ZeePedia

THE EFFICIENT MARKET HYPOTHESIS (EMH)

<< COMMON STOCK: ANALYSIS AND STRATEGY
Behavioral Finance >>
img
Investment Analysis & Portfolio Management (FIN630)
VU
Lesson # 23
MARKET EFFICIENCY
The fair price function of the capital markets provides assurance that investors can sell stock
at the going price and not be taken to the cleaners. A discussion of fair pricing inevitably
leads to the efficient market hypothesis (EMH), the theory supporting the nation that market
prices are in fact fair. The EMH is probably the single most important paradigm in finance.
Like technical analysis, market efficiency is a controversial part of finance. In an efficient
market security prices are based on the available information so as to offer and expected
return consistent with their level of risk. While most professors are convinced that the
markets are quite efficient and that free lunches are as scarce as Ty Cobb baseball cards,
some professional money managers believe otherwise. Capital market prices are presumed
to be fair because they are the equilibrium result of the analyses of many people, each of
whom is seeking to increase personal wealth. When a listed stock is put up for sale,
hundreds of people can bid for it. The markets ensure that the seller trades with the highest
bidder. Conversely, a buyer is confronted with numerous potential sellers, and the system
ensures that the buyer's order is matched up with the best price, which from the buyer's
perspective is the lowest price. The greater the number of participants and the more formal
the marketplace, the more an investor is assured of a good (fair) price.
THE EFFICIENT MARKET HYPOTHESIS (EMH):
To speak intelligently about the efficient market hypothesis a person must under-stand what
the hypothesis says and what it does not say. Efficiency can be categorized by both type and
degree.
Types of Efficiency:
The two types of efficiency are operational efficiency and informational efficiency.
Operational efficiency is a measure of how well things function in terms of speed of
execution and accuracy. At a stock exchange, operational efficiency is measured by such
factors as the number of orders lost or filled incorrectly and the elapsed time between the
receipt of an order and its execution. All market participants are concerned with these
matters, but the EMH does not refer to this type of efficiency.
Informational efficiency is a measure of how quickly and accurately the market1 reacts to
new information. New data constantly enter the marketplace via economic reports, company
announcements, political statements, or public opinion surveys, to name a few sources.
What does all this information mean? Is raising unemployment in the United Kingdom good
or bad for holders of U.S. Treasury bonds? How about a company's announcement that it
intends to split its stock five for one? Suppose the price of gold jumps $10 an ounce in one
day, what effect, if any, is this event likely to have on stock prices?
We know security prices adjust rapidly and accurately to the news without the need to
digest it very long. Sometimes the speed of adjustment is remarkably fast. For instance, the
author was once sitting in a brokerage firm punching up his current stock on the Quotron
machine. One of his holding was common stock in MGM Grand Hotel. The stock was
trading at $10¼. At that very moment, across the room, the bell rang and the red light
flashed on the Dow Jones News Service monitor, indicating hot news. The headline read,
"Fire at the MGM Grand Hotel". In the seconds it took the author to walk from the service
142
img
Investment Analysis & Portfolio Management (FIN630)
VU
monitor back to the Quotron machine, the stock fell to $7½, which is approximately where
it remained the rest of the day.
One need not be a Mensa member to realize that a hotel fire is bad news. In an
informationally efficient market, prices are going to react fast, just as they did in the MGM
situation. An investor cannot expected to read about the fire in The Wall Street journal the
following day and think, "Well, I'll bet that hammers the stock; I'd better sell," and then
expect to find that the market is still trying to sort out the news. Prices would have dropped
long ago.
Because the market is efficient, the meaning of the news is discovered quickly, and prices
adjust. Students in an investments course are sometimes disappointed to learn that simply
taking a stock market course does not ordain them with the power to read the financial
pages and fluently pick stocks that will double in price by next week. Things do not work
that way.
