International Review of Business Research Papers

Vol.3 No.2 June 2007, Pp. 362 - 375

362

Stock Return Volatility in Emerging Equity Market (Kse): The
Relative Effects of Country and Global Factors


Mohammad Faisal Rizwan* and Safi Ullah Khan**



This paper focuses on the role of macroeconomic variables and global
factors on the volatility of the stock returns in an emerging market like
Pakistan. The paper uses two multivariate models, multivariate EGARCH
and Vector Auto Regressive (VAR) models to investigate the effect of
exchange rate, interest rate, industrial production, money supply, Morgan
Stanley Capital International (MSCI) World Index and 6-months LIBOR
on stock prices in Pakistans equity market. The estimate shows that
domestic macroeconomic variables have varying degrees of importance
in explaining the relationship between stock returns and volatility in
Karachi Stock Exchange. The empirical results also show that the two
global factors, MSCI World Index and 6-months LIBOR, variables used in
this paper explain the stock returns in KSE. An important conclusion
drawn from the results is that macroeconomic variables exhibit
asymmetric effects on returns volatility. Overall, the results show that
Pakistans stock market is partially integrated as shown by the significant
role of both country and global factors.


Field of Research: Finance


1. Introduction

Studies on the link between macroeconomic variables and stock returns are broadly
divided into two groups based on the level of market integration. The first strand
promotes the view that markets are generally integrated and, as a result, global risk
factors are more important in explaining returns volatility than country factors. Most of
the studies that emphasize the role of global risk factors have modeled returns as linear
function of global risk factors (Ferson and Harvey, 1994; Dumas and Solnim, 1995;
Harvey, 1995). A list of most popular factors examined in the integrated market
approach includes industrial production, unexpected inflation, changes in expected
inflation, real interest rates, term structure risk and price of crude oil.


________________________________________

*Faisal Rizwan , Government College University, Faisalabad, Pakistan.
Email: rizwan_faisal2000@yahoo.com
**Safi Ullah Khan , Kohat University of Science and Technology, Kohat, Pakistan. Email:
safiullah75@yahoo.com

Rizwan & Khan
363

Ferson and Harvey (1993a, 1993b) examined all these variables, measured as global
aggregates, as potential risk factors for global asset pricing models. The few studies
that looked at the role of global risk factors on the stock returns of emerging markets
have also implicitly or explicitly emphasized the relevance of market integration in their
analyses. Harvey (1995), for example, tested the relationship between returns in the
emerging markets and a set of global variables including world inflation, world GDP,
world oil prices and trade weighted world exchange rate. Results of his study suggests
that standard global asset pricing models, which assume complete integration of capital
markets, fails to explain the cross section of average returns in emerging markets.
Bekaert and Harvey (1997) use liberalization dates to examine the behavior of volatility
in emerging markets. The time series and cross-sectional models analyze the reasons
that volatility is different across emerging markets, particularly with respect to the timing
of capital market reforms. This study finds that capital market liberalization often
increase the correlation between local market returns and the world market but do not
drive up local market volatility.

The proponents of complete market segmentation highlight the overriding role of
domestic macroeconomic variables in their investigation of return volatility. Again most
of these studies are predominantly focused on developed markets (Chen et al., 1986;
Fama; 1990; Jorion, 1991; Ely and Robinson, 1997). It has been argued that in less
integrated markets, correlation with the world portfolio is weaker and according to
Harvey (1998, 2000) only local market variance explains the cross-section in emerging
markets.
Chen, Roll and Ross (1986) studying whether innovations in macroeconomic variables
affect stock returns, find that the spread between long and short run interest rates,
expected and unexpected inflation, industrial production and the spread between high
and low grade bonds were found to systematically affect the stock market returns.
FAMA (1990) finds that U.S. stock returns and its aggregate real activity are correlated.
Ely and Robinson (1997) concludes that stocks do maintain their value relative to
movements in overall price indexes and this conclusion generally does not depend on
whether the source of the inflation shock is from the real or monetary sector.

