Friedman, the late Nobel Prize–winning monetarist who never left academia, and Buffett, who spent his entire career in the arena building companies and backing future winners en route to becoming the most successful investor of all time, have more in common than you might think. Friedman was an ardent champion of capitalism and so is Buffett; both loved math and considered becoming actuaries; and each has rejected the theory that the stock market is “efficient,” insisting that buying frenzies and panic selloffs often send prices soaring far above, or cascading well below, the enterprises’ fundamental values.

Shortly before his passing in 2006, Friedman told this writer, “Over long periods, the stock market’s value does track companies’ earnings. But in the short term, the market is far from efficient.” Buffett couldn’t agree more. He’s famous for staying calm and scooping up bargains in tough times when the Wall Street crowd is fleeing in a stampede.

Friedman believed from long observation, and Buffett reaped billions betting, that equities are often excessively pricey or overly cheap. What’s more, both cited favorite metrics that measure when markets are drastically under- or overvalued. For Friedman, the ruling measure was corporate earnings as a share of GDP. When that ratio went out of whack because profits spiked, he told me, the jump was unsustainable. “Profits can’t keep absorbing more and more of national income,” declared Friedman. For that to happen, they’d need to outgrow wages, investment, and other GDP components for long periods, a dynamic that Friedman told me couldn’t last.

If profits rose to devour an unusually high portion of the nation’s output, he said, the boom in business that pushed them up would eventually push them back down. Why? Because the surge in sales that initially raised profits would force companies to hire additional workers, and the competition for those workers would raise payrolls and depress margins.

By implication, Friedman was saying that if stock prices soar because earnings are racing far faster than the overall economy, watch out! The trend is bound to reverse, pounding profits and, in all probability, dragging down equity prices.

For Buffett, the yardstick is the value of all U.S. equities compared to GDP. As he told Fortune in 2001, “It’s probably the best single measure of where valuations stand at any single moment.” The market cap to national income metric is often called “the Buffett Indicator.”

The Friedman and Buffett measures are closely related, and both are based on bedrock economics. Essentially, the two tests are saying that the very workings of an open, competitive economy make it extremely difficult for companies to generate outsize profits for long periods—with exceptions for great brands and highly regulated, price-protected sectors such as pharmaceuticals. In general, when a company keeps raising prices and garnering fat margins on semiconductors or sports equipment, those rich profits are a magnet for competitors to jump in. Rivals that can make and sell the same stuff better and cheaper slash prices to win customers, and margins eventually retreat to their historic benchmarks.

Friedman is talking specifically about how economic gravity governs profits. The Buffett metric also looks at earnings but encompasses a second component, valuation or price/earnings multiples. Either one of them operating separately, or the pair in tandem, can swell the S&P’s value to excessive levels of GDP. The first driver is a jump in earnings to historically lofty, unsustainable levels. The second is the lift from crazy ebullience that causes investors to pay bubble-like prices for each dollar of profits, sending P/Es soaring. Or the S&P 500’s market cap can get dangerously inflated by a combination of the two. That was the case in early 2020 before the pandemic struck. Since prices are nearly back to those levels, it’s natural to question how they can advance from here without a huge jump in profits, an outsize P/E that flashes red, or the two working together.

Let’s examine what the Buffett and Friedman measures are telling us about the remarkable exuberance that’s defying the steep economic downturn caused by the COVID-19 pandemic. At the close on Thursday, July 22, the S&P 500 stood at 3236, a shade above its near record of 3131 at the end of 2019. The index now sits just 4.4% below its all-time high of 3386 on Feb. 19. To buy an S&P index fund or a diversified portfolio of large-caps, investors must believe they’ll be a lot higher in 12 or 24 months. That’s especially true since the markets been careening through lurching swings, so that funds and individuals want to be well rewarded for what promises to remain a rough ride.

Let’s start by deploying the Buffett Indicator. It incorporates both GDP and aggregate value of equities, for which we’ll use the S&P 500, accounting for most of that total, as a proxy. The Congressional Budget Office projects that national income will tank from an annualized $21.729 trillion in Q4 2019, to $19.246 trillion in Q2 and $20.361 in Q3, drops of 11.4% and 6.3% respectively. In the CBO’s forecast, output doesn’t return to year-end 2019 levels until the end of Q4 2021. (Its figure for that quarter is virtually flat, with Q4 2019 at $21.75 trillion.) That’s around 18 months from now. It’s conservative to assume that investors would want a 10% gain over that year-and-a-half, or 6.5% on an annual basis, to brave the rapids.

