Financial advisers often maintain that long-term investors allocate
more aggressively to equities. But according to research by Nicholas
Barberis, associate professor of finance, investors would be wise
to hedge their bets.
Financial advisers today generally suggest that long-term investors
aggressively allocate assets to stocks. If you follow a human capital
argument, for example, younger people should invest more in stocks
than should older people. Human capital is an individual's future
wage earnings - a very important asset - and many economists think
of these fixed annual wages as bonds. So if younger people already
have a sizeable position in bonds through their human capital,
they should counterbalance this with a heavy financial investment
in stocks.
Then there is the mean reversion rationale, which says
that historically, bull markets have generally been followed by
bear markets, and vice versa. As stock market fluctuations offset
each other over time, it makes sense for younger investors to
put more into stocks than bonds. Even if the stock market were
to crash, there would be a high likelihood that losses for younger
investors could be recovered over the long term.
"There's this popular advice that people who have many years left to live should
put more of their money into stocks. Because there's some mean
reversion in stocks, in the long run, they're not that risky,"
said Nicholas Barberis, associate professor of finance.
Barberis analyzed the validity of this advice in his paper "Investing for
the Long Run When Returns Are Predictable." Though he concluded
that young investors should indeed invest more heavily in stocks
than should older investors, the effect is not as large as many
people think. The reason, he warns, is that "the evidence of cycles
in the stock market is actually very weak."
With such tenuous evidence that the stock market reverts to the mean after huge
upswings or downswings, Barberis believes that many advisers may
be encouraging investors to hold overly aggressive positions in
equities.
Calculating Long-Run Risk
According to Barberis, the
statistical evidence for mean reversion consists of two data points.
The stock market went up in the 1920s and down in the 1930s, then
up in the 1950s and 1960s and down in the 1970s.
In conducting his own research, Barberis used monthly data spanning 523 months,
from June 1952 to December 1995. The stock index he employed was
the value-weighted index of stocks traded on the New York Stock
Exchange (NYSE), as calculated by the Center for Research in Security
Prices (CRSP) at the University of Chicago. In calculating excess
returns, Barberis used U.S. treasury bill returns provided by
Ibbotson Associates.
"Our analysis shows that sensible portfolio allocations for short- and long-term investors can be different
in the context of predictable returns," Barberis wrote. Mean reversion
in returns lowers the variance of long-horizon returns. "This
makes equities appear less risky at long horizons, and hence,
more attractive to the investor," said Barberis.
But when he looked at the actual historical data on asset returns, Barberis
found that the true extent of mean reversion in returns was fairly
small. Long-term investment in the stock market is still sound
advice, according to Barberis, but the mean reversion rationale
behind it is not as powerful as many people think. "I found that
people with a 20-year horizon would want to put 15 percent more
into stocks than people with a one-year horizon, which is a lot
less aggressive than some advisers recommend," said Barberis.
