Speaking to Wall Street in late April, Coca-Cola CEO James Quincey followed the example of a growing number of his peers and did something previously unusual for a Fortune 500 leader: He threw up his hands. Forecasting the beverage giant’s results in the months ahead, the chief executive declared, was well-nigh impossible. “We recognize that these are truly unprecedented times,” said Quincey, on the company’s first-quarter earnings call. “Given the great uncertainty of the current environment, we feel it’s prudent to hold off providing fiscal-year 2020 guidance.”

This spring, more than 100 companies in the S&P 500 that regularly provide earnings guidance—including IBM, Intel, and Kimberly-Clark—have said they won’t attempt to forecast results for 2020. Putting a number on just how badly the coronavirus crisis will batter their businesses is too hard to calculate with any degree of accuracy.

Given that lack of visibility at publicly traded companies, you can be forgiven for getting flummoxed by the confusing outlook for corporate profits. The stock market certainly has been flashing conflicting signals. After hitting an all-time high in mid-February, the S&P 500 collapsed 34%, with investors evidently fearing the pandemic would cause a long, steep fall in earnings. But then a massive April rally—the best month for stocks since 1987—drove the market’s valuation back to a historically high level. Current prices imply that, in 18 months or so, profits will be better than ever.

Here’s the overriding question vexing both companies and investors: Where will profits stand when the economy returns to generating the same output of goods and services as before the outbreak?

Estimates of when GDP will fully recover vary widely. A reasonable forecast is Bank of America’s view that GDP will regain 2019 levels by the end of 2021. Keep in mind that the BofA forecast is optimistic: Economic production shrank at a 4.8% annual rate in the first quarter of 2020—the biggest one-quarter hit since the Great Recession. And Wall Street economists foresee a steep double-digit drop in the second quarter.

Even if GDP follows that best-case trajectory, the likelihood is that earnings will be significantly lower when the economy finally recovers than at their record peak last year. The reason is twofold. First, key industries such as airlines, energy, and commercial real estate will suffer such severe structural damage that they’ll take far longer to return to their old profitability. Second, big companies won’t be nearly as profitable as in recent years, when a confluence of low labor costs and a buoyant consumer swelled margins to unsustainable levels.

A sharp reversal

Earnings were rocketing up before the pandemic. Last year, total operating profits for the S&P 500 hit $1.3 trillion, up 44% from 2016. Over the same span, GAAP earnings for the Fortune 500 rose from $890 billion to $1.2 trillion—an all-time high. The two benchmarks have a lot of overlap: In any given year, about 330 companies are on both the S&P 500 and the Fortune 500, the latter of which includes non–publicly traded companies. By digging into what analysts predict for the S&P, we can extrapolate the future direction of the Fortune 500 as well.

A careful look at the profit mix of the S&P 500’s 11 industry sectors reveals that much of the gains in recent years came from a handful of industries. From 2016 through 2019, the share of profits earned by the 66 “financials” in the S&P 500—led by JPMorgan and Bank of America—jumped from 18% to 25%. And the portion from communication services, a category encompassing 26 companies including Facebook, Alphabet, and Comcast, more than tripled from 3% to 10%. Meanwhile, industrials, consumer staples, and consumer discretionary companies—sectors that are home mostly to old-economy stalwarts—dropped from a combined one-third of total earnings to 24%.

For the first two months of 2020, the earnings express kept rolling. “The halt to the trade war in late 2019 gave profits new impetus,” says Amanda Agati, chief investment strategist at PNC Financial Services Group. “We were expecting 10% gains in earnings per share this year.” Then COVID-19 swept around the globe.

The coronavirus shutdown in the U.S. precipitated the deepest drop since the Great Recession in polls of earnings estimates conducted by S&P and FactSet. According to ¬FactSet’s report from early May, analysts are forecasting a 13.6% earnings-per-share (EPS) decline in the first quarter, compared with a year earlier, and a 40.6% stumble in the second. The consensus is that a second-half rebound will leave profits 19.7% lower at year-end than at the close of 2019. The Street predicts that earnings for financials will plunge 38%, from $252 billion to $158 billion. And industrials are forecast to shrink from $125 billion to $73 billion—a 42% drop. Energy has been a minor contributor to overall earnings in recent years, adding just $52 billion, or 3.8%, of the S&P total in 2019. And the tumble in oil prices this year—from $60 per barrel in January to $12 in late April—will obliterate that figure. Analysts expect a $4.9 billion loss in energy this year.

It’s a near certainty that even those dire forecasts are too rosy. Analysts are always overly optimistic, and the parade of bad news is so relentless that estimates are falling at record speed. Since the start of April, FactSet’s forecast for the first quarter has sunk sixfold—from negative 3.3% to that negative 19.7%. Consider, for example, Delta’s warning that its revenues will crater 90% in the second quarter. It take a lot of upside surprises to balance out those kinds of collapses.

A more realistic take comes from Savita Subramanian, managing director at Bank of America Merrill Lynch, who predicts a 29% drop in S&P earnings per share for 2020. As the economy rebounds, she reckons, profits will lag—in large part because of the changes in how companies and consumers spend when the crisis ends. With the unemployment rate surging to highs not seen since the Great Depression, families will be more cautious about spending on everything from restaurants to cars. Executives, having seen how productive their employees can be while working from home, will rethink the need for maintaining big, expensive offices. “That’s what our analysts are hearing from companies they cover, and what our private bankers are told by entrepreneurs running small businesses,” says Subramanian. The trend is likely to hammer rents and profits in commercial real estate.

Airlines will be slow to regain their pre-outbreak altitude, especially now that the entire business world has been conditioned to meet via Zoom. “Leisure travel will probably go back to normal,” says Subramanian. “But business travel will fall. Executives will reconsider the need to go to China four times, or Europe twice a year.”

Naturally, sectors catering to working and shopping from home will be big beneficiaries, and their gains will partly offset the damage. It’s happening right now. Microsoft’s Teams video collaboration service, for instance, has jumped from 44 million to 75 million daily active users since mid-March. Microsoft has also benefited from rising sales of its cloud services and networking technology for stay-at-home workers. The demand helped boost the software giant’s operating income 25% in the first quarter.

But a boost from digital winners won’t be sufficient to cover the broader-¬based weakness in profits—at least in the short term. The problem is margins. Mark Zandi, chief economist at Moody’s Analytics, predicts much lower profitability from the overseas sales that contributed over 40% of the S&P 500’s total revenue last year. China’s growth has slowed, he notes, and Europe and emerging markets are likely to rebound far more slowly than the U.S. A bigger pool of Americans looking for work will slow growth in labor costs, but not enough to counter the lower prices that businesses from airlines to restaurants to hotels will need to charge to lure back customers.

The bottom line

So where will earnings settle if GDP indeed returns to last year’s heights by the end of 2021? In the fourth quarter of 2019, operating margins clocked in at 11.4%. That’s almost three points higher than the median over the past decade. Let’s keep it simple, and project that profitability returns to slightly above average, at 9% of sales. In that scenario, the S&P 500 would earn 20% less than in 2019, when EPS hit $163. My estimate is profits will land at $130 by the end of 2021.

That outcome would be a big disappointment to Wall Street. The analysts surveyed by FactSet in May are projecting EPS for the S&P 500 of $168 a share in 2021, and the S&P poll says $165. But math and logic suggest both predictions are farfetched at best. Profits existed in a magical age until just a few months ago. It’s unlikely to return anytime soon. 

A version of this article appears in the June/July 2020 issue of Fortune with the headline “The trillion-dollar question: How far will profits fall?”