Empirical studies that have investigated the level of integration have generally found
that emerging stock markets are only partially integrated. Since the markets are neither
perfectly segmented nor perfectly integrated, investigating the role of macroeconomic
variables at the two extremes may lead to inconclusive results. Hence, the third
research strand calls for consideration of both domestic and global variables in the
study of emerging markets.
The current study examines the relevance of macroeconomic variables in the
framework of partial integration. In other words, the study investigates the effects of
both country and global variables on the volatility of emerging market stock returns.

This study investigates the following questions regarding the relationship between
macroeconomic factors and stock returns in Pakistans emerging equity market. First,
are the mean effects and volatility shocks of macroeconomic variables transmitted to
Karachi Stock Market (KSE).And if so, are the shocks persistent. The presence of
Rizwan & Khan
364

persistence in volatility clustering may signify inefficiency of the market, and provide the
possibility of arbitrage. Second, whether macroeconomic shocks and global factors
have any significant asymmetric effects on the returns volatility of the equity market.
Ross (1989) points out that volatility is directly linked to the rate of information flow.
Hence, positive and negative macroeconomic shocks may be transmitted differently.

This study is important for domestic and foreign investors in terms of risk management
and portfolio diversification strategies for many reasons. First, the accurate
measurement of the volatility of financial data is important because economic agents
make decisions based on the perceptions that high levels of volatility of financial tend to
cause the general erosion of investors confidence and a flow of capital away from the
stock markets (Schwert, 1989, Ballie and DeGennaro, 1990). Second, answers to these
questions may be important for decision-making by national policy makers as well as
our understanding of how such decisions affect the stock market of the country.
Analyzing specific macroeconomic fundamentals can also help us understand the
extent to which differences in performance of stock market is justified by economic
fundamentals. In addition, the predictability of volatility is important in designing optimal
asset allocation decisions as well as dynamic hedging strategies for options and futures
(Baillie and Myers, 1991). Therefore understanding the fundamentals underlying
volatility is also important for the valuation of stocks and their corresponding derivative
products in stock market.


2. Data and Descriptive Statistics

Karachi Stock Exchange, established in 1950, is the largest and the most active of the
three stock exchanges of Pakistan. Other regional exchanges are in Islamabad and
Lahore which are relatively inactive. KSE currently lists 662 companies with a market
capitalization of over Rs. 52 billion. The KSE-100, a capital weighted index, represents
major blue chips companies and is fairly representative of the market.

This study uses monthly returns on stock index. The month-end closing values of the
KSE-100 index were obtained from online database of Yahoo finance for the period
from July 2000 to Jun 2005. The market returns are calculated by taking first difference
of logs of two consecutive month-end closing values.

2.1 Country and Global Variables

The domestic macroeconomic variables used for the study are Money supply, Monthly
CPI, Industrial production, Exchange rate and Interest rate, while 6-month LIBOR and
Morgan Stanley Capital International (MSCI) All Countries World Index are used as
global variables. The data on these macroeconomic variables were obtained from
monthly publication of International Monetary Fund (IMF). Data on all macro economic
variables cover the period July 2000 to Jun 2005.



Rizwan & Khan
365

2.2 Descriptive Statistics

To assess the distributional properties of the monthly stock returns various descriptive
statistics are reported in Table 1 that includes mean, variance, standard deviation,
kurtosis, skewness. This study covers a period from July 2000 to June 2005. The mean
monthly return is 2.55 percent and the standard deviation is 8.7 percent which indicates
high volatility of returns that characterize the market as one of the emerging markets.
The deviations from the normality in the returns can be seen from the excess positive
kurtosis which implies that the underlying distribution has long right tails. Kurtosis of
normal distributions is said to be equal to 3. If the kurtosis exceeds 3, the distribution is
peaked relative to normal (i.e., leptokurtic). Therefore, stock returns examined in the
study exhibit non-normality.


