Volume I Issue 2(2) Winter 2010

MACROECONOMIC DETERMINANTS OF STOCK RETURNS IN

PAKISTAN: THE CASE OF KARACHI STOCK EXCHANGE
Nadeem SOHAIL
University of Sargodha, Pakistan
[email protected]
Hussain ZAKIR
University of Sargodha, Pakistan
[email protected]

Abstract
The stock market is one of the imperative indicators of the economy. This study strived to explore the
impact of five macroeconomic variables on General Index in the long run and short run. In order to investigate the
long run and short run relationships Johansen cointegation technique and VECM was applied. This study used
monthly data from November 1991 to June 2008 for analysing General Index. The study revealed that consumer
price index, and real effective exchange rate, and industrial production index had a positive impact on stock
prices in Pakistan while money supply and three month treasury bills rate affected stock prices negatively in the
long run. The VECM demonstrated that it took more than eight months to eliminate the disequilibrium. The
variance decompositions exposed that consumer price index and money supply showed greater forecast error
than real effective exchange rate, industrial production index, and three month treasury bills rate for General
Index.
Keywords: stock returns, macroeconomic variables, cointegration, Variance decompositions
JEL Classification: E20, E22, E32, E52, G11, G12

1. Introduction
The managed and well-structured stock markets encourage and channelize the savings into the
productive projects, which promotes economic activities in an economy. The General Index represents the overall
changes in stock prices of the ordinary shares of all listed companies at Karachi Stock Exchange (KSE).
Hence this index contained comprehensive picture of Karachi stock market. As Karachi stock exchange
accounting for nearly 70 percent of Pakistani stock market, therefore, General index is a leading ind icator
that presents true picture of the stock market in Pakistan. The Karachi Stock Exchange showed vigorous
performance from 2002 to 2004. However, KSE turn into extremely volatile in 2007- 2008. In the last three
quarters of the year 2008, the KSE had to endure severe crisis. The Board of Directors of KSE laid floor for stock
prices in August, 2008 to control the speedy diminution of the stock prices. This floor was removed in the end of
December, 2008. The major reason for prompt fall in stock prices was capital flight because of instability in
economic policies and political uncertainty. Therefore, this study intended to explore the macroeconomic factors
affecting the stock returns in Pakistan and to suggest policy recommendation to minimize risk in investing in
shares.
Eva & Stenius (1997), and Kearney & Daly (1998) investigated relationship between the conditional
volatility of stock market and volatility of macroeconomic variables in Finland and Australia respectively. The
results showed that macroeconomic variables were the most important determinants of the volatility in stock
prices. Ibrahim & Yusoff (2001), Nishat & Shaheen (2004), Patra & Poshakwale (2006), Sohail and Hussain
(2009), and Humpe & Macmillan (2009) examined impact of macroeconomic variables on stock indices. Ibrahim
and Yusoff (2001) discovered that the stock prices have immediate positive responses to monetary expansion but
negatively related in the long run. Nishat and Shaheen (2004) established positive impact of industrial production
and negative effect of inflation on stock prices. In the Athens stock exchange, Patra and Poshakwale (2006)
explored short run and long run connections between stock returns and macroeconomic variables and
established that inflation, trading volume and money supply had a equilibrium association with the stock prices.
Humpe and Macmillan (2009) explored that US stock prices were affected by inflation and the long-term interest
rate negatively while, industrial production had a positively impact on US stock prices, while; in Japan, stock
prices were negatively related to the money supply and inflation rate, however, positively related to industrial
production. This study was an effort to explore the impact of five macroeconomic variables i.e. interest rate,
inflation, industrial production index, exchange rate, and money supply on General Index.

