BANGLADESH RESEARCH PUBLICATIONS J OURNAL

ISSN: 1998-2003, Volume: 6, Issue: 2, Page: 225-233, November - December, 2011

Review Paper
THE RANDOM WALK OF DHAKA STOCK EXCHANGE (DSE)
RETURNS: EVIDENCE AND IMPLICATIONS

Mohammed Abu Rayhan
1
, Rokeya Sultana
2
and Syeedul Al-Amin
* 3

Mohammed Abu Rayhan, Rokeya Sultana

and Syeedul Al-Amin (2011). The Random Walk of Dhaka
Stock Exchange (DSE) Returns: Evidence and Implications. Bangladesh Res. Pub. J . 6(2): 225-233.
Retrieve from http:/ / www.bdresearchpublications.com/ admin/ journal/ upload/ 09277/ 09277.pdf

Abstract
The main focus of the study is to detect the random walk of the monthly
stock returns of the DSE. The study covers the twenty three-year-long period
commencing from J anuary 1987-March 2010. Tests used in this study are
the well known runs test and the unit root test. The study revealed that
monthly DSE price index follows random walk but monthly DSE returns dont
follow it. The study also revealed that monthly DSE returns follow
Autoregressive conditional heteroskedasticity (ARCH) properties. The study
recommended to introduction of derivatives securities, enlistment of shares
of giant local and multinational corporations, Stopping frequent change of
decisions by the policy maker, Off-loading government shares, introducing
funds managed by professional money managers, ensuring proper
functioning of arbitrageurs, disclosing all price sensitive information publicly,
to adopt internationally accepted auditing and accounting and so forth
could establish the DSE as barometer for the countrys economic
dynamism.
Key Words: Random Walk, DSE, Runs test, Unit root test, ARCH, Price index.
Introduction
Dhaka Stock Exchange (DSE) is the countrys leading stock exchange. On
April 28, 1954 the DSE was first incorporated as the East Pakistan Stock Exchange
Association Limited. However, formal trading began in 1956 with 196 securities
listed on the DSE with a total paid up capital of about Taka 4 billion (Chowdhury,
1994). On J une 23, 1962 it was renamed as Dhaka Stock Exchange (DSE) Limited.
After 1971, the trading activities of the Stock Exchange remained suppressed until
1976 due to the liberation war and the economic policy pursued by the then
government. The trading activities resumed in 1976 with only 9 companies listed
having a paid up capital of Taka 137.52 million on the stock exchange
(Chowdhury, 1994). As of August, 2010 there were 450 Securities listed on the DSE
with a market capitalization of $ 50.28 billion which is 43.65% of GDP (GDP
projected 6, 90,000 Crore Taka at the end of fiscal year 2009-10). The Dhaka Stock
Exchange (DSE) is registered as a Public Limited Company and its activities are
regulated by its Articles of Association and its own rules, regulations, and by-laws
along with the Securities and Exchange Ordinance, 1969; the Companies Act,
1994; and the Securities and Exchange Commission Act, 1993. As per the DSE
Article 105B, its management is separated from the Council. The executive power
of the DSE is vested with the Chief Executive Officer (CEO). The CEO is appointed
by the Board with the approval of the SEC. The Dhaka Stock Exchange is a self

