Chapter 11: The Efficient Market Hypothesis 1. If markets are efficient, what should be the correlation coefficient between stock returns for two non-overlapping time periods?
If markets are efficient, then the correlation coefficient between stock returns for two non periods (the autocorrelation) should be zero. If not, one could use returns from one overlapping periods period to predict returns in later periods and make abnormal profits. Note that we do observer some non-zero return autocorrelations. See page 358 to 359 in the text. 2. A successful firm like Microsoft has consistently generated large profits for years. Is this a violation of the EMH?
No. The EMH implies that Microsoft’s investors, over any period, have earned the correct riskadjusted return on their investment. 3. “If all securities are fairly priced, all must offer equal expected rates of return.” Comment.
No. Expected rates of return vary according to premium earned from incurring risk. The CAPM implies that the risk-premium of a security is solely a function of its share of market risk (as measured by its market-beta), while other theories/models allow for other “priced” risk-factors. For example, according the results of tests of the Fama-French Three Factor Model (FF3F Model), a security’s risk -premium -premium (it’s return in excess of the risk -free -free return) is a function of its market risk (measured by market beta), its relative size (measured by SMB bet a) and its relative price (its book value of equity to its market value of equity measured by HML beta). bet a). Other risk factors such as exchange rates or commodity comm odity prices may also be important to determining the correct risk premia. 4. Steady Growth Industries has never missed a dividend payment in its 94-year history. Does this make it more attractive to you as a possible purchase for your stock portfolio?
Not more (or less) attractive. The EMH implies the the predictability of its dividend is reflected in the stock ’s price. ’s price. A predictable dividend would likely lower the perceived risk and therefore therefore decrease the required return and therefore increase the price. 5. At a cocktail party, your co-worker tells you that he has beaten the market for each of the last 3 years. Suppose you believe him. Does this shake your belief in efficient markets?
Assume “beat the market” means a return in excess of the the return on the S&P 500. It may be that case that the coworker bought bou ght high risk stocks (CAPM β > 1) and held h eld them in an up market. In this case the coworker might have earned a return commensurate commensurate with the risk incurred. To determine if the coworker earned a risk-adjusted excess excess return calculate calculate the α. α.
1
If you use just the market return, then it is called the CAPM α (also called Jenson’s α): R P = α + β R M + ε If you use the market return, the small stock premium (SMB) and value stock premium (HML), then it is called the Fama-French α: R P = α + βM R M + βSMB SMB + βHML HML + ε Recall that SMB in each time period equals the return to a portfolio of small stocks less the return to a portfolio of big stocks and is therefore the small stock premium. HML in each time period equals the return to a portfolio of high BM (value stocks) less the return to a portfolio of Low BM (growth stocks) and is therefore the value stock premium. If the portfolio, be requirement, held only oil service stocks, and therefore incurred oil risk R P = α + βM R M + βSMB SMB + βHML HML + βOil R Oil + ε In any case, a positive and significant α indicates excess return after accounting for risk . If α is not positive and significant, and the portfolio’s return is in excess of the market return then it may be due to some type of increased risk (β) and not “skill” (α). 6. “Highly variable stock prices suggest that the market does not know how to price stocks.” Comment.
The EMH states price changes are caused by new information which market participants use to update expectations and then rebalance portfolios. Volatile prices caused by volatile underlying economic conditions and volatile changes in expectations is consistent with the EMH. 7. Why are the following “effects” considered efficient market anomalies? Are there rational explanations for any of these effects? (a) P/E effect. (b) Book-to-market effect. (c) Momentum effect. (d) Small-firm effect.
