9.1 Efficient Market Hypothesis
Empirical Evidence Supporting the EMH
COACH'S REMARKS: This section contains a lot of text. For this exam, focus on identifying the evidence that supports each version of the EMH. Multiple studies have been performed in recent years to determine the validity of each form of the efficient market hypothesis.
Weak-Form EMH Study #2: Brealey, Meyers, and Allen
A study was performed by Brealey, Meyers, and Allen (2017), in which the authors investigated price changes in four stocks for consecutive days in 1991-2014. For each stock, they created a scatterplot with the return of day tt on the x-axis and the return of day t+1 on the y-axis. Below is a diagram of the S&P 500 returns similar to the scatterplots created in the study: In the study, all four plots showed no distinct pattern in the points, similar to the S&P scatterplot above, with the concentration of points being around the origin, and no bias towards any quadrants. This indicates the return in the market on one day has no effect on the next day's return. In addition, the autocorrelation coefficients were all very close to 0, implying there is no relationship between price returns on consecutive days.
Semi-Strong-Form EMH
According to the semi-strong form of the EMH, there should be no pattern in stock prices for a period of time after the release of new information regarding a company. Theoretically, there should be no pattern in stock prices before the release of new information as well. However, there are often differences in levels of information access or leakage before the official release. To test the semi-strong form of the EMH, researchers investigated stock prices after major news releases, such as earnings announcements, merger and takeover intentions, and the release of macroeconomic conditions. In order to determine the price return impact of the new information, the return on the stock must be adjusted in order to remove the portion of the stock price change due to systematic factors (e.g., market return). This can be done by subtracting the expected stock return from the actual stock return, thus isolating the return on the stock due to nonsystematic factors.
LƯU Ý CỦA COACH
Among the list of "other anomalies," the size effect and the value effect are most important. Make sure you know them well. Also, because of the size effect, the value effect, and the momentum effect, there is indeed some evidence that markets may not be efficient. Thus, a stock's beta may not be an adequate measure of a firm's systematic risk, which would make the CAPM incomplete. The FFC model exploits the size effect and the value effect and includes them as two of the factor portfolios. Thus, in an FFC model, in addition to the market portfolio, the model considers the momentum vs. reversal effect, size effect, and value effect.
NỘI DUNG KHÁI NIỆM CƠ BẢN
An efficient market is a market in which security prices adjust rapidly to reflect any new information. Competition between investors in a competitive market will tend to produce an efficient market. In this market, security prices reflect all past and present information. If markets are efficient, then it is difficult to find inaccurately priced securities. Superior returns cannot be attained, and thus passive strategies are best. If markets are inefficient, securities may be inaccurately priced. Thus, it is possible to achieve superior returns by trading in these mispriced securities. In this market, an active strategy may outperform a passive strategy. In Section 9.1, we will discuss three forms of the efficient market hypothesis: weak, semi-strong, and strong.
Empirical Evidence Against the EMH: Market Anomalies
COACH'S REMARKS:This section contains a lot of text. For this exam, your goal should be to identify and describe each anomaly.
Trading on News or Recommendations
Do profitable trading opportunities exist after big news announcements or a stock analyst's recommendations? Takeover Offers. When a firm is the target of a takeover offer, the offer price for the firm being acquired is typically significantly higher than its current stock price. The target's stock price will typically jump on the announcement, but it may not jump completely to the offer price. While this seemingly creates a profitable trading opportunity, there is still uncertainty as to whether the deal will occur at the offered price (or even occur at all). The following diagram from the Berk/DeMarzo text illustrates the average responses to takeover announcements: From the diagram above, note that after the initial jump in the stock price at the time of the announcement, target stocks do not appear to generate abnormal subsequent returns on average. However, stocks that are ultimately acquired tend to appreciate and have positive alphas, while stocks that are not acquired tend to depress and have negative alphas. Hence, an investor could profit from correctly predicting the outcome.
LƯU Ý CỦA COACH
Exam problems on this topic will most likely be conceptual-type word problems. It is important to note the following logic: If something supports the stronger form of the EMH, then it also supports the weaker forms. For example, if something supports the semi-strong form, then it must also support the weak form. If something violates the weaker form of the EMH, then it also violates the stronger forms. For example, if something violates the semi-strong form, then it must also violate the strong form. However, the reverse is not necessarily true: If something supports the weaker form of the EMH, we cannot conclude that it also supports the stronger forms. If something violates the stronger form of the EMH, we cannot conclude that it also violates the weaker forms.
