NPTEL Course

Course Title: Security Analysis and Portfolio Management

Course Coordinator: Dr. Jitendra Mahakud

Module-4
Session-8
Testing Market efficiency
In this section we will look forward for specific statistical test that are used to examine
whether the data support for the acceptance of certain kind of market efficiency in a
particular market. The empirical literature available in this area suggests that, the
degree of market efficiency varies from one market to another market depending up on
the market structure and informational efficiency in a particular market.

8.1.Tests of Weak-Form Efficient Market Hypothesis (EMH)

Academicians and researchers have formulated two groups of tests of the weak form of
EMH. The first category involves statistical tests of independence between rates of
return and second entails a comparison of risk-return results from trading rules that
make investment decisions based on the past market information relative to the results
from buy and hold policy1. The Statistical Tests used to test the weak form EMH are (i)
Autocorrelation test and (ii) Run Test.
(i) Autocorrelation test:

Autocorrelation tests of independence measure the

significance of positive or negative correlation in returns over time. Does the rate of
return on day t correlate with the rate of return on day t 1, t 2, or t 3? Those who
believe that capital markets are efficient would expect insignificant correlations for all
such combinations.
Example:
Consider the following closing prices of ABC Limited during the month of July 2004
and July 2007: Do they follow a random path? Check their independence employing
auto-correlation test. It has been assumed that the investor you buy stock at the
beginning of the test period and hold it to the end.
Trading Days July 2007
3
4
5

Closing price
2016.85
2040.85
2058.00

Trading Days July 2004

2
3
6

Closing price
785
780
770

6
7
10
11
12
13
14
17
18
19
20
21
24
25
26
27
28
31

2042.10
2026.50
2092.10
2170.00
2270
2327
2285
2240
2213
2207
2225.90
2254.75
2290
2330
2315.5
2284.75
2241
2295.8

8
9
10
13
14
15
16
17
20
21
22
23
24
27
28
30
31
---

795
770
780
775
780
780
770
761.75
764.75
764.25
771.25
774.75
764
768
774
775.75
781.5
---

Solution:
Price Change
in (X)
-5
-10
25
-25
10
-5
5
0
-10
-5.25
3
-0.5
7
3.5
-10.75
4
6
1.75
5. 75
X= -3.5

Price Change
in (Y)
24
17.5
-15.9
50
77.9
100
57
-42
-45
-27
-6
18.9
28.85
35.25
40
-14.5
-30.75
-43.75
54.8
Y= 278.95

X2

Y2

XY

25
100
625
625
100
25
25
0
100
68.0625
9
0.25
49
12.25
115.56
16
36
3.06
33.06
X2=1967.25

576
294.1225
252.81
2500
6068.41
10000
3249
1764
2025
729
36
357.21
832.32
1242.56
1600
210.25
945.56
1914.06
3003.04
Y2=37599.35

-120
-175.5
-397.5
-1250
779
-500
285
0
450
222.75
-18
-9.45
201.95
123.375
-430
-58
-184.5
-76.56
315.1
XY=-838.33

Mean of X =(X bar) = -3.5/ 19 = -0.184, (X bar)2 = 0.034, Y (bar) = 278.95/ 19 =14.68,
(Y bar)2 = 215.55

Using the formula given above:

-838.34 19 (0.184) (14.68)
b= ------------------------------------ = -787.02 / 1966.604 = -0.4002
1967.25 19 (0.034)
a = 14.68 (-0.4002) (-0.184) = 14.61
Putting the values of a and b we can calculate the r as follows:
14.61 (278.95) + (-0.04002) (-838.34) -19 (215.5
r2 =

-------------------------------------------------------------- = 0.0094
37599.35 19 (215.55)

r = 0.097
Since the correlation between the prices in two different periods is small we can
conclude that the price move in a random fashion and the weak form of market
efficiency is valid in this context.
(ii) Run Test: the runs test considers a series of price changes over a certain period of
time (suppose from t1 t10 ) and each price change is either designated a plus (+) if it is
an increase in price or a minus () if it is a decrease in price. For instance the result is a
set of pluses and minuses as follows: t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 . A run exist
when two consecutive changes are the same (i.e., ++ or --). When the price changes in a
different direction, such as +- or -+ the run ends and a new run may begin. To test for
independence, the number of runs for a given series of price changes compared with
the number in a table of expected values for the number of runs that should occur in a
random series.
To test the independence of the prices, we require:
Total Number of Runs:

