I’ve talked quite a bit about valuations lately. Specifically: How far and how fast they’ve come in.
2022’s stock selloff is a multiple compression story. So far. This week suggested it might soon become an earnings recession story, though. Abysmal reports from a who’s who of retailers augured poorly for the US consumer.
As JonesTrading’s Mike O’Rourke put it, “companies that were once thought infallible are stumbling daily.” That, he warned, means “investors can’t have faith in current 2022 earnings estimates.”
The figure (below, from Deutsche Bank) gives you a sense of the valuation shakeout.
It could easily have further to go. We’re just now back in the middle of the 2015-2019 range.
If the selloff does roll on, and profit growth does decelerate, US shares could have considerable downside, as detailed earlier this week by Deutsche Bank’s Binky Chadha and Parag Thatte.
One obvious question going forward is this: What multiple is appropriate for stocks given current macro realities? If you ask BofA, the answer isn’t the 21st century multiple, around 19x.
“EPS and GDP forecasts are heading lower,” the bank’s Michael Hartnett said. “The debate now shifts to the correct multiple.”
As the figure (above) shows, Hartnett reckons a more “traditional” approach might be warranted. “We say the 20th century multiple of 14x is more appropriate for stagflation,” he wrote, in the latest installment of his popular weekly Flow Show series.
That’s a trailing multiple. If you apply it to 2021 EPS, you get 2,926 for the S&P. Bottom-up consensus for 2022 currently expects 10% EPS growth, to $230. A “20th century” valuation gives you SPX 3,220.
Since 1948, the median peak-to-trough drop for S&P earnings around recessions is 13%. Assuming i) the recession starts now and ii) a drop in earnings consistent with that seen around historical downturns, applying Hartnett’s “20th century” multiple gives you SPX 2,550, for a peak-to-trough drawdown of around 46%.
Incidentally, that’d be right around where Jeremy Grantham suggested the S&P would trade if the worst case scenario he outlined earlier this year were to unfold.
It’d also be consistent with a casual observation from Zoltan Pozsar. “At 2,500, we’d lose not only all of the post-pandemic gains, but would eat into some of the pre-pandemic gains too,” he wrote earlier this month. “And if something indeed happened to the supply of labor post-pandemic (and some of that’s wealth related), then to cool price pressures, maybe a pre-pandemic wealth level is appropriate indeed.”
The S&P fell into a bear market Friday, before trimming losses into the close. The index notched a seventh consecutive weekly decline, the longest such stretch since 2001. The Dow’s weekly string of losses, at eight, was the longest since 1923.
Market decline: feature, not bug . . .
“There has been a lot of repricing in markets,” Bullard said. “Part of that is due to the Fed, but part of it might also be what were the prices before the downturn occurred. You would expect with the Fed raising rates that all these assets, trillions of dollars worldwide, would have to be repriced.”
https://www.bloomberg.com/news/articles/2022-05-20/fed-s-bullard-says-front-loading-could-lead-to-rate-cuts-by-2023
Whatever the bottom is, my guess is we’ll get there faster than prior cycles might suggest. This is the first tightening cycle I have personally experienced in which, one might suspect, asset repricing is not only not a concern to the Fed, but one of the Fed’s desired outcomes.
“Don’t fight the Fed” again, but in reverse this time.
Nice summary.
And hats off to H for telegraphing- I read his lines and also in between his lines, but I did not like what he was saying.
I missed selling close to the top- but I will definitely (still) be in at the bottom.
In my psyche, I have always had a harder time selling/letting go. “Getting in” is so much easier (for me). As in life, will work on shifting gears. GLTA.
In the Graham and Dodd security analysis text I taught from in the 70s, the average P/E was reckoned about 13-15x and the authors were beginning to recognize that growth stocks might need higher multiples. They related these multiples to AAA bond rates and growth. The BofA chart shows the 20th century ave at about 14x, right in G&D’s range. That 3200 figure somehow feels right. Hang on. At that level a lot of folks wanting to retire early will hit a big speed bump if any of this is right.
“Since 1948, the median peak-to-trough drop for S&P earnings around recessions is 13%”. It appears this 13% drop is being applied to an S&P 500 earnings number of $209 to get to a recession scenario trough of ~$182. Note that this recession scenario trough level is ~16% HIGHER than the annual 2019/pre-pandemic RECORD earnings of ~$157. I’d suggest higher inflation/costs, higher interest rates AND a recession would take S&P 500 earnings much lower than ~$182 (and likely lower than the pre-pandemic record earnings of ~$157).