The past: a 7% real return. The future: 3%.
Stocks have returned a glorious 7% annually over the past century (total return, net of inflation). Continuing on the same course, they’d deliver very comfortable golden years to you. But they won’t do that. The stock market is now poised to deliver not even half its historical return.
I’m not predicting a crash, or calculating what the market will do next year. No one can do that. I’m saying that returns from stocks will be, over the next several decades, quite disappointing.
My sad forecast is not built on any gloomy view of where the economy is going. Nor does it assume any slackening in corporate growth. It’s simply an arithmetic inevitability having to do with earnings yields.
This is bad news for retirees and future retirees. If you are still working, plan on working longer and saving more of your paycheck. If you’re retired, live frugally.
You might think that past returns over a long period—like 100 years—would be a sensible place to start with a forecast of future returns. But investing doesn’t work that way. What matters to today’s buyer of a stock or bond is not what it did in the past but what it earns for the buyer at today’s purchase price.
A bond buyer can’t miss this. The yield to maturity on 30-year inflation-protected Treasurys is now 1%. If you buy one and hold until 2047 you can be quite certain of getting that 1%—no more, no less. The fact that in the past 30 years or past century Treasury bonds delivered a higher real return is irrelevant.
Stocks, of course, don’t lock in a return the way a default-proof bond does. But they are constrained by the arithmetic of their earning power, just as bonds are constrained by their coupons. With stock prices high, that arithmetic doesn’t look pretty.
The S&P 500 index has been hovering near 2600. Corporate earnings over the 12 months to June 2017 came in at $104 per index unit. That’s a 4% earnings yield. You buy a share for $100 and you get a claim on $4 of profit that you can either spend or reinvest.
What happens to stock owners over time? Let’s start by making two optimistic assumptions. One is that corporate profits would keep up with the cost of living even if corporations distributed all their profits as dividends. The other assumption is that the S&P’s price/earnings ratio remains forever at its present lofty level of 25.
Now let’s see what the total real return would be in a world where corporations did distribute all their earnings. You take your $4 dividend and buy more equity. The share price and earnings per share stay constant in real terms, but your share count compounds at a 4% rate. Sometime later you sell your stake at the same 25 times earnings at which you got in.
In this scenario your annual real return will be exactly 4%. That’s probably shy of what your retirement planning counts on.
At this point the bulls will shout out that corporate earnings growth has far outpaced inflation. Yes, it has. Earnings on the S&P have quadrupled in real terms over the past century. But that’s true only because corporations reinvested a lot of their earnings.
Well, you can’t reinvest the same earnings twice. Any dollars of profit that corporations are plowing back themselves are dollars you aren’t getting as dividends to be plowed back via your brokerage account.
My grim 4% return number doesn’t change a bit if the corporation does all the reinvesting for you, by buying in shares or buying new lines of business. It doesn’t change if you and the corporation share the work of reinvesting.
If the corporation is doing all the plowing back—an Alphabet, not an Enterprise Products Partners—then your share count would stay put but your real share price and earnings per share would compound at 4%. And when you sold out at 25 times earnings you’d have the identical 4% annual return.
Note: I’m talking about the collective $22 trillion equity market, and what happens to someone who owns a slice of it via, say, Vanguard Total Stock Market (VTI). I’m not making any predictions about Alphabet.
Now let’s take a closer look at the assumption that collective earnings would go sideways in real terms if all earnings were distributed as dividends. That assumption is indeed too optimistic. It turns out that corporate America would have had to reinvest a fourth of its earnings over the past century in order to keep its earnings constant in real terms. (In fact it reinvested more than a fourth, enabling the S&P, as noted above, to quadruple earnings per share.)
This somewhat depressing news may come as a surprise. Accounting rules demand that earnings be reported net of capital consumption. Exxon Mobil’s profits are what’s left after an allowance for the depletion of its wells, Norfolk Southern’s after the cost of replacing railroad ties. Why isn’t the cash flow from depreciation enough to keep things going? Why does business also need a reinjection of profit peeled off the bottom line?
One reason is that depreciation charges are too meager because they are not adjusted for inflation. Another reason, perhaps more important in a high-tech economy, is that business has to innovate just to stand pat. Back when Kodak was a blue chip it scrupulously depreciated its machinery. But the real problem was not wear and tear. It was the coming irrelevance of film.
Go back to our $4 earnings yield on $100 invested. With the first dollar merely allowing you to stand in place, your real growth shrinks to 3%. Assuming P/E ratios go neither up nor down, 3% is going to be your real return.
Why did our ancestors do so much better? Because stocks used to be cheap. That meant each reinvested earnings dollar bought a lot of incremental earning power.
Since 1917 earnings yields have averaged 7.4%. Add to the 7.4% a 1.4% annual kicker from the expansion of P/E multiples (which quadrupled over the century). Now subtract the one-fourth of earnings needed to stay in place. The end result is a compound annual return equal to the observed 7%.
The 3% forecast assumes, optimistically, that P/E ratios remain on a permanent high plateau. Robert Shiller, of Irrational Exuberance fame, is skeptical that they will do that. His recent bearishness has not yet been vindicated, but it might be, starting maybe next week or maybe in 2023. If multiples contract, you’ll get less than 3% a year over the next quarter century.
Tighten your belt.
Originally published at Forbes