Number of Positive Price Changes from t1 t10 :

n1

Number of Negative Price Changes from t1 t10 :

n2

Once we have the data the mean and standard deviation of the mean are calculated by
using the formula given below:
2 n1 n2
r = ------------- + 1,
n1 + n2

2 n1 n2 (2 n1 n2 - n1- n2)
r =

--------------------------(n1 + n2)2 (n1 + n2 -1)

]1/2

At a given level of significance, we calculate the upper and lower limits and check
whether the number of runs observed from the test falls within the limits or not. If it is
between the limits we conclude that the prices are random or independent of each other,
otherwise not.
Illustration 5.2: (all the data are continuous)
Trading
days
April2007
1
2
3
4
5
8
9
10
11
12
15
16
17
18
19
22
23
24
25
26
29
30

Stock
price
881.5
856.5
859.75
918.5
1010.25
1072.25
1074.5
1123.5
1203.5
1256.25
1341.5
1469.75
1450.25
1306.5
1258
1332.75
1315.75
1351
1407
1547.75
1521
1484.25

Trading
days in
May2007
1
2
3
6
7
8
9
10
13
14
15
16
17
20
21
22
23
24
27
28
29
30
31

Stock
price
1458
1425
1350.75
1447.5
1448.25
1505.25
1511.25
1588.25
1560
1576
1638.5
1648.5
1710.75
1703
1665
1665.75
1550.75
1508
1460
1316.25
1229
1324.75
1205.75

Trading
days in
June2007
3
4
5
6
7
10
11
12
13
14
17
18
19
20
21
24
25
26
27
28

Stock
price

Trading
days in
July2007
1
2
3
4
5
8
9
10
11
12
15
16
17
18
19
22
23

1295.5
1292.25
1191.5
1179.5
1279
1280.25
1247.5
1186.5
1207.75
1323
1366.75
1312
1412.75
1420.5
1375.5
1377.5
1304.5
1395
1498.75
1549.75

Stock price
1520
1489.5
1397.75
1425
1487
1536.5
1500.5
1481.25
1437.25
1460.75
1385.5
1401.25
1381.25
1487.25
1471.25
1452.75
1442.5

Solution:
Days

Price

1
2
3
4
5
8
9
10
11
12
15
16
17

881.5
856.5
859.75
918.5
1010.25
1072.25
1074.5
1123.5
1203.5
1256.25
1341.5
1469.75
1450.25

PC

+
+
+
+
+
+
+
+
+
+
-

Run

Days

Price

PC

1
2
3
6
7
8
9
10
13
14
15
16
17

1458
1425
1350.75
1447.5
1448.25
1505.25
1511.25
1588.25
1560
1576
1638.5
1648.5
1710.75

+
+
+
+
+
+
+
+
+

Run

10

Days

Price

PC

Run

Days

Price

PC

3
4
5
6
7
10
11
12
13
14
17
18
19

1295.5
1292.25
1191.5
1179.5
1279
1280.25
1247.5
1186.5
1207.75
1323
1366.75
1312
1412.75

+
+
+
+
+
+
+

16

1
2
3
4
5
8
9
10
11
12
15
16
17

1520
1489.5
1397.75
1425
1487
1536.5
1500.5
1481.25
1437.25
1460.75
1385.5
1401.25
1381.25

+
+
+
+
+
-

17

18
19
20
21

Run

27
28

29
30
31
32
33

18
19
22
23
24
25
26
29
30

1306.5
1258
1332.75
1315.75
1351
1407
1547.75
1521

+
+
+
+
-

20
21
4
22
5
23
24
6
27
28
29
Cont.
30
31
Note: PC= Price Change

1703
1665
1665.75
1550.75
1508
1460
1316.25
1229
1324.75
1205.75

+
+
-

20
21
24
25
26
27
28

11
12

13

1420.5
1375.5
1377.5
1304.5
1395
1498.75
1549.75

+
+
+
+
+

22
23
24
25

1487.25
1471.25
1452.75
1442.5

18
19
22
23

+
-

26

14
15

Total Number of Runs (r) = 35

Number of Positive Price Changes (n1) = 43
Number of Negative Price Changes (n2) = 38
2 n1 n2
r = ------------- + 1,
n1 + n 2
2 n1 n2 (2 n1 n2 - n1- n2)
r =

1/2

2 (43) (38)
---------------- = 41.34
43 + 38

2 (43) (38) [ 2(43) (38) -43-38]

]1/2

--------------------------= ------------------------------------= 4.45

2
2
(43 + 38) (43 + 38 -1)
(n1 + n2) (n1 + n2 -1)
Assuming that the distribution follows the normal distribution if we test at 5 percent
significance level then Z = 1.96.
Lower Limit = 41.34 1.96 (4.45) = 32.62
Upper Limit = 41.34 + 1.96 (4.45) = 50.06
Since the observed number of runs (35) falls within the lower and upper limit, we can
conclude that the prices are independent at 5 percent level of significance.

