Research

Research Relationship between Stock Market Volatility . . .

RELATIONSHIP BETWEEN STOCK

MARKET VOLATILITY AND
MACROECONOMIC VARIABLES:
EVIDENCE FROM PAKISTAN
Muzammil Hussain1, Bedi-uz-Zaman2 and Nisar Ahmad3

Abstract

This study explores the relationship between stock returns

volatility and macroeconomic variables in Pakistan. This study
has used monthly observations covering the period from 2001-01
to 2011-06. First, Exponential Generalized Autoregressive
Conditional Hetroskedasticity (2, 2) model is used to analyze the
volatility in stock returns. Graph of news impact curve shows that
higher risk is contributed toward negative shocks in stock market
as compared to positive shocks of the same magnitude. In the next
step the researcher has explored the macroeconomic determinants
of stock market volatility through ARDL approach because
variables are I (0) in addition to I (1). Results from ARDL approach
revealed that macroeconomic variables are responsible factors in
explaining stock market volatility. Inflation, real exchange rate
and oil prices are found encouraging factors of stock market
volatility while Industrial sector output and real supply of money
affects the volatility negatively.

Keywords: Stock Market Volatility, Inflation, Real Exchange Rate,

EGARCH and ARDL
JEL Classification: E 440

1- Department of Economics, University of Sargodha, Pakistan

2- Department of Economics, University of Sargodha, Pakistan
3- Department of Economics, University of Sargodha, Sub-Campus
Bhakkar, Pakistan

723 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

Introduction

A stock market is an organization or institution recognized

for dealing in securities whether integrated or not. The stock exchange
provides a physical place for investors to trade their stocks and it is a
significance source to raise funds. Affiliation between the stock
markets is increasing obviously from the preceding decade with the
incorporation of general economies in the course of international trade,
flow of capital and technological advancement (Chan et al., 1998).
Volatility is a variable which swamps major monetary tools and performs
a vital job in the numerous fields of finance. Volatility in stock returns
refers to deviation in stock prices varies throughout a time (Zafar et
al., 2008). Stock market progress relies on the health of financial system,
macroeconomic solidity and also disturbs with the outside markets
(Aliyu, 2012). Officer (1973) is one of the pioneer researchers who
linked stock price volatility with economic indicators. He found a high
volatility in the times of great depression in 1930s.

Generally stock markets of both developed and

underdeveloped economies are volatile (Choo et al, 2011). The
investors are interested to identify the nature of volatility. Different
studies have shown asymmetric relationship between stock returns
and volatility. Good news and bad news have different impacts on
volatility (Campbell and Hentschel, 1992). Karolyi (2001) argued that
decrease in stock prices lead to increase volatility in stock market.
Bollerslev et al. (1994) and brooks (2008) found that negative shocks
contributed to more volatility as compared to positive shocks of the
identical magnitude.

Higher stock market volatility in recent years has enhanced

the conversation on stock price movements in developed countries in
general and particular in case of developing countries including
Pakistan. Unlike the established equity markets of developed
economies, the stock markets of Pakistan begin to broad quickly and
responsive to issues like changes in economic activities, political

PAKISTAN BUSINESS REVIEW JAN 2015 724

Research Relationship between Stock Market Volatility . . .

environment and macroeconomic variables. Several researchers

underlined their consideration on the stock markets of emerging
economies because stock markets of Asia provide good opportunities
for foreign investment (Chiou-wei, 2011). Therefore, modeling stock
market volatility is a very essential aspect in the developed as well as
emerging economies.

The dynamics of stock market volatility is somewhat

comprehensive and has a strategic meanings. Due to immense
compass of volatility various issues of stock markets are discussed
by researchers. General objective in this research is the investigation
of the relationship between stock return volatility and macroeconomic
variables. However, the following specific objectives are also framed
in this study.

To investigate whether the nature of volatility is symmetric

or asymmetric (impact of good news and bad news on stock market
volatility).

To explore the relationship between stock market volatility and

macroeconomic variables in Pakistan.

The evaluation of major assumptions of stock market and

looking at the Pakistan history provides help for making choice about
the appropriate variables and building econometric model to estimate
the determinants of stock returns volatility in Pakistan. After review
of information and objectives the following hypothesis are set:

H1; Stock return shocks have asymmetric effect on stock market

volatility.
H1; There exists a significant relationship between stock market
volatility and macroeconomic variables in Pakistan.