Still, the market is not completely efficient. It still rewards people who process the news
better than the next person. For one thing, not everyone has access to the same news, nor
does everyone receive the news in a timely fashion. Because of this discrepancy, market
participants commonly talk about three forms of the EMH, each of which is based on the
availability of a different level of information.
The efficient market hypothesis is one of the most important paradigms in finance.
The efficient market hypothesis deals with informational efficiency, which is a measure of
how quickly and accurately the market digests new information. It is well established that
the market is informationally efficient.
Degrees of Informational Efficiency:
1. Weak form Efficiency:
The least restrictive form of the EMH is weak form efficiency, which states that future stock
prices cannot be predicted by analyzing price from the past. In other words, charts are of no
use in predicting future prices.
According to the weak form of the EMH, how a stock arrived at its current price is
irrelevant. It could have followed the route of Stock A, or it could have behaved like Stock
B. The only thing that matters is the current price. Any information contained in the past
price series is already included in the current price.
The realization is a difficult pill for most people to swallow. A survey of a variety of people
would reveal that virtually everyone would identify stock B as clearly a better buy than
stock A. After all, B is "rising" while A is "falling". Who would want to buy a declining
stock?
143
img
Investment Analysis & Portfolio Management (FIN630)
VU
The point that is missed in this logic was made earlier: past prices do not matter; future ones
do. Everyone has access to past price information2. According to the EMH, so many people
are looking at these same numbers that any "free lunches" have already been consumed.
The current price is a fair one that takes into account any information contained in the past
price data.
Human nature is prone to extrapolate the past into the future. Business Week conducted a
poll3 in late 1999 asking investors their views on the stock market. Fifty-eight percent
indicated they believed the stock market was "very" or "somewhat" overpriced, but 52
percent of the respondents believed that stocks would be higher a year later. Respondents
aged 18 to 24 were most bullish, with 63 percent predicting the market would be higher in
2000. Oddly, though, 67 percent of this same group predicted a market crash in the coming
year.
Charting is a topic discussed in hundreds of books. In the same way we look for identifiable
forms in the clouds or in star constellations, our brains are creative enough to find patterns
in a sequence of stock prices. Technical analysts learn "important" patterns through folklore
or their own imagination.
Look at the four graphs in Figure 9-2. Are these random patterns, or is one or more of them
revealing something? Some technical analysts would look at Chart A and see a stock that
has been unsuccessful in penetrating a "resistance level" at 0. Its failure to rise above this
point after several attempts is followed by a major downturn in the stock price.
Chart B shows a pattern that looks appealing. This stock is on a sustained rise. Chart C
shows a bearish situation. Here a stock has penetrated its support level at 0, resulting in a
significant decline to the -5 area. A technical analyst would call this a breakout on the
downside. Chard D shows congestion in the -2 to +1 range, followed by a sharp break to a
new equilibrium level around -4.
What do these patterns mean? Would an investor be more inclined to buy one of these
stocks than the others? Is one clearly inferior to the others? Actually, each of these figures
was created using the random number generating function of Lotus 1-2-3. These are four
successive Lotus graph; each graph has a different seed number to start its series. In all four
graph, each observation is either one unit greater than the previous observation or one unit
144
img
Investment Analysis & Portfolio Management (FIN630)
VU
smaller, and each of the two possible out-comes had a 50 percent probability of occurring.
Are these graphs useful in predicting what Lotus will select next? Probably not.
Past prices do not matter; future ones do. Weak form efficiency
145
img
Investment Analysis & Portfolio Management (FIN630)
VU
Table 9-2 shows the procedures for the test.
146
img
Investment Analysis & Portfolio Management (FIN630)
VU
when they rise x percent from a subsequent low. Because anyone can calculate these
realized percentages, filter rules should not work if the markets are weak form efficient.
Fama also investigated the performance of filter rules, as have numerous other researchers.
The results are similar to those of the autocorrelation tests. Occasionally one reads reports
of successful filters, but they still prove uneconomic when the effect of transaction costs is
included.
