The Stock Market Is Repeating a Pattern Not Seen in Over 2 Decades. History Says This Could Come Next.

If you’d put $1,000 into the Nasdaq Composite index on March 10, 2000, your investment wouldn’t have regained its starting value until April 23, 2015. That’s roughly 15 years of negative returns, highlighting the potentially devastating impact of buying stocks at the peak of a bubble.

This example is pertinent because the Nasdaq and S&P 500 are once again near all-time highs, driven by generative artificial intelligence (AI). Let’s explore the historical parallels between the contemporary boom and the dot-com bubble of the late 1990s to decide where stocks are headed during the next few years.

Stocks repeat an alarming historical pattern

The starkest warning about potential market overvaluation comes from the cyclically adjusted price-to-earnings (CAPE) ratio. This metric compares inflation-adjusted corporate earnings over 10 years to reduce the impacts of short-term fluctuations. And right now it stands at 41, well above its historical average of 17.4 and a level only surpassed during the dot-com bubble when it hit an all-time high of 44 in 1999.

The good news is that this isn’t exactly an apples-to-apples comparison. As a backward-looking metric, the CAPE ratio doesn’t reflect earnings growth rates, which help to justify current stock prices. And this is important because, unlike the dot-com bubble, which was driven by unprofitable tech companies, the current AI boom is led by established and profitable behemoths, including Nvidia and Micron.

However, underneath the surface, these companies’ huge profits rest on less-than-ideal foundations. As infrastructure providers, they rely on their clients being able to turn their hardware into profitable consumer-facing tech services. However, the frontier labs face an uncertain future as they navigate competition, regulation, and spiraling costs.

Frontier models are on shaky ground

While most people agree that AI can transform how people live and do business, there is still uncertainty about how effectively these algorithms can be monetized. In fact, the high gross margins currently enjoyed by infrastructure companies like Nvidia (75%) and Micron (85%) make it much harder for frontier-model providers to create sustainable business models.

According to leaked internal documents obtained by the Financial Times, leading AI company OpenAI had a loss of $21 billion against $13 billion in revenue for the full year 2025, with much of this due to an immense research and development budget as it fights to stay at the cutting edge of AI model development. It is important to note that the company also has substantial future cash requirements, with management expecting to burn through almost $280 billion by the end of 2030, with much of that going to computing power and other forms of AI infrastructure.

It may be only a matter of time before Wall Street decides it is no longer willing to pour money into backing frontier model companies, especially as cheap Chinese rivals continue to catch up to their technical capabilities.

The political powder keg

The market looks shaky. But the increasingly volatile U.S. political situation could end up being the straw that breaks the camel’s back. Under President Donald Trump, the White House has embarked on a series of inflationary policies, including wide-ranging tariffs on top trading partners and a military operation in Iran, which has led to the partial blockage of the Straight of Hormuz, a major chokepoint for the global oil trade.

In response, the Federal Reserve has been forced to raise its interest rates to a target range of 3.75% to 4.00% in an effort to fight inflation. U.S. bond yields are also rising, with the 10-year Treasury near 5%.

Government debt represents the risk-free rate in the U.S. economy, so when Treasury yields rise, so do capital costs for people and businesses. This will make it even more expensive for AI companies to build their data centers while making investors demand higher returns for their capital. Both factors could contribute to bringing down tech valuations. Investors can weather the storm by diversifying their portfolios outside the sector.