725 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

Capital Markets in Pakistan

The role of stock market is very important for any country.

An efficient asset market improves the pecuniary strength of country
by means of valuable management of resources. Stock market in
Pakistan started in 1947 with the establishment of Karachi Stock
Exchange (KSE). The other two stock exchange markets operating in
Pakistan are Lahore stock exchange (LSE) and Islamabad stock
exchange (ISE). LSE was set up in 1970 and ISE came into existence in
1989. KSE is the largest and most liquid exchange market operating in
Pakistan where approximately 85% trade takes place. Nearly 670
companies are scheduled in KSE comprising a market capitalization of
more than US$95.18 Billion (Rafique and Rehman, 2011). The companies
belong to KSE correspond to different sectors of economy.

KSE 100 index is the most important standard to evaluate

prices at Karachi stock exchange. The index is constructed with the
stocks of 100 companies. It is a capital weighted index and the
companies with maximum market capitalization are chosen. KSE
provides data for a reasonable time period and it is also known as a
well recognized market of emerging economy. The KSE was declared
as the best performing world stock market in 2002 (Business Week)4.

Literature Review

According to Fisher’s Hypothesis, the market rate of interest

included the projected inflation and expected real rate of interest (Fisher,
1930). As nominal rate of interest and rate of inflation moved one-to-
one, then, real rate of interest was not affected by a permanent change
in inflation rate in the long-run. Thus, it was concluded that stock
returns and rate of inflation moved in the same direction. The
relationship between exchange rates and stock returns is based on a
simple financial theory. When the domestic currency decrease in

4-The international magazine

PAKISTAN BUSINESS REVIEW JAN 2015 726

Research Relationship between Stock Market Volatility . . .

value against foreign currencies, prices of export products decline,

and consequently, the volume of the country’s export will increase
a devaluation of the domestic currency has a negative relationship
with return.

Liljeblom and Stenius (1997) examined the relationship

between macroeconomic volatility and stock market volatility for
Finland. GARCH model was used for the estimation of conditional
volatility at stock market. This study found a significant linkage
between macroeconomic volatility and stock market volatility and a
predictive power in both direction.

Beltratti and Morana (2002) studied the relationship between

stock market volatility and macroeconomic volatility. The variables
used in this study were S&P 500 index, industrial production, CPI,
federal fund rate, treasury bills and ten-year treasury bonds. A
significant finding of this research was that there exist a significant
relationship between stock market and macroeconomic volatility.

Chowdhury et al. (2006) explored how predicted

macroeconomic volatility is related to the stock market volatility in
Bangladesh. The study used monthly data on stock prices, CPI,
exchange rate and industrial production for 1990:01-2004:12. The study
used GARCH model and found that there exists unidirectional causality
from industrial production volatility to stock return volatility. This
study also found direction of causality from stock return volatility to
inflation volatility.

Saryal (2007) examined the impact of inflation on the

conditional stock market volatility in Turkey and Canada. GARCH (1,
1) and Quadratic GARCH model were employed for modeling stock
market volatility. This study found that the greater the inflation rate,
the higher is the volatility in the stock market.

727 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

Wang (2010) examined the linkage among macroeconomic

volatility and stock market volatility for China. The study used monthly
data on real stock return, real GDP, CPI and interest rate from the
period 1992 to 2008. This study investigated the time-series relationship
using EGARCH and lag-augmented VAR models. The study found no
causal relationship between stock market volatility and real GDP
volatility, bilateral causality between stock market volatility and
inflation volatility and uni-directional causality from stock prices to
interest rate.

Choo et al. (2011) tried to examine the relationship of stock

market volatility with some macroeconomic variables in Japan. The
study used daily data on Nikkei 225 index, currency exchange rate, oil
price and gold price from May 1997 to July 2009. The study employed
ten dissimilar models for forecasting. The study found no impact of
macroeconomic variable on the volatility of Japan’s stock market and
better results were obtained through simple GARCH (1, 1) model and
suggested that GJR GARCH model was better in predicting volatility
than simple GARCH (1, 1) model. This study recommended that
macroeconomic variables did not improve the forecasting accuracy of
GARCH (1, 1) which showed that macroeconomic uncertainty did not
explain the volatility of Nikkei 225 Index.

