can quickly turn things around.

Long-term investors should view a potential market crash as a buying opportunity to scoop up quality stocks for a discount. The period before the crash can be used to accumulate cash or lower-risk assets such as bonds and preferred stocks, to help diversify your portfolio.