*Corresponding Author s Email: syeedul101@yahoo.com
1
Lecturer, Faculty of Business, ASA University Bangladesh, ASA Tower, Shyamoli, Dhaka-1207
2
Sr. Lecturer, Department of Business Administration Manarat International University, Bangladesh.
3
Lecturer, Management Studies, Comilla University, Comilla, Bangladesh.
Rayhan, et al.
http://www.bdresearchpublications.com/journal/
226
regulated non-profit organization. It has provisions for 500 members though at
present the number of members is 230. Membership is open to the foreigners as
well. The Exchange has a 24 member Council, of which 12 are elected from the
members and the other 12 are nominated as non-member from different apex
bodies. Trading is done through automated on-line system every day except
Friday, Saturday and other government holidays. There are five markets in the
system: (1) Public Market: Only trading of market lot share is done here through
automatic matching. (2) Spot Market: Spot transactions are done here through
automatic matching which must be settled within 24 hours. (3) Block Market: A
place where bulk quantities of shares are traded through pick and fill basis. (4)
Odd Lot Market: Odd lot scripts are traded here based on pick and fill basis. (5)
OTC Market: Dhaka Stock Exchange (DSE) has launched the over-the-counter
(OTC) market-a separate trading floor to facilitate trading of the non-listed and
de-listed companies. Shares of companies that have been de-listed from the
premier bourse will be placed on the OTC market in the first phase. In the next
phase, shares of non-performing and non-operational companies that will be
removed from the main board will be traded on the new OTC floor.
All transactions in public market of a day, after netting, are settled and
cleared through the DSE Clearing House due on 3rd and 5th working day
respectively, calculated from date of trading. Members shall be allowed to carry
out transaction of foreign buyers and/or seller involving a custodian bank to be
settled directly between the member through the custodian bank within the fifth
day subsequent to the trading day, i.e., T + 5 in respect of the transactions carried
out on each trading day with intimation to the clearing house.
Over the passage of time some significant developments took place in
DSE. One important development in the capital market in 2004 was the initiation
of electronic settlement through the Central Depository System (CDS) in J anuary
2004. In order to prevent market manipulation by in-house officials of listed
companies, SEC banned the purchasing or selling of shares of a company by its
owners during an interim period (from the date of the financial year closure and
the day of approval of accounts by the companys board). De-listing of 13
companies in August 2004 by DSE due to their repeated failure in complying with
the listing rules was also an important step toward bringing discipline in the stock
exchange. The trading of Bangladesh Government Treasury Bonds (BGTBs) started
in DSE from J anuary 2005. However, in order to temper the rising trend of stock
index and control excess liquidity in the capital market, SEC temporarily
suspended the credit facility extended by brokers to their clients. At the same
time, SEC also increased the members trade margin requirements by reducing
the free trading limit from Tk. 10 million to Tk. 5 million. All the market barometers
significantly rose during 2010 reflecting regained investors confidence after 1996
stock market bubble. The increasing trend of DSE-Gen index in J anuary 2007
crossed the 3,000 mark for the first time and it crossed 8187 points for the first time
with a significant fluctuation at the end of 2010 (Table 1: Market highlights). Two
DSE Branch offices were established at Chittagong and Sylhet. The members from
these two branches are getting facility of full redundant connectivity using BTTB
DDN and Ranks-ITT. DSE has already established two network hub rooms (DSE
POP) in Dhanmondi and Uttara.
DSE Returns: Evidence and Implications
http://www.bdresearchpublications.com/journal/
227
Table 1: DSE market highlights (As on 31
st
March, 2010)
Indicators DSE
No. of companies 240
No. of mutual funds 23
No. of Debentures 8
No. of treasury bonds 164
No. of corporate bonds 2
Total No. of listed Securities 437
(Fig. in Million)
No. of shares of all listed companies
No. of certificates of all listed mutual funds
No. of debentures of all listed debentures
No. of all listed Govt T-bond
No. of all listed Corporate bonds
Total No. of tradeable securities
4336
945
0.41
3.85
4.34
5290
(Fig. in Million Taka)
Issued Capital of all companies 166,577.00
Issued Capital of all Mutual funds 9,766.00
Issued debentures 140.00
Total Issued capital 566,632.00
Total market capitalization 2,275,558.00
All share price index 4573.81
Source: DSE annual report 2008-2009
This study aims at examining the random walk of stock returns of the DSE.
The study also focuses on the characterization of data set and suggesting some
policy implications to overcome the existing anomalies in DSE. Two hypotheses
have been developed to reach the conclusion in this regard.
Hypotheses of the study are:
Ho: The returns series follow the random walk;
H1: The returns series do not follow the random walk;
Statistical Terms and Methodology of the Study
A time series is a sequence of data points, measured typically at successive times
spaced at uniform time intervals. To develop more accurate time series and
forecasting model, first of all we need to find out that the data are stationary or
not. Line charts or time series plots are particularly effective for business and
economic data because we can show the change or trends in a variable over
time.
Anderson-Darling test is a test that a given sample of observations arises from
some specified theoretical probability distribution. For testing the normality of the
data, the test statistic is
2
1
1
1
(2 1){log log(1 )}
n
n i
i
n i
A i z z
n
+
=

= +

n

Where
(1) (2) ( ) n
x x x
are the ordered observations,
2
s
is the sample variance
and
( ) i
i
x x
z
s