3. Econometric Models

As shown by descriptive statistics in Table 1, the distribution of the return series does
not follow a normal distribution; the nature of the distribution suggests volatility
clustering, where large returns are followed by large returns and small returns tends to
be follow by small returns, leading to contiguous periods of volatility and stability. Such
data require appropriate econometric modeling techniques in order to ensure proper
interpretation and conclusion.
Engle (1982) developed a model to capture time-varying variance- the Autoregressive
Conditional Heteroscedastic (ARCH) approach. The basic ARCH model has led to other
related formulations describing the evolution of the variance of time series, such as the
Generalized ARCH (GARCH) model. Among the various formulations of the GARCH
models, the EGARCH approach has been identified as the most appropriate model for
stock indexes. Moreover, parameter restrictions are not necessary because EGARCH
models the log of conditional variance, thereby guaranteeing that the variance will be
positive. Furthermore, Engle and Ng (1993), report that asymmetric models such as
EGARCH provide the best forecast of volatility. Based on the conceptual underpinnings
of the study, the hypothesized questions and the past performance of the exponential
GARCH model, this study uses a multivariate EGARCH model.

4. Unit Root tests

It is important to check whether the data series is stationary before using the Vector
Auto Regressive (VAR) and the EGARCH Models. Engel and Granger (1987) suggest
different unit root tests. Augmented Dickey Fuller (ADF) test is believed to be the most
powerful test among them. The ADF test examines the unit root of the observed data by
taking the unit root (non stationarity) as the null hypothesis. The rejection of H
o
implies
that the series X
t
is stationary. Table 2 reports the results of the ADF test. Since the
ADF test statistic for monthly stock returns (7.83) is less than the critical value, the null
hypothesis of a unit root is rejected. Therefore, monthly stock returns series are
stationary. The result of the ADF tests for the countrys macroeconomic and the global
factors are also less than the critical values, and are therefore, stationary.
Rizwan & Khan
366

5. Multivariate EGARCH Model

The multivariate Exponential Generalized Auto Regressive Conditional
Heteroscedasticity (EGARCH) model used in this study is based on one estimated by
Christofi and Pericli (1999). Let R
t
be the logarithmic market return at time t for j
macroeconomic variables, given that j=1, 2, 3, 4, 5, 6, 7 (1= Exchange rate, 2=
Domestic interest rate, 3= Industrial production, 4= Money supply, 5= Monthly CPI, 6=
MSCI world index, 7= 6-months LIBOR),
t
and
2
t
the conditional mean and conditional
variance respectively, and k is the lag length.
t
is the innovation at time t (i.e.
t
= R
t
-
t
)
and Z
t
the standardized innovation (i.e. Z
t
=
t
/
2
t
) at time t. The model is specified as
follows:
) 1 (
2
1
) (
7
1
0
+ + =

= =
t k t
k
k j
j
t
R c | |
) 2 ( ) ln( exp
2
1
2
1
) (
7
1
0
2

(

+ + =

=

=
t
k
k t k j
j
T
X o o o o

( ) | | ) 3 (
1 . 1 . 1 . 1 .
+ =
t j j t j t j t j
Z Z E Z X o
Equation (1) models the conditional mean return of country and is specified as a
function of the lagged values of its exchange rates, interest rates, industrial production,
money supply, the MSCI world index, CPI and the 6-months LIBOR rates. The residuals
from first-order autoregressions fitted to these variables are used as proxies for shocks
to returns. This specification is meant to investigate the first question of the study
whether the mean shocks of macroeconomic variables are transmitted to the stock
market returns. The coefficient
j
measures the degree of effects across
macroeconomic variables. A significant
j
coefficient would imply that variable j leads
market or equivalently, that current returns in variable j can be used to predict future
returns in market.
The hypothesis tested is that each macroeconomic variable is important in explaining
the returns of a country. So the null hypothesis here is that:
H
o
:
i
= 0, for all i
Eq. (2) describes the conditional variance process as an extended EGARCH process,
which allows the testing and measuring of the asymmetric impact of its own, and, the
domestic and global macroeconomic variables past standardizes innovations on the
conditional variance of a market. The specification here examines the research question
regarding the transmission of macroeconomic volatility shocks and whether they are
persistent over time. The coefficient
j
in Eq. (2) captures the effect of innovations from
macroeconomic variable j to market. This coefficient specifically measures the
significance of past volatilities (standardized residuals) in each macroeconomic variable
on the conditional variance of the equity returns.
Significant
j
implies that the past volatilities of a macroeconomic variable impact the
conditional volatility of the equity returns. The null hypotheses tested here are that past
volatilities in the js do not impact the conditional variance of the equity returns.
H
o
:
i
= 0 for all i
The coefficient in Eq. (2) measures the persistence in volatility. A high value suggests
that an information shock tends to persist for some time into the future. The presence of
persistence in volatility clustering may imply the inefficiency of the market. The variable
Rizwan & Khan
367

