The 20-Year Signal: Preparing for the Predicted Reversal of AI-Fueled Stock Market Gains

There is no question that artificial intelligence has transformed the stock market as we know it over the past three years, all but single-handedly driving the S&P 500 to record after record. Companies that are sitting at the forefront of this technology, like NVIDIA (NASDAQ:NVDA), Microsoft (NASDAQ:MSFT), Google (NASDAQ:GOOGL), and Amazon (NASDAQ:AMZN), have delivered returns that seemed impossible just a few years ago.

Most importantly, the AI technology is real, and so far, the enthusiasm for this technology has been mostly justified, all while the long-term potential (and dangers) seems incredible. For investors who have stayed the course, portfolios have grown in a massive way over the last 36 months. But, and there is always a but, beneath this celebratory growth, there is a warning signal that has emerged.

This indicator, which looks at the close relationships between stock prices and underlying economics, hasn’t flashed at current levels in over 20 years, not since the dot-com bubble that burst in the early 2000s. The last time this signal appeared, it was followed by a multi-year bear market that erased roughly half of the market’s overall value. While history doesn’t always repeat itself, it does at least suggest that we need to look at these readings and consider, as investors, what the next smart move really is.

AI Mania Drives Stocks to Historic, Record Highs

The artificial intelligence revolution has, without question, captivated investors in a way that almost no other technology has. The belief that AI is going to and has already started to automate work, revolutionize manufacturing, and create entirely new industries is just scratching the surface of what investors are thinking about.

For companies that are in the AI world, all of this speculation and promise has led to explosive revenue growth. For example, NVIDIA’s data center business has become one of the fastest-growing divisions in corporate America, and Microsoft’s AI integration across basically its entire product lineup is driving enterprise adoption. In other words, this technology is actually driving real revenue.

This technological transformation has driven stock prices to levels that would have seemed absurd three years ago. The S&P 500 has surged almost 80% since late 2022, and NVIDIA has seen its market cap grow by trillions. The Mag 7’s technology stocks represent a whopping 30% of the entire S&P 500 market value, a level not seen since the 1990s. Valuations have also exploded as the forward P/E of the S&P 500 now hovers around 21 to 22, 40% above historical averages.

The optimism that is driving these valuation numbers is assuming that AI will continue to deliver double-digit earnings growth for years to come. More importantly, it also assumes that AI adoption won’t have any obstacles to adoption in the coming years and that companies building AI infrastructure will continue generating returns to justify their current high stock prices and valuations.

This pattern of pricing in future success isn’t new, as the late 1990s boom also seemed a trajectory where investors were basically paying today for profits they expected companies to deliver in the next five years. However, as the world knows, the technology of the internet boom from the 1990s succeeded, it’s how you’re reading this today, but many of the stocks crashed because valuations got too far ahead of any financial reality.

A Rare Signal Not Seen in Two Decades: What the Indicator Means

The warning signal that is flashing today is the Shiller CAPE ratio, a number that is an adjusted price-to-earnings ratio that measures stock prices relative to average inflation-adjusted earnings over the past 10 years. This metric attempts to smooth out short-term earnings volatility and give a longer-lens look at valuation extremes. As of December 2025, the Shiller CAPE sits above 35, a level that has only been seen twice in modern history: the Great Depression market crash and the 1999-2000 tech bubble peak.

The last time the US saw the CAPE ratio reach these current levels was in early 2000, right before the stock-market crash. At that peak, investors were paying extreme premiums for stocks based on the expectation (think hope) that internet companies would continue to just deliver on their unprecedented growth. History reminds us that the reality of what took place was that these companies did not enjoy unstoppable growth and that the market declined 49% over a 2.5-year period, the NASDAQ fell 78%, and it took 7 years to recover.

The CAPE ratio’s predictive power comes from understanding long-term perspectives, and unlike traditional P/E ratios that can be distorted by temporary earnings spikes or collapses, the CAPE ratio looks with a wide lens at a full decade of earnings. When this ratio does hit 30, it can be a historical signal that stocks are priced for perfection and have almost no room for disappointment.

What makes the current signal particularly troubling is that it isn’t just existing in a silo. Instead, it’s appearing alongside three other late-cycle indicators like three consecutive years of double-digit S&P 500 gains, forward P/E ratios at 40-year highs, extreme concentration in a handful of mega-cap stocks, and retail investor enthusiasm that feels eerily like the dot-com bubble. While no single indicator is going to be definitive, the combination of all three suggests that a major reversal could be on the horizon, and an entire generation of investors has never experienced what happens when valuations are this stretched.

History’s Verdict: A Reversal is Coming – How Smart Investors Respond

History is pretty unambiguous about what can happen when an extreme CAPE reading is taking place. For one, you might be in a position that creates a once-in-a-generation buying opportunity. At the 1929 stock peak, stocks fell 89% over three years, and again in 2000, you had the 49% S&P 500 drop over 2.5 years. What a drop might look like in the near future is unknown, but the pattern that occurred during these two previous instances of extreme valuations resolving through major corrections seems to be all too familiar to what we’re going through now with AI.

For investors, this could be the strategic opportunity they have been waiting for rather than a situation full of panic. The mistake during previous bubbles wasn’t recognizing that stocks were not expensive, but it was remaining fully invested at those prices and then panic-selling at the bottom. The smart approach today is about recognizing that current valuations create a situation where the downside potential far exceeds any upside of staying fully invested.

Building a cash reserve now so investors can “buy the dip,” as the saying goes, is arguably the best move, to the tune of upward of 30% of a portfolio being liquidated for cash or put into short-term bonds. If the market does decline by 40-50% from current levels, this cash can be invested into companies like Microsoft, Amazon, and NVIDIA at valuations these companies haven’t experienced in years. This could give investors a chance to double their position size.

The key here is going to be the right amount of discipline, as history shows that bear market pain isn’t the actual decline itself, it’s the psychology of trying to understand how to buy at the bottom. If the market falls 30%, there might be a voice going on that says it’s going to 50% and if it falls 50%, total economic collapse might be the voice on the news. In other words, investors should have capital reserved for buying during these panics so they can ignore the noise and acquire quality companies at steep discounts and prepare for the next decade’s best returns.