CryptocurrencyInvestingBanksReal Estate

Stocks are now overpriced by 22%, according to this rule

April 21, 2020, 2:48 PM UTC

Investors are brimming with fresh optimism that’s brought one of the greatest rallies ever. But if they’re right, then future earnings will need to be to be terrific. Don’t fall for the wishful thinking. The coronavirus crisis will hammer profits that were already in a bubble, and due for a fall.

Since hitting a low of 2237 on March 23, the S&P 500 surged 638 points or 28.5% by the close on Friday, April 17, regaining well over half of its drop from the record high of mid-February. That extraordinary jump means Wall Street believes that future earnings are looking a lot better now than they did three weeks ago.

But current prices point to a level of profits that defies reality. Stock prices reflect the stream of profits companies are expected to generate for many years come. Depressed results in 2020 and 2021 caused by the COVID-19 outbreak are far less important than where earnings settle when they’ve recovered to what we’ll call “normal” levels, meaning a benchmark where they start expanding again with the newly-growing economy.

So based on the S&P’s close at 2875 on April 17, where do investors put the normal earnings for America’s big cap stocks? Over the past 60 years, the market’s median price-to-earnings multiple stands at 17.4. But the PE has been a lot higher over the past three decades, averaging 21.3, a legacy of record-low interest rates that may or may not continue, and perhaps, excessive enthusiasm. We’ll start with the average of the two, 19.4, and round it up to 20-times, estimating that a magic wallet sprouting $100 a year in earnings should sell for $2000.

Deploying the 20 PE is Part 1 of what we’ll call the “Tully-20 Test.” If the “normal” profits you get by dividing the S&P by 20 look unreasonably high, then stock prices are inflated.

At our 20 PE, the S&P’s sturdy earnings-per-share, when the U.S. gets chugging again, should be $143.75 (2875 divided by our PE of 20). But wait, that’s higher than the $139.47, based on four quarters of trailing GAAP earnings, where the S&P finished 2019! That mark was an all-time record, a gain of 48% from the close of 2016. It also capped something of a golden interlude for profits. Operating margins averaged 11.1% in 2019, 1.9 points higher than the margin in 2016, and sales over the last three years jumped by almost a quarter, at an annual pace of 7.3%.

Those trade winds have since turned into a tornado. It’s impossible to predict how hard the pandemic will hit profits. But we know that projections are a moving target that’s rapidly getting worse. On April 9, analyst polled by Refinitiv posited a drop of 8.5% for the year. A week later, their estimate had almost doubled to 15.5%. Look for forecasts to keep dropping.

Once again, what matters most is where earnings land when the turmoil ends. The bull run means that investors believe that EPS will surge past the 2019 peak to $144, and increase from there. But earnings were in something of a bubble at the end of 2019 due to the unusual confluence of margins and sales growth that far exceeded their historic levels. We don’t know for sure if big caps earnings will quickly rebound to $144, but we do know where they need to go in the three years after the good times return to reward investors. Assuming companies continue paying 40% of their profits in dividends, and 60% in buybacks and cash reinvested in plants and products, earnings should wax at 5% a year––a pretty good number in an economy projected to grow at a long term “real” rate of 2%-2.5%, or 4%-4.5% including inflation.

Investors’ view that profits will replant the flag at greater heights than the 2019, golden age record, is probably a fantasy. So the S&P flunks Part 1 of the Tully-20 Test. Let’s move on to Part 2. If $144 is an inflated starting point for a comeback, what’s a reasonable base? And where does that put a reasonable value for the S&P?

An excellent guide is the Cyclically Adusted Price Earnings ratio developed by economist Robert Shiller. The CAPE uses a 10-year average of inflation-adjusted profits to smooth their erratic course, including spikes like the one before the virus struck. The CAPE’s most recent reading for “adjusted” profits is $107.

Of course, that figure will decline somewhat with the blow from the pandemic. But since Shiller’s methodology spreads earnings over a long period, the downward adjustment is most likely to be just a few dollars. I’ll also add 10% to the Shiller number because using a price level ten years old to adjust earnings over the next decade makes the current number look artificially low. After fiddling with the CAPE’s figure, I calculate a base for earnings per share, the footing where they’re most likely to settle when the U.S. emerges from the trough, at around $118.

The Tully-20 Rule reaches its judgment by applying the PE of 20 to that benchmark. That formula yields a fair value for the S&P of 2360 (20 times $118). The S&P actually dipped 5% below that number on March 23. But as of April 17, the index floats 22% above where the Tully-20 rule put fair value.

Conclusion: The latest party was a bash. Get ready for the hangover.

More must-read finance coverage from Fortune:

—5 veteran investors on how to approach the coronavirus stock market
—These countries’ stock markets have been hit the hardest by the coronavirus
—China’s next coronavirus crisis: What happens after a country closes its economy
—This time, the banks were ready: How the Big Four prepared to survive the coronavirus
How the American economy can recover from the coronavirus pandemic
—Listen to Leadership Next, a Fortune podcast examining the evolving role of CEO
—VIDEO: 401(k) withdrawal penalties waived for anyone hurt by COVID-19

Subscribe to Fortune’s Bull Sheet for no-nonsense finance news and analysis daily.