181

Journal of Advanced Studies in Finance

The residue of the paper is planned as follows. Section 2 illustrates data and methodology to discover the
impact macroeconomic variables and stock prices. In section 3, empirical results are discussed while section 4
concludes.
2. Data and Methodology
Impact of macroeconomic on stock returns in Pakistan was investigated by using monthly time series. This
study incorporated General Index related to Karachi Stock Exchange (for Stock Prices) and five macroeconomic
variables i.e. Consumer Price Index ( for Inflation), Industrial Production Index (for GDP), M 2 (for Money Supply),
Real Effective Exchange Rate (for Exchange Rate), and Three Month Treasury Bills Rate (for Interest Rate). The
main data sources were International Financial Statistics (IFS), Statistical Bulletins published by State Bank of
Pakistan, and Publications by the Federal Bureau of Statistics. The data about the real effective exchange rate,
consumer price index, and three months treasury bills rate were retrieved from IFS CD-Rom. To investigate the
impact of macroeconomic variables on General index, this study used the data of 200 months from November
1991 to June 2008. The log forms of variables were used in this study and description of the variables were given
as below:
GINDEX=General Index
CPI=Consumer price index
IP=Index of industrial production
M2= Money supply (Broader money)
REER=Real effective exchange rate
TTBR=Three month treasury bills rate
2.1. Stationary test
The macroeconomic and financial time series variables are usually non-stationary at levels (Hill et al.,
2001). In case of non stationary time series, mean or the variance or both the mean and the variance reliant on
time and this variance approaches to infinity with the passage of time while if a time series is stationary, then
shocks are supposed to be transitory (Asteriou & Hall, 2006). Hence, it was obligatory to test the stationarity of
the series involved in the study. Following three tests were applied to test the stationarity:
a. Augmented Dickey Fuller test (Dickey and Fuller, 1981)
b. Phillips Perron test (Phillips and Perron, 1988)
c. KPSS (Kwiatkowski, Phillips, Schmidt. and Shin, 1992)
2.2. Cointegration Test and Vector Error Correction Model
To recognize the long-run relationships between the variables, Cointegration test was applied. Johansen
and Juselius (1990) cointegration procedure was applied to find the long run relationships. This approach
provides Trace Value test statistics, and maximum Eigen Value to perceive number of cointegrating vectors. This
technique gives structure for cointegration test in the framework of VAR procedure. The technique was elucidated
as below:
k

xt Ao Aj xt j t

(1)

j 1

Where; xt is an (n x 1) vector of non stationary I(1) variables, Ao is an (n x 1) vector of constants, Aj is a

(n x n) matrix of coefficients , k is the no. of lags, and t is assumed to be a (n x 1) vector of error terms. This
VAR procedure could be reformulated into a vector error correction model (VECM) as demonstrated below:
k 1

xt Ao j xt j xt k t
j 1

Where;
k

Aj
i j 1

and

Aj
i j 1

182

(2)

Volume I Issue 2(2) Winter 2010

is the divergence machinist and I is an (n x n) identity matrix,. The Trace and the Maximum Eigen Value
tests helped to discover the cointegrating vectors.
2.3. Variance decomposition
To explore causality in the short run between stock returns and macroeconomic variables, the vector
autoregressive (VAR) procedure by Sims (1980) was employed and variance decomposition technique was used
to explain relationships between five macroeconomic variables and General Index. In equations (3) and (4),
Bayesian VAR model in first differences was expressed.
k

i 1

j 1

X t 1 11 (i)X t i 12 ( j )Yt j xt
k

i 1

j 1

(3)

Yt 2 21 (i) X t i 22 ( j )Yt j t

(4)

In the language of VAR, s (the stochastic error terms) are known as shock or innovations.
2.4. Model
In this study, following model was used to discover the long run relationship between General index and
macroeconomic variables.
GINDEX = 1CPI+ 2IP+ 3 REER + 4 M2 + 5TTBR + t

The following model was estimated to explore the short run dynamics between variables and the long run
relationships.
P
P
P
.
GINDEX t 1 1U|t 1 1i CPI t 1 1i IPt 1 1i REERt 1
i 1
i 1
i 1
(5)
P

i 1

i 1

1i M 2t 1 1i TTBRt 1 t

3. Empirical results
3.1. Stationarity test
In the time series analysis, it was mandatory to test the stationary of time series. The study applied three
different procedures to test the stationarity of the data. All three tests were unanimous in the results and indicated
that all the series were found stationary at first difference but non stationary at level as revealed in Table1.