Aliyu (2012) investigated the inflationary force in explaining

volatility and returns of stock market. Additionally, this study explored
the effect of asymmetric shocks using quadratic GARCH model.
Variables of interest in this study comprised CPI and stock price index
of the relevant country. The investigation covered the time of 1998:01
to 2010:05 and 1999:12 to 2010:05 for the Nigeria and Ghana respectively.
GARCH (1, 1) model was used to model the volatility in the stock
markets of both states. Outcomes showed that the impact of bad news
on stock volatility was larger than good news for the case of Nigeria
stock exchange (NSE), while for Ghana stock exchange (GSE) results
were different. Study also proved the inflation as an important
determinant of stock volatility in both the markets.

PAKISTAN BUSINESS REVIEW JAN 2015 728

Research Relationship between Stock Market Volatility . . .

Falkberg (2012) analyzed the impact of macroeconomic

variables on volatility and returns of Standard and Poor’s (S & P) 500
index. The other seven variables used with the exception of seasonal
dummies were: default spread, inflation volatility, Industrial production
volatility, slope of the yield curve, implied volatility and volatility of
3-months treasury bills. The data used for this study was collected
on monthly basis ranges from 1957:01 to 2011:08. Empirical outcomes
were obtained by using VAR, granger causality test and some
additional econometric techniques. The results also showed the
presence of seasonal patterns plus asymmetric volatility. According
to the results of this study no relationship was found between
macroeconomic volatility and stock market volatility.

Research Methodology and Data Description

The formulation of accurate econometric methodology and

apposite data handling are believed as the heart of every research
work. Sample selection bias and selection of inappropriate estimation
technique lead to biased results. The study used secondary data to
scrutinize the determinants of stock market volatility in Pakistan.
Equation 1 is the general form of model estimated in this research.
HT=b0+b1(GMP)t-i+b2(RM2)t-i+b3(OP)t-i+b4(REER)t-i+b5(INF)t-i+ut……………… 1

Here, stock returns volatility (HT) is the dependent variable

in this study and the other independent variables are growth rate of
industrial production (GMP), real supply of money (RM2), oil prices
(OP), real effective exchange rate (REER) and inflation rate (INF).
Data on these variables is used on monthly basis and it ranges from
2001:01 to 2011:06. The researcher has collected statistics from different
data sources like statistical bulletin5 (State Bank of Pakistan), hand
book of statistics on Pakistan economy6 (2011) and International
Financial Statistics (IFS).

5-http://www.sbp.org.pk/reports/stats_review/bulletin2011
6-http://www.sbp.org.pk/department/stats/PakEconomy_HandBook/2011

729 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

Modeling Conditional Variance

Engle (1982) in macroeconomic analysis originated

unbalanced variance for a few categories of data. The case of
uncertainty in stock returns is calculated via variances, and it varies
with time. The researchers focus on hetroskedasticity in dealing with
time series investigations.

Autoregressive Conditional Hetroskedasticity (ARCH) Model

Engle (1982) recommended ARCH model as a choice to the

typical time series handling. The model allows the conditional variance
to vary with time and implies that residual variance at present time rely
on the precedent squared error term. ARCH (q) model examines the
mean and variance as follow:

n
πt = π0 + πi Xt-i +εt …………………. 2
i=1

Xt-i and signify k x 1 vector of independent variables and

coefficients respectively. Ɛt t is independently distributed residual
term.

q
ht = γ0 + γjε2t-j ……..………………………. 3
j=1

Equations 2 and 3 are mean and variance equations

respectively. One shortcoming of the ARCH model according to Engle
(1995) is that it resemble extra moving average pattern than auto
regression.

PAKISTAN BUSINESS REVIEW JAN 2015 730

Research Relationship between Stock Market Volatility . . .

Generalized ARCH (GARCH) Model

Bollerslev (1986) introduced GARCH model. Model permits

conditional variance to depend on its own lag value. Bollerslev says
that volatility depends on both AR and MA terms. GARCH (p, q)
model can describe as in equation 4:
p q
ht = γ0 + δi ht-i + γ jε2t-j ………………………4
i=1 j=1

Exponential GARCH (EGARCH) Model

Nelson (1991) planned EGARCH model. Variance equation

of EGARCH model can be expressed in different ways. The model is
superior to GARCH model because it ignores the non-negativity
constraint and it doesn’t impose any constraint on the parameters.
EGARCH also explores the impact of bad innovation that is very
important in financial markets.

q q p
εt-j εt-j
logh t = γ + α j +β j + δilog(h t-i ) ……… 5
j=1 ht-j j=1 h t-j i=1