The evidence against predictive chart patterns and valuable filter rules is so powerful that
anything new about this topic seldom appears in the finance literature. People who review
articles for the various academic journals have seen overwhelming evidence that the market
is weak form efficient, and editors rarely want to devote space to another article confirming
decades of prior work in this area.
2. Semi-strong Form:
The weak form of the EMH states that security prices fully reflect any information
contained in the past series of stock prices. Semi-strong form efficiency takes the
information set s step further and includes all publicly available information. The semi-
strong form of the EMH states that security prices fully reflect all relevant publicly
available information.
A plethora of information holds potential interest to investors. In addition to past stock
prices, economic reports, brokerage firm recommendations, investment advisory letters, and
so on all contain a myriad of details about what affects business performance and stock
value. While no one sees every one of these items, the market does, and prices move as
people make decisions to buy and sell based on what they learn from the information set
available to them.
The news item about the MGM Grand Hotel fire was not past price data, but it was publicly
available and the stock did decline because of it. According to the semi-strong form of the
EMH, this behavior is exactly what is expected.
Semi-strong form efficiency states that security prices reflect all publicly available
information.
Tests of Semi-strong Form Efficiency:
Extensive academic research supports the semi-strong version of the efficient market
hypothesis. The literature devotes much more attention to tests of semi-strong form
efficiency than to weak form tests. Studies have investigated the extent to which people can
profit by acting on various corporate announcements such as stock splits, cash dividends,
and stock dividends. While and occasional research paper shows that small profits could
have been made in a particular case, the general result is consistent: The market reacts to
public information efficiency, and investors will seldom outperform the market averages by
analyzing public news, especially if they must pay commissions to buy and sell.
We recognize that the market is pretty efficient. We have seen time after time that when we
get the word that a Wall Street firm is now recommending a stock that stock is already up a
point-and-a-half. The market is that efficient. As soon as anyone gets wind of a firm's
recommendation-boom-people are buying it, and that stock's price goes up. By then, the
value [of the information] is diminished.
147
img
Investment Analysis & Portfolio Management (FIN630)
VU
-Michael J. C. Roth
Executive Vice President
USAA Investment Management
In fact, academic evidence indicates that active portfolio managers (those who frequently
change their portfolios to include "better" stocks) tend to subtract value rather than add it.
Most people try to beat the Standard & Poor's Stock Index by picking better stocks or
moving to better sectors of the market. But studies show it is a game that underperforms.
-Leonard H. Wissner
Chief Investment Officer
Ward & Winsser Capital Management
Researchers at The Wall Street Journal, in conjunction with Zack's Investment Research,
report that for the 5-year period ending 30 September 1999 the stock recommendations of
only 3 of the 15 major brokerage firms managed to outperform the 205.6 percent earned by
the S&P 500 index.7 In the first nine months of 1998, 88 percent of the actively managed
U.S. mutual funds trailed the performance of the S&P 500. Most people would expect that if
anyone could analyze the market better than average, the well-trained, experienced analysts
at the major investment houses could. These "experts" did not do well during this period.
Their substandard performance is discouraging but surprisingly common.
One particularly famous study by Ball and Brown deals with the market's reaction to
corporate earnings announcements.8 This research reported that stock prices react favorably
to better-than-expected earnings, and vice versa. However, they also reported that security
prices seemed to anticipate the news as much as a year prior to the announcement, and that
by the time the actual earnings were made public and investor has little opportunity to
capitalize on the news.
Another noteworthy event study looked at the market reaction to the death of a corporate
chief executive officer. Interestingly, market prices declined when the CEO was a
professional manager. But when the CEO was the company founder, the death was
associated with an increase in the stock price. This finding may mean the market was
encouraged about the prospect of a new "professional" company leader.
Many investors view stock splits favorably, as mentioned earlier in the book. Companies
announce splits in advance of the actual split date. Can investors earn abnormal profits by
buying shares that are about to split? Fama is also associated with the classic study on this
topic. The study found several things. First, companies usually increase their dividends
when they split their shares. If the firm fails to do so, the market reacts adversely to this
preceding the split, but once the split is announced they cease to accrue any further. To
profit from the split, investors would have had to have bought the shares months in advance
of the split date. Once the split is announced, the free lunch is gone.