Variables

Table 1. Unit Root Analysis

Augmented Dickey-Fuller test
Phillips-Perron Test
statistic
Statistics
Null Hypothesis: Variables
Null Hypothesis: Variables
are Non-stationary
are Non-stationary
Level

GINDEX
CPI
IPI
REER
M2
TTBR

First Difference

Level

-0.399
-0.148
2.106
-1.904
0.295
- 2.172

First Difference

Kwiatkowski-PhillipsSchmidt-Shin test statistic

Null Hypothesis: Variables
are stationary
Level

-11.964*
-.398
-11.958*
0.871
-4.273*
-0.695
-11.291*
1.702
-13.446*
-2.285
-21.744*
1.755
-11.346*
-1.383
-11.233*
1.498
-3.107*
-0.734
-15.749*
1.749
-5.249*
-1.609
-10.276*
0.707
MacKinnon (1996) critical values
5% Level
- 2.875
- 2.875
10% Level
-2.574
-2.574
* &** shows that the coefficient are significant at significance level 5% and 10% respectively.

183

First Difference
0.247*
0.368*
0.266*
0.065*
0.190*
0.112*
0.463
0.347

Journal of Advanced Studies in Finance

3.2. Cointegration Analysis
In this study, to find the long run relationship between the General index (GINDEX) and five
macroeconomic variables i.e. CPI, IP, REER, M2, and TTBR Johanson and Juselius (1990) cointegration
technique was applied after confirming the stationarity of the series.
The stationarity test results were shown in the Table 1 illustrated that every variable involved in the study
was found integrated of order one. Hence, the Johansen and Juselius (1990) cointegration procedure was used
to discover the association between the macroeconomic and general index in the long run. In the first step,
appropriate lag length was determined by using Schwarz Bayesian Criteria (SBC), which showed that the
appropriate lag length was equal to one. To discover the no. of long run cointegrating vectors (relationships)
between the variables, both the Maximal Eigen values and Trace statistic were used. Using Pantula principle, the
model with Unrestricted intercept and no trend was selected. The results for both Maximal Eigen values and
Trace statistic were shown in Table 2 and Table 3 respectively. The Maximal Eigen statistic identified two
cointegrating vector while Trace statistic recognized three cointegrating vectors. As the Trace statistic is
considered more robust than the Maximal Eigen statistics (Cheung and Lai, 1993), therefore, in this study, three
cointegrating vectors were assumed to find the long-run relationships among the variables.
Table 2. Unrestricted Cointegration Rank Test (Maximum Eigen value)
Hypothesized
Max-Eigen
No. of CE(s)
Eigen value
Statistic
None *
0.215
47.82
At most 1 *
0.184
40.26
At most 2
0.126
26.67
At most 3
0.063
12.96
At most 4
0.042
8.51
At most 5
0.005
1.06
Max-eigenvalue test indicates 2 cointegrating eqn(s) at 5% significance level
* denotes rejection of the hypothesis at 5% significance level

0.05
Critical Value
40.08
33.87
27.58
21.13
14.26
3.84

Prob.**
0.006
0.008
0.065
0.456
0.329
0.303

Table 3. Unrestricted cointegration rank test (trace)

Hypothesized
Trace
No. of CE(s)
Eigen value
Statistic
None *
0.215
137.29
At most 1 *
0.184
89.47
At most 2 *
0.126
49.21
At most 3
0.063
22.54
At most 4
0.042
9.57
At most 5
0.005
1.06
Trace test indicates 3 cointegrating eqn(s) at 5% significance level
* denotes rejection of the hypothesis at 5% significance level

0.05
Critical Value
95.75
69.81
47.85
29.79
15.49
3.84

Prob.**
0.000
0.001
0.037
0.269
0.315
0.303

3.3. Long Run Relationship

The normalized coefficients were calculated by normalizing the first cointegrating equation on GINDEX as
shown in Table 4.
Table 4. Normalized coefficients
GINDEX
1
S.E.
t-value

CPI
-16.491
-8.189
2.014

IP
-16.256
-2.311
7.035

REER
-7.143
-7.962
0.897

M2
14.125
-4.471
-3.159

The first normalized equation was estimated as below;

GINDEX = 16.491CPI + 16.256IP +7.143REER 14.125 M2 - 1.843LTTBR
184

TTBR
1.843
-0.782
-2.358

(6)

