Model Behavior
Are Wall Street's market models passe or are stocks headed
for fall?
By ANDREW BARY
The formerly trusty mathematical models used by Wall Street strategists to
gauge the appeal of stocks have broken down in the past year, as shares have
risen sharply in the face of higher interest rates. A model used for the past 15
years by Byron Wien, chief domestic strategist at Morgan Stanley Dean
Witter, now suggests that the Standard & Poor's 500 Index is overvalued by
a record 54%. Wien's model says the S&P 500 should be trading at around
945, way below its recent level of 1455.
Other models flash similar danger
signals. The "Greenspan model,"
unveiled by the Federal Reserve
chairman in 1997, calculates that the
S&P is a whopping 67% above its
appropriate level. It simply compares
projected profits for S&P 500
companies with the yield on the
10-year Treasury note.
"The model served me well until last summer," Wien says. "It used to be that
20% overvaluation was the threshold" portending a market setback. "But we
blew right through that in 1999."
Wien points out that his model says equities are more pricey now than they
were prior to the 1987 market crash, when the S&P was 40% overvalued.
Until recently, the model had a pretty good predictive record, pegging stocks
as expensive in 1987 and in the summer of 1998, and as attractive from 1993
through 1996.
The failure of the various rate-based models in the past year is causing Wien
and others to reassess their usefulness and questioning the assumptions
underpinning them. Wien says he's thinking about tweaking some of the
inputs, but not abandoning the model.
His model and others used by Street
strategists seek to measure the relative
appeal of stocks and bonds using
three key inputs: the level of interest
rates, projected corporate profit
growth and a risk premium for stocks.
Lower rates, higher profit growth and
lower risk premiums all are bullish for
equities.
In the past, higher rates generally have
crimped the stock market by inducing
investors to switch to bonds. But
during the past year, stocks, as
measured by the S&P 500, are up
20%, while bonds have endured one
of their worst bear markets, pushing
up the yield on the 30-year Treasury
bond to around 6.70% from about
5%. This yield now significantly
exceeds the so-called earnings yield
on stocks, which is about 4%.
Earnings yield is the inverse of the
market's P/E ratio. The wide
bond/stock yield gap suggests that
bonds look very attractive relative to
stocks. So why do stocks keep going
up while bonds suffer?
A key reason is earnings. Profit growth has been robust, and expectations are
that earnings will stay strong for the foreseeable future. It's estimated that
S&P profits were up an impressive 17% in 1999, following the disappointing
4% gain in 1998, according to First Call. Moreover, Wall Street is betting
that torrid profit growth of 1999 will continue, with S&P earnings rising
17.3% annually over the next five years, based on analysts' estimates
compiled by I/B/E/S. The five-year profit growth projection has gone up
significantly in recent years. As recently as 1996, the long-term estimate was
around 12%.
Wien notes that his model assumes profit growth
of 10% over the next 10 years. "People used to
tell me my projection was too high," he says. "Now they're telling me it's too
low." Actual profit growth for the S&P 500 in the past five years has been
around 11%, but the long-term growth rate has been closer to 7%.
The argument for using a high earnings-growth assumption rests partly on the
changing complexion of the S&P, which now features a 30% weighting in
technology stocks, up from 18% as recently as late 1998. Bulls are betting
that tech companies will generate substantial profit gains for the foreseeable
future. The current expectation is that the tech sector of the S&P will produce
26% annual profit growth in the next five years.
And Tom McManus, strategist at Banc of America Securities, argues that
tech investors aren't fazed by higher interest rates because they figure the
profit outlooks for Microsoft, Cisco, Qualcomm or Yahoo don't hinge on
whether rates are 7% or 5%. Using traditional analysis, high-P/E tech stocks
should be vulnerable to higher rates because of the diminished current value of
their distant earnings.
The Wien model also assumes a 2%
risk premium for stocks, meaning
investors should demand two
percentage points of additional return
on stocks relative to bonds to
compensate for the higher risk of
holding stocks.
Wien says that when he developed
the model, critics said his risk
premium was too low. Now, the issue
is whether stocks deserve any risk
premium at all based on their strong
historical performance.
Wien believes some risk premium is
merited because the return on
government bonds is virtually
guaranteed.
It's worth emphasizing that the Wien
model and others are quite sensitive to
changes in assumptions. For instance,
if no risk premium is assigned to stocks, the Wien model shows that the S&P
is just 5% overvalued. If the profit growth rate is lifted to 11% from 10%, the
S&P goes from being 54% overvalued to 39% overvalued.
Finally, the S&P actually is 16% undervalued if the earnings growth rate is
boosted to 15% and the risk premium is cut to 1% from 2%.
Wien, for one, isn't buying it. "The prevailing view is that we'll never have a
serious market correction," he says. "The economy will never have a serious
recession, and the Fed has figured out how to keep things under control.
These three things are implicit in the market now."
E-mail comments to editors@barrons.com |
