[quote=pri_dk][quote=davelj]S&P earnings grow at about 5.5% per year over the VERY long term. That’s 3% inflation + 2.5% real growth. The 2.5% real growth comes from 1.5% population growth + 1% productivity growth (sometimes the two growth rates are reversed for periods). [Enigmatic reference to “animal spirits” removed for brevity.] [/quote]
Earnings growth is not the same as return on equity. Earnings growth is the rate of growth of the return.
Take a portfolio with a 10% / year ROE (S&P historical average is at least that) and 3% earnings growth. Using the original share price as a baseline, the portfolio will effectively return around 10.3% the next year, around 10.7% the following year and so on.
Even a very small rate of consistent earnings growth will yield huge payouts in the long run.
Wall Street analysts do lie. But figures do not.
[/quote]
There are several problems with your ROE analysis. If you buy a company at book value and it earns a 10% ROE annually and retains all of its earnings for 10 years and then sells for book value, then you will have generated a 10% IRR over the period. But you bought and sold the company at book value.
When you’re buying the S&P500 you’re typically paying several multiples of book value, so this 10% “return” (just to use 10% as an example) is on a cost basis that you don’t enjoy. You paid a big premium to “par,” so to speak. And that’s before subtracting out intangibles, which are considerable in the S&P. So focusing on ROE where the S&P is concerned may not be of much help.
A stock (or index) is a bond. The stock just has coupons (dividends/earnings) which are more volatile than a bond’s, and a terminal payoff that’s less certain and generally further off into the future than that of a bond. But the value of each is just the estimated present value of the cash flows that each will throw off to its holders.
The formula for calculating the approximate future return from the S&P is elegantly simple:
Beginning Dividend Yield + Annualized Earnings Growth + Annualized Change in the P/E Ratio
So – just to use some random numbers – if you start with a 3% dividend yield, assume 5% earnings growth, and think the P/E will expand from 13x to 16x over a 10-year period, the math looks like this:
3% + 5% + (16/13-1)/10, or
3% + 5% + 2.3% = 10.3%
This formula isn’t precise because it ignores any change in the rate of dividend reinvestment over the period and the change in the P/E ratio isn’t compounded. Nevertheless, this formula will get you VERY close to the actual return number if your assumptions are correct, which is why so many folks use it. And this formula applies to stocks as well. The challenge, of course, is inputting the proper assumptions.
So, using today’s numbers and assuming that earnings and profit margins normalize from today’s depressed levels over the next decade the math is as follows for the S&P500:
2.9% + 9.5% – 2.8% = 9.6%
(I’m assuming that 2009 S&P EPS are $45, but grow 5.5% annually over 10 years from current normalized EPS of $65, which puts normalized EPS at $111 in 2019. I’m further assuming that the current 19.4x P/E ratio declines to a normalized 14x, which accounts for the -2.8% at the end of the equation.)
So, the expected returns over the next decade – assuming everything reverts back to trend (the mean) – for a pure buy-and-hold investor are as follows:
Importantly, these are “expected” returns. That stock return might be WAY off because reversion to the mean may happen sooner or later than precisely one decade. But it represents the midpoint of possible future return distributions.
Interestingly, the 6.5% expected equity risk premium (9.6%-3.1%) is extremely high by historical standards. In a totally normalized world (normalized inflation, profit margins and valuations) of 3% inflation, a 2.5% real bond premium and a 3% equity risk premium, the 10-year Treasury would yield 5.5% and the expected return on stocks would be 8.5%. It’s pretty unusual to see a 6.5% equity risk premium (versus the 10-year Treasury). But that doesn’t mean it can’t get a hell of a lot wider before it mean reverts on a more permanent basis… (think S&P 600 or below). Also, the high current equity risk premium is more a function of abnormally low bond yields than of abnormally high expected future returns from stocks.
Anyhow, you can monkey around with the assumptions (I’ve just thrown some numbers in here), but the formula will at least give you a close approximation of your EXPECTED return (although your expectations may be dashed!) based on whatever assumptions you choose to use.