Volume I Issue 2(2) Winter 2010

The results revealed that stock prices responded positively to consumer price index (CPI) which implied
that in long run, equities served against inflation. The positive relation between stock prices and consumer price
index were similar to the results provided by Abdullah and Hayworth (1993) and Ratanapakorn and Sharma
(2007). Hence, results further confirmed that stock assets provided shield against inflation. The association
between GINDEX and industrial production index (IP) was found significantly positive as it was reported by many
researchers (Fama, 1981; Abdullah and Hayworth, 1993; Ibrahim and Yusoff, 2001; Nishat and Shaheen, 2004;
Cook, 2007; Ratanapakorn and Sharma, 2007; Shabaz et al., 2008; Liu and Sinclair, 2008; Humpe and
Macmillan, 2009; and Sohail and Hussain, 2009). The Real effective exchange rate (REER) showed insignificant
positive relation with General index similar results were reported by Smyth and Nandha, (2003) for four Asian
countries, namely Pakistan, India, Bangladesh, and Sri Lanka. Sohail and Hussain, (2009) also reported positive
impact of REER on stock prices in case of Lahore Stock Exchange. A negative relationship was seen between
General index and money supply. The results were similar to the study of Humpe and Macmillan (2009). The
results revealed that three month treasury bills rate (TTBR) had a negative and significant long run impact on
stock prices. These results were similar to the previous studies (Nishat and Shaheen, 2004; Humpe and
Macmillan, 2009) but it was in contrast with the results of Ratanapakorn and Sharma (2007) and Sohail and
Hussain, (2009).
3.4. Vector Error Correction Model
The vector error correction mechanism was adopted to find out the short run dynamics of the model. The
results of VECM were exposed in Table 5. The coefficients attached with ecm1 (-1), ecm2 (-1) and ecm3 (-1)
illustrated disequilibrium adjustment speed. The adjustments in GINDEX were only due to the first error correction
term i.e. ecm1. Equation 7 showed that GINDEX adjusted by 11.5 percent in one month due to the coefficient of
ecm1 (-1). This implied that more than eight months (1/0.115= 8.695) were required to remove the disequilibrium.
DGINDEX = O.009 +0.137DGINDEX (-1) 0.559 DCPI (-1) +0.057DIP (-1) 0.028DREER (-1)
0.086DM2 (-1) 0.122DTTBR (-1) 0.115 Vecm1 (-1) 0.009 Vecm2 (-1) + 0.047Vecm3 (-1)

(7)

Table 5. Results of Vector Error Correction Model

Variables
Vecm1(-1)
t-value
Vecm2(-1)
t-value
Vecm3(-1)
t-value
D(GINDEX(-1))
t-value

D(GINDEX)
-0.115***
(-1.81)
-0.009
(-0.13)
0.047
(1.09)
0.137***
(1.86)

D(CPI)
0.006*
(3.98)
0.019*
(3.45)
-0.001
(-0.16)
-0.009
(-1.56)

D(IP)
0.037
(1.45)
-0.396*
(-4.42)
-0.397*
(-6.84)
0.174***
(1.75)

D(REER)
0.012*
(2.98)
0.030**
(2.06)
0.013
(1.32)
-0.022
(-1.33)

D(M2)
0.014*
(3.09)
0.029***
(1.84)
-0.006
(-0.61)
-0.020
(-1.48)

D(TTBR)
0.033***
(1.69)
0.120***
(1.76)
0.006
(0.137)
-0.050
(-0.65)

D(CPI(-1))
t-value

-0.559
(-0.59)

0.109
(1.42)

-0.370
(-0.29)

-0.170
(-0.81)

-0.149
(-0.67)

1.444
(1.49)

D(IP(-1))
t-value

0.057
(1.09)

-0.002
(-0.35)

0.147**
(2.039)

0.008
(0.65)

0.027**
(2.17)

-0.004
(-0.08)

D(REER(-1))
t-value
D(M2(-1))
t-value
D(TTBR(-1))
t-value

-0.028
(-0.08)
-0.086
(-0.28)
-0.122***
(-1.81)

-0.049***
(-1.87)
0.011
(0.45)
0.005
(0.96)

0.249
(0.57)
0.696***
(1.67)
-0.037
(-0.41)

0.272*
(3.85)
0.046
(0.67)
-0.011
(-0.71)

0.102
(1.35)
-0.160**
(-2.19)
-0.011
(-0.7)

-0.265
(-0.8)
-0.045
(-0.14)
0.306*
(4.38)

C
t-value

0.009
(0.97)

0.006*
(7.43)

-0.004
(-0.341)

0.000
(0.04)

0.016*
(6.87)

-0.009
(-0.87)

