Monthly
Discussion
NYSE and the
Efficient Market Hypothesis
Two-and-half
years ago the market was booming and investors' greatest fear was whether it
would crash or not. There was justification for such a fear. The equities
market had been steadily growing for almost ten years outperforming by far all
other types of investment. Such market behavior had been associated with a
crash. In September 1999 The Economist published a graph pointing out
similar behaviors during the decade that preceded the 1929 crash and the 1989
crash of the Japanese stock market, see Exhibit 3.
Exhibit
3. Share prices rising and leading into
a crash, as reported in The Economist on September 25, 1999,
(trimesterly data normalized to the dates shown in the graph legend). These
markets grew by more than a factor of 4 in less than 10 years. The red circles
represent recent data.
But the market did not crash. Instead it embarked on
a rather flat (with downward overtones) trend. This stagnation has been going
on for 10 trimesters now. Impatient questions arise about how long it will
continue. According to our long-term forecast in Exhibit 1 there is no end in
sight. I am not saying that the market will stay flat forever. I am
simply aware of the fact that for many of us "forever" translates to
a modest number of years!
What has happened? The market has always been
reliable for what concerns the long term. From grandfathers to shrewd market
analysts the traditional advice has always been "Don't panic. If you hold
out long enough, the market will reward you." In fact, the rule has been
that during any 20-year period stocks will outperform bonds. Reliable data over
the last 75 years confirm that rule for American markets. But there can only be
three non-overlapping 20-year periods since 1926, which undermines the
statistical validity of this confirmation. Moreover, there is evidence that in
stock markets of other countries, and in older times of the American stock
markets, there have been periods where 40 consecutive years were necessary for
stock investments to outperform bond investments.*
And there is a more intriguing and
disturbing issue raised by The Economist: the ultimate
application of the Efficient Market Hypothesis (EMH).
EMH says that the market is very
efficient in assimilating information. Following the observation of any
financial pattern—that small company stock outperform those of large ones, that
gains of winter seasons outperform those of summer seasons, that
low-price/high-yield stocks outperform the DOW, and so forth—investors exploit
the pattern promptly and eventually the pattern dies away. The corresponding
rule or advice for investors becomes useless.
The pattern in question could simply be that stocks
outperform bonds over 20 years; the corresponding advice is "if you hold
out long enough, the market will reward you." Have investors
exploited this pattern sufficiently to render it useless? There is some
evidence in that direction. Record numbers of investors have rushed to the
stock market during the last ten years and most of them are stubbornly holding
out waiting for the market upturn.
EMH applied to the very concept of the stock market
reminds me of the snake that eats its own tail. Even if the consequence is only
that we'll have to wait more than 20 years for risk-free returns from stocks to
become superior to those from bonds, there is reason to consider readapting our
investment portfolios.
However, for the frequent trader there is good news.
Low growth periods
are
naturally accompanied with large fluctuations (see discussion of the
"winter season" in Conquering Uncertainty). Strategies designed for either
bull or bear markets need to be revamped so that they exploit the large and
frequent up-and-down excursions that characterize flat markets. Betting in both
directions could prove a profitable rule, until EMH catches up with this one
too.