A trading rule that can be used to earn risk-adjusted abnormal returns is sometimes called an “Efficient Market Anomaly.” The P/E effect and Book-to-Market effect are names for the observed higher returns earned by value stocks. The Small-firm effect is a name for the observed higher returns earned by small stocks. Are these violations of the EMH or are these example of returns earned for realized risk? Over time and across stocks, value stocks and small stocks have positive and significant CAPM α’s. Growth 2
stocks and large stocks have negative and significant CAPM α’s. This might suggest a strategy for risk-adjusted excess returns over time. However, it is possible that the CAPM (a single factor index model) does not fully account for all risk factors which affect a returns. A firm that is riskier may have a lower price and therefore a lower P/E multiple. If this is true, the holders of value stocks are just being compensated for incurring the extra risk. The same analysis can be applied to small stocks. Some studies show stock price exhibiting positive and significant autocorrelation with past performance over short to intermediate horizons and negatively and significant autocorrelation over long horizons. Possible explanations are that investors underreact in the short term to new information which causes positive returns to be followed by positive returns and negative returns to be followed by negative returns. In the long term this would cause some stocks to have an incorrectly high relative price (and therefore be a growth stock) and others to have an incorrectly low price (and therefore be a value stock). The subsequent corrections to these incorrect prices would cause the value stock to earn more that the growth stock and be consistent with long-term negative autocorrelation. The short term return persistence (positive autocorrelation) could also be caused b y a bandwagon effect or herding behavior. 8. If prices are as likely to increase as decrease, why do investors earn positive returns from the market on average?
Over the long term, there is an expected upward drift in stock prices based on their risk-adjusted expected rates of return. The expected return over any single day is very small. For example 10% per year is only 10%/365 = 0.0274% per day, so that on any day the price is virtually equally likely to rise or fall since the long-term upward drift component is so small relative to other factors that cause daily returns. However, over longer periods, the small expected daily returns accumulate, and upward moves are indeed more likely than downward ones. We can see this in the annual arithmetic mean returns for different classes of securities covered in Chapter 5 and the discussion of technology implementation and productivity increases in your notes. 9. Which of the following most appears to contradict the proposition that the stock market is weakly efficient? Explain. (a) Over 25% of mutual funds outperform the market on average. (b) Insiders earn abnormal trading profits. (c) Every January, the stock market (at least small stocks) earns abnormal returns.
Twenty-fve percent of mutual funds outperforming the market and insiders earning abnormal returns on private information would not violate the weak form. Therefore only (c), a predictable and easily followed rule to earn abnormal returns would violate the weak form.
3
Note that the stock market does not rise predictable every January. It did for a while – then it was thought that only small firms rose in January… then the whole January Effect stopped. See page 361 in the text or this USA Today article which indicates that the January Effect has migrated to December. Another possible reason for the discovery of this January Effect anomaly is data mining . If market participants look hard enough for anomalies (price or return patterns), they will find them. Computer power and data are cheap. It is easy to look for patterns. It does not mean there is any recurring effect. 11. Suppose that, after conducting an analysis of past stock prices, you come up with the following observations. Which would appear to contradict the weak form of the efficient market hypothesis? Explain. (a) The average rate of return is significantly greater than zero. (b) The correlation between the return during a given week and the return during the following week is zero. (c) One could have made superior returns by buying stock after a 10% rise in price and selling after a 10% fall. (d) One could have made higher-than-average capital gain s by holding stocks with low dividend yields.
Not (a). Compensation for incurring risk is consistent with the EMH. Not (b). Zero autocorrelation is consistent with the EMH. Not (d). The statement is observed higher than average capital gains (not returns) on stocks with low dividend yields. Stocks that don’t pay dividends reinvest profits (retain earnings) and the stock’s price increases. Statement (c) is describing positive autocorrelation (also called serial correlation and is an example of a classic simple filter rule which should not be able to produce superior returns if markets are weak-form efficient. (Note that this is not an observed result.) 12. Which of the following statements are true if the efficient market hypothesis holds? (a) It implies that future events can be forecast with perfect accuracy. (b) It implies that prices reflect all available information. (c) It implies that security prices change for no discernible reason. (d) It implies that prices do not fluctuate.
Only (b) is true. This is the definition of an efficient market. 13. Respond to each of the following comments. (a) If stock prices follow a random walk, then capital markets are little different from a casino.