R2
F-statistic

0.09
2.11

0.19
5.01

0.23
6.23

0.19
4.93

0.09
2.15

0.18
4.65

( ) shows t values, * shows the coefficient significant at significance level 1%, ** shows the coefficient significant
significance level at 5%l, *** shows the coefficient significant significance level at 10%

185

Journal of Advanced Studies in Finance

3.5. Variance Decompositions
To explore the degree of exogeneity of the variables involved in this study, variance decomposition was
estimated. The variance decomposition also confirms regarding the relationship among the variables. It illustrates
the share of the forecast error of one variable as a result of changes in the other variables. Hence, the relative
significance of each variable can be determined which causes oscillations in the other variable (Ratanapakorn
and Sharma, 2007). The results were reported in Table 6 for variance decomposition at 1, 6, 12, 18, and 24
months horizon. The results suggested that after 24 months, that relative to the other variables, GINDEX
explained less erogeneity. As 50 percent variance in GINDEX resulted by itself. Out of five macroeconomic
variables CPI and M2 showed maximum proportion of forecast error for GINDEX i.e. 18.66 and 17.06 percent
after 24 months. Changes in left over variables i.e. IP, REER, and TTBR showed a little forecast variance in
GINDEX i.e. 5.73 percent, 3.70 percent, and 4.71 percent respectively. The value of variance forecast error
explained by all macroeconomic variables increased along with the passage of time.
Table 6. Variance decompositions
VDC of
Months
S.E.
GINDEX
CPI
IP
1
0.07
100.00
0.00
0.00
GINDEX
12
0.25
75.49
9.09
5.13
24
0.35
50.14
18.66
5.73
1
0.01
1.42
98.58
0.00
CPI
12
0.02
6.88
89.17
1.57
24
0.03
26.84
64.85
0.87
1
0.11
0.01
3.50
96.49
IP
12
0.16
1.79
7.64
77.69
24
0.17
3.59
8.65
69.96
1
0.02
0.60
1.48
0.04
REER
12
0.04
6.11
0.92
8.49
24
0.04
12.83
2.01
8.06
1
0.02
1.81
3.62
3.37
M2
12
0.05
14.99
1.10
4.62
24
0.08
24.11
0.58
3.84
1
0.08
1.48
0.22
0.02
TTBR
12
0.28
4.59
3.48
0.79
24
0.39
15.86
3.16
0.54
Note: Cholesky Ordering: GINDEX, CPI, IP, REER, M2, TTBR

REER
0.00
2.44
3.70
0.00
1.07
3.52
0.00
6.84
7.79
97.88
79.11
66.59
0.44
3.00
5.99
0.02
0.50
1.32

M2
0.00
7.09
17.06
0.00
0.09
1.99
0.00
5.04
8.94
0.00
1.18
1.14
90.76
75.39
61.62
0.02
3.22
3.87

TTBR
0.00
0.75
4.71
0.00
1.23
1.94
0.00
1.00
1.07
0.00
4.18
9.37
0.00
0.90
3.85
98.25
87.42
75.26

4. Conclusion
This study exerted to explore the impact of macroeconomic variables on stock returns Pakistan in the long
run and short run. The study explored that there were three cointegrating vectors among the studied variables.
This implies there were three long run relationships among variables. The results revealed that Real Effective
Exchange Rate, Consumer Price Index and Industrial Production Index showed positive impact on stock prices in
Pakistan. While, Three Month Treasury Bills Rate and Money supply affected stock prices negatively in the long
run.
The VECM described that it took more than eight months for the adjustment of disequilibrium of the
previous period. The results of Variance Decomposition revealed that General index explained nearly 50 percent
variance in GINDEX resulted by itself while CPI, IP, REER, M2, and TTBR explained 18.6 percent, 5.7 percent,
3.7 percent, 17.1 percent, and 0.75 percent variance in General index respectively. Consumer Price Index
produced the maximum variance in GINDEX among the macroeconomics variables.
The study suggested by adopting a suitable monetary policy by the State Bank of Pakistan and setting
appropriate fiscal measures by the Government of Pakistan to manage the rate of inflation, the volatility in stock
prices can be reduced. Stock markets can be developed by adopting Industrial production promoting measures
as it has a positive significant impact on stock prices in Pakistan. The long run positive impact of exchange rate
on General index suggested that for the development of stock market in Pakistan, exchange rate should be
managed carefully keeping in view the elasticises of exports and imports that will lead to stability in stock market.
The monetary authorities should take care in executing monetary policies particularly to affect movements in the