Many tests of semi-strong form efficiency use the event study methodology. In an event
study, a phenomenon occurring at a known point in time, such as a stock split or the
announcement of corporate earnings, is designated as time zero. Two days prior to the event
is day "T minus two," while two days after would be "T plus two." In the typical conduct of
an event study, a researcher would gather a sample of firms showing one or more instances
of the event of interest. Security returns before and after the event would be collected.
Depending on the researcher's hypothesis, the data might be collected for monthly, weekly,
148
img
Investment Analysis & Portfolio Management (FIN630)
VU
daily, or even intraday returns ­ how far before the event and how long after would also
depend on the particular study. Typically, the length of the before and after periods is the
same.
Regardless of the month or year in which the event occurred, each stream of returns is then
"lined up" so that each company's event corresponds to day zero. For instance, the first
observation might be a split that occurred on March 4, 1988. The second company might
have had a split on August 15, 1992. Both of these dates would be time zero in the event
study. In both instances the following day would be day T plus one.
3. Strong Form Efficiency:
The most extreme version of the EMH is strong form efficiency. This version states that
security prices fully reflect all public and private information. In other words, even
corporate insiders cannot make abnormal profits by exploiting their private; inside
information about their company. Inside information is formally called material, nonpublic
information.
Section 16 of the Securities Exchange Act of 1934 defines an insider as "an officer or
director of a public company, or an individual or entity owning 10 percent or more of any
class of a company's shares." The law requires insiders to report their holdings of corporate
securities within 10 days of becoming an insider. They must also report subsequent
transactions in these securities for themselves or a member of their family by the tenth day
of the month following the trade.
The evidence does not support this form of the EMH. Insiders can make a profit on their
knowledge, and every year people go to jail, get fined, or get suspended from trading for
doing so. Inside information gives an unfair advantage that can be used to extract millions
of dollars out of the marketplace. Where did these millions in profit come from? They came
from the pockets of individual investor who did not have access to the confidential
corporate news. Society does not feel this advantage is fair; consequently, insider trading is
illegal.
The Enforcement Division of the Securities and Exchange Commission is responsible for
detecting and prosecuting insider trading violations. People sometimes believe they can stay
at arm's length from the law by passing the inside information to a relative, who passes it to
a friend, who passes it to someone else who then acts upon it to the benefit of all parties
concerned. This strategy seldom works unless the trades are small. (Investigating small
potential violations is not economically feasible). All brokerage accounts are computerized,
and it is a relatively simple matter to screen for unusual account activity surrounding
mergers, important corporate announcements, and similar events.
The Insider Trading Sanctions Act of 1984 permits the courts to impose civil penalties of up
to three times the profit gained or loss avoided because of the use of material, nonpublic
information, plus it provides for a criminal fine of up to $100,000. It also precludes
corporate officers, directors, and anyone owning 10 percent or more of a firm's equity
securities from making a profit on a purchase and sale of the company's equity within a six-
month period. These people also may not sell the company's equity short.
The Insider Trading and Securities Fraud Enforcement Act of 1988 is related legislation.14 It
increased criminal fines to $1 million, raised the maximum jail term to ten years, and
required firms involved in the securities business to implement programs to prevent insider
trading by their employees. The real teeth of the law comes from holding a firm liable if its
149
img
Investment Analysis & Portfolio Management (FIN630)
VU
employees engage in insider trading, and providing a 10 percent bounty to encourage
informants to come forward.
The strong form of EMH states that security prices fully reflect all relevant public and
private information.
Tests of Strong Form Efficiency:
Strong form tests are more difficult to conduct because it would be hard to do so without
breaking the law. We can, however, find evidence to support the potential value of inside
information.