Semi-Strong-Form EMH- TIẾP
For each firm, these abnormal returns were compared prior to the takeover announcement, at the time of the announcement, and after the announcement. Three months prior to the announcement, the stock price gradually increased, with a small positive abnormal return occurring throughout the period. This was caused by investor suspicion, and perhaps even leaked information. At the time of announcement, the stock price instantaneously jumped, providing a large positive abnormal return. Finally, after the announcement, the abnormal returns dropped to zero, and no further trend was present in the stock price. The diagram below illustrates the cumulative abnormal returns:
COACH'S REMARKS
For this exam, you should know which of the anomalies above is a result of underreaction or overreaction. Also, note that the FFC model (discussed in Section 8.2) exploits the momentum effect and includes it as one of the factor portfolios.
The Performance of Fund Managers
Fund Manager Value Added. Evidence shows that the average professional fund manager is able to identify opportunities in the financial market that would yield profit. However, not all managers can, and most incur trading costs greater than their ensuing profits. The median mutual fund actually destroys value. Nevertheless, the mutual fund industry still has positive value added because skilled managers manage more money and add value to the whole industry. Returns to Investors. Investors might think they will benefit by identifying and investing in the funds that yield profit, but evidence shows, on average, they will not. On average, an investor does not profit more from investing in an actively managed mutual fund compared to investing in passive index funds. The value added by a fund manager is offset by the mutual fund fees. Studies have also found that the superior past performance of funds was not a good predictor of future ability to outperform the market.
LƯU Ý CỦA COACH
Here's an example of the momentum effect vs. reversal effect. Assume McDonald's announces a new sandwich that will increase revenues and profits. If investors: Underreact to this new information, then the stock price will not increase as much as it should, given the increase in profits. Thus, the stock price will continue to increase over time as the actual profit increase emerges. This creates momentum in McDonald's stock as the stock price will keep increasing. Overreact to this new information, then the stock price will increase too much. Eventually, when profits are reported, investors will realize they have overstated the impact of the new sandwich and the stock price will fall. This is a reversal because the stock price is reversing its direction.
COACH'S REMARKS.
If past price patterns could be used to predict future price patterns, then investors could make easy profits. However, in competitive markets, easy profits do not last. As investors try to take advantage of the information in past price patterns, prices adjust immediately until superior profits from analyzing past prices disappear. As a result, all the information in the past prices will be reflected in today's stock price, not tomorrow's.
Weak-Form EMH Study #1: Kendall
In 1953, Kendall performed a study hoping to find regular cycles in stock prices. Instead, he discovered that prices followed a random walk model. The random walk model states that stock prices behave as though successive values differ by a random number. Therefore, in the random walk model, past stock prices have no bearing on future prices. Below is a diagram illustrating the random walk model:
Forms of Market Efficiency
In 1970, Dr. Eugene Fama, a Nobel-Prize winning economist, developed a framework for describing the degree of a market's efficiency. There are three versions of the efficient market hypothesis (EMH), and they differ in their assumption of what information will be reflected in current security prices. Below is a table summarizing his framework.
NỘI DUNG CHƯƠNG 9
In Section 9, we discuss several theories explaining how markets operate: Section 9.1: Efficient Market Hypothesis (EMH). We focus on what information is available to investors and how they react given that information. Section 9.2: Behavioral Finance. We explore how investor behavior and psychology influences the market.
Strong-Form EMH
In order to test the strong form of the EMH, investors must evaluate all information regarding a company that can be acquired from a "painstaking analysis." This information often can only be identified or inferred by experts, and it includes a review of the financial state of the company, its industry, and the economy as a whole. Under the strong form of the EMH, professional portfolio managers should not be able to consistently realize superior returns to individual investors and the market.
Empirical Evidence Against the EMH: Market Anomalies- PHẦN 1
In the past, the EMH has been assumed to reflect reality remarkably well. However, discoveries of market anomalies have given evidence to the contrary. When a market anomaly exists, an investor can profit from it and achieve returns greater than the risk-adjusted opportunity cost of capital. Before looking at some examples of market anomalies, let's list some cautions about market anomalies: Anomalies found in past data cannot be assumed to persist into the future. Data mining can yield false market anomalies because if you dig deep enough into a large data set, you eventually find unusual patterns. For many anomalies, unless an investor uses a computing algorithm to engage in high-frequency trading, the investor cannot profit from the anomaly.
