Chapter 9

EFFICIENT MARKET HYPOTHESIS

The Collective Wisdom

Outline• Random Walk

• What is an Efficient Market

• Empirical Evidence on Weak-form Efficient Market Hypothesis

• Empirical Evidence on Semi-strong Form Efficient Market Hypothesis

• Empirical Evidence on Strong-form Efficient Market Hypothesis

• What is the Verdict

• Implications for Investment Analysis

Random Walk

• Maurice Kendall found that stock prices followed a random

walk, implying that successive price changes are

independent of one another.

• A number of researchers have employed ingenious methods to

test the randomness of stock price behaviour.

• Academic researchers concluded that the randomness of stock

prices was the result of an efficient market.

What Is An Efficient Market

• An efficient market is one in which the market price of a security is an unbiased estimate of its intrinsic value

• Market efficiency is defined in relation to information that is reflected in security prices. Fama distinguishes three

levels of market efficiency.

• Weak-form efficiency

• Semi-strong form efficiency

• Strong-form efficiency

Stock Market Efficiency

Strong

Semi-strong

Weak

Misconceptions

1. Emh.. Implies… market has perfect forecasting.2. As prices tend to fluctuate they cannot reflect fair value.3. Inability of institutional portfolio managers to achieve superior investment performance implies that they lack competence.4. The random movement of stock prices suggests that the stock market is irrational.

Foundations of Market Efficiency

Any one of the following three conditions leads to market efficiency

• Investor Rationality

• Independent Deviations from Rationality

• Effective Arbitrage

Profound Impact of a Simple Idea

•Receptive climate for three seminal developments in financial

theory: (a) MM theories, (b) CAPM, (c) OPM

•Important practical developments like indexing, asset

securitisations, and performance measurement

•Growing respect for free markets

Empirical Evidence On Weak-Form Efficient

Market Hypothesis

• Serial correlation test

• Runs test

• Filter rules test

Returns over Short Horizons

Empirical studies done in 1950s, 1960s, and 1970s using daily or monthly returns of individual securities overwhelmingly confirmed the unpredictability of stock returns. So much so that Michael Jensen asserted in 1978 that “The efficient market hypothesis is the best – established fact in all of social science.” Recent research, however, has cast a shadow over the random walk hypothesis. For example, Andrew Lo and Gaig Mackinlay found that : (a) Weekly portfolio returns are strongly positively correlated but the returns to individual securities generally are not .(b) While the daily and weekly returns are strongly positively correlated, the serial dependence for monthly, quarterly, or yearly returns is virtually zero.

Returns over Long Horizons

While returns over short horizons are characterised by minor positive serial correlation, returns over long horizons seem to be characterised by pronounced negative serial correlation. To explain the latter result, a “fads hypothesis” has been proposed, which says that stock prices tend to overreact to news. Such overreaction results in positive correlation over short horizons and negative correlation over long horizons. Put differently, prices overshoot in the short run but correct in the long run. Thus, market prices display excessive volatility in comparison with intrinsic value. Robert J. Shiller and others argue that the presence of mean reversion and excessive volatility in prices imply market inefficiency. Fama and others dispute it. They contend that changing market risk premiums over time, and not market inefficiency, result in mean reversion and excessive volatility.

Empirical Evidence on Semistrong - Form efficient Market Hypothesis

To test this form of market efficiency, specific investment

strategies are examined to see whether they earn excess return. Since excess return represents the difference between the actual return and the expected return, implicit in a test of market efficiency is some model of the expected return. The expected return may be based on the capital asset pricing model or the arbitrage pricing model or some other model. Hence, a test of market efficiency is really a joint test of market efficiency and the model for expected return. If there is evidence of excess return, it may mean that the market is inefficient or that the model used to calculate the expected return is wrong or both. This seems to be an insoluble problem. However, if the results of a study do not vary with different models of expected return, one can argue that the results can be attributed to market inefficiency and not model misspecification. 