=


Where
( )
2 1
2
1
2
x
u
x e d

=

u
. The null hypothesis of normality is
rejected for large values
2
n
A
. Critical values of the test statistic are available.
Rayhan, et al.
http://www.bdresearchpublications.com/journal/
228
Cramer-von Mises test is a test of whether a set of observations arise from a
normal distribution. The test statistic is
2
2
1
(2 1) 1
2 1
n
i
i
i
W z
n n
=

= +

2

Where the are found from the ordered sample values
i
z
(1) (2) ( ) n
x x x
as
2 1
( )
2
1
2
i
x
x
i
z e

=

dx
. Critical values of can be found in many sets of statistical
tables.
2
W

J arque-Bera (J B) test of normality is an asymptotic, or large-sample, test. The test
statistic is
2 2
( 3)
6 24
S K
JB n

= +



Where sample size,
S
=Skewness coefficient, and
n =
K
=Kurtosis coefficient. The
J B test of normality is a test of the joint hypothesis that and
S
K
are 0 and 3. Under
the null hypothesis, the test statistic follows chi-square distribution with 2 degrees
of freedom.
Shapiro-Wilk test is a test that a set of random variables arise from a specified
probability distribution. Most commonly used to test for departures from the
normal distribution and the exponential distribution. The test statistic is
W
2
(1)
2
( )
1
( )
1
( )
n
i
i
x x
n
W
n
x x
=

Where
(1) (2) ( ) n
x x x
are the ordered sample values and
x
is their
mean. Critical values of
W
based on the simulation studies are available in many
statistical tables.
Weakly stationary can be tested by the correlogram of a time series, which
is a graph of autocorrelation at various lags. For stationary time series, the
correlogram tapers off quickly, whereas for non-stationary time series it dies off
gradually. For a purely random series, the autocorrelation at all lags 1 and greater
are zero.
Stationary can also be checked by finding out if the time series contains a
unit root. The Dickey-Fuller (DF) and augmented Dickey-Fuller (ADF) tests are used
for this purpose. If the time series contains a unit root, the data may be follow
random walk model. Random walk is a motion of a particle that moves the
discrete jumps with certain probabilities from point to point.
Suppose is a white noise error term with mean zero and variance .
Then the series is said to be random walk if
t
u
2

t
Y
t t t
u Y Y + =
1
.
In order to examine the serial dependence we conduct runs test which is
based on positive and negative runs from the mean.
Autoregressive model (AR) is a model used primarily in the analysis of time series in
which the observation, , at time , is postulated to be a linear function of
previous values of the series. The first order autoregressive model is of the form
t
Y
t
DSE Returns: Evidence and Implications
http://www.bdresearchpublications.com/journal/
229
t t t
u Y Y + =
1

.
Autoregressive Conditional Heteroskedasticity (ARCH) models are specifically
designed to model and forecast conditional variance. The variance of the
dependent variable is modeled as a function of past value of the dependent
variable. ARCH model is introduced by Engle (1982). ARCH model is especially
useful in analyzing financial time series, such as stock prices, inflation rates, and
exchange rates. A distinguish feather of this model is that the error variance may
be correlated over time because of the phenomenon of volatility clustering.
In this study, stock return is defined as the difference in the logarithm of
prices, that is,
1
ln ln
t t
r p p
t
=
, where In pt and In pt-1 are current period and
one-period lag log prices respectively. Stock returns defined above are nominal
only, and it doesnt take into account the effects of dividends, inflation and
exchange rates. The data set used in this study consists of twenty three-year-long
period commencing from J anuary 1987-March 2010, obtained from various issues
of the DSE monthly bulletin, national dailies and internet.
Empirical Findings
In order to test for random walk in the monthly return data, the study
selected two common and popular statistical techniques: the first one is the unit
root test and the second one is the runs test.
Before testing the presence of random walk, the study examines various
descriptive statistical properties (i.e. mean, variance, skewness, kurtosis, normality
and so forth) of the monthly DSE return series. Table 2 represents that overall
monthly DSE mean return is positive. Sample estimates of skewness and excess
kurtosis for monthly DSE returns are both large and positive. This indicates that
returns have mass in the tail (positively skewed) than that of normal distribution. In
symmetric distribution the value of skew ness and kurtosis are 0 and 3 respectively.
Number of Months
265 241 217 193 169 145 121 97 73 49 25 1
M
o
n
t
h
l
y

