Is It 1996 Or 1999? Can Stock Market Double By 2030 Like It Did When Dot-Com Vauations Hit 60-year Highs

It doesn’t take brains to spot a bubble, but it’s almost impossible to know whether you’re in 1996 or 1999, and that’s where most money is made.

On December 5, 1996, Fed chair Alan Greenspan famously mused about “irrational exuberance” in stock prices. Many thought that warning marked the top, and stocks fell the next day.

But 1996 turned out to be just the early innings of the dot-com mania, and those who bailed in late 1996 missed one of the most impressive five-year S&P 500 stretches on record. By mid-2000, the S&P 500 had more than doubled. Even at the dot-com trough, anyone who stayed in was still in the green.

By almost every yardstick, the market looks expensive today, including two of the most-watched indicators.

The forward P/E — Wall Street’s go-to measure that shows what you’re paying today for a dollar of expected earnings over the next 12 months — is at its highest since the dot-com bust, excluding a few brief spikes during Covid.

Then there’s the Shiller P/E, Buffett’s favorite valuation measure. It compares prices to real 10-year average earnings, offering a longer-term perspective that isn’t distorted by rosy earnings forecasts.

Just like forward P/E, the Shiller P/E today hovers at the highest level since the dot-com era.

Anecdotal data doesn’t help either, especially around AI. Some engineers are landing $100 million packages, and there’s even a rumor of a $1 billion hiring bonus, the biggest on record, matching the GDP of some countries.

And then there’s last week’s Oracle episode, one of the most eye-popping single-day tech surges of the decade. After it touted a multifold jump in AI infrastructure backlog, the near-trillion-dollar company jumped over 30% in a day.

The catch is The Wall Street Journal reported last Wednesday that OpenAI has pledged to pay Oracle $300 billion over roughly five years. If accurate, that would imply OpenAI accounts for nearly all of Oracle’s current long-term backlog.

Layer on the politically flavored, pick-a-number pledges for data-center build-outs from Big Tech and trade partners, including the UAE’s touted $1.4 trillion plan for U.S. “AI infrastructure,” and you have a house of cards built more on promises than earnings.

In fact, by D.A. Davidson analyst Gil Luria’s estimate, top AI software companies — including OpenAI, Salesforce, and Adobe — have only generated tens of billions in revenue from AI.

Much like the fiber build-out of the internet era, today’s AI investments are largely FOMO-driven and front-loaded on the assumption that AI’s exponential adoption curve will hold.

No line captures this “blank check” spending better than Google CEO Sundar Pichai’s remark during Alphabet’s earnings call last August.“When you go through a curve like this, the risk of underinvesting is dramatically greater than the risk of overinvesting.”

Some of that overinvestment will eventually have to bleed out, but when — and where — only time will tell. As Wells Fargo’s Ohsung Kwon put it, “Music stops when AI capex stops. Enjoy the party.”

But AI isn’t the only version of “this time is different” today. As Howard Marks wrote in a recent memo, American exceptionalism is also at play. There’s “The belief that, for most investors, there really is no alternative to the U.S. markets.”

Then there’s the wealth effect and the “buy the dip” mentality of a new generation of investors.

“The last sustained market correction ended in early 2009, which means it’s been more than 16 years since risk-taking was seriously punished and buying the dips didn’t work. In practice, no one under about 35 — pros or amateurs — has ever lived through a prolonged bear market,” he said.

So the question isn’t whether it’s a bubble. The real, trillion-dollar question is whether it’s 1996 or 1999. Remember, the Shiller P/E sat at a 60-year high from 1995 to 2000, one of the best stretches in stock market history.