In the variance equation, , and are the parameters. On the

left side of equation log of series is taken to compose exponential
leverage effect. The model is symmetric for:
Here, if it represents more impact
of negative news than positive.
Time Series Analysis

This study first checked stationary of data in order to avoid

the prospect of spurious results. Advancement in econometrics with
the passage of time expose that some of the time series are non
stationary and to scrutinize such data with ordinary least square
(OLS) leads to incorrect conclusion. Box and Jenkins (1970, 1976)
731 PAKISTAN BUSINESS REVIEW JAN 2015
Relationship between Stock Market Volatility . . . Research

devised regressions at first difference stationary data with the rationale

of spurious outcomes and supposed that by differentiating a series
again and again non-stationary series transforms into stationary series.
This method is defeat of costly information and for that reason
Davidson et al., (1978) considered it the foremost weakness of this
procedure.
Dickey and Fuller provided augmented edition of test to remove the
problem of autocorrelation. They used additional lag term of dependent
variable to solve the problem. AIC and SBC are used to determine the
optimal lags. Three possible shapes of ADF test are given below. The
diversity in the three equations is of elements and where an
intercept term is and represent trend in a series.
n
ΔYt = αYt-1 +  β k ΔYt -k + ε t
k =0
n
ΔYt = δ 0 - αYt-1 +  β k ΔYt-k + ε t
k=0

n
ΔYt = δ 0 - αYt-1 + δ 2 t +  β k ΔYt-k + ε t
k =0
Autoregressive Distributed Lag (ARDL) Approach for Cointegration

ARDL model is initiated by Pesaran and Shin (1999) and

broaden by Pesaran et al. (2001). The model is useful for several
reasons. It is not necessary for all the variables to be I(1) like Johansen
technique. The model is appropriate if some variables are I(0) and
others I(1). Estimation of ARDL involves two major stages. It tests
long run relationship at initial stage and in the second stage long run
and short run coefficients are estimated.

The General Form of Unrestricted ECM model in ARDL (p,q,r,x,y,z)

formulation
p q r x y
dH Ti  a 0   B i dH Tt  i   C dG M P
i ti   D i dR M 2 t  i   E i dop t  i   Fi dRER t  i 
i 1 i0 i0 i0 i 0
z

 G dINF
i0
i t i  1H Tt  i   2 G M Pt  i   3 RM 2 t  i   4 O Pt  i   5 RER t  i   6 INFt  i  u t

PAKISTAN BUSINESS REVIEW JAN 2015 732

Research Relationship between Stock Market Volatility . . .

Here,
 “d” is the first difference operator

 The coefficients of first fraction such as Bi,Ci, Di,Ei,Fi and

Gi, correspond to the short run dynamics

 The coefficients θ1,θ2, θ3, θ4, θ5 and θ6 stand for the long run
relationships between the variables

 And ut for white noise error term

Long run relationship is investigated using bound test under the

procedure of Pesaran et al. (2001) its mechanism is based on F-test.
If cointegration found in the general form of unrestricted ECM model
in ARDL (p,q,r,x,y,z) formulation, then subsequent long-run model is
projected:

p q r x y z
HTt = a1 + Bi HTt-i + Ci GMPt-i +Di RM2t-i + EiOPt-i + FRER
i t-i + Gi INFt-i + ut
i=1 i=0 i=0 i=0 i=0 i=0

If the study found long-run relationship between the

variables, the next step is to estimate short-run coefficients. The
following ECM model is applied to estimate short-run relationship
between the variables.

p q r x
dHTt = a1 + b1 (ecm) t-1 +  Bi (dHT) t-i +  Ci (dGMP) t-i + Di (dRM 2 ) t-i +  E i (dOP) t-i
i=1 i=0 i=0 i=0
y z
+ Fi (dRER) t-i +  Gi (dINF) t-i + u t
i=0 i=0

Econometric Results and Explanations

The study first checked descriptive statistics of monthly

stock returns. The main purpose to check this statistics is to know
about the sequential features of data. Table 1 explains the descriptive
statistics for stock return variable.

733 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

Table 1:
Descriptive Statistics of Stock Returns

Variable Mean Median Std.dev Skewness Kurtosis Jarque- Probability

Bera

SR 0.0172 0.0218 0.1908 -0.7667 10.5891 312.2123 0.0000

Statistics of skewness -0.7667 shows that data is skewed to

left side. Value of kurtosis greater than 3 indicates leptokurtic
distribution. Large figure of Jarque-Bera test and probability value do
not accept the null hypothesis of normal distribution and confirms
high volatile stock return at 1% significant level.