Empirical Evidence On Semi-Strong Efficient

Market Hypothesis

• Possible to earn superior risk-adjusted return by trading

on information events? (Event study)

• Possible to earn superior risk-adjusted return …by trading

on an observable characteristic of a firm? (Portfolio study)

Event Study1. Identify the announcement date of the event

Announcement date

2. Collect returns data around the announcement date Rj -n Rj,O Rj, +n

-n TO +n

3. Calculate the excess returnERj t = Rj t - BETAj x Rmt

4. Compute the average and the standard error of excess returns across all firms

5. Assess whether the excess returns around the announcement date are different from zero

Average excess return T Statistic for excess return on day t = Standard error

Results of Event Studies

• Stock Splits

• Announcement of Accounting Changes

• Corporate Events

• Unexpected World Events and Economic News

Results of Portfolio Studies

• Price-Earnings Ratios

• PEG Ratios

• Size Effect

• BV/MV Ratios

Gradual Adjustments To Earnings Announcements

Portfolio Study1. Define the variable (characteristic) on which firms will be

classified

2. Classify firms into portfolios based upon the magnitude of the variable

3. Compute the returns for each portfolio

4. Calculate the excess returns for each portfolioE Rji = Rji - BETAj x RMt

5. Assess whether the average excess returns are different across the portfolios

Returns By P - E Multiple Class

18.00%

16.00%

14.00%

12.00%

10.00%

8.00%

6.00%

4.00%

2.00%

0.00% LOWEST 2 3 4 5 6 7 8 9

HIGHEST

Time Series Analysis

The time series analysis starts with the assumption that in an

efficient market long-term historical rates of return represent the

best estimate of future rates of return. The studies of this genre

determine whether any public information will provide superior

estimates of returns over different time horizons, short-term or long

term.

 

Results of Time Series Analysis

•A Positive relationship exists between dividend yield and future long-run stock market returns.

•Stock prices adjust gradually, not rapidly, to announcements of unanticipated changes in quarterly earnings.

•Stock returns tend to be negative over, the period from Friday’s close to Monday’s opening.

•Stock prices seem to rise more in January than in any other month of the year.

•The volatility of stock prices is too large to be justified by the volatility of dividends.

•Market interest rates move in a normal range.

Empirical Evidence On Strong-Form

Efficient Market Hypothesis

Empirical evidence broadly suggests the following:

• Corporate insiders earn superior returns, after adjustment for risk.

• Mutual fund managers, on average, do not earn superior returns after adjustment for risk.

Other Evidence

• Price overreactions

• Calendar anomalies

• Excess volatility

• Normal range of interest

The Crash Of 1987October 19, 1987 DJIA 23%

P ≃ IV Appears less appealing

Difficulty . . valuing equities …

3100 =

0.16 - 0.13

375 =

0.16 - 0.12

Difficulty in valuing equity stocks … two implications :

1. Investors typically price an equity stock in relative terms

2. Almost impossible . . test the hypothesis … p ≃ iv

Absolute efficiencyvs

relative efficiency

Empirical Evidence on Strong-Form efficient Market Hypothesis

The strong-form efficient market hypothesis holds that all available

information, public as well as private, is reflected in the stock prices.

To test this hypothesis, researchers have analysed the returns earned

by certain groups who have access to information which is not

publicly available or ostensibly possess greater resources and

expertise to intensively analyse information which is in the public

domain. More specifically, researchers have tested this form of EMH

by analysing the performance of corporate insiders, stock exchange

specialists, security analysts, and professional money managers.

 

Empirical Evidence on Strong-Form efficient Market Hypothesis

• Corporate insiders consistently enjoy superiors risk-adjusted returns.

• Stock exchange specialists (or market makers)seem to earn superior risk-adjusted returns.

• While there are some superior security analysts, the majority of them seem to make recommendations that are not helpful

• Professional money managers do not, on average, earn a superior risk-adjusted return.

Semi-efficient Market Hypothesis

The EMH has a cousin, the semi-efficient market hypothesis

(SEMH). SEMH holds that some stocks are priced more efficiently

than others extending this idea, one may argue that the market

perhaps has several tiers. Put differently, there is a pecking order of

efficiency.

Micro Efficiency and Macro Inefficiency

Paul Samuelson has argued that modern markets show considerable micro

efficiency because the minority that spots deviations from micro efficiency

can make money by exploiting those deviations and, in doing so, they

eliminate persisting inefficiencies. In contrast, Paul Samuelson

hypothesised that markets display considerable macro inefficiency in the

sense that aggregate indexes of security prices remain below or above

various definitions of fundamental values for long periods of time. There

seems to be substantial evidence in support of Samuelson’s dictum where

inefficiency is defined as predictability of future (excess) returns.