Business Week publishes a column called "Inside Wall Street," in some respects similar to
The Wall Street Journal's "Heard on the Street" column. Stock prices often react to news in
these types of articles, but they should not react until the publication hits the streets. The
magazine is not released to the public until 5:15 PM on Thursdays, so news of stocks
mentioned in the "Inside Wall Street" column in not public information during the Thursday
trading day.
Unusual trading activity in a number of stocks mentioned in the column over a period of
five months led McGraw-Hill; the publisher of Business Week, to inform officials at the
Securities and Exchange Commission as well as at the exchange.15 Table 9-4 shows the
suspicious price movement.
For seven issues of the magazine over this period, stocks mentioned in the article rose an
average of 11.54 percent compared to an average rise of 0.12 percent in the Standard &
Poor's 500 stock index. The large Thursday rise and increased trading volume was
compelling evidence that someone was trading ahead of the public distribution of the
magazine. This act was illegal trading on inside information.
The Semi-Efficient Market Hypothesis:
The essence of the semi-efficient market hypothesis (SEMH), a cousin to the EMH, is the
notion that some stocks are priced more efficiently than others. This idea is appealing to
many market analysts. Consider two very different companies such as IBM and a
hypothetical start-up firm called Triple-Scan Video. Everyone has heard of international
Business Machines, which trades on the New York Stock Exchange and many regional
exchanges. Thousands of portfolios contain its shares, and virtually all security analysts
watch it. The likelihood of realizing an unusual gain in the shares of IBM is extremely
small. The stock is priced fairly, and investors who buy some will likely earn a long-term
return consistent with the stock's level of risk.
What about Triple-Scan Video? According to the SEMH, fewer people are watching this
company, which implies a greater likelihood these shares will be undervalued. In other
words, the stock might not be priced as efficiently as the shares of IBM or other well-known
companies.
This idea is sometimes used in support of the thesis that the market has several tiers. The
first tier contains IBM, GM, Exxon, and other large firms. The second tier might contain
lesser-known but well-established companies such as those that trade on the American
Stock Exchange or the NASDAQ National Market System. The third tier might be
companies such as Triple-Scan Video. Another tier might be pink sheet stocks. The further
down the tier list an investor goes, the less efficient the pricing, or so the reasoning goes.
150
img
Investment Analysis & Portfolio Management (FIN630)
VU
It is probably safe to say that most students of the market are generally sympathetic to the
logic of the semi-efficient market hypothesis. It is not possible to follow every security.
Analysts need to follow the big names, and simply do not have time to research the ever-
expanding list of emerging companies.
The essence of the semi-efficient market hypothesis is the notion that some stocks are
priced more efficiently that others.
Security Prices and Random Walks:
The efficient market hypothesis states that the current stock price fully reflects relevant
news information. While some of the news is expected, much of it is unexpected. The
unexpected portion of the news, by definition, arrives randomly ­ the essence of the notion
that security prices follow a random walk because of the random nature of the news. Some
days the news is good, some days it is bad. Specifics of the news cannot be predicted with
great accuracy.
Substantial uncertainty even surrounds news that seems reasonably predictable. An article
in Forbes reported the result of a study showing that over the period 1973-1990, the average
error made by security analysts in forecasting the next quarter's corporate earnings for the
firms they covered was 40 percent. On an annual basis, the average error was never less than
25 percent. From 1985 to 1990, the average error was 52 percent, indicating that the
analysts' forecasting ability had not improved over the period.
It is perfectly possible for analysts to disagree in an efficient market. As an example, on
November 30, 1998, the firm Van Kasper and Company reaffirmed its "Strong Buy"
recommendation on Transocean Offshore (RIG, NYSE). That same day Janney
Montgomery Scott changed its recommendation from "Moderate Buy" to "Strong Sell."
When the news relevant to a particular stock is good, people adjust their estimates of future
returns upwards or they reduce the discount rate they attach to these returns. Either way the
stock price goes up. Conversely, when the news is bad, the stock price goes down.