R
e
t
u
r
n
s
.8
.6
.4
.2
-.0
-.2
-.4
-.6
Number of Months
265 241 217 193 169 145 121 97 73 49 25 1
p
r
i
c
e

i
n
d
e
x
5000
4000
3000
2000
1000
0

Fig (1) Fig (2)
Fig (1): Correlogram of Monthly DSE Price Index (J anuary 1987-March 2010)
Fig (2): Monthly DSE Returns (J anuary 1987-March 2010)
Source: DSE annual report 2008-2009

But in figure 2, time series plot suggest that monthly DSE Returns contain a large
number of outliers especially during the period of bubble in 1996. So the
distribution of the DSE monthly returns is fat or heavy- tailed. The value of Kuiper
(V), Cramer-von Mises (W2), Watson (V2), Anderson-Darling (A2), and J arque-Bera
normality statistic are significantly indicating that the monthly DSE Price Index is
not normally distributed which is shown in table 3. In symmetric distribution the
value of skew ness and kurtosis are 0 and 3 respectively. But time series plot
Rayhan, et al.
http://www.bdresearchpublications.com/journal/
230
t
suggest that monthly DSE Returns contain a large number of outliers. So the
distribution of the DSE monthly returns is fat or heavy- tailed. The value of Kuiper (V),
Cramer-von Mises (W2), Watson (V2), Anderson-Darling (A2), and J arque-Bera
normality statistic are significantly indicating that the monthly DSE Price Index is not
normally distributed.
Table 2: Descriptive Statistics of Monthly DSE returns,
1
ln ln
t t
r p p

=

Number of observations 278
Mean 0.01
Median 0.00
Standard Deviation 0.096329
Skewness
0.970662 (0.146911)
Kurtosis
6.495657 (0.293821)
Maximum 0.57
Minimum -0.39
Source: DSE annual report 2008-2009
Table 3: Normality test of Monthly DSE returns
1
ln ln
t t
r p p
t
=