Verdict

• True, the efficient market hypothesis, like all theories, is an imperfect and limited description of the stock market. However, here does not, at least for the present, seem

to be a better alternative.

• Merton miller : “it is closer to being a ‘paradigm’ than a mere hypothesis, bringing a common and coherent explanatory framework to a wide variety of seemingly unrelated phenomena. Like all scientific paradigms, it will survive until displaced by a better one. At the moment, at least no better paradigm is in sight”

Data Mining

The advocates of efficient market hypothesis argue that it is not surprising

that several anomalies and puzzles have been discovered. When data is

mined extensively, one is bound to find a number of patterns. As Bradford

Cornell put it: “Even a set of random numbers generated by a computer

will appear to have some pattern after the fact. Those patterns, however,

are spurious and will not be replicated in another generation of random

numbers. Many scholars feel that the same is true of stock prices.” William

Sharpe puts it more vividly: “If you torture the data long enough it will

confess to any crime.” As an extreme example of data mining, Leinweber,

who sifted through United Nations CD-Rom, found that historically the

best predictor of the S&P 500 was butter production in Bangladesh!

Difficulty in Exploiting

Even if inefficiencies exist, it is difficult to take advantage of them.

As Richard Roll says: “Over the past decade, I have attempted to

exploit many of the seemingly most promising ‘inefficiencies’ by

actually trading significant amounts of money according to a trading

rule suggested by the ‘inefficiencies’ … I have never yet found one

that worked in practice, in the sense that it returned more after cost

than a buy-and-hold strategy.”

  

 

Impossibility of Informationally Efficient Markets

 

In a classic paper published in 1980, Grossman and Stieglitz pointed

toward the “impossibility of informationally efficient markets.”

Their argument is fairly straightforward and goes as follows: If

market prices reflect all information about stocks, no one would do

equity research (as it involves cost) and everyone will simply accept

market prices as the best estimates of intrinsic values. And if no one

does equity research to obtain and analyse information about

companies, how can market prices reflect all information about

stocks?

Bottom Line on Market Efficiency

What is the bottom line on market efficiency? Jay Ritter suggests that it is useful to classify events into two categories, high – frequency events and low-frequency events. High-frequency events occur often and the market is efficient with respect to them. That is why it is hard to identify a trading strategy which is reliably profitable and mutual funds have difficulty outperforming their benchmarks. Low frequency events occur infrequently and the market seems to be inefficient with respect to them. Here are some examples of massive mispricing: • The stock price and land price bubble of Japan in the 1980s

• The stock market crash of October 1987.

• The TMT (technology, media, and telecom) bubble of 1999-2000.

Implications For Investments• Substantial evidence in favour of randomness suggests that technical analysis is of dubious value.

• Routine and conventional fundamental analysis is not of much help in identifying profitable courses of action

• The key levers for earning superior rates of returns are:

• Early action on any new development.

• Sensitivity to market imperfections and anomalies.

• Use of original, unconventional, and innovative modes of analysis.

• Access to inside information and its sensible interpretation

• An independent judgment that is not affected by market psychology.

Summing Up

• Stock prices appear to follow a random walk. The

randomness of stock prices is the result of an efficient market

• It is useful to distinguish three levels of market efficiency :

weak form efficiency, semi-strong form efficiency, and

strong form efficiency.

• The weak form efficient market hypothesis says that the

current price of a stock reflects all information found in

the record of past prices and volumes.

• The semi-strong form efficient market hypothesis holds

that stock prices adjust rapidly to all available public

information.

• The strong form efficient market hypothesis holds that all available information, public and private is reflected in

stock prices.

• Empirical evidence seems to provide strong support for weak- form efficiency, mixed support for semi-strong form efficiency, and weak support for strong-form efficiency.

• The efficient market hypothesis is an imperfect and limited description of the stock market. however, at least for the present, there does not seem to be a better alternative.

• The key implications of the efficient market hypothesis are that technical analysis is of dubious value and routine fundamental analysis is not of much help.