Many people misunderstand what the random walk idea really means. It does not say that
stock prices move randomly. Rather, it says that the unexpected portion of the news arrives
randomly, and that stock prices adjust to the news, whatever it is.
In a famous analogy, a drunk staggers from lamppost to lamppost with a point of departure
and a target destination. The path of the drunk shows a trend from one post to the next.
Along the way, however, the path is erratic. The drunk wanders right and left, perhaps
occasionally out into the street or into a building wall. The precise route cannot be
predicted. The same is true of a security price and its consequent return. Over the long run,
security returns are consistent with what we expect, given their level of risk. In the short
run, however, many ups and downs seem to cloud the long-run outlook. The stock price
behaviors shown in the four charts of Figure 9-2 are random walks. Each succeeding
observation is just as likely to be up as down.
ANOMALIES:
This section reviews several important market anomalies that financial researchers actively
explore. In finance, the term anomaly refers to unexplained results that deviate from those
expected under finance theory, especially those related to the efficient market hypothesis.
Familiar anomalies include the low PE effect, the small firm effect, the neglected firm
effect, the January effect, and the overreaction effect.
151
img
Investment Analysis & Portfolio Management (FIN630)
VU
The Low PE Effect:
Numerous academic studies have uncovered evidence that stocks with low PEs provides
higher returns than stocks with higher PEs. This tendency is called the low PE effect. The
studies show this result even after accounting for risk differentials, which seems to be in
direct conflict with the capital asset pricing model and the theory behind it.
Some evidence indicates that low PE stocks outperform higher PE stocks of similar
risk.
Low-Priced Stock:
Many people believe that certain stock price levels are either too high or too low.
Equivalently, they believe the price of every stock has an optimum trading range. By
finance theory, the stock price should be merely a marker and, by itself, be of no value in
comparing firms. The size (and value) of a piece of pie depends on the number of pieces
into which the pie is cut.
Still, folklore surrounds stock prices. As early as 1936, the academic literature showed that
low-priced common stock tended to earn higher returns as stock with a high price. 18 In the
classic investment book by Graham and Dodd, the authors state, "It is a commonplace of the
market that an issue will rise more steadily from 10 to 40 than from 100 to 400."
Consider the following question: Is it easier for a stock to rise from $5 to $6 than it is for it
to rise from $50 to $60? Most people who play the market believe it is. If it is the case
(which theory and empirical evidence dispute), then every firm whose stock sold for $50
should split ten for one so that its share price would advance faster.
The Small Firm and Neglected Firm Effects:
Like the low PE effect, the small firm effect and the neglected firm effect are two important
market anomalies that influence the stock selection of some investors (and some portfolio
managers).
The Small Firm Effect:
The small firm effect recognizes that investing in small firms (those with low capitalization)
seems to, on average; provide superior risk-adjusted returns. Solid financial research
supports this hypothesis. Important papers on this topic include those by Reinganum20 and
by Banz.
The obvious implication of the small firm effect is that portfolio managers should give
small firms particular attention in the security selection process. We do not know why the
small firm effect exists, but it seems to persist. In the past, some anomalies tended to
disappear soon after they were reported. The small firm effect is still with us.
The Neglected Firm Effect:
The neglected firm effect is a cousin to the small firm effect and the semi-efficient market
hypothesis. Institutional investors are sometimes limited to larger capitalization firms. As a
consequence, security analysts do not pay as much attention to those firms that are unlikely
portfolio candidates. One paper by Arbel, Carvell, and Strebel investigated 510 firms over a
ten-year period and found, as expected, that the smaller firms outperformed those widely
152
img
Investment Analysis & Portfolio Management (FIN630)
VU
22
held by institutions. The authors postulated that institutions might perceive more risk with
the smaller firms, and hence they ignore them. In a related paper, Arbel and Strebel showed
other evidence that the attention of security analysts does affect the way shares are priced,
and that if analysts neglect a firm it has a systematic impact on the share value.