Stock prices tend to follow a “random walk” in the short run. Long term returns tend to be positive and compensate the hold for the risks incurred form holding the security.
4
(b) A good part of a company's future prospects are predictable. Given this fact, stock prices can't possibly follow a random walk.
Some part of a company’s future prospects may be predictable. Some parts may not be predictable. One way to think about it is that market participants’ prize predictability so securities with claims on the prospects of more predictable companies will hav e higher prices and lower returns. (c) If markets are efficient, you might as well select your portfolio by throwing darts at the stock listings in The Wall Street Journal.
The EMH predicts market participants will earn a return commensurate with risk. This implies that security selection is still important in order fit a portfolio to the investor’s desired risk level. 14. Which of the following would be a viable way to earn abnormally high trading profits if markets are semistrong-form efficient? (a) Buy shares in companies with low P/E ratios. (b) Buy shares in companies with recent above-average price changes. (c) Buy shares in companies with recent below-average price changes. (d) Buy shares in companies for which you have advance knowledge of an improvement in the management team.
If prices reflect all available information (semi-stong form efficient) then abnormally high trading profits (risk-adjusted excess returns) are only available from advanced (inside?) information. 16. “If the business cycle is predictable, and a stock has a posit ive beta, the stock's returns also must be predictable.” Respond.
The statement, as written, is true. The problem is that business cycles are not predictable. 20. In a recent closely contested lawsuit, Apex sued Bpex for patent infringement. The jury came back today with its decision. The rate of return on Apex was r A = 3.1%. The rate of return on Bpex was only r B = 2.5%. The market today responded to very encouraging news about the unemployment rate, and r M = 3%. The historical relationship between returns on these stocks and the market portfolio has been estimated from index model regressions as: Apex: rA = 0.2% + 1.4rM Bpex: rB = -0.1% + 0.6rM On the basis of these data, which company do you think won the lawsuit?
Given market performance, predicted returns on the two stocks would be: Apex: 0.2% + (1.4 x 3%) = 4.4% Bpex: -0.1% + (0.6 x 3%) = 1.7% Apex underperformed and Bpex outperformed so Bpex won the lawsuit. 5
22. Dollar-cost averaging means that you buy equal dollar amounts of a stock every period, for example, $500 per month. The strategy is based on the idea that when the stock price is low, your fixed monthly purchase will buy more shares, and when the price is high, fewer shares. Averaging over time, you will end up buying more shares when the stock is cheaper and fewer when it is relatively expensive. Therefore, by design, you will exhibit good market timing. Evaluate this strategy.
The term “Dollar-Cost Averaging” can refer to two strategies. The first way dollar-cost averaging works is this: Today buy $10,000 of GE at $40 per share (buy 250 shares). Later if the price decreases (or maybe because the price has decreased and now my $40 shares are losers) buy another $10,000 of GE at $20 per share (buy another 500 shares). Total shares = 750 Total cost = $20,000 Average cost per share = $20,000/750 = $26.67 The idea is to “turn around” my bad decision to buy GE at $40 since the “aggregate” GE position will be profitable at a price above $26.67. Implicit in the decision to pay $20 is that the price must go up again, since “what goes down must go up.” This may not be the correct way to think about it. The purchase of the $40 shares and the purchase of the $20 shares are separate investment decision and should be evaluated separately. If you like GE at $20, buy it. That fact that you already own some at $40 is irrelevant. The second way “Dollar-Cost Averaging” works is as described in the text: I want buy $20k of GE so buy $5,000 every month four months. I’ll end up buying more shares when the price is “low” and fewer shares when the price is “high.” Implicit in this strategy is the notion that stock prices “ fluctuate around a normal ” level. Otherwise, there is no meaning to statements such as “when the price is high.” How do we know, for example, whether a price of $25 today will turn out to be viewed as high or low compared to the stock price six months from now? 23. We know that the market should respond positively to good news and that good-news events such as the coming end of a recession can be predicted with at least some accuracy. Why, then, can we not predict that the market will go up as the economy recovers?