X
j,t-1
is specified in Eq. (3) and is intended

to capture the asymmetric effect of past
standardized innovations, Z
j,t-1
(Z
j,t-1
=
j,1-t
/
j,t-1
), on current volatility. This specification
is intended to investigate, whether there are asymmetric effects in the transmission of
the volatilities of macroeconomic variables. The term |Z
j.t-1
| - E(|Z
j.t-1
|) measures the size
effect, while the term
j
Z
j.t-1
measures the sign effect. If the lagged values of the market
and/or macroeconomic variable advance and decline impacts volatility symmetrically,
the coefficient
j
would not be expected to be significant. If declines in macroeconomic
variable j (Z
j,t-1
>0) are followed by higher (lower) volatility than the variables advances
(Z
j,t-1
>0), then
j
would be expected to be negative (positive) and significant. Assuming
that
j
is positive, the larger the deviation of past standardized innovation from its
expected value, the larger the impact (positive or negative) on the current variance. The
null hypothesis tested is that:
H
o
:
j
= 0 and
j
= 0 for
j
> 0 and
j
> 0


6. Empirical Results of the EGARCH Model

The exponential GARCH results for the relative effects of country and global factors on
stock returns are reported in Table 3. It reports the empirical results for Karachi Stock
Exchange past returns, modeled as lagged returns in the GARCH model are positive
and significant in explaining current returns. This result is consistent with the volatility
clustering phenomena of most emerging markets. That is, higher periods of returns are
followed by higher returns and vice versa. Table 3 also reports that the industrial
production coefficient is positive but insignificant. In fact most previous studies have
found low or no effect of industrial activity on security prices. Hardouvetis (1987), for
instance, concludes that financial markets respond primarily to monetary and not real
activity news. So the failure to detect any significant effect may be due to the fact that
the impact horizon is short under this model. The money supply variable is negative and
significant (-1.56308). The reason may be that Pakistan has experienced higher levels
of inflationary trends that tend to depress stock prices. The coefficient for CPI is
negative and significant (-4.63263). The reason may be that higher inflation uncertainty
increases risk, the discount factor and so leads to decline in stock prices. The table also
reports that the stock market is positively impacted by the global factor; MSCI World
Index. The coefficient for MSCI world Index is (0.638144) positive and significant. This
finding suggests that Pakistans stock market have increased interaction with the global
economy ever since it opened up for foreign investors. While coefficients for exchange
rate and interest rate are negative but insignificant. This is because currency exchange
rate and interest rate in emerging markets like Pakistan is controlled by government
policies. This situation means that the exchange rate and interest rate are less
predicting factors compared to the situation in developed markets. The fundamental
economic factors such as exchange rate, deposit interest rate, have been found to be
more predicting factor in case of developed stock markets, this is due to the fact that
fundamental factors are tied and controlled in the emerging economies and may be less
relevant to predict stock volatility.


Rizwan & Khan
368

7. Analysis of the Conditional Variance Equation

Table 4 reports the estimated coefficients of the conditional equations and shows MSCI
World Index, Money Supply, CPI and LIBOR have significant impact on the stock price
volatility. The implication of this result is that increases in the volatilities of these country
and global variables leads to increases in the volatility of the Pakistans stock market.
While the coefficients of the exchange rate, interest rate and industrial production are
not significant in explaining the volatility in the stock market. The reason may be that in
emerging markets, investors in financial markets respond mainly to monetary news
instead of real activity news. Similar studies of interest rate volatility on Brazil and
Argentinas market also found insignificant results for interest rates volatility. The money
supply (-1.733) and the CPI (-0.0057) variables are negative and significant. A possible
explanation may be that high levels of inflation tend to depress stock prices. The table
also reports that the coefficients of MSCI World Index, Exchange rate and interest rate
not significant in explaining stock market volatility.


