Investments: Analysis

and Behavior

Chapter 6- Efficient Market

Hypothesis

©2008 McGraw-Hill/Irwin
Learning Objectives
 Understand the role of randomness and luck investment
performance.
 Identify the levels of market efficiency.
 Characterize the time series of stock returns.
 Avoid gamblers fallacy and data snooping problems.
 Recognize that price bubbles challenge market
efficiency.

6-2
Short-term Speculation:
Good, or Lucky?
 A coin-flipping contest
6 billion people pay $1 each to join
 Heads you stay in, tails you are out
 After one round, 3 billion are still in
 After ten rounds, about 6 million are still in
 Imagine, flipping 10 heads in a row.
 People begin to believe they are good at flipping, not lucky.
 After 20 rounds, around 6,000 people left
 Locals become heroes!
 But half of these falter in the next round
 After 25 rounds, 180 flippers are remain
 If the game stopped now, each would receive $33.3 million
 These people write books about their technique and strategy

6-3
  After the 30th round, only 6 remain
 Each would get $1 billion of the game stopped
 It would probably take 32 rounds to end with a single winner
 The odds of flipping heads 32 times in a row is roughly one in six
billion.
 Is the winner good a flipping? Lucky?

 Thereare millions of investors, analysts, portfolio

managers, advisors, etc. participating in the investment
process.
 Many will appear to be top performers (at least for the short-
term)
 Who is lucky and who is good? How can you tell?

6-4
Market Efficiency
 The price for any given stock is effectively “fair”
 = the expected net present value of all future profits
 Discounted using a fair risk-adjusted return
 Need
 Large number of buyers and sellers
 Free and readily available information
 Essentially identical securities
 Uninhibited trading
 If there are bargains available, investors would bid
up the price buying those stocks until the stock is no
longer a bargain.
 If markets are efficient, then it would be difficult for an
investor to consistently beat the market.

6-5
Efficient Market Hypothesis
 EMH
 Security prices fully reflect all available information.
 New information arrives in an independent and random fashion
 Current stock prices reflect all relevant risk and return information
 Investors rapidly adjust stock prices to reflect new information
 Levels of Efficiency (based on Information)
 Weak-form: prices reflect all stock market information
 Prices, volume, patterns, trading rules, etc.
 Semistrong-form: current prices reflect all public information
 Accounting statements, economic activity, old news stories
 Strong-form: current prices reflect all public and private info

6-6
Prices do react quickly!
 After the close of the stock market on Monday, April 18, 2005, Texas Instruments
announced that it beat Wall Street's expectations for the first quarter of 2005 and
posted a profit of $411 million, which is 12 percent more than in the year-ago quarter.
TI had been expected to post single-digit profit growth.

6-7
 Are Daily Returns Predictable?
Prior Following
Date Open High Low Close Change % Day Day
A. Dow Jones Industrial Average Big Up Days
24-Jul-02 7698.5 8243.1 7489.5 8191.3 489.0 6.35% -1.06% -0.06%
29-Jul-02 8268.0 8749.1 8268.0 8711.9 447.5 5.41% 0.95% -0.37%
8-Sep-98 7964.9 8103.7 7779.0 8020.8 380.5 4.98% -0.55% -1.94%
16-Mar-00 10139.6 10716.2 10139.6 10630.6 499.2 4.93% 3.26% -0.33%
15-Oct-02 7883.2 8304.6 7883.2 8255.7 151.4 4.80% 0.35% -2.66%

28-Oct-97 7190.9 7553.6 6933.0 7498.3 337.1 4.71% -7.18% 0.11%

1-Oct-02 7593.0 7964.2 7558.4 7938.8 346.9 4.57% -1.42% -2.31%

24-Sep-01 8242.3 8733.4 8242.3 8603.9 368.1 4.47% -1.68% 0.65%

5-Apr-01 9527.2 9969.9 9527.2 9918.1 402.6 4.23% 0.31% -1.28%
11-Oct-02 7540.7 7919.6 7540.7 7850.3 316.3 4.20% 3.40% 0.35%
Averages 4.86% -0.36% -0.78%