The market responds positively to new information. If an eventual recovery is anticipated, then (according to the EMH) the recovery is already reflected in stock prices. Only a better-thanexpected recovery should affect stock prices. 26. Good News, Inc., just announced an increase in its annual earnings, yet its stock price fell. Is there a rational explanation for this phenomenon?
The market may have anticipated an even greater increase in annual earnings. Compared to the expectations, the announcement was a disappointment.
6
Additional Question: What are the three forms of the Efficient Market Hypothesis (EMH) and what should we expect to observe if a form holds (is true) or does not hold?
Strong Form of the EMH: Prices reflect AL L information. If it holds: You can earn risk-adjusted returns abnormal returns (“beat the market”) only if you are lucky. So we should not observe people beating in greater number than would be expected if it were random You cannot earn risk-adjusted returns abnormal returns even with inside information. Prices will not react to announcements.
Semi-Strong Form of the EMH: Prices reflect ALL AVAILABLE or PUBLIC information. If it holds: Prices will react to announcements. You can beat the market using inside (and therefore illegal) information Lucky people will beat the market but not in numbers exceeding what we expect to see if it were random. You cannot earn risk-adjusted abnormal returns even if you work hard and spend money to try to find mispriced securities.
Weak Form of the EMH: Prices reflect only EASILY AVAILABLE or CHEAP information. If it holds: Prices will react to announcements. You can beat the market using inside (and therefore illegal) information Lucky people will beat the market but not in numbers exceeding what we expect to see if it were random. You can beat the market using difficult to obtain information (spending lots of time or money to acquire the information) or superior analysis.
There is good evidence that the strong form is not true (or “does not hold”) since we see prices move when information which was previously known only to insiders is announced. We also observe that insiders can earn risk-adjusted abnormal returns (beat the market) when th ey act on inside (and illegal) information. The EMH only says - within the bounds of increasingly strong assumptions about the information processing of investors - that assets are fairly priced. An implication of this is that - on average - the typical market participant cannot earn excessive profits from a particular trading strategy. However, that does not mean that some individual investors won’t outperform the market over some investment horizon. Certain investors who do well for a period of time get a lot of attention from the financial press, but the scores of investors who do not do well over the same period of time generally get considerably less attention.
7
The question is: Are results of investors like Warren Buffet evidence that markets are not efficient? This is difficult to answer. It may be the case that he is just lucky and markets are efficient. It may be the case that he just works really really hard and is really really smart. So maybe his results are evidence that markets are weak-form efficient, meaning that you cannot earn superior risk-adjusted returns (“beat the market”) looking only at cheap easily-available information, but that markets are not semi-strong efficient, meaning you can beat the market using public information only if you work hard and are smart. CFA PROBLEMS
1.
b.
Semi-strong form efficiency implies that market prices reflect all publicly available information concerning past trading history as well as fundamental aspects of the firm. The full price adjustment should occur just as the news about the dividend becomes publicly available.
2.
a.
3.
d.
If low P/E stocks tend to have positive abnormal returns, this would represent an unexploited profit opportunity that would provide evidence that investors are not using all available information to make profitable investments.
4.
c.
In an efficient market, no securities are consistently overpriced or underpriced. While some securities will turn out after any investment period to have provided positive alphas (i.e., risk-adjusted abnormal returns) and some negative alphas, these past returns are no t predictive of future returns.
5.
c.
A random walk implies that stock price changes are unpredictable, using past price changes or any other data.
6.
d.
A gradual adjustment to fundamental values would allow for the use of strategies based on past price movements in order to generate abnormal profits.
7.
a.
8.
a.
Some empirical evidence that supports the EMH: (i) professional money managers do not typically earn higher returns than comparable risk, passive index strategies; (ii) event studies typically show that stocks respond immediately to the pu blic release of relevant news; (iii) most tests of technical analysis find that it is difficult to identify price trends that can be exploited to earn superior risk-adjusted investment returns.
b.