8.1.1. Tests of Trading Rules

Mechanical trading systems, such as filter rules and moving averages, provide a more
sensitive test for randomness, because they do not depend on the pattern or cause of the
price changes. Advocates of efficient market hypothesized that investors could not
derive above normal profits above a buy-and hold policy using any trading rule that
depend solely on past market information. More trading in a security should promote
the market efficiency. Most popular trading technique is Filter Rule i.e. an investor
trades a stock when the price change exceeds a filter value set for it (Buy-and-Hold
Strategy). If the behaviour of the stock price changes is random, filter rules should not
outperform a simple buy-and-hold strategy.

34
35

8.2. Tests of the Semi Strong Form of Market Efficiency

The two more popular methods in this regard are:

Event studies: That examine how fast stock prices adjust to specific significant
economic events.

Portfolio Study or Return Prediction Study: That examines the possibility of

earning superior risk-adjusted returns by trading on an observable characteristic
of a firm like P/E ratio, dividend yield etc.

8.2.1. Event Study

Steps involved with this process:

Identify the event to be studied and pin point the date on which the event was
announced

Collect returns data around the announcement date

Calculate the excess returns by period around the announcement date for each
firm in the sample

Compute the average excess returns across all firms

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

Results of Available Earlier Event Studies:

Stock split studies show that splits do not result in abnormal gains after the split
announcement, but before

Initial public offerings seems to be under priced by almost 18%, but that varies
over time, and the price is adjusted within one day after the offering

Announcements of accounting changes are quickly adjusted for and do not seem
to provide opportunities

Stock prices rapidly adjust to corporate events such as mergers and offerings

Example:
Two companies, Cargo Mutas Ltd. and Telco Farma Ltd., have increased their level of
cash dividends for the year 2010-2011. A financial analyst at a mutual fund wanted to
test the consistency of the semi-strong form of market efficiency. He calculated the
characteristic lines for a period of six years on a monthly basis up to four months before

the announcement took place. The relationship between the returns on these four
companies and the market are given by:
Cargo Mutas Ltd.

Telco Farma Ltd.

R C a rg o t 1 .7 0 % 1 .0 5 R m t

RTelcot 1.53% 1.08Rmt

Periods

Actual Return

Market Return

RC arg ot

RTelcot

Rmt

13.47

13.61

11.00

11.90

12.04

10.00

13.50

13.67

11.15

12.97

13.12

10.88

13.43

13.60

10.90

12.50

12.62

10.05

13.09

13.25

11.05

14.51

14.71

12.15

14.80

14.99

12.67

Solution:

Periods
4
3
2
1
0
1
2
3
4

Actual Return
13.47
11.90
13.50
12.97
13.43
12.50
13.09
14.51
14.80

Periods
4
3
2
1
0
1
2
3
4

Actual Return
13.61
12.04
13.67
13.12
13.60
12.62
13.25
14.71
14.99

Cargo Mutas Ltd.

Market Return
Expected Return
11.00
13.25
10.00
12.20
11.15
13.40
10.88
13.12
10.90
13.15
10.05
12.25
11.05
13.30
12.15
14.45
12.67
15.00
Telco Farma Ltd.
Market Return
Expected Return
11.00
13.41
10.00
12.33
11.15
13.57
10.88
13.28
10.90
13.30
10.05
12.38
11.05
13.46
12.15
14.65
12.67
15.21

Abnormal Return
0.22
-0.30
0.10
-0.15
0.28
0.25
-0.21
0.06
-0.20
Abnormal Return
0.20
-0.29
0.10
-0.16
0.30
0.24
-0.21
0.06
-0.22

Expected return is calculated by substituting market returns in the characteristic line

equation. For the market return of 10.88%, the expected return is given by:

RC arg ot 1.70% 1.05Rmt 1.70% 1.05(10.88) 13.12%

The abnormal return is given by the difference between the actual return and the
expected return. Now, we have to calculate the average abnormal returns (AAR) for
each of the week before and after the dividend were announced:
Period