6-8
 Prior Following
Date Open High Low Close Change % Day Day
B. Dow Jones Industrial Average Big Down Days
27-Oct-97 7608.3 7717.4 7150.1 7161.2 -554.2 -7.18% -1.69% 4.71%
17-Sep-01 9294.6 9294.6 8755.5 8920.7 -684.8 -7.13% 0.00% -0.19%
31-Aug-98 8079.0 8149.0 7517.7 7539.1 -512.6 -6.37% -1.40% 3.82%
14-Apr-00 10922.9 10922.9 10173.9 10305.8 -617.8 -5.66% -1.81% 2.69%
19-Jul-02 8356.7 8356.7 7940.8 8019.3 -390.2 -4.64% -1.56% -2.93%
20-Sep-01 8375.7 8711.4 8304.5 8376.2 -382.8 -4.37% -1.62% -1.68%
27-Aug-98 8377.9 8448.7 8062.2 8166.0 -357.4 -4.19% -0.92% -1.40%
12-Mar-01 10638.5 10638.6 10138.9 10208.3 -436.4 -4.10% -1.97% 0.81%
3-Sep-02 8659.3 8659.3 8282.9 8308.1 -355.4 -4.10% -0.09% 1.41%
27-Sep-00 7996.0 7997.1 7664.9 7701.5 -295.7 -3.70% 1.98% -1.42%
-5.14% -0.91% 0.58%
Note: Data are for the nine-year period from 7/1/97 to 12/31/05.

6-9
 Figure 6.3 Daily Returns Are Noisy and Random Around A Mean of Zero,
from 1/1/97 to 12/31/05.

A. DJIA Daily Returns

15.00%

10.00%

5.00%
C. Nasdaq Daily Returns
0.00%
15.00%

-5.00%
10.00%
-10.00%
5.00%
Date
0.00%

-5.00%

-10.00%
Date

6-10
Random Walks and Prediction

 Random Walk Theory

 Stock prices movements do not follow any patterns
or trends
 Past price action cannot be used to predict future price
movements.
 Subsequent price changes represent arbitrary departures
from previous prices.
 Random walk with a drift
 Stock prices do tend to increase, on average, over time.

6-11
Yet, millions of people examine charts and search for
patterns and trends.

 The Human Brain is well suited to seeing patterns

 Even when the data is random and no pattern exists!
 Gambler’s fallacy
 Popular, but erroneous, belief that some self-correcting process
impacts random events.
 After 5 fair coin flips of “heads” in a row, many people act as if they
believe the probably of a heads in the next flip is different from 50%
 Lottery players examine previous numbers picked. So do Keno players.
 Data-Snooping
 Patterns may appear in random data if enough different tests are
examined.
 Much data is available and computers can quickly crunch it.
 Back testing and out of sample tests can help determine whether a
pattern may repeat in the future or is simply an artifact of the data.
6-12
 From Dogs to Fools
 Dogs of the Dow
 Strategy identifies the 10 highest-dividend paying firms in the DJIA
 1992 book, Beating the Dow by O’Higgins and Downes
 Buying the Dogs at the beginning of each year is shown to beat the
buy-and-hold strategy of just owning the 30 stocks in the Dow by over
4% per year.
 Value-oriented strategy
 Dow Five
 Buy the 5 lowest priced dogs
 Beats the Dow by 8% per year
 No basis in theory

 Motley Fools’ adaptation

 Foolish Four: Of the Dow Five, throw-out the lowest price
stock and double-up on the second lowest price stock.
 Purported to beat the Dow by 12% per year
 Is the data tortured enough?
 Motley Fool eventually abandoned this strategy when it both
failed to perform well and other investors started trading against
those following the strategy 6-13
Investing in an Efficient Market
 Two investors are walking down the street when one
spots a $100 bill on the sidewalk. He points it out. The
companion says, “It must not be a real $100 bill or
someone would have already picked it up.”
 If markets are efficient, we should not bother to look for
bargains. However, if no one is looking for bargains, how can
markets be efficient?