8. Analysis of Asymmetric Component

The EGARCH proposed by Nelson (1991) estimates the conditional variance as a
function of standardized innovations and allows the conditional variance to respond
asymmetrically to positive and negative innovations. The estimates from the multivariate
EGARCH model shows that the country and global variables do not impact stock prices
symmetrically as has been assumed by some previous studies. Table 4 (5.3 reports that
the symmetry effect hypothesis is rejected in case of Pakistans stock market
suggesting that macroeconomic variables do exhibit asymmetric effect. This asymmetric
coefficient is (-.7223747) negative and is significant at 5% level and the negative sign
indicates that negative news about macroeconomic variables affect stock prices more
than positive news. The results also suggest that stock market sensitivity to
macroeconomic news depends on the size and sign of the surprise in the news. The
negative sign of the asymmetry coefficient suggests that bad economic news is followed
by higher stock market volatility than good economic news.



9. Multivariate Vector Auto Regressive Model

Developed by Sims in 1990, the VAR model is based on multivariate time series
analysis, in which a variable, R
t
, is stated as a function of the past history of R
t
and the
past history of other variables (R
2
.R
n
) that influence R
t
. The following multivariate
VAR model is used following Granger (1969).
i i t
n
i
j
m
i
i t i t
K R R c | o + + + =

= =

1 1
1

Where R
t
is monthly return at time t for stock index., and
i,

j,
are parameters, m and n
are the lag lengths for monthly stock index returns and the macroeconomic and global
factors respectively to be used in the equation. The above-mentioned Granger (1969)
Causality test is designed to examine whether time series move one after the other or
Rizwan & Khan
369

contemporaneously. When they move contemporaneously, one provides no information
for characterizing the other. If some of the
j

values are statistically not zero, then K is
said to Granger-cause stock returns. If
j
parameters are statistically equal to zero,
then economic variables are not impacting the stock returns. A standard F-test can be
applied to test the null hypothesis that economic variables and the global variables fail
to Granger cause the stock returns.
H
o: 0 =
j

, for all j macroeconomic variables
Using the Akaikes (1969) Final Prediction Error criterion for determining the auto-
regressive lag length, the equation is estimated for m = 1 and n = 1.

Table 5 reports the VAR coefficients for the response of the stock market to the shocks
from the country and global factors. The results show that MSCI World Index, Money
Supply, CPI, LIBOR are important variables in explaining stock price performance, as
the coefficients of these variables are significant. These variables significantly impact
the performance of the stock prices.


10. Conclusion and Implications

A number of previous studies have reported that a relationship between macroeconomic
variables and equity market returns. However, most of these studies have typically
focused on developed market. This paper extends the literature by investigating
whether there are relationships between macroeconomic variables and the stock
returns in an emerging market like Pakistans stock market. This paper uses two
multivariate models, multivariate EGARCH and Vector Auto Regressive (VAR) models
to investigate the effect of exchange rate, interest rate, industrial production, money
supply, Morgan Stanely Capital International (MSCI) World Index and 6-month LIBOR
on stock prices in Karachi Stock Exchange. The estimate shows that domestic
macroeconomic variables have varying degrees of importance in explaining the
relationship between stock returns and volatility in KSE. The empirical results show that
the two global, MSCI World Index and 6-months LIBOR, variables used in this paper
explains the stock returns in KSE. An important conclusion drawn from the results is
that macroeconomic variables exhibit asymmetric effects on returns volatility. In
particular the results show that for KSE the past returns, modeled as lagged returns in
the GARCH model are positive and significant in explaining current returns. This result
is consistent with the volatility clustering phenomenon of most emerging markets.
The estimated coefficients of EGARCH models show that stock returns respond
significantly to money supply, CPI and LIBOR and MSCI World Index. Similar results
are found for the VAR model.
The estimated coefficient of EGARCH conditional equation shows that only money
supply, CPI and LIBOR volatility in Pakistan has significant impact on stock price
volatility. The implication of this result is that an increase in money supply and CPI
volatility in Pakistan leads to an increase in volatility of the stock market. The industrial
production coefficient as reported by VAR is positive but not significant. Previous
studies have found low or no effect of industrial activity on security prices. Hardouvelis
Rizwan & Khan
370