Strong-Form EMH- TIẾP
Many studies evaluating the returns of professionally managed funds have shown "expert analysts" do not achieve superior returns to the market. In fact, Brealey, Meyers, Allen (2017) evaluated the performance of actively managed mutual funds from 1971 to 2013 and discovered that those mutual funds only beat the Wilshire 5000 index 40% of the time. Another study by Malkiel (1995) revealed that the top performing fund managers in one year only have a 50% chance to beat their reference index the following year. Because mutual funds also require fees and expenses to run, professionally managed funds become unattractive to individual investors. Without fees, mutual funds, on average, tend to match their respective benchmarks. However, when fees are accounted for, the funds significantly underperform their reference indices. These findings, based on numerous studies on the performance of fund managers, are well known in the investment industry and are consistent with the strong form of the EMH. This has led to a movement towards passive investment strategies through investing in index funds, resulting in significantly lower fees and elimination of possible underperformance.
LỜI GIẢI VÍ DỤ 1
SOLUTION: The momentum strategy suggests that there is a positive serial correlation in stock prices. Thus, price patterns exist and can be exploited by analyzing past price information. This contradicts the weak form of the EMH, which asserts that it is impossible to consistently attain superior profits by analyzing past returns. Recall that if something violates a weaker form of the EMH, then it also violates the stronger forms. In this example, the momentum strategy violates the weak form of the EMH, so it also violates the stronger forms (the stronger forms include the semi-strong form and the strong form). Thus, conclude that the momentum strategy contradicts all three forms of the EMH.
Semi-Strong-Form EMH
Semi-strong-form EMH assumes that current security prices fully reflect all publicly available information, including past market data. Public information, such as those found on the Internet or in the financial press, is factored into market values. Prices will adjust immediately upon the release of any public announcements (earnings, mergers, etc). Proponents of this hypothesis assert that it is impossible to consistently attain superior profits by analyzing public information because the information is already fully built into the security price. Note that a semi-strong-form efficient market is also weak-form efficient.
Empirical Evidence Against the EMH: Market Anomalies- Calendar/Time Anomalies
Some market anomalies are calendar-based or time-based. The investor appears to be more likely to achieve superior returns during specific times of the year, week, or day. January effect. Returns have been higher in January (and lower in December) than in other months. It is believed that investors, for end-of-year tax purposes and spending needs, sell portions of their portfolios in December, leading to a price drop. The high returns in January result from an increase in buying as a response to the lower prices. Monday effect. Returns have been lower on Monday (and higher on Friday) than on other days of the week. In fact, the Monday returns are often negative. Monday returns are measured from Friday close until Monday, so one would expect returns from the 3-day weekend to be higher than returns from Friday. Some have proposed that this phenomenon, also called the weekend effect, results from bad news occurring over the weekend and investors' pessimism as they head back to work. Time-of-Day effect. Returns are more volatile close to the opening and closing hours for the market. Also, the trading volumes are higher during these times.
Trading on News or Recommendations- TIẾP
Stock Recommendation. When a stock recommendation is given at the same time that news about the stock is released, the initial stock price reaction appears correct. The stock price increases in the beginning, then it flattens out. When a stock recommendation is given without news, the stock price seems to overreact. The stock price surges the following day, then it falls compared to the market. The resulting alpha is negative in the next several weeks. These relationships are shown clearly in the following diagram from the Berk/DeMarzo text.
Strong-Form EMH
Strong-form EMH assumes that security prices fully reflect all information, including public and private information. This includes all information obtainable from an in-depth analysis of the company, industry, and economy, or any information acquired from private sources. Proponents of the hypothesis assert that there are only lucky and unlucky investors, but no one (not even company insiders) can consistently attain superior profits based on any information. Consequently, there is no need to analyze any information, and thus the passive investment strategy would work best. Note that a strong-form efficient market is also semi-strong and weak-form efficient.
Weak-Form EMH Study #3: Poterba and Summers
Study #3: Poterba and Summers Poterba and Summers (1988) further explored the random walk model and examined the proportionate variances of returns. In theory, the variance of returns should increase proportionally to the intervals measured. For example, the variance of annual returns should be 12 times larger than monthly returns and 252 times larger than daily returns, given 252 trading days in the year. In all, this idea worked as an approximation, although there were some notable deviations depending on the size of time periods compared: In comparisons of short time periods of differing lengths, the variance of the longer period was slightly larger than the theory suggested. This supports a short-term momentum in stock prices. In comparisons of long time periods of differing lengths, the variance of the longer period was slightly smaller than the theory suggested. This indicates stock prices tend to reverse. Despite the two bullet points above, the vast majority of empirical research appears to be consistent with the weak form of the EMH. The evidence supports that stock prices reflect all information from previous stock prices, thus past prices cannot be used to predict future stock prices.
