|
|
|
Stocks losing steam after 200 years on top
Conventional wisdom
says long-term investments belong in the stock market. But a new study
by two finance experts suggests stocks’ winning ways may be over. Can
your retirement portfolio survive 3.5% yearly returns?
By Dan Fisher, Forbes
For more than a century stocks have given the patient investor a higher
return than bonds. The difference is called the "equity risk premium."
That's supposed to be the reward you get for taking on the higher risk
of stocks. It's a huge number, on the order of 5% a year, as calculated
from the past 75 years of data by research firm Ibbotson Associates.
If that doesn't sound huge, try compounding it for a working lifetime
(5% a year for 40 years gives you a 600% increment). Or put it this way:
The equity risk premium says that stocks have compounded four times as
fast as bonds in real (that is, inflation-adjusted) terms. From such historical
data comes the conventional wisdom that your retirement savings should
be heavily skewed toward stocks.
Beware, say two finance experts in an alarming
study recently published in the Financial
Analysts Journal. They say that history -- or at least the interpretation
usually put on stock market history -- is bunk. If they are right, the
risk premium in coming decades won't be anything like 5% and you shouldn't
count on the S&P 500 ($INX)
to provide you with a fat and comfortable retirement. Let's tour their
findings and, if you accept them, consider an alternative to abandoning
equities. Need Help?
The guys doing the study, Robert Arnott and
Peter Bernstein, are not Ivory Tower theorists. Arnott manages $15 billion
at First Quadrant L.P. in Pasadena, Calif. Bernstein is a New York-based
economist who advises institutional investors. Neither expects S&P
500 index funds to do particularly well over the next several decades.
200 years of 7% returns
Let's start with the long-run historical return
on stocks. Including dividends and discounting for inflation, the return
over the past two centuries has averaged 7% a year. Early-19th-century
data are a little sketchy, but such a conclusion about the long-term performance
of stocks is well accepted.
Arnott and Bernstein deconstructed the 7% return.
A large part of it -- close to 5% -- came from dividend yields; a smaller
portion, 1.4%, from earnings growth. A remaining sliver came from an expansion
in price/earnings multiples. Note that we're talking averages here, and
there was much variation in the components over time.
In early years, dividends were fat while P/E
multiples didn't go anywhere. In the 1982-99 bull market, in contrast,
yields collapsed but stock market investors more than made up what they
lacked in dividends with a wild expansion of P/E multiples.
Need Help?
Where do we go from here? A naive extrapolation
of past into future tells you the 7% will keep on coming. Arnott and Bernstein
look, component by component, at the total real return and arrive at a
grimmer conclusion.
Where’d all the yield go?
That 5% dividend yield is history, they note.
The S&P 500 is now yielding only 1.5%. A bull would argue that dividends
have gone out of style, with companies (like Microsoft) plowing
their profits into share repurchases and internal growth.
Arnott and Bernstein have an answer to this
argument. They studied what happened to earnings plowed back rather than
paid out and found that companies did a miserable job of investing them.
When the dividend payout ratio -- the percentage
of earnings paid out as dividends -- climbed above 50% in the late 1950s,
subsequent 10-year earnings growth was between 2% and 4% a year. When
the payout ratio fell below 50% in the mid-1970s, subsequent earnings
growth plunged into negative figures. The payout ratio hit its second-highest
level since World War II in 1991, at 70%; one of the nation's biggest
booms followed.
Arnott tested the theory to see if this was
simply a case of earnings reverting to the mean. For sure, managers are
loath to cut dividends when earnings fall and rarely raise them to the
full extent that earnings rise. So one year's high payout ratio might
just reflect low earnings, and vice versa. But Arnott's statistical tests
indicated that the natural tendency of good earnings to follow bad was
a far less powerful predictor of 10-year earnings growth than the payout
ratio itself.
"We're coming off peak earnings in 2000
with the lowest payout ratios ever," Arnott says. "Those earnings
didn't generate growth; they were just wasted money."
Need Help?
The problem, he says, is that managers of public
companies squander profits on empire-building projects that fail to earn
a decent return. Investors would be better off if those managers distributed
corporate earnings as dividends, which investors could put into new enterprises
that grow faster.
"There's an incredible arrogance in management
thinking their 10th-best idea is better than their shareholders' first-best
idea," says Arnott.
Buybacks aren’t a good substitute for dividends
What about the argument that share repurchases
have replaced dividend checks as a way of distributing corporate cash?
Stock buyback programs were much in vogue in the 1990s, but Arnott says
those programs, for the most part, bought back the shares issued to executives
through options.
Perhaps the most surprising statistic to turn
up in this study is the meager 1.4% real growth in earnings per share,
a number Arnott and Bernstein calculated from data going back to 1871.
That is far less than the 3% yearly growth in the economy or even the
1.6% growth in per capita gross domestic product, a good measure of productivity.
That's a puzzling result if the stock market is supposed to mirror the
economy, let alone outperform it.
Part of the reason, their study concludes, is
the dual nature of the economy: Many of the fastest-growing companies
are outside the public markets and therefore not available to stock investors.
There's one more component to stock market returns,
and that is the expansion in P/E multiples. Could this continue to enrich
stockholders? Unlikely. The market is already expensive by this measure,
at 25 to 42 times last year's earnings (depending on how you count "nonrecurring"
writeoffs). If anything, the P/E is likely to go in the other direction,
back toward its historical norm of 14.
How am I expected to retire on that?
Absent any new boosters, Arnott and Bernstein
say, investors today should not count on more than the paltry earnings
and dividend growth they discovered in their study: a little less than
per capita gross domestic product, which they say is unlikely to rise
above 2% anytime soon. Add earnings growth to the meager dividend yield
and you get a total real return not much better than 3.5%. Note you can
get a risk-free real return of 3.5% on inflation-indexed Treasurys. That's
how the 5% risk premium may have vanished, at least on most S&P 500
stocks.
Is there a way out of this bind for investors
who need to build wealth for retirement and other worthy goals? Arnott
and Bernstein suggest that investors seek out investments that pay cash
-- real estate investment trusts, timber partnerships and traditionally
generous companies like General Motors and J.P. Morgan Chase.
There is one surefire way the risk premium could
return, Arnott says: P/E ratios (and thus stock prices) could collapse,
perhaps as much as 50%. If that happened, dividend yields would rise and
investors could once again look forward to high real returns.
"Those who argue long-term returns will
be better than 3.5% because yields will go up might be right," Arnott
says. "But the process will be unpleasant." Need
Help?
Interest Rates
|
