The Stock Market Is Repeating a Pattern Not Seen in Decades: History Says This Will Come Next

During the past 100 years, the S&P 500 has returned an average of 10% annually. However, during the past three years, the market has soared by 72% — doubling the typical expected annual growth rate because of optimism surrounding the generative artificial intelligence (AI) megatrend.

If history is anything to go by, this period of elevated growth won’t last forever. In fact, there is a real possibility that stock market returns could slow or even swing negative during the coming years. Let’s dig deeper to find out what might happen next for the major indexes.

History tends to rhyme

The current generative AI boom has some interesting similarities with the dot-com bubble, which occurred in the late 1990s and saw the S&P 500 index peak at an all-time intraday high of 1,552.87 on March 24, 2000. This followed a return of 21% in 1999, 28.5% in 1998, and 33% in 1997.

Just like today, the above-average returns were being driven by a new technology (the internet) that promised to transform the way people lived and worked. While it eventually delivered on expectations, it didn’t happen soon enough for the companies trading at valuations that already priced in years of explosive growth that couldn’t be matched by reality. By October 2002, the S&P 500 had lost an eye-watering 49% of its value.

The bulls will argue that the AI boom is fundamentally different because, unlike the dot-com bubble, it is being driven by stable and profitable technology giants that are sophisticated enough to know what they are doing. While this is a strong point, it doesn’t make a crash impossible.

Why AI could crash

Instead of inflated valuations, the current bubble is driven by inflated spending. Analysts at Goldman Sachs expect global AI-related capital expenditures (capex) to top $1 trillion this year, mostly for data center equipment like graphics processing units (GPUs), memory chips, and other infrastructure.

The problem is that demand for consumer-facing large language models (LLMs) might not live up to expectations. The tech giants funding the AI build-out could be left sitting on a mountain of rapidly depreciating assets that will never recoup their bloated purchase prices. Furthermore, the companies that supply this hardware (like Nvidia and Micron) could see a sharp reduction in their growth rates and profit margins.

Image source: Getty Images.

The example of Cisco Systems offers a striking historical parallel. The company boomed by providing the routers and switches that powered early internet companies. But when internet hype faded, and clients suddenly cut back on capex spending, the seemingly bulletproof company lost a whopping 88% of its peak value in just two years.

Goldman Sachs estimates that about half of the S&P 500’s earnings growth is coming from AI capex. A sudden end to the party could have substantial effects on the performance of the whole index.

What comes next?

Periods of substantially above-average S&P 500 growth often end with years of below-average or negative growth as the index reverts to its mean of about 10%. The current AI boom shows strong similarities to the dot-com bubble, raising the possibility of another big bust.

Still, there is no need to panic. Timing the market is notoriously difficult and can lead to investors missing out on the elevated returns that tend to happen late in a bubble. Furthermore, the years of negative returns seem to be coming less often as the U.S. government and Federal Reserve show an increasing willingness to intervene through aggressive stimulus. Instead of selling everything, investors should look at a potential future downturn as an opportunity to buy quality stocks at a discount.