The implication is the same as with the small firm effect: Neglected firms seem to offer
superior returns with surprising regularity. When the Arbel, Carvell, and Strebel paper was
published in 1983, the authors closed by stating that the effect was "unlikely to persist over
time." Neglected firms continue to be an important research area that we have not yet
figured out.
Market Overreaction:
Another area of current research interest lies in the observed tendency for the market to
overreact to extreme news, with the general result that systematic price reversals can
sometimes be predicted. For instance, if stocks fall dramatically, they have a tendency to
perform better than their betas indicate they should in the following period. De Bondt and
Thaler have written several important papers dealing with this subject.
Experimental psychologists know that people often rely too heavily on recent data at the
expense of the more extensive set of prior data. At a race track, for instance the betting
pattern on the following race, even if the handicapper was largely inaccurate on previous
races. In their studies, De Bondt and Thaler found "systematic price reversals for stocks that
experience extreme long-term gains or losses: Past losers significantly outperform past
winners."
Brown and Harlow found that the overreaction is stronger to bad news than to good news
during the period of their study. After an especially large drop, security returns over the
following period were unusually large and persistent.
The January Effect:
Another well-known anomaly is called the January effect. Numerous studies show
persuasive evidence that stock returns are inexplicably high in January, and that small firms
do better than large firms in January.
In Richard Roll's Journal of Portfolio Management study, the begins by reporting, "For 18
consecutive years, from 1963 through 1980, average returns of small firms have been larger
than average returns of large firms on the first trading day of the calendar year." Comparing
AMEX stocks, which are generally smaller firms, with those on the NYSE, Roll found that
the average return differential was 1.16 percent in favor of the small firms, and that the t-
statistic for significance of the difference was a whopping 8.18.
Several explanations of this phenomenon have been proposed. Branch proposes that the
superior January performance comes from tax loss trading late in December.27 A better
explanation is probably provided by Rogalski and Tinic, who provide evidence that the risk
of small stocks is not constant over the year, and tends to be especially high early in the
year. The reason for this higher risk phenomenon is itself unexplained. Kiem explains this
result by reporting another anomaly. For some reason, stocks tend to trade near the bid price
at the end of the year and toward the ask price at the beginning of the year. In any event,
January tends to be a good month for the stock market.
Some analysts will argue that this effect should really be called the "November through
January" effect because these three months stand out for their good performance. Time
153
img
Investment Analysis & Portfolio Management (FIN630)
VU
30
magazine recently reported on a study by the Hirsch Organization that shows since 1950
the S&P 500 index was up an average 1.7 percent in November, 1.8 percent in December,
and another 1.7 percent in January. The next best month was April at 1.4 percent.
(September is the only month that is negative, down 0.2 percent on average.)
Other studies find evidence of a January effect in securities other than common stock. Chen
documents the presence of the effect with high, medium and speculative grades of preferred
stock.31 Wilson and Jones do so for corporate bonds and commercial paper.32 Gay and Kim
look at seasonality in the futures markets.34 The January effect is a pervasive result that
puzzles many people.
Further, some people consider the first five trading days of January to be a harbinger of how
stocks will perform for the rest of the year. Since 1950 in only three years (1966, 1973 and
1990) has the S&P 500 been higher at the end of the first week of the year, but lower by the
end of the year. Some people refer to this as the January indicator to distinguish it from the
January effect. It is probably not an especially useful indicator. The market is usually up for
the year, and the historical data indicate a bad first week for the S&P 500 does not predict a
down year for the market.
Stock returns are inexplicably high in January, and small firms' stocks do better than large
firms' in this month.
The Weekend Effect:
The weekend effect is the observed phenomenon that security price changes tend to be
negative on Mondays and positive on the other days of the week, with Friday being the best
of all.35 This persistent result does not yet have a satisfactory explanation. Some
behaviorists claim that people are upbeat on Fridays, and this attitude translates into stock
market optimism. Monday is a down day in other ways, so it might as well be a down day
for the market, too, or so the thinking goes. Whatever the cause, the weekend effect remains
as anomaly. Once again, however, the effect is too small to be economically significant.