Modeling Stock Market Volatility

The first step in modeling volatility is to estimate mean

equation and variance equation simultaneously. Mean equation is
estimated through autoregressive moving average (ARMA) model.
EGARCH (2, 2) model is estimated in the next step for the estimation of
stock return volatility. The upper part in table 2 shows results of mean
equation while the lower part explains variance equation. The model
shows significant results for both mean and variance equation. Here
the positive coefficient of shows that 1% increase in the volatility
at current time being followed by the 0.6299% increase in volatility in
the last period.

PAKISTAN BUSINESS REVIEW JAN 2015 734

Research Relationship between Stock Market Volatility . . .

Table 2:
Results of EGARCH (2, 2) Model of Stock Returns

Coefficient Std. Error Z-statistics Prob.

Mean equation
Π0 -0.3055 0.0423 -7.2367 0.0000*

Π1 15 .2879 1.8145 8.4254 0.0000*

Π2 1.3830 0.1742 7.9407 0.0000*

Π3 0.0109 0.0471 0.2268 0.0000*

Π4 0.2756 0.0481 5.7425 0.0000*

Π5 1.4125 0.1601 8.8279 0.0000*

UT (-1) -15.3643 1.8148 -8.4662 0.0000*

Variance equation
-11.4567 0.5017 -22.8387 0.0000*
0.6299 0.1075 5.8622 0.0000*
0.1846 0.1371 1.3463 0.1782
0.3639 0.1935 1.8815 0.0599**

-0.5117 0.1335 -3.8331 0.0001*

-0.8472 0.0698 -12.1474 0.0000*

-0.7629 0.0604 -12.6368 0.0000*

R-squared F-statistic 4.8399
0.3602 Pro b.( F-statistic) 0.0001
Adjusted R-squared AIC -1.0257
0.2686 SBC -0.6987
SE of regression
0.1778
Durbin-Watson stat
2.1312
Note *,**,*** shows 1%,5%,10% significance level respectively

Only the coefficient of is insignificant in explaining volatility..

The negative sign link with the term shows more pressure of
depressing shocks than the optimistic shocks. The model confirms
the asymmetric volatility and statistically significant at 99%
confidence level. Once EGARCH (2, 2) model is estimated for stock

735 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

return variable after that researcher makes GARCH variances series to

evaluate stock return volatility.

Nature of Stock Return Volatility

A graphic illustration of asymmetric instability toward

positive and negative shocks is specified with “news impact curve”
launched by Pagan and Schwert (1990). News impact curve conspires
the next episode volatility (ht) that might occur from a mixture of
negative in addition to positive values of t”1. Graph of this curve is
shown in the figure 1. Here sig2 is the name for HT series. Figure 1
show more impact of bad news on stock market volatility as compare
to good news of the same magnitude. The results are reconciled with
the earlier studies of (Bollerslev et al.1994 and Choo et al., 2011).

Figure 1:
News Impact Curve of Stock Return Volatility
2.5

2.0

1.5
IG
S 2

1.0

0.5

0.0
-5 0 5 10

ADF test is performed at level and first difference with trend

and intercept. Results of ADF statistics are specified in table 3. Table
shows mix results for the different regressors. Results of ADF test
confirmed that some variables are integrated of order one I (1) while
other of order zero I (0). Hence, ARDL approach devised by (Pesaran
et al., 2001) is used to find the relationship among the variables. The
relationship is investigated in three steps.

In the first step study applies Bound test to find out the long
run relationship between the variables. Results of Bound test shows
that in our principle model F (HT/GMP, RM2, OP, RER, INF) value of F- statistics
7.183 is greater than upper critical bound and it rejects the null

PAKISTAN BUSINESS REVIEW JAN 2015 736

Research Relationship between Stock Market Volatility . . .

hypothesis of no cointegration and this recommend existence of

cointegration at 5% significance level.

Table 3:
Results of Unit Root Test
ADF Test Statistics
Regressors
Level 1s t D ifference
HT -3.8963** -8.4859*

GMP -7.2777* -7.1069*

RM 2 -1.3463 -5.9915*

OP -3.3273*** -4.8638*

REER -3.4342*** -5.7071*

INF
-4.4151* -7.4508*
*,**,*** shows 1%,5%,10% significance level
respectively

Results of long run estimates using AIC are reported in table 4. The
results are consistent with the Fisher’s hypothesis that in the long
run inflation and interest rate affect the stock returns.

