Internet a safer source of
advice: ANALYSTS' FORECASTS:
By ALPESH B PATEL
Since
1945, the
have
outperformed all other years with anaverage gain of 17 per cent.
To benefit
from this, wewant stock ideas from highly trained and
educated
full-time researchers who spend
hundreds of hours tracking companies and
meeting
their management. So, is it time
again to trust equity research analysts
from
the big institutions for stock
ideas? Remember the old accusations -
that these
analysts misled investors by issuing
flattering research reports to drum
up
other business for their banks.
The recent
trust-building changes in the equity research industry
have been
significant: a Dollars 1.4bn (Pounds
880m) settlement in fines and Wall
Street
firms agreeing to spend Dollars
450m to fund independent research. So
are equity
analysts' stock picks performing
outstandingly well under the new
regime? It
doesn't look like it. Bloomberg
tracked the picks of 459 analysts at the
top
investment banking firms. Only 18
analysts would have made or saved
their
clients money by outperforming the
S&P 500. In fact much publicised
remedies,
such as independent research
departments to remove analysts' bullish
bias, miss
their target altogether.
Recent
research notes three possible explanations for analyst bias:
"career
concern perspective"
(analysts' optimism is somehow rewarded by
employers);
"selection
bias" (analysts tend to cover the stocks they can recommend
over
stocks they cannot); and "behavioural bias" (analysts tend to like the
stocks
they cover). "If you listen
to the Enron debate," the research author
says, "all
the senators and congressmen
are going for the first reason; but the
analysts
point to the second and third
explanations of this bias."
Furthermore,
US companies have a seemingly uncanny ability to beat
earnings
forecasts set by "expert"
analysts who speak beforehand to the company
management. A suspiciously high 82 per
cent of the companies that
reported
earnings by mid-February this year
met or beat analysts' earnings
estimates,
according to First Call. Beating
analysts' figures was the most watched
measure
of corporate success in the
late 1990s. The suspicion is that analysts
are
guided by companies to set
beatable estimates so that when the real
figure is
released the stock receives a boost.
Clearly,
we cannot simply trust analysts because of recent investment
banking
changes. So, what should the
private investor do for stock ideas as the
market
promises to add to its gains of 25
per cent since March? Forty pages of
investment research are compelling to
a private investor, whatever they
say, and
whatever warnings about your
financial health are on the cover.
Instead,
trust machine over man, (or web over woman). Why would you
want to
follow an analyst of whom you have
never heard, whose performance record
is not
provided, and whose human
fallibility may lead to subconscious if not
overt
bias?
The
alternative is far more compelling. Websites can crunch numbers
of more
companies than an analyst ever could
and deliver stock picks based on
the proven
criteria of market legends such as
Benjamin Graham (the man who taught
Buffett), Marty Zweig and Peter Lynch. For example, www.validea.com
shows that
if you followed its
"Benjamin Graham" stock analysis you would have
outperformed
the S & P 500 by 31 per
cent annually since 1999.
We all
know past performance is not necessarily a guide to future
performance. But if ever it could be,
it would be in the case of proven
performers such as Benjamin Graham,
not a 30-something banker.
Furthermore, the
outstanding www.investars.com
ranks analysts' track records - something
banks
and the
track in
a particular sector? Which
analyst has the best track record for a
particular
stock? You won't get that in a
stock research report.
Before the
internet, such information would never have reached
private
investors. Unfortunately, private
investors still do not know that they
can
reach it. These sites do focus on
US stocks but
buy US
stocks as easily as
with
better information. Doubtless much
analyst research is excellent - but
without
web tools investors just can't
tell which. The best approach may be to
use
analyst research for the data it
provides, not the "buy" or "sell"
rating. Focus
on the serious fundamentals
such as price-earnings. Use the tables
comparing the
stock to its peers. And discount
the opinion and arguments the analyst
makes and
focus on the data.
*Analysing the Analysts: Career Concerns and Biased Earnings
Forecasts,
Harrison Hong and Jeffrey Kubik, Journal of Finance, Feb 2003
www.stanford.edu/hghong/analysts-Jf.pdf