The Efficiency of the Market Portfolio- TIẾP
The assumption of rational expectations is less rigid than that of homogeneous expectations. If we assume investors have rational expectations, then all investors correctly interpret and use their own information, along with information from market prices and the trades of others. The market portfolio can be inefficient (and thus it is possible to beat the market) only if a significant number of investors: * Do not have rational expectations (thus information is misinterpreted). * Care about aspects of their portfolio other than expected return and volatility (thus they are willing to hold portfolios that are mean-variance inefficient). In this section, we start by examining whether there is any evidence that individual or professional investors can outperform the market. We then discuss the reasons why the market portfolio might not be efficient.
Other Anomalies
The following are anomalies that make even less sense than those previously mentioned. 1. Siamese twins. Two stocks with claims to a common cash flow would be exposed to identical risks. Intuitively, they should have similar prices, but they do not. 2. Political cycle effect. For a given political administration, its first year and last year yield higher returns than the years in between, possibly because the market is anticipating new policies. 3. Stock split effect. A stock split happens when a company divides its outstanding shares into multiple shares in order to lower the trading price of the stock, thus making it more attractive to investors. The stock's market capitalization stays the same, so the net effect should be zero since each investor's portfolio is worth the same before and after the stock split. Surprisingly, returns are higher before and after the company announces the stock split. 4. Neglected firm effect. This refers to lesser-known firms yielding abnormally high returns. Market analysts are less likely to study very small companies than larger companies. Thus, these "neglected" firms tend to be very small companies. 5. Super Bowl effect. Historical data shows that for the year following the Super Bowl, the stock market is more likely to do better if the winner is an NFC team. If an AFC team wins, the market is likely to do worse. Though this has a statistical basis, it is extremely unlikely that there is a correlation between which division wins the Super Bowl and how the stock market performs. This effect is most likely a result of data mining. 6. Size effect. Small-cap companies have outperformed large-cap companies on a risk-adjusted basis. Small stocks (those with small market capitalizations) tend to have positive alphas, and thus they tend to be above the SML. Big stocks (those with large market capitalizations) tend to have low or negative alphas, and thus they tend to be near or below the SML. 7. Value effect. Value stocks (below average price-to-earnings and market-to-book ratios) have consistently outperformed growth stocks. Value stocks (those with high book-to-market ratios) tend to have positive alphas, and thus they tend to be above the SML. Growth stocks (those with low book-to-market ratios) tend to have low or negative alphas, and thus they tend to be near or below the SML.
Empirical Evidence Against the EMH: Market Anomalies- Underreaction or Overreaction Anomalies
The semi-strong-form EMH implies that all public information is calculated into a stock's current price. However, some anomalies point to investors reacting disproportionately to new information. New-Issue/IPO puzzle. Overreaction to new issues pushes up stock prices initially. There is often a frenzy to buy stocks when a private company goes public because investors try to take advantage of an opportunity to earn big returns. However, within a few years, the returns fall below those of a comparable portfolio. Earnings announcement puzzle. A study was performed using data from 1972 to 2001 to analyze stock performance after unexpectedly good earnings and unexpectedly bad earnings were announced. The study showed that the 10% of stocks of companies with the best earnings earned approximately 1% more each month in the next six-month period than the 10% of the stocks of companies with the worst earnings. Although a substantial adjustment occurs prior to and at the announcement date, an adjustment also occurs after the announcement. Evidently, the investors underreacted to the earnings announcement. As a result of these slow price adjustments, companies that display the largest positive earnings surprises subsequently display superior stock return performance, whereas poor subsequent performance is displayed by companies with low or negative earnings surprises. Momentum effect vs. reversal effect. The weak form of the EMH uses the random walk model. However, many studies have identified momentum and reversal effects, suggesting that stock prices are not purely random or unrelated to past data. The momentum effect asserts that there is a positive serial correlation in stock prices, meaning rising stock prices continue to rise and falling prices continue to fall, as investors underreact to new information. The reversal effect suggests a negative correlation in stock prices and that a mean reversion in stock prices is in effect as investors overreact to new information.