We’ll get to assessing that ratio in a minute. But first let’s invoke the spirit of Friedman and look at earnings. What level of profits are needed to boost the market to 3560 in 18 months, and what are the chances they can get that big? To explore that question, I’ll introduce a new participant, the author, who, though awed to join this august company, adds his own measure. It’s called the “Tully 20 Rule.” That yardstick posits that 20 is a reasonable price/earnings multiple to project for the S&P 500 once the economy gets back on track. It’s lower than the level of the past 20 years, but much higher than the 60-year average of under 17.

At a 20 P/E, the S&P companies would need to earn $178 a share to achieve the 3560 mark that would deliver a 10% return over the next 18 months. That’s 28% higher than the record of $139.47 achieved at the close of 2019.

Keep in mind that S&P profits have collapsed so fast that the Wall Street analysts polled by S&P forecast that for all of 2020, earnings per share will hit $93, a drop of 33% from their record of $139.47 the end of 2019. The analysts further expect that by the close of 2021, when the CBO predicts that national income returns to late 2019 levels, EPS will reach $147, just above the roughly $140 mark of two years before.

To get to 3560 in 18 months, a rise the bulls view as a gimme, the S&P would need to generate 21% higher EPS than even the analysts, who are notoriously optimistic, are forecasting. Here’s the killer and the figure that would dismay Friedman. Total S&P earnings as a share of GDP would have to hit 6.7%. At the end of 2019, EPS as a share of GDP was already high by historical standards at 5.3%. In the run-up to the Great Financial Crisis in July 2007, that ratio was 5.5%; in the tech bubble of mid-2000, the number was 4.3%. We’ll dismiss the “earnings will ride to the rescue” scenario right now. The notion that profits will be almost 30% bigger heading into 2022 than in the great times of late 2019, and also far surpass the shares in previous frenzies, is a fantasy. It won’t happen.

It’s clear that the S&P won’t come close to hitting 3560 at our projected P/E of 20. Earnings can’t do the job. In theory the S&P could deliver a 10% gain from here—but only if the future P/E is far higher than our assumption of 20. To be sure, that would require an epic performance, considering how far EPS has fallen in this deep recession. If earnings rebound to nearly $140, their all-time high posted in Q4 2019, the S&P would ring the bell if the P/E ratio went to 25.4 (3560 divided by $140). That’s 27% above what we consider a reasonable projection of 20. The S&P has never posted a multiple that high for a single quarter in the past decade, and the average over that span is 16% lower at 21, a figure that’s already well above the historic norm.

Of course, it’s theoretically possible that the EPS could jump to $178, or that the P/E could climb to 25-plus, or that some combination of high EPS and an elevated P/E could lift the S&P by 10% over the next 18 months. The clincher is the Buffett Indicator. In any of those cases, the destination is the same, an S&P at 3560 that brings its value versus national income to 135%. The Buffett Indicator rates the probability that will happen as extremely low.

Why? Because a S&P to GDP of 135% is wildly out of line with almost any past precedent. It would represent a new normal that’s too abnormal to happen. In the mid-2007 boom, the S&P’s value stood at 97% of GDP. The highest figure notched in recent decades was 123%, registered twice, once at summit preceding the tech meltdown in June of 2000, and again at the end of 2019. The average since 1975 is around 90%.

The scenario that many on Wall Street are selling, that the S&P has plenty of room to rise from here, runs headlong into both the Buffett Indicator and the Friedman earnings-to-GDP metric. It gets an F on both tests. A better bet is that profits return to something like their late 2019 level of $139 in 18 months and garner a 20 P/E. That would put the index at 2780, 14% below where it is today, and 22% under the 3560 S&P level that would provide a 10% gain from here. Forget the “this time it’s different” story about “the Fed put,” unstoppable momentum in big tech, and hoards of cash on the sidelines. Believe the measures preferred by the two sages, one still with us, one for the ages, that are flashing the same signal that a reckoning is coming.