Table 4:
The long run results: ARDL (2, 0, 2, 2, 2, 1) selected based on AIC
Dependent variable is HT

Regressors Coefficient Standard T- [Prob.]

Error Ratio
GMP -0.0018 0.0275 0.0665 [0.947]

RM2 -0.0463 0.0274 -

1.6901 [0.094]***
OP 0.0215 0.0143 1.5027 [0.136]

RER 0.2777 0.0794 3.4983 [0.001]*

INF 0.0133 0.0061 2.2337 [0.028]**

C -0.8529 0.4177 - [0.044]

2.0421
*,**,*** shows 1%,5%,10% significance level respectively

737 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

Table 4 shows a significant relationship of real money supply, real

exchange rate and inflation rate with stock return volatility. The study
also found a significant impact of inflation on stock market volatility.
This finding is reconciled with the previous findings of Saryal (2007)
made for Turkey and Canada and Aliyu (2011) for Ghana and Nigeria
but conflicts with Rashid et al. (2011). Positive impact of exchange
rate on stock market volatility is found. Bilson et al. (2001) conclude
that a depreciation of the home currency has a harmful impact on
returns. Vardar et al, 2012 obtained same results of exchange rate for
Istanbul stock exchange.

After testing long run relationship ECM approach is utilized

for short run dynamics. ECM value is significant at 1% significance
level. Here coefficient of ECM is (-0.6445) and shows a meaningful
pace of adjustment. It means if there is disequilibrium in the long run
then due to shocks in the short run nearly 64% of adjustment takes
place in one year.

Table 5:
ECM Representation for the selected ARDL model
ARDL (2,0,2,2,2,1) based on AIC

Regressor Coefficient Standard T- [Prob]

Error Ratio
dHT1 0.2787 0.0891 3.1278 0.002]

dGMP -0.0012 0.0178 0.0664 [0.947]

dRM21 0.4498 0.1101 4.0842 [0.000]

dROP1 -0.0476 0.0181 -2.6371 [0.010]

dRER1 -0.3391 0.1213 -2.7948 [0.006]

dINF 0.0083 0.0026 0.3197 [0.750]

dINPT -0.5497 0.2771 -1.9842 [0.050]

Ecm(-1) -0.6445 0.0807 -7.9813 [0.000]

R- Squared R-Bar-Squared
0.4748 0.4034
Mean of Dependent Variable F( 10, 107)
0.4036 9.311[0.000]
S.D. of Dependent Variable AIC
0.0183 327.9161
DW-statistic SBC
1.9858 307.1362

PAKISTAN BUSINESS REVIEW JAN 2015 738

Research Relationship between Stock Market Volatility . . .

Diagnostic tests are performed in order to check accuracy of

model. Results of both LM version and F version statistics are given
in table 6. In the final step CUSUM and CUSUMSQ tests are applied
to observe the stability in parameters.

Table 6:
Diagnostic Tests

Test Statistics LM Version F Version

A:Serial Correlation CHSQ(12)= 12.7713[0.386] F(12, 91)= 0.9204[0.530]

B:Functional Form CHSQ(1) = 1.9947[0.158] F(1 , 102)= 1.7539[0.188]

C:Normality CHSQ( 2) = 0.3 957[0.821] Not applicable

D:Heteroskedasticity CHSQ(1) = 0 .9923[0.319] F(1, 116)= 0.9837[0.323]

Graphs of both these tests are shown in the figure 2 and 3.

The figure shows that recursive residual are bounded by the critical
boundaries and do not reject null hypothesis of stability at 5%
significance level.

Figure 2:
CUSUM Test

739 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

Figure 3:
CUSUMSQ

Conclusion and Policy Implications

The present research is a step toward investigation of

determinants of stock return volatility in Pakistan. Volatility in stock
returns is estimated through EGARCH model and then the news impact
curve is drawn to check the asymmetric behavior of stock returns. It is
found that macroeconomic indicators are important in explaining stock
market volatility. In particular the variables like inflation, real exchange
rate and money supply are found significant determinants of stock
market volatility. Coefficient of ECM (-0.642) shows a meaningful
adjustment pace toward equilibrium. CUSUM and CUSUMSQ tests
show that estimated parameters are stable at 5% level of significance.
This study clearly suggests that reduction in inflation rate can reduce
the volatility of stock market. Investor must consider monetary policy
of the country in which he is interested to make investment. The results
has also proved the equation of exchange i.e. MV=PY. The study also
suggests that the investors should look at exchange rates and inflation
as the major basis of risk involved in stock market. Financial managers
and strategy makers should consider macroeconomic aspects
whenever they formulate or implement financial stability.

