Good piece in Fortune 8/17
pathfinder.com
"What Analysts and Cattle Have in
Common
Don't Trust Wall Street
Matt Siegel
Risk taking is often associated with youth--one
reason, for instance, that teenagers pay so much
for auto insurance. But when it comes to Wall Street,
it turns out that the youngest and least experienced
securities analysts tend to be the most risk averse.
That's according to a recent report by three
economists who studied "sell side" analysts. They
found that the longer an analyst spends in the
business, the more his forecasts deviate from those
of others covering the same company. More
experienced analysts are also more likely to publish
their forecasts first, and once they do, they're less
likely to change them. In short, as analysts age,
they become more willing to stick their necks out.
Stock analysts as a group appear to engage in herd
behavior in part because they're constantly
evaluated against their peers, says Harrison Hong,
one of the report's authors. When forecasting
earnings, young analysts figure it's better to fit in
with the crowd--even if the crowd is wrong--than to
risk being off on their own. That's because a few
notable failures can crush reputations. Later, when
analysts are more established, they can stray from
the pack without risking as much. Herd behavior is
prominent in almost all work settings, says Bengt
Holmstrom of MIT, who pioneered the theory of
reputation-based behavior. But stock analysts make
particularly good subjects for study because they
produce uniform, quantifiable outputs, like earnings
forecasts.
What makes Hong's study interesting is that it could
have turned out quite differently. Common sense
suggests that there should be some herd impulse
among analysts, just as there should be among any
group of people competing. But it's not clear
whether the effects should diminish or intensify with
age. For instance, one might argue that older
analysts should be more cautious than their more
callow counterparts, since they have a lot more to
lose. Similarly, younger analysts might intentionally
stray further from the herd, since they can gain
celebrity status by taking a risky position that turns
out to be correct, Hong says.
Sure, it's obvious to Wall Street watchers that
analysts tend to behave like cattle--when one
analyst switches a recommendation on a stock,
others often follow. However, that might happen
simply because the analysts are all working from
the same data. For instance, when housing starts
decline, real estate analysts will trim their forecasts.
That's not "herd behavior"--that's just being smart.
But by focusing on how analysts' behavior changes
with age, the study factored out this type of
clustering, concentrating on behavior that occurs
when analysts stick to a consensus for the sake of
blending in. Goldman Sachs' David Fleischer, one of
the top-ranked natural-gas and pipeline analysts,
says that Hong's "older is bolder" interpretation
makes sense to him. In 1991 he was the only
analyst on Wall Street to publish a buy
recommendation on Columbia Gas--an apparently
curious decision, given that the firm was in the
middle of a Chapter 11 bankruptcy proceeding. Why
did he do it? "It was because I understood the
industry," says Fleischer, 50. "It wouldn't have
ruined my career if I was wrong.... if a new kid was
wrong on something like that, they'd say, 'How could
you be so stupid, to recommend a company in
bankruptcy?' " (Turned out Fleischer was right. The
stock has since risen from $12 to $56.)
Having convincingly demonstrated the correlation
between herding and age, Hong now plans to slice
the data again to see exactly which analysts cleave
most closely to the herd--those at mainstream
investment banks or those at boutiques. Whatever
the results, one lesson for individual investors is
sure to survive: Use analysts' reports as sources of
insight and raw data, but take the earnings
forecasts with more than a grain of salt. Young
analysts looking out for their careers will always tend
to herd. Investors need to make sure they don't get
trampled."
Issue date: August 17, 1998 |