Calculation of Average Abnormal Returns (AAR)

(1/4) ( 0.22+0.20)

(1/2) (-0.30-0.29)

(1/2)(0.10+0.10)

(1/2)(-0.15-0.16)

(1/2)(0.28+0.30)

(1/2)(0.25+0.24)

(1/2)(-0.21-0.21)

(1/2)(0.06+0.06)

(1/2)(-0.20-0.20)

AAR
0.21
-0.295
0.1
-0.155
0.29
0.245
-0.21
0.06
-0.21

The cumulative average abnormal returns (CAAR)

= (0.21-0.295+0.1+-0.155+0.29+0.245-0.21+0.06-0.21) = 0.003889
Since as per our analysis the value of CAAR is close to zero, we can conclude that
markets are efficient in the semi strong form.

8.2.2. Portfolio Study or Return Prediction Study

This method follows the following steps:

Define the variable (characteristic) on which firms will be classified

Classify firms into portfolios based upon the magnitude of the variable

Compute the returns for each portfolio

Calculate the excess returns for each portfolio

Assess whether the average excess returns are different across the portfolios

However there are two implications for the results of abnormal return in the
portfolio study: (i) abnormal returns could occur because either markets are
inefficient, or (ii) market model is not properly specified and provides incorrect
estimates of risk and expected returns. In other words it tests the joint hypothesis of
market efficiency and the testing of equilibrium asset pricing models.

Results of prior Portfolio Study in the context of several developed and emerging
markets:

Low P/E stocks experienced superior risk-adjusted results relative to the market,
whereas high P/E stocks had significantly inferior risk-adjusted results. Publicly
available P/E ratios possess valuable information regarding future returns. This
is inconsistent with semi strong efficiency.

Studies have hypothesized an inverse relationship between the PriceEarnings/Growth Rate (PEG) ratio and subsequent rates of return. This is
inconsistent with the EMH.

Several studies have examined the impact of size on the risk-adjusted rates of
return. The studies indicate that risk-adjusted returns for extended periods
indicate that the small firms consistently experienced significantly larger riskadjusted returns than large firms. Firm size is a major efficient market anomaly
and a serious challenge to the market efficiency hypothesis.

Ratios of Book Value of a firms Equity to Market Value of its equity (BV/MV)
have shown a significant positive relationship with future stock returns. This
results are also inconsistent with the EMH

In total along with the findings of several market anomalies like sz,
BV/MV,PE,PEG studies on predicting rates of return for a cross-section of
stocks indicates markets are not semi strong efficient.

8.3. Tests of the Strong-Form Efficient Market Hypothesis

Strong-form EMH contends that stock prices fully reflect all information, both public
and private. This implies that no group of investors has access to private information
that will allow them to consistently earn above-average profits. The tests are based on
the group of investors.
Testing Groups of Investors:

Corporate insiders: Insiders include major corporate officers, directors, and

owners of 10% or more of any equity class of securities.

Stock exchange specialists: Specialists have monopolistic access to information

about unfilled limit orders

Security analysts: There is evidence in favour of existence of superior analysts

who apparently possess private information

Professional money managers: Trained professionals, working full time at

investment management.

Evidences for Strong Form: Mixed results, but much support

Tests for corporate insiders and stock exchange specialists do not support
the hypothesis (Both groups seem to have monopolistic access to important
information and use it to derive above-average returns)

8.4. Market Efficiency and Corporate Managers

Market efficiency has implications for corporate managers as well as for
investors. This takes a lot of the "gamesmanship" out of corporate management. If a
market is efficient, it is difficult to fool the public for long and by very much. for
instance, only genuine "news" can move the stock price. It is hard to pump-up the stock
price by claims that are not verifiable by investors. "Fake" news will not move the price
-- or if it does, the price will quickly revert to the pre-announcement value when the
news proves hollow. Publicly available information is probably impounded in the price
already. This is hard for some managers to believe. Another lesson: accounting tricks
don't fool anybody. Don't worry about timing accounting charges and don't worry about
whether information is revealed in the footnotes or in the statements. An efficient
market will quickly figure out the meaning of the information, once it is made public.

____________________________________________________________
_
Additional Readings:

Alexander, Gordon, J., Sharpe, William, F. and Bailey, Jeffery, V.,

Fundamentals of Investment, 3rd Edition, Pearson Education.
Bodie, Z., Kane, A, Marcus,A.J., and Mohanty, P. Investments, 6th Edition,
Tata McGraw-Hill.

