In one study of the medical field, doctors were asked to diagnose
which patients had pneumonia. The doctors were told to select only
the cases where they were 90 percent sure of the diagnosis. After
the doctors completed their exams and diagnoses, follow-up stud-
ies showed that only 50 percent of the patients selected actually
had pneumonia.
It’s no better in the investment arena. In one study, analysts
were asked to specify a trading range for a given stock. (This was
before the days of dot.com stocks, when a reasonable range would
be, say, $5 to $500 for a given week.) Like the doctors in the earlier
study, the analysts were told to be 90 percent sure that they were
right about the range. In this case, the analysts did a little better
than the doctors. One could have expected about 10 percent of the
stocks to violate the rule and move outside the range, but in fact
35 percent of the stocks traded outside it. The conclusion? Again,
overconfidence. We think we know more than we actually do. (Does
this ever happen to you? You are at a party and someone asks what
band recorded “Wooly Bully.” No one knows—but you do. You
are absolutely certain that it was The McCoys, so you clear your
throat modestly and enlighten everyone. Oh, it’s so great. You are
Mr./Ms. Know-it-all. That is, until some smarty-pants surfs the Net
on his Palm Pilot and announces, “No, it was Sam the Sham and
the Pharaohs.” Damn. I hate that.)
In the face of our rampant overconfidence, one of the best ego
deflators in the investment world is Mark Hulbert, publisher of
the Hulbert Financial Digest. He runs many rigorous statistical tests
to see which investment newsletters are demonstrating true exper-
tise in picking stocks and which are just lucky. Of the newsletters
that he follows, 89 have been around for 10 years or more. Hulbert
subjected the performance of these newsletters to a “runs test” to
see if any of the funds could be credited with a hot hand for the
period. In fact, three of the newsletters passed the test. Are these
the true market savants? Does this mean that we should rush out
and subscribe? Although many investors do exactly that, the an-“A” Is for Assume Nothing 16314-25 ware 163 1/19/01, 1:15 PM