The Stock Market Just Did Something for the 2nd Time in 100 Years, and History Says What Comes Next

The cyclically adjusted price-to-earnings (CAPE) ratio has now closed above 40 for three months running. In more than a century of stock market data, that has happened exactly one other time: the dot-com bubble that eventually saw the S&P 500 (^GSPC +1.66%) lose nearly 50% of its value from March 2000 to October 2002.

So, should you be worried?

What the CAPE ratio tells us about market valuations

The CAPE ratio is basically a smoothed-out version of the ordinary price-to-earnings ratio (P/E), in which the price of a stock — or the entire market — is divided by its earnings per share (EPS). The difference is that the CAPE ratio takes the level of the S&P 500 and divides it by the average of its inflation-adjusted earnings over the previous 10 years.

That 10-year average is critical. A single bad year or one blowout quarter can send a regular P/E ratio all over the place, but the decade-long time frame takes out the noise when measuring earnings growth over time.

Why three months above 40 echoes the dot-com bubble

While the CAPE reaching such heights is concerning in and of itself, the three-months-in-a-row part is particularly so. It really drives home just how expensive the market is, and it makes the parallel to 1999 more stark.

Back then, the S&P 500 had roughly tripled over five years on enthusiasm for the internet, which investors believed would change the world. It did, but not before the bottom fell out. The market got ahead of the technology’s profitability and paid the price.

This belief that a new technology made it “different this time” led investors to ignore the signs and pile into companies with wild valuations just because they had “.com” at the end of their name.

Why this time is actually different

Now, while this obviously seems to rhyme with today, there’s a genuine argument that this time is materially different.

The dot-com highs were built on companies that were, in a lot of cases, barely companies at all, many with no profits — some without even revenue to speak of.

Today’s market is dominated by tech giants that make enormous amounts of money. Microsoft, Nvidia, Alphabet, and Amazon have the balance sheets and the income statements to justify a good chunk of what investors are paying.

If earnings keep climbing at anything like the pace they have, it’s not hard to picture the CAPE drifting back toward normal without stock prices needing to correct (even if they slow their pace of growth).

The hidden flaw in the AI investment thesis

There’s merit to this, but it’s ignoring a pretty major flaw in the AI ecosystem: All of this, by and large, rests on the frontier model builders at the heart of the boom — companies like OpenAI and Anthropic — and these companies are losing money on a scale we’ve just never seen before.

Image source: Getty Images.

Yes, they are also growing revenue at a pace never seen on this scale, but the costs are rising with it. This is a fundamental problem — one that still hasn’t been solved.

There’s something that gets lost in the conversation, in my opinion: A technology can be truly revolutionary — fundamentally society-altering — and still be a bad business.

Consider flight. Few innovations have had a more profound effect on society, yet the airline industry is famously a tough business, to put it lightly.

Warren Buffett once told his shareholders, “If a farsighted capitalist had been present at Kitty Hawk, he would have done his successors a huge favor by shooting Orville down.” The airline business, he said, is “the worst sort of business” because it “grows rapidly, requires significant capital to engender the growth, and then earns little or no money.”

I think we may find AI to be a similar sort of business — at least the business of creating frontier models.

What smart investors should do right now

Now, of course, I could very well be wrong — plenty of smart people would clearly disagree with me. And, even if I am proven right in the end, when the music stops is the real question. The CAPE spent nearly a year and a half above 40 before the market peaked in 2000.

So, where does all this leave you? Not running for the exits. It’s a cliche, I know, but I don’t think it’s a tired one: Time in the market beats timing the market. That’s the real lesson from history.

Instead, treat this as a reason to look hard at what you actually own. Are you holding companies in your portfolio that make money today, or do they need an impressive story to keep their stock price afloat? If the AI narrative shifts, can the businesses in your portfolio adapt?