Why The Market Portfolio Might Not Be Efficient
There are several reasons why the market portfolio might not be efficient: Proxy Error. The true market portfolio includes every type of tradable assets in the economy, such as bonds, real estate, and art. However, due to the lack of competitive price data, the market proxy cannot include most of these investments. So, even if the true market portfolio is efficient, the proxy may be inaccurate. Furthermore, standard proxies like the S&P 500 may be inefficient relative to the true market. Behavioral Biases. While sophisticated investors hold efficient portfolios, other investors may be subject to systematic behavioral biases (which we will discuss in Section 9.2.2) and therefore hold inefficient portfolios. Since the market portfolio is the combined holdings of biased investors and sophisticated investors, the resulting market portfolio might not be efficient. Alternative Risk Preferences. Some investors focus on risk characteristics other than the volatility of their portfolio, and they may choose inefficient portfolios as a result. For example, some investors seek investments with an unlikely but extremely high payoff, and they are willing to assume the diversifiable risks from holding these investments. Non-Tradable Wealth. Investors are exposed to other significant risks outside their portfolio. For example, an investment banker is exposed to financial sector risk, while a computer programmer is exposed to tech sector risk. These risks are due to their work (human capital) and are not tradable. The banker and the programmer may choose to invest less in their respective sectors to offset the inherent exposures from their human capital, thus deviating from the market portfolio. Note that just because the market portfolio may not be efficient, this does not rule out the possibility that another portfolio may be efficient (or more efficient).
Semi-Strong-Form EMH- TIẾP NỮA
This study is consistent with the semi-strong form of the EMH, as the stock price reacted instantaneously after the news announcement. Thus, an investor late to receive the information cannot "buy-in" and make a profit.
Bubbles and Irrational Exuberance
Though they are not market anomalies, bubbles also violate market efficiency. The main characteristic of a bubble is the market value of the asset significantly deviates from its intrinsic value. A bubble is fueled by investors' irrationally exuberant trading behavior and is characterized by a surge in asset price that is not warranted by earnings and dividends. The bubble is sustained by more investors joining the frenzy. The bubble bursts when there are no more investors willing to buy. Huge losses ensue for those who buy or sell too late. Three examples of bubbles are: The Japanese stock and real estate bubble. In the late 1980s, Japan's stock and real estate prices soared to great heights. The Nikkei, the stock market index for the Tokyo Stock Exchange, increased by about 300%. By the late 2000s, stock prices and real estate prices had fallen by 80% and 87%, respectively. The dotcom bubble (also known as the technology bubble). In the late 1990s, when many Internet-based companies were founded, the NASDAQ Index increased about 580%. During the early 2000s, the bubble collapsed. The U.S. housing bubble. Prices of homes in the United States nearly tripled from 1996 to 2006, fueled by banks overextending credit to borrowers and investors pushing prices up. By 2009, the price of many homes had dwindled to 30% of their peak value.
Semi-Strong-Form EMH PHẦN 2
Two researchers, Keown and Pinkerton (1981), performed a study on 17,000 firms that were targets for takeover attempts. When a company is taken over, a premium is typically paid on the current stock price, thus the acquiring company pays more than the fair market value of the stock price. This causes the stock price of the target firm to increase dramatically at the time of the takeover announcement.
The Efficiency of the Market Portfolio
Under the CAPM assumptions, investors have homogeneous expectations, and the market portfolio is an efficient portfolio. All stocks are on the SML and have an alpha of zero. Hence, all investors should hold the market portfolio (combined with risk-free investments). This investment advice does not depend on the quality of an investor's information or trading skill. Even naive (i.e., relatively uninformed) investors with less information and less than average trading skill can guarantee himself the average return simply by holding the market portfolio. However, in the real world, it is possible for sophisticated (i.e., informed) investors to profit by taking advantage of non-zero alpha stocks. For example, sophisticated investors can profit by buying stocks with positive alphas or selling stocks with negative alphas, both of which will raise the portfolio's Sharpe ratio, and thus improve the portfolio's performance. Note that in order for investors to profit by buying a positive-alpha stock, there must be other investors who are willing to sell it. Thus, this implies that not all investors have homogeneous expectations, which contradicts one of the CAPM assumptions.
Weak-Form EMH
Weak-Form EMH Weak-form EMH assumes that current security prices reflect all past market data, which refers to all historical price and trading volume information. Since current security prices have factored in all historical data, future returns on a stock should be independent of its past returns or patterns of past returns. Investors cannot predict future price changes by extrapolating prices or patterns of prices from the past. In other words, price changes are random. Proponents of weak-form EMH assert that it is impossible to consistently attain superior profits by analyzing past returns.
