A thought experiment I like to do is mull what S&P 500 earnings would have been today if we hadn’t had a decade or more of cheap credit. Earnings have grown, on average, 9% a year for years in the U.S., but that has been goosed to an unknown degree by credit.
What would the world look like otherwise? Well, one way to back into an (admittedly simplistic) answer is pick a date in the past, and then project forward from there at some more reasonable economy-wide earnings growth rate.
Fair enough. But what would that be? Well, corporate earnings can’t grow forever faster than GDP or their share of GDP rapidly becomes crazy-large, like all of GDP in a few decades from the current starting point of 7% of GDP. The upshot: The trend growth in U.S. GDP is a reasonable floor, so let’s call it 3.4%. With that in mind, somewhere between 3.4% and 9% earnings growth is a reasonable place to start in projecting forward to what an ex-leverage S&P 500 earnings number might look like today.
Using the preceding, and having picked 1987 as the starting date for the exercise (which isn’t completely ad hoc), here is what I end up with under various growth and earnings multiple scenarios S&P 500. The upper table is 1987 S&P 500 earnings projected forward to 2008-2010 under various conditions; the lower table is an array of resulting S&P 500 targets.
In short, depending on when you start, what ex-leverage growth rate you pick, and what trough multiple you apply, you can get to S&P 500 targets in 2009 from 268 (!) to 1135 (!!). To narrow things down, say you pick something mid-range for ex-leverage growth, and apply a 12x earnings multiple on the S&P, which seems fairly conservative, you get to a 612-754 target.
Of course, the preceding assumes at least two other important things. First, it assumes that markets don’t overshoot, which they almost always do. Second, it assumes that global trade stays close enough to trend that such comparisons are meaningful, an assumption that I think highly unlikely given the rapidity which worldwide trade is now tumbling.
Is this scientific? Of course not. It is, however, a way for you to at least capture the relevant earnings/growth/multiple dimensions in a way that shows you some possible implications for the overall market.
Originally published at Paul Kedrosky’s blog and reproduced here with the author’s permission.
4 Responses to “What is the New S&P 500 Normal?”
A very nice attempt to put things in perspective! Factoring productivity growth would make it much more useful.
Thank you. Very enlightening. And how about this: At what level would the S&P 500 be today if it had always returned its average compound rate of 10.5 since (say) 1960? That number should help to tell if any given market level is high or low.I can’t find this anywhere, nor do I know how to compute it. Perhaps you do.
If I understand the above chart correctly: The postulated annual rates of earnings growth in Row 1 (3.4% to 8.4%) are simple annual rates. Base year 1987 Earnings across Row 2 are $16. Doing the math using the numbers in the Column 2 as an example (which assumes the lowest annual earnings growth of 3.4% per annum), I find Mr. Kedrosky’s total earnings growth number thru 2008 (32.4) to be roughly 201.8% compounded, thru 2009 (33.5) to be roughly 208.66% compounded, and thru 2010 (34.6) to be roughly 215.76% compounded. The foregoing corresponds to a compound average annual rate of growth during the 23 year period from 1987 thru 2010 of about 9.3% – relatively close to your 1960-present S&P compound average rate of 10.5%.
Very informative.Thank you.