The Oracle of Omaha has long held to a simple maxim when it comes to whether the stock market is undervalued, fairly valued, or overvalued. His thesis: The total value of U.S. stocks, over the long term, can’t outpace the growth of businesses as reflected in the GDP. So when the ratio of S&P 500 to national income diverts hugely from the norm, it is bound to swing the opposite way and “revert to the mean.” During the Dot Com bubble, in a 2001 story in Fortune that he penned, Buffett highlighted a chart in the text displaying that at the craze’s peak in March 2000, that number, now known as the “Buffett Indicator,” reached a heady 200%.

“The message of the chart,” he wrote, “is that if the relationship [between the total value of equities and GDP] drops to 70% or 80%, buying stocks is likely to work out very well for you. If it approaches 200% as it did in 1999 and 2000, you are playing with fire.” Indeed, the S&P had already fallen over 20% by the time the Buffett story appeared, and by mid-2022 retreated by almost one half from its peak, taking the Buffett Indicator below 80%. As the Buffett formula predicted, the tech rampage’s aftermath proved a great moment to buy.

Which brings us to the present. Since the stock market decline prompted by the surprise start of the Iran war, the S&P 500 has rebounded to a near all-time record of 7165. Here’s the shocker: The Buffett Indicator now stands at 227%, a figure that’s around one-sixth higher than what he identified as the prepare-for-a-roasting zone. A reading this elevated comes with two problems. First, corporate profits have been waxing much faster than GDP. The bulls claim that trend justifies today’s valuations, and that EPS can keep rolling in double digits while national income trudges at a nominal 5% or so. The argument’s dubious: Profits are now 12% of GDP versus an historic average of 7% to 8%. In our highly competitive economy, fat margins attract competitors seeking a share of the action, so they push down prices and expand volumes to steal market share from the profit-rich incumbents. Extraordinary earnings growth generally doesn’t stay extraordinary. As the late Nobel-winning economist Milton Friedman told this writer, “Corporate earnings as a share of national income cannot rise beyond their historic share of GDP for long periods.”

Second, stocks have also gotten far more expensive relative to their profits. The S&P 500’s price/earnings ratio based on forecast Q1 GAAP net earnings exceeds 28. That’s two-thirds higher than the 100-year average of around 17. Best bet: Both profits and P/Es trend back toward normal, taking the Buffett Indicator, and the S&P, downwards with them.

How bad could the drop be, based on the past instances of an astronomical Buffett Indicator? Once again, the decline from the dotcom driven 200% mark that prompted the Buffett piece was about half. In November 2021, the Indicator reached just over that fearsome benchmark, then tumbled 19%.

In the Fortune article, Buffett warned that if investors expected shares to roar higher when his Indicator was hovering at those historic highs, “the line would have to go straight off of the chart,” meaning the optimists were banking on a suspension in economic gravity. Right now, the bulls are in charge, and they’re predicting the Buffett Indicator that’s already hit uncharted territory will push further into the “playing with fire” realm. Buffett’s thesis does not predict when the market swings back into balance, just that it will eventually, and when it does all investors will feel the pain.