The battered S&P 500 index is not pricing in a recession, according to DataTrek Research.
“At 4,000, the recession odds imbedded in S&P are close to zero,” said DataTrek co-founder Nicholas Colas in a note emailed Tuesday. “By our math, 50:50 odds of a recession equate to an S&P at 3,525.”
The S&P 500
a stock benchmark measuring the performance of large U.S. companies, has dropped more than 16% this year after closing Monday at 3,991.24. That marked its lowest closing value since March 31, 2021, which was the last time the index ended below 4,000, according to Dow Jones Market Data.
The U.S. stock market has tumbled this year amid fears over high inflation, rising interest rates, the Russia-Ukraine war, China’s COVID-19 lockdowns and a slowing economy. “If we do actually get a typical economic downturn, then the S&P should trade for right around 3,000,” according to Colas.
“Recent volatility simply says investors think the window of opportunity to get back on the right track is closing,” he said. But the window “is not shut yet,” Colas wrote, “otherwise, the S&P would be at 3,500 (50:50 recession odds) or even lower.”
DataTrek looked at the current “earnings power” of the S&P 500, pegging it at $218 per share. That would be “peak earnings” if the U.S. is heading into a recession, according to the note.
“Recessions hit earnings, of course, but by varying amounts,” said Colas.
While “standard recessions” cause an average 26% drop in peak-to-trough earnings, the S&P 500 saw a peak-to-trough earnings decline of 57% during the “Great Recession” ending in 2009, according to the note.
So a “garden variety economic contraction” would put S&P 500 earnings at $161 a share, or a 26% drop from the current $218 a share, DataTrek calculated. A 50% chance of recession translates into $190 per share, the note shows.
DataTrek looked at the troughs of price-to-earnings ratios around the past three recessions, estimating an average multiple of 18.5 based on troughs in 2020, 2009 and 2002. Colas tossed out the recession-related trough seen in 1990 as “market valuations were generally much lower than now due to higher interest rates.”
With recession odds 100% baked in at $161 a share, the S&P 500 would be trading at a price-to-earnings ratio of 24.8, according to DataTrek. That compares with a multiple of 18.3 based on $218 per share and no chance of recession, and a ratio of 21.1 at $190 a share and a 50 percent chance of an economic contraction.
So the S&P 500 “needs to go to 3,525,” based on the average 18.5 multiple seen in prior economic downturn troughs, “just to discount 50:50 odds of a recession,” the note shows. The S&P 500 would trade around 3,000 in a typical recession based on that same multiple and earnings at $161 a share.
“This is not a prediction, but rather a crude but historically defensible approach to assessing where the S&P ‘should’ trade if recession fears continue to grow,” according to DataTrek.
The U.S. stock market ended mixed Tuesday, with the Dow Jones Industrial Average
falling 0.3% to book a fourth straight day of losses. The S&P 500 rose about 0.2% to close at 4,001.05 while the technology-laden Nasdaq Composite
gained 1%. All three major benchmarks had suffered a sharp selloff on Monday.
Market volatility remained elevated Tuesday. The CBOE Volatility index
traded around 33, after closing Monday at about 34.8, according to FactSet data. That’s above the 200-day moving average of about 22 and 50-day moving average of 26.5.
Investors are now digesting fresh data on inflation, which has been a top economic concern after running around a 40-year high.
The consumer-price index rose 0.3% in April for an annual rate of 8.3%, slowing from March when it increased to a 12-month pace of 8.5%, according to a report Wednesday from the U.S. Bureau of Labor Statistics. Core CPI, which excludes food and energy, rose 0.6% in April, exceeding expectations. But the annual pace of core CPI also slowed, falling to 6.2%, from 6.5% in the 12 months through March.
The S&P 500, Dow and Nasdaq were each trading up Wednesday morning, according to FactSet data, at last check.