Some evidence that is difficult to reconcile with the EMH concerns simple portfolio strategies that apparently would have provided high risk-adjusted returns in the past. Some examples of portfolios with attractive historical returns: (i) low P/E stocks; (ii) high book-to-market ratio stocks; (iii) small firms in January; 8
(iv) firms with very poor stock price performance in the last few mon ths. Other evidence concerns post-earnings-announcement stock price drift an d intermediate-term price momentum.
9.
c.
An investor might choose not to index even if markets are efficient because he or she may want to tailor a portfolio to specific tax considerations or to specific risk management issues, for example, the need to hedge (or at least not add to) exposure to a particular source of risk (e.g., industry exposure).
a.
The efficient market hypothesis (EMH) states that a market is efficient if security prices immediately and fully reflect all available relevant information. If the market fully reflects information, the knowledge of that information would not allow an investor to profit from the information because stock prices already incorporate the information. i. The weak form of the EMH asserts that stock prices reflect all the information that can be derived by examining market trading data such as the history of past prices and trading volume. A strong body of evidence supports weak-form efficiency in the major U.S. securities markets. For example, test results suggest that technical trading rules do not produ ce superior returns after adjusting for transaction costs and taxes. ii. The semistrong form states that a firm’s stock price reflects all publicly available information about a firm’s prospects. Examples of publicly available information are company annual reports and investment advisory data. Evidence strongly supports the notion of semistrong efficiency, but occasional studies (e.g., those identifying market anomalies such as the small-firm-in-January or book-to-market effects) and events (such as the stock market crash of October 19, 1987) are inconsistent with this form of market efficiency. However, there is a question concerning the extent to which these “anomalies” result from data mining. iii. The strong form of the EMH holds that current market prices reflect all information (whether publicly available or privately held) that can be relevant to the valuation of the firm. Empirical evidence suggests that strong-form efficiency does not hold. If this form were correct, prices would fully reflect all information. Therefore even insiders could not earn excess returns. But the evidence is that corporate officers do have access to pertinent information long enough before public release to enable them to profit from trading on this information.
b.
i. Technical analysis involves the search for recurrent and predictable patterns in stock prices in order to enhance returns. The EMH implies that technical analysis is without value. If past prices contain no useful information for predicting future prices, there is no point in following any technical trading rule. ii. Fundamental analysis uses earnings and dividend prospects of the firm, expectations of future interest rates, and risk evaluation of the firm to determine proper stock prices. The EMH predicts that most fundamental analysis is doomed to failure. According to semistrong-form efficiency, no investor can earn excess returns from trading rules based on publicly available information. Only analysts with unique insight achieve superior returns. 9
In summary, the EMH holds that the market appears to adjust so quickly to information about both individual stocks and the economy as a whole that no technique of selecting a portfolio using either technical or fundamental analysis can consistently outperform a strategy of simply b uying and holding a diversified portfolio of securities, such as those comprising the popular market indexes.
10.
c.
Portfolio managers have several roles and responsibilities even in perfectly efficient markets. The most important responsibility is to identify the risk/return objectives for a portfolio given the investor’s constraints. In an efficient market, portfolio managers are responsible for tailoring the portfolio to meet the investor’s needs, rather than to beat the market, which requires identifying the client’s return requirements and risk tolerance. Rational portfolio management also requires examining the investor’s constraints, including liquidity, time horizon, laws and regulations, taxes, and unique preferences and circumstances such as age and employment.
a.
The earnings (and dividend) growth rate of growth stocks may be consistently overestimated by investors. Investors may extrapolate recent growth too far into the future an d thereby downplay the inevitable slowdown. At any given time, growth stocks are likely to revert to (lower) mean returns and value stocks are likely to revert to (higher) mean returns, often over an extended future time horizon.
b.
In efficient markets, the current prices of stocks already reflect all known relevant information. In this situation, growth stocks and value stocks provide the same risk-adjusted expected return.
10