Bhole, L.M., and Mahakud, J. (2009), Financial institutions and markets.5th Edition, Tata McGraw Hill
(India).

Fisher D.E. and Jordan R.J., Security Analysis and Portfolio Management, 4th
Edition., Prentice-Hall.
Jones, Charles, P., Investment Analysis and Management, 9th Edition, John
Wiley and Sons.

10

Prasanna, C., Investment Analysis and Portfolio Management, 3rd Edition,

Tata McGraw-Hill.
Reilly, Frank. and Brown, Keith, Investment Analysis & Portfolio
Management, 7th Edition, Thomson Soth-Western.

____________________________________________________________
_
Additional Questions with Answers
Session-8: Testing Market efficiency
______________________________________________________________________
__
1.
Explain Weak-Form of efficient market hypothesis (EMH) test.
Ans.

Current prices reflect all security-market information, including the historical

sequence of prices, rates of return, trading volume data, and other market-generated
information. This implies that past rates of return and other market data should have no
relationship with future rates of return
Tests of Weak-Form EMH: Statistical tests of independence between rates of return

Autocorrelation tests: price change in one period is correlated with the price
change in some other period

Runs tests: To test a series of price changes for independence, the number of
runs in that series is compared to see whether it is statistically different from the
number of runs in a purely random series of the same size.

Filter Rules Test: A filter rule is a trading rule regarding the actions to be taken
when shares rise or fall in value by x%. Filter rules should not work if markets are
weak form efficient.
2.
How to explain Tests of Trading Rules?
Ans.

More trading in a security should promote the market efficiency.

Most popular trading technique is Filter Rule i.e. an investor trades a stock
when the price change exceeds a filter value set for it (Buy-and-Hold Strategy).

If the behaviour of the stock price changes is random, filter rules should not
outperform a simple buy-and-hold strategy.
3.
How to tests the Semi strong Form of Market Efficiency?
Ans.: Semi strong form efficiency states that security prices reflect all publicly
available information.
Semi strong Form of Market Efficiency Test takes care of the following approaches:
I.
Event studies: That examine how fast stock prices adjust to specific significant
economic events

Identify the event to be studied and pin point the date on which the event was
announced

11


Collect returns data around the announcement date

Calculate the excess returns by period around the announcement date for each
firm in the sample

Compute the average excess returns across all firms

Assess whether the excess returns around the announcement date are different
from zero
II.
Portfolio Study or Return Prediction Study: That examines the possibility of
earning superior risk-adjusted returns by trading on an observable characteristic of a
firm like P/E ratio, dividend yield etc.

Define the variable (characteristic) on which firms will be classified

Classify firms into portfolios based upon the magnitude of the variable
Compute the returns for each portfolio
Calculate the excess returns for each portfolio
Assess whether the average excess returns are different across the portfolios

4.
What do you mean by the size effect?
Ans.

Several studies have examined the impact of size on the risk-adjusted rates of
return
The studies indicate that risk-adjusted returns for extended periods indicate that the
small firms consistently experienced significantly larger risk-adjusted returns than large
firms
Firm size is a major efficient market anomaly
Attempts to explain the size anomaly in terms of superior risk measurements,
transactions costs, analysts attention, trading activity, and differential information have
not succeeded.
The P/E studies and size studies are dual tests of the EMH and the CAPM
Abnormal returns could occur because either: markets are inefficient or market model
is not properly specified and provides incorrect estimates of risk and expected returns
Size and Ratio of Book Value of a firms Equity to Market Value of its equity
(BV/MV) dominate other ratios such as E/P ratio or leverage
5. What are the other evidences against strong form of Market Efficiency?
Ans.
Strong-Form Efficient Market Hypothesis:
Strong-form EMH contends that stock prices fully reflect all information, both public
and private
This implies that no group of investors has access to private information that will
allow them to consistently earn above-average profits
Evidences towards Strong Form Market Efficiency
Mixed results, but much support
Tests for corporate insiders and stock exchange specialists do not support the
hypothesis (Both groups seem to have monopolistic access to important information
and use it to derive above-average returns)

12

Contradictory Evidence in the form of various stock market anomalies resulting above
average return:

The Low PE Effect

Low-Priced Stocks
The Small Firm and Neglected Firm Effects
Market Overreaction
The January Effect
The Weekend Effect
The Persistence of Technical Analysis

13