Method Value Probability
Kuiper (V) 0.790443 0.0000
Cramer-von Mises (W2) 18.37295 0.0000
Watson (U2) 18.36884 0.0000
Anderson-Darling (A2) 88.04601 0.0000
J arque-Bera 511.4173 0.0000
Source: DSE annual report 2008-2009
By Correlogram of Monthly DSE Price Index shown in figure 3, monthly DSE price index
is non-stationary.
Autocorrelation Partial Correlation AC PAC Q-Stat Prob
.| *******| .| *******| 1 0.865 0.865 210.25 0.000
.| ****** | .| . | 2 0.739 -0.037 364.23 0.000
.| ***** | .| * | 3 0.652 0.083 484.40 0.000
.| ***** | .| * | 4 0.597 0.085 585.66 0.000
.| **** | .| . | 5 0.545 0.000 670.37 0.000
.| **** | .| * | 6 0.513 0.082 745.79 0.000
.| **** | .| . | 7 0.486 0.021 813.55 0.000
.| *** | .| . | 8 0.455 0.003 873.30 0.000
.| *** | .| . | 9 0.429 0.030 926.57 0.000
.| *** | .| . | 10 0.398 -0.020 972.64 0.000
.| *** | .| . | 11 0.370 0.008 1012.5 0.000
.| *** | .| . | 12 0.340 -0.012 1046.3 0.000
.| ** | .| . | 13 0.313 -0.004 1075.1 0.000
.| ** | .| . | 14 0.282 -0.027 1098.5 0.000
.| ** | .| . | 15 0.251 -0.019 1117.2 0.000
.| ** | .| . | 16 0.228 0.011 1132.6 0.000
.| ** | .| . | 17 0.211 0.003 1145.9 0.000
.| ** | .| . | 18 0.200 0.019 1157.9 0.000
.| * | .| . | 19 0.192 0.016 1169.0 0.000
.| * | .| . | 20 0.186 0.013 1179.4 0.000
.| * | .| . | 21 0.182 0.019 1189.4 0.000
.| * | .| . | 22 0.176 0.005 1198.8 0.000
.| * | .| . | 23 0.176 0.035 1208.2 0.000
.| * | .| . | 24 0.174 0.008 1217.5 0.000
.| * | .| . | 25 0.170 -0.001 1226.4 0.000
.| * | .| . | 26 0.163 0.002 1234.5 0.000
.| * | .| . | 27 0.155 -0.004 1242.0 0.000
.| * | .| . | 28 0.148 0.001 1248.8 0.000
.| * | .| . | 29 0.143 0.003 1255.2 0.000
.| * | .| . | 30 0.139 0.003 1261.3 0.000
.| * | .| . | 31 0.131 -0.018 1266.7 0.000
.| * | .| . | 32 0.118 -0.022 1271.1 0.000
.| * | .| . | 33 0.106 -0.009 1274.7 0.000
.| * | .| . | 34 0.095 -0.008 1277.5 0.000
.| * | .| . | 35 0.087 0.002 1279.9 0.000
.| * | .| . | 36 0.081 0.002 1282.0 0.000
Fig (3): Correlogram of Monthly DSE Price Index (J anuary 1987-March 2010)
Source: www.dsebd.org
DSE Returns: Evidence and Implications
http://www.bdresearchpublications.com/journal/
231
Autocorrelation Partial Correlation AC PAC Q-Stat Prob
.| ** | .| ** | 1 0.204 0.204 11.716 0.001
.| . | .| . | 2 0.016 -0.027 11.787 0.003
.| . | .| . | 3 0.053 0.058 12.597 0.006
.| . | .| . | 4 0.031 0.009 12.867 0.012
*| . | *| . | 5 -0.096 -0.108 15.514 0.008
*| . | .| . | 6 -0.087 -0.050 17.689 0.007
.| . | .| * | 7 0.048 0.076 18.357 0.010
.| . | .| . | 8 -0.003 -0.022 18.360 0.019
Fig (4): Correlogram of Monthly DSE Returns (J anuary 1987-March 2010)
Source: www.dsebd.org
From Correlogram of Monthly DSE Returns shown in figure 4, we found that
monthly DSE returns follow the properties of stationary. Again, from the unit root
test, there is enough evidence to conclude that monthly DSE Price Index,
(J anuary 1987-March 2010) follows random walk. But Monthly DSE Returns
(J anuary 1987-March 2010) does not follow random walk i.e. they are serially
dependent which are shown in table 5 and table 6 respectively. In addition, the
results obtained from runs test shown in table 4 suggest that we can reject the null
hypothesis (Ho) as the expected number of runs (139) is lower than the observed
number of runs (280). Therefore, we can conclude that the DSE monthly returns
are serially dependent at 95% confidence interval.
Table 4: The run test for monthly DSE stock returns (J anuary 1987-March 2010)
95%
confidence
intervals of
runs
Number of
observations
Expected
number
of runs
Number
of
positive
runs
Number
of
negativ
e runs
Total
number
of
observed
runs
Standard
deviation
of runs