 Studies show that, on average, mutual funds and

investment newsletter writers do not beat their
benchmark.
 Evidence for EMH

 What should investors do if markets are efficient?

6-14
Challenges to EMH
 Excessive Volatility
 Why is the market so volatile?
 Dividends are not volatile at all.

 Bubbles
 Dramatic increases and declines in the
stock market

6-15
 Figure 6.4 Real Stock Prices and Present Values of
Subsequent Real Dividends
Source: Robert Shiller, 2003, “From Efficient Market Theory to Behavioral Finance,” Journal
of Economic Perspectives, 17(1), Figure 1.

10000

Real Stock Price (S&P 500)

1000
PDV, Interest

PDV, Constant Discount

100 PDV,

10
1860 1880 1900 1920 1940 1960 1980 2000 2020

Year
6-16
 From January 2, 1985, at 11,543, the Nikkei 225 soared
to a closing high of 38,916 on December 29, 1989.
 This represents a gain of 237.1% in the Nikkei over a 5-year
period, and a stunning 27.5% compound annual rate of return.
 Then, the bubble burst and the bottom fell out of the Japanese
equity market. Fifteen years after the Japanese market peak, in
December, 2004, the Nikkei stood at 10,796. That’s 72.3%
below the December, 1989 peak.

 From a (split-adjusted) level of 125 on January 31, 1985,

the Nasdaq 100 soared to 4,816.35 on March 24, 2000.
 This represents a 15 1/4-year return of 3,753.3%, and an
amazing compound return of 27.1% per year.
 Then the Nasdaq plunged, losing over 80% of its value by 2002.

6-17
 Figure 6.6 Will Post-crash Nasdaq 100 Valuations Languish For A Decade or
More?

5,000 50,000
4,500 45,000
4,000 40,000
3,500 35,000
3,000 30,000
Nasdaq 100

Nikkei 225
Nasdaq 100
2,500 25,000
Nikkei 225
2,000 20,000
1,500 15,000
1,000 10,000
500 5,000
0 0

Year

6-18
Why might the market not be efficient?
 Investors make decisions influenced by
emotions and psychological biases.
 If large groups of investors become too
optimistic or pessimistic, they may move
prices
 Investor mood
 How investor mood can impact
expectations…

6-19
 If prices reflect a dividend discount model:
PV = D1 / (k – g)
 But expectations become biased:
E(P) = D1 / (k – E(g))
 So prices can deviate from true value by:

E ( P) k g

PV k  E(g)

6-20
 Example:
 If the annual market return over a long period of time is
expected to be 12% and the long-term expected growth
rate of stock market firms is 4%, the Dow Jones
Industrial Average is fairly valued at 10,000. If Optimistic
investors believe the long-term growth rate is 5%, how
far would the DJIA be expected to fall?
 The stock market should become over-valued as
E ( P) k g 12  4 8
   1.14
PV k  E ( g ) 12  5 7
 The DJIA would be expected to rise to 1.14 × 10,000 =
11,400, or a 14% rise. However, it would also be 14%
overvalued.

6-21
Can investment fraud occur in an efficient market?
 Microcap Fraud
 Microcap stocks, or penny stocks,
often trade as pink sheets on the
OTC Bulletin Board
 Low liquidity means the price can be
susceptible to false press releases of
exaggerations or lies, and “pump and
dump” schemes.

 Fraud on the Internet

 Easy place to spread false “news”
and “unbiased” opinion.

6-22
Red Flags!

 Assets Are Large But Revenues Are Small

 Unusual Accounting Issues
 Thin Public Float
 SEC Trading Suspensions
 High Pressure Sales Tactics

6-23
EMH
 The EMH is still hotly debated today.
 In Support:
 Short-term prices are unpredictable
 Price adjust quickly and pretty accurately
 Professional investors don’t seem to beat the market, on
average.
 Against:
 Market is too volatile
 Stock market bubbles exist
 Investor mood may drive prices away from fair value
 Investment fraud
 The next chapter examine some interesting
anomalies that also put the validity of the EMH into
question
6-24