(1987), for instance concludes that financial markets respond primarily to monetary and
not real activity news.
The measure of persistence of volatility shocks in the EGARCH model suggests that
news information in this market does not die quickly but is persistent over time. The
symmetry effect hypothesis is rejected suggesting that macroeconomic variables do
exhibit asymmetric effect in Pakistans stock market. The negative signs indicate that
negative news about macroeconomic variables in Pakistans stock market affect stock
prices more than positive news do. The result suggests that bad economic news is
followed by higher stock market volatility than good economic news. The finding is
consistent with the findings in some other studies. Overall, the results show that
Pakistans stock market is partially integrated as shown by the significant role of both
country and global factors. The above conclusions have important implications for both
investors and policy makers.

11. Implications

The results reported in this study suggest that macroeconomic variables are important
with regard to stock market volatility in Pakistan. International investors can therefore,
improve their portfolio performance by considering the stability in economic
fundamentals as determinants of stock market volatility. Policy makers, on the other
hand, can concentrate their efforts to attain stability in economic fundamentals in order
to reduce volatility and minimize investor uncertainty. Policy makers should be sensitive
to asymmetric effects of volatility in the market. Effective macroeconomic management,
market transparency and availability of information can provide a market environment
with low asymmetric information and its related problem.



References

Baille, R. T., Myers, R. 1991. Modelling commodity price distributions and estimating
the optimal futures hedge, Journal of Applied Econometrics 6,109-124.

Harvey, C.R., 1997, Emerging equity market volatility, Journal of Finance 55, 565-613.

Bekaert, G., 1995, Market integration and investment barriers in emerging equity
markets. World Bank Economic Review 9, 75-107.

Ballie, R. T., DeGennaro, R. P. 1990, Stock returns and volatility, Journal of Financial
and Quantitative Analysis 25, 203-215.

Chen, N., Roll, R., Ross, S. 1986, Economic forces and the stock market, Journal of
Business 39, 383-403.

Christofi, A., Pericli, A., 1999, Correlation in price changes and volatility of major Latin
American stock markets. Journal of Multinational Financial Management 9, 79-93.

Rizwan & Khan
371

Dumas and Solnik, 1995 Solnik, B., 1995, The world price of foreign exchange risk,
Applied Economics Letters 3, 121-123.

Ely, D.P., Robinson, K. J, 1997, Are stock a hedge against inflation? International
evidence using a long-run approach, Journal of International Money and Finance 16,
141-167.

Engle, R. F. Ng, V. K. 1993, Measuring and testing the impact of news on volatility,
Journal of Finance 48, 1749-1778.

Fama, E. F. 1990, Stock returns, expected returns, and real activity, Journal of Finance
1089-1109.

Ferson, W. E. and Harvey C.R., 1994a, Sources of risk and expected returns in
international equity markets, Journal of Banking and Finance 18, 775-803.

Ferson W. E., Harvey, C.R. 1993, The risk and predictability of international equity
returns, Review of Financial Studies 6: 527-566.

Ferson W. E., and Harvey, C.R. 1994b, An exploratory investigation of fundamental
determinants of international equity returns. In Frankel, J.A. Internationalization of
Equity Markets, Univrsity of Chisago Press (ISNB 0-226-26001-1), pp. 59-148.

Granger, C. W. J. 1981, Some Properties of Time Series Data and Their Use in
Econometric Model Specification, Journal of Econometrics, 16: 121-30

Hardouvelis, G. 1987, Macroeconomic information and stock prices, Journal of
Economic and Business, 131-139.

Harvey, C. R. 1995, The risk exposure of emerging markets, World Bank Economic
Review 9, 19-50.