The Persistence of Technical Analysis:
Market efficiency tests, especially those dealing with the weak form; have routinely found
that any evidence of market inefficiency cannot be profitably exploited after including the
effects of transaction costs. Still, an immense amount of literature is printed each year based
in varying degrees on technical techniques that, if the EMH is true, should be useless.
Even finance professors seem less than totally committed to the EMH paradigm. In a
national survey pf investment professionals, 40 percent of the respondents with a Ph. D. in
finance felt that advance-decline lines (a popular technical analysis tool) were "useful" or
"very useful." One-fourth of the respondents agreed that "charts enhance investment
performance."
Needless to say, we do not fully understand the theory or practice of technical analysis. Its
techniques are generally imprecise and do not lend themselves to rigorous statistical testing.
Certain phenomena from the clinical psychology literature seem to be at least partially
operative in the stock market. At a casino craps table, for instance, shooters throw the dice
harder when trying for a high number. Low numbers, of course, require an easier toss. Even
though no connection can be made between the random number that occurs and the strength
of the toss, the shooters experience a psychological illusion of control. Similarly, humans
suffer from hindsight bias. With trading techniques, people tend to remember their
154
img
Investment Analysis & Portfolio Management (FIN630)
VU
successes and displace their failures. One investor, for example, owned 200 shares of a
common stock, his only investment. On days when the stock was up a point, he would brag,
"I made $200 today." On days when the stock was down, he said nothing. Later, when the
stock rose a point again, in his view he "made another $200." He made that same $200
many times.
Final Thoughts:
The U.S capital markets are informationally and operationally quite efficient from the
individual investor's perspective. They are the envy of much of the world, and many
developing second and third world cultures emulate them. Portfolio managers, however, are
hired and fired largely on the basis of realized return, and a few basis points can make a
significant difference in the progression of their careers. Inefficiencies that may be
economically insignificant to a retail customer after considering trading fees may be much
more attractive to an institutional investor.
Eugene Fama and Kenneth French recently reported updated research on the nature of asset
pricing that may eventually be helpful in explaining market anomalies. Expanding on their
earlier work, they find "anomalies largely disappear in a three-factor model," where security
prices are determined by the excess return on the market portfolio, the difference between
the returns on portfolio of small stocks and a portfolio of large stocks, and finally the
difference between the returns on a portfolio of high book-to-market stocks and low book-
to-market stocks. Importantly, though, they state, "The three-factor risk-return relation is,
however, just a model. It surely does not explain expected returns on all securities and
portfolio."
In sum, much is not yet known about asset pricing. The markets are not perfect. Still, the
vast number of securities traded on the exchanges, the rapid introduction of new financial
products, and the globalization of world economies provide a fair, but complicated financial
battleground.
Summary:
The efficient market hypothesis (EMH) relates to informational efficiency and the fair
pricing function as opposed to operational efficiency. The essence of the EMH is that so
many people watch the marketplace that few if any individuals can consistently make
windfall profits by picking stocks better than the next person.
There are three forms of the EMH. The weak form says that past prices, or charts, are of no
value in predicting future stock price performance. The semi-strong form says that security
prices already fully reflect all relevant publicly available information. The strong form of
the EMH includes private, inside information as well. Considerable empirical research
supports the semi-strong form; however, we know that insiders can make illegal profits.
The random walk theory does not state that security prices move randomly. Rather it
maintains that the news arrives randomly, and that in accordance with the EMH security
prices rapidly adjust to this random arrival of news.
Anomalies are occurrences in the market that are inexplicable by finance theory. Stocks
with low PEs tend to show unusually higher returns; January is a good month for the stock
market; and small firms tend to do especially well in January. Technical analysis is
diametrically opposed to the efficient market hypothesis, yet it has many advocates,
including well-educated finance professors and practitioners.
155