186

Volume I Issue 2(2) Winter 2010

stock market, because soft monetary policy elevate stock prices in the short-run leading to adverse results in the
long-run. The study also recommended that three months treasury bills rate should be kept appropriately low so
that it cannot affect stock returns adversely.
6. References
1. Abdullah, D.A., Hayworth, S.C. (1993), Macroeconometrics of stock price fluctuations, Quarterly Journal of
Business and Economics, 32:1, 49-63.
2. Asteriou, D. and Hall, S.G. (2006), Applied Econometrics, 288.
3. Cheung, Y-W. and Lai, K. (1993) A fractional cointegration analysis of purchasing power parity, Journal of
Business & Economic Statistics, 11, 103 -112.
4. Cook, S. (2007), Threshold adjustment in the long-run relationship between stock prices and economic activity,
Applied Financial Economics Letters, 3:4, 243 246.
5. Dickey, D. A. and W. A. Fuller (1981). Likelihood ratio statistics for autoregressive time series with a unit root
Econometrica, 49, 1057 - 72.
6. Eva, L. and Stenius, M. (1997), Macroeconomic volatility and stock market volatility: empirical evidence on
finish data, Applied Financial Economics, 7:4, 419 426.
7. Fama, E.F. (1981), Stock returns, real activity, inflation and money, American Economic Review, 71,545-65.
8. Hill, R. C., Griffiths, W., and Judge, G. (2001), Undergraduate Econometrics, 2nd edition, by Wiley, New York,
pp. 335 -338.
9. Humpe, A. and Macmillan, P. (2009), Can macroeconomic variables explain long-term stock market
movements? A comparison of the US and Japan, Applied Financial Economics, 19:2,111 -19.
10. Ibrahim, M. H. and Yusoff, W. S. W. (2001),Macroeconomic variables, exchange rate, and stock price a
Malaysian perspective, IIUM Journal of Economics and Management, 9:2, 141- 163.
11. Johansen, S. and Juselius, K. (1990), Maximum likelihood estimation and inference on cointegration with
applications to the demand of money, Oxford Bulletin of Economics and Statistics, 52, 169 210.
12. Kearney, C. and Daly, K. (1998), The causes of stock market volatility in Australia, Applied Financial
Economics, 8:6, 597-605.
13. Kwiatkowski, D., Phillips, P. C. B., Schmidt, P. and Shin, Y. (1992), Testing the null hypothesis of trend
stationanity, Journal of Econometrics, 54,159-178.
14. Liu, X. and Sinclair, P. (2008) Does the linkage between stock market performance and economic growth vary
across Greater China?, Applied Economics Letters, 15:7, 505 508.
15. Nishat, M. and Shaheen, R. (2004), Macroeconomic factors and Pakistani equity market, The Pakistan
Development Review, 43:4, 619-637.
16. Patra, T. and Poshakwale, S. (2008), Long-run and short-run relationship between the main stock indexes:
evidence from the Athens stock exchange, Applied Financial Economics, 01-10.
17. Phillips, R. C. B. and Perron (1988), Testing for a Unit Root in Time Series Regression. Biometrika, 335-346
18. Ratanapakorn, O. and Sharma, S. C. (2007), Dynamic analysis between the US stock returns and the
macroeconomic variables, Applied Financial Economics, 17:5,369-377.
19. Shahbaz, M., Ahmed, N., and Ali, L. (2008), Stock market development and economic growth: Ardl Causality
in Pakistan, Journal of Finance and Economics, 14, pp. 182-195.
20. Sims, C. (1980), Macroeconomics and Reality, Econometrica, 48, pp.1-48.
21. Smyth, R. and Nandha, M. (2003), Bivariate causality between exchange rates and stock prices in South
Asia, Applied Economics Letters, 10:11, pp. 699 -704.
22. Sohail, N. and Hussain, Z. (2009), Long-Run and Short-run Relationship between Macroeconomic Variables
and Stock prices in Pakistan: The Case of Lahore Stock Exchange, Pakistan Economic and Social Review,
Volume 47, No. 2 (Winter 2009), pp. 183-198

187

Copyright of Journal of Advanced Studies in Finance is the property of ASERS Publishing and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.