Harvey, C. R. 1998, The future of investment in emerging countries. NBER Reports
(Summer), 5-8.

Harvey, C. R. 2000, The derivers of expected in international markets, Emerging
Markets Quarterly 3: 32-49.

Johnson, P. 1991, The pricing of exchange rate risk in the stock market, Journal of
Financial and Quantitative Analysis 26: 363-376.

Nelson, D.B. 1991, Conditional heteroscedasticity in asset returns: a new approach,
Econometrica 59: 347-370.

Ross, S.A. 1989, Information and volatility: The no-arbitrage Martingle approach to
timing and resolution irrelevancy, Journal of Finance 44: 1-17.
Rizwan & Khan
372


Schwert, G.W. 1989, Why does stock market volatility change over time? Journal of
Finance 44 (5): 1115-1154.
Rizwan & Khan
373

Appendices


Tables


Table 1
Summary Statistics of Monthly KSE Returns

Mean 0.025528
Median 0.021724
Maximum 0.241114
Minimum -0.15455
Std. Dev. 0.087079
Skewness 0.29688
Kurtosis 2.771848




Table 2: Augmented Dickey Fuller Test Results for stock returns and
macroeconomic variables

KSE Returns -7.838
*
1% Critical Value -3.569
MSCI World Index -6.717
*
5% Critical Value -2.924
Money Supply -8.429
*
10% Critical Value -2.597
Exchange Rate -3.904
*

CPI -6.219
*

LIBOR -3.941
*

Interest Rate -7.792
*

* shows significance at 1%










Rizwan & Khan
374

Table 3: Mean Effects of Macroeconomic Variables on Stock Returns

) 1 (
2
1
) (
7
1
0
+ + =

= =
t k t
k
k j
j
t
R c | |


Coefficient Std. Error t-Statistic
Past Returns 0.022553
**
0.012718 1.77323
Industrial
Production
0.00076 0.000826 0.920069
CPI -4.63263
*
1.362635 -3.39976
Exchange Rate -1.36637 0.969568 -1.40925
Interest Rate 0.003296 0.158029 0.020855
LIBOR 0.51002
*
0.115744 4.406457
Money Supply -1.56308
*
0.516928 -3.02378
MSCI World Index 0.638144
*
0.213747 2.985511
C 0.129876
*
0.032461 4.000937
R
2
0.213745
Note:
**
shows significance at 10% and
*
at 5% level.























Rizwan & Khan
375


Table 4: EGARCH RESULTS

) 2 ( ) ln( exp
2
1
2
1
7
1
0
2

(

+ + =

=

=
t
k
k t k
j
T
X o o o o

Co-eff. Std. Err. t-statistic
MSCI World Index 0.31387 0.26243 1.2
Money Supply -2.32176
*
0.79826 -2.91
Exchange Rate 0.19775 0.98254 0.2
CPI -3.64433
*
1.67786 -2.17
LIBOR 0.40508
*
0.14652 2.76
Interest Rate -0.06653 0.14909 -0.45
Asymmetric effect -0.72237
*
0.34156 -2.11
-0.84674
*
0.22810 -3.71

0
-9.94651
*
1.27097 -7.83
ARCH + GARCH terms 0.70
Note:
*
shows significance at 5% level.

Table 5: Results of Vector Auto Regressive Model

i i t
n
i
j
m
i
i t i t
K R R c | o + + + =

= =

1 1
1


F-Statistic 2.81
*
R
2
0.31
Coef. Std. Err t
1
o 0.116614 0.3588285 0.32
KSE Returns -0.11335 .1136981 -1.00
MSCI World Index 0.678133
*
.2638073 2.57

Money Supply -1.73345
*
.7685238 -2.26

Exchange Rate 0.006148 .0067167 0.92
CPI -0.0057
*
.002718 -2.10

LIBOR 0.320005
**
.1677306 1.91

Interest rate 0.00212 .0070671 0.30
Industrial Production 9.92E-05 .0002094 0.47
Note:
*
shows significance at 5% and
**
at 10% level.