with computing power becoming so cheap.^16 Throw in a load of vari-
ables to explain stock price movements and you are sure to find some
spectacular fits—like over the past 100 years stocks have risen on every
third Thursday of the month when the moon is full!
The representative bias has been responsible for some spectacularly
wrong moves in the stock market, even when the situations seem re-
markably similar. When World War I broke out in July 1914, officials at
the New York Stock Exchange thought it was such a calamity that the ex-
change closed down for five months. Wrong! The United States became
the arms merchant for Europe; business boomed, and 1915 was one of
the single best years in stock market history.
When Germany invaded Poland in September 1939, investors
looked at the behavior of the market during World War I. Noting the fan-
tastic returns, they bought stocks like mad and sent the market up by
more than 7 percent on the next day’s trading! But this was wrong again.
FDR was determined not to let the corporations prosper from World War
II as they had from World War I. After a few more up days, the stock
market headed into a severe bear market, and it wasn’t until nearly six
years later that the market returned to its September 1939 level. Clearly,
the representative bias was the culprit for this error, and the two events
weren’t as similar as people thought.
Psychologically, human beings are not designed to accept all the
randomness that is out there. It is very discomforting for many to learn
that most movements in the market are random and do not have any
identifiable cause or reason. Individuals possess this deep psychological
need to know why something happens. That is where the reporters and
“experts” come in. They are more than happy to fill the holes in our
knowledge with explanations that are wrong more often than not.
Dave:I can relate personally to this representative bias. I remember that
before I bought the technology stocks in July 2000, my broker compared
these companies to the suppliers providing the gear for the gold rushers
of the 1850s. It seemed like an insightful comparison at the time, but in
fact the situations were very different. It is interesting that my broker,
who is supposed to be the expert, is subject to the same overconfidence
that I am.
IC:There is actually evidence that experts are even more subject to over-
confidence than the nonexperts. The so-called experts have been trainedCHAPTER 19 Behavioral Finance and the Psychology of Investing 327
(^16) For a reference to data mining, see Andrew Lo and Craig MacKinlay, “Data-Snooping Biases in
Tests of Financial Asset Pricing Models,” Review of Financial Studies, vol. 3, no. 3 (Fall 1999), pp.
431–467.