We’ve reached the “P/E ratios are irrelevant” stage. And not a moment too soon, considering how shrill the valuation warnings have been lately.
“On a statistical basis, the 20.4x S&P 500 fwd PE puts the market over 1.5x standard deviations above the average of 15.4x and three-quarters of the way to the Tech Bubble high of 25x”, BofA writes, in a quant strategy piece dated Thursday.
That’s probably a generous assessment. Depending on your outlook for the assumed earnings rebound following the collapse accompanying the virus lockdowns, that multiple could be far (far) higher. And that’s to say nothing of other valuation metrics that some critics cite in branding this market among the most unattractive in history.
So, are P/E ratios irrelevant? It feels like a dangerous thing to say. And BofA admits as much, even as they see some validity in claims that because we’ve “veered off the grid” when it comes to historical precedent, comparing today’s multiple to yesteryear could be misleading.
“Despite it being disconcerting that the irrelevance of P/Es was a refrain of the early 2000s, right before the Tech Bubble burst, a lack of comparable periods is a valid point”, the bank says, before suggesting that “a fair value approach may better incorporate today’s ultra-low rate environment”.
But even that’s fraught with uncertainty.
“During the post-GFC period, the S&P 500 ERP has ranged from 394 – 695bp, which would yield a range from 1882 to 2958 using normalized 2021 earnings of $155”, BofA says, adding that the “median ERP since 1986 excluding the Tech bubble is about 350bp, which would yield 3269 for the S&P 500”. “Fair” appears to be roughly 10% lower than current levels.
(BofA)
You can draw your own conclusions there, but I suppose what I would note is that when it comes to 2021 earnings, nobody really knows what to expect.
In addition to the wild cards around the virus, betting markets now put the odds of Democrats flipping the Senate at roughly 50%. If you subtract the projected earnings hit from a hypothetical rollback of the Trump tax cuts from Goldman’s downside scenario for 2021 EPS (i.e., in order to generate a kind of “absolute worst-case“), you end up with a forward multiple of nearly 32, as of last week’s closing levels.
The good news is, that’s come down a bit during this week’s selloff, but you get the point: If P/Es are irrelevant, it’s good news, because on anything other than a rosy projection for a profits rebound, valuations are stretched.
Stretched as equities may be, though, BofA reminds you that it’s all relative.
“Stocks are trading above average on every valuation metric we track except bonds”, the bank said Thursday. “Relative to bonds, stocks have not been this attractive since the 1950s”.
Don’t tell Stanley. Or David. Or Jeff. Or that little voice in your head that’s telling you not to buy equities at the onset of a literal depression.
Watch what they do not what they say…..repeated daily at a place I once worked……
Valuation is less relevant at market bottoms because E is depressed by recession. 2. FV PE is a function of yields. Lower yields justifies higher PE. 21 is fair given current UST yields. Next year SPX earnings peaked at 195. Lets slice it by 35% which gives us 126.75, a 21 multiple is 2661. Thus SPX is a little rich but not outrageous IMHO. It is worth noting that the trough in earnings might come sooner, so the 2021 number might be too negative. If 156 is closer to the truth, then 21 times is 3276. Just to emphasize, valuation is not a strong driver early in the cycle. It is not worth obsessing over.