Z
-
t
e
s
t

Lower Upper
278 139.8201 59 58 117
8.31079
8
-2.75 124 156
Source: www.dsebd.org
Table 5: The Unit root test of Monthly DSE Price Index, (J anuary 1987-March 2010)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic 0.977463 0.9132
Test critical values: 1% level -2.573367
5% level -1.941978
10% level -1.615931
*MacKinnon (1996) one-sided p-values.
Source: www.dsebd.org
Table 6: Autoregressive Conditional Heteroskedasticity test of Monthly DSE Returns
(J anuary 1987-March 2010)
Variable Coefficient Std. Error t-Statistic Prob.
LNDSE(-1) 0.212229 0.058801 3.609270 0.0004
ARCH Test:
F-statistic 12.67432 Probability 0.000437
Obs*R-squared 12.20239 Probability 0.000477
Source: www.dsebd.org
Moreover, Autoregressive model may be applied also to monthly DSE
returns prediction. But we need to justify whether the error variance are
correlated over time because of the phenomenon of volatility clustering., So
autoregressive conditional Heteroskedasticity (ARCH) test is more appropriate for
this data set for prediction. In this paper, ARCH LM test suggest that ARCH effect
Rayhan, et al.
http://www.bdresearchpublications.com/journal/
232
exists in the data set. Therefore, monthly DSE returns follow Autoregressive
Conditional Heteroskedasticity properties.
Conclusions and Policy Implications
This study examines the random walk in the monthly stock return index of
the Dhaka Stock Exchange (DSE). The study found that there is enough evidence
to conclude that monthly DSE Price Index follows random walk whereas monthly
DSE Returns does not follow the same i.e. they are serially dependent.
However, before discussing the policy implications of findings of the study,
it is wise to consider some major developments and anomalies at the DSE. Over
the passage of time some significant developments took place in DSE. One
important development in the capital market in 2004 was the initiation of
electronic settlement through the Central Depository System (CDS) in J anuary
2004. In order to prevent market manipulation by in-house officials of listed
companies, SEC banned the purchasing or selling of shares of a company by its
owners during an interim period (from the date of the financial year closure and
the day of approval of accounts by the companys board). De-listing of 13
companies in August 2004 by DSE due to their repeated failure in complying with
the listing rules was also an important step toward bringing discipline in the stock
exchange. The trading of Bangladesh Government Treasury Bonds (BGTBs) started
in DSE from J anuary 2005. However, in order to temper the rising trend of stock
index and control excess liquidity in the capital market, SEC temporarily
suspended the credit facility extended by brokers to their clients. At the same
time, SEC also increased the members trade margin requirements by reducing
the free trading limit from Tk. 10 million to Tk. 5 million. All the market barometers
significantly rose during 2010 reflecting regained investors confidence after 1996
stock market bubble. The increasing trend of DSE-Gen index in J anuary 2007
crossed the 3,000 mark for the first time and it crossed 8187 points for the first time
by the end of 2010. The market is more or less captive in the hands of some
influential brokers. In addition, sluggish monitoring by the regulatory bodies,
political instability, and price manipulation by listed companies their through
window-dressing or through direct participation in tractions have made the
market more peculiar. During 1994 through 1996, the most eventful period in the
history of the DSE, rumors, unethical speculation coupled disguised demand
created supply through transactions among some influential brokers on the very
nose of the regulatory bodies, made the DSE a wrong barometer of the countrys
economic condition Small and nave investors without any ability to interpret
market signals properly were attracted t the market on the expectation of a
continuous uptrend in the share prices. The regulators remained silent to this
speculative price hike, which could not be backed by any change in underlying
fundamentals.
In order to transform the DSE into a correct barometer of the countrys
economic condition, the government and the SEC should take following actions:
1. Introduction of varieties shares with derivatives to attract nave and small
investors and to minimize risk exposure.
2. Fundamentals should be the basis of asset valuation and information about
the market is efficiently disseminated so that investors cannot make
abnormal gains.
3. Enlisting the giant national and multi-national corporation because Small
participation of these reputed firms through IPO (initial public offer) can
attract institutional as well as foreign investors and raise the depth (liquidity)
of the market.
4. The Govt. should off-load the shares of committed state-owned enterprises
DSE Returns: Evidence and Implications
http://www.bdresearchpublications.com/journal/
233
immediately to meet up the growing need of investors and to balance
between the market forces demand and supply.
5. The policy makers need to introduce funds managed by professional money
managers, so that the hopeless small and nave investors become more
confident about equity investments.
6. In order to ensure proper functioning of arbitrageurs, the SEC should
introduce strictly supervised (so that arbitrageurs cannot manipulate) short
selling at the DSE short selling plays important and constructive roles in the
equity markets by providing liquidity and pricing efficiency.
7. Like other emerging market, the DSE poor in information dissemination. To this
end the listing rules should make companies responsible to disclose all price
sensitive information, which has a significant impact on the companys share
price. The SEC should also ensure that companies are making price sensitive
information available to the market as a whole. If companies give any price
sensitive information to substantial share holders or other parties they are
dealing with, such insiders should not be allowed to deal in a companys
securities before the information is made public.
8. Stopping the frequent change of decisions by the policy makers.
Finally, tight disclosure requirements accompanied by internationally
accepted auditing and accounting standards will certainly create confidence
among local and foreign investors to commit their funds to emerging stock
markets like DSE.
References
Chowdhury, A. R. (1994), Statistical Properties of Daily Return from the Dhaka Stock
Exchange, Bangladesh Development Studies, Vol. XXII, No. 4.
DSE annual report 2008-2009
Engle, R, (1982), Autoregressive conditional heteroscedasticity with estimates of the
variance of UK inflation, Econometrica, 50, 987-1007.
MacKinnon, A. (1996). Theory of Econometrics, 2nd ed., Macmillan Press Ltd., Hampshire.
www.dsebd.org