PAKISTAN BUSINESS REVIEW JAN 2015 740

Research Relationship between Stock Market Volatility . . .

References

Aliyu, S. U. R. (2012). Does inflation have an impact on stock returns

and volatility? Evidence from Nigeria and Ghana. Applied
financial economics. 22, 427-435.
Beltratti, A. and Morana, C., 2004. Breaks and persistency:
Macroeconomic causes of stock market volatility. Journal of
econometrics, 131,11-27.
Bilson C. M., Brailsford T. J and Hooper V. J. (2001). Selecting
macroeconomic variables as explanatory factors of emerging
stock market returns. Pacific basin financ. J, 9, 401-426.
Bollerslev, T. (1986). Generalized autoregressive conditional
hetroskedasticity. Journal of econometrics. 31, 307-327.
Bollerslev, T., Engle, R.F. and Nelson, D.B. (1994). ARCH models,” in
Engle, R.F. and McFadden, D., Eds. The hand book of
econometrics, 4, 2959-3038.
Box, G. E. P. and Jenkins, G. M. (1976). Time series analysis: Forecasting
and control (revised edition), Holden Day, San Francisco.
Box, G.E.P. and Jenkins, G.M. (1970). Time series analysis: Forecasting
and control, Holden-Day, San Francisco
Campbell, J. Y. and Hentschel, L. (1992). No news is good news. An
asymmetry model of changing volatility in stock returns.
Journal of financial economics. 31, 281-318:
Chan, L. K. C., Karceski, J. and Lakonishok, J. (1998). The risk and
return from factors, Journal of financial and quantitative
analysis, 33, 159–88.
Chiou-wei, S. (2011). The macroeconomic determinants of stock price
volatility: Evidence from Taiwan, South Korea, Singapore and
Hong Kong.8,114-134.
Choo, W., Lee, S. and Ung, S. (2011). Macroeconomics uncertainty
and performance of GARCH models in forecasting Japan stock
market volatility. International journal of business and social
science. 2, 200-208.
Chowdhury, S. S. H., Mollik, A. T. and Akhter, M. S. (2006). Does
predicted macroeconomic volatility influence stock market
volatility? Evidence from the Bangladesh capital market.
Working Paper at Department of Finance and Banking
University of Rajshahi, Bangladesh.
Davidson, J. E. H., Handry, D. F., Srba, F. and S. Yeo, (1978),
“Econometric modeling of the aggregate time series relationship

741 PAKISTAN BUSINESS REVIEW JAN 2015

Relationship between Stock Market Volatility . . . Research

between consumers expenditures and income in the United

Kingdom”, Economic journal, 88, 661-692.
Dickey, D. A. and Fuller, W. A. (1981). Likelihood ratio statistics for
autoregressive time series with a unit root. Econometrica, 49,
1057-72.
Diebold, F. X. and Yilmaz, K. (2008). Macroeconomic volatility and
stock market volatility, worldwide. NBER Working Paper series,
No. 14269.
Engle, R. (1982). Autoregressive conditional hetroskedasticity with
estimates of the variance of United Kingdom inflation.
Econometrica, 50, 987-1008.
Engle, R. F. and Granger, C. W. J. (1987), Cointegration and error
correction: representation, estimation and testing.
Econometrica, 55, 251-276.
Falkberg, A. (2012). Do macroeconomic variables affect stock market
volatility? Business and Economics.
Govt. of Pakistan (2011). Handbook of statistics on Pakistan economy.
State bank of Pakistan, Karachi.
Govt. of Pakistan, “Statistical bulletin, different issues”, State bank of
Pakistan, Karachi
Griffiths, W. E., Hill, C. R. and Judge, G.G. (2001). Undergraduate
econometrics, Second edition, John Wiley and Sons, Inc.
Karolyi, G. A. (2001). Why stock return volatility really matters,
Institutional investor journals series, Prepared for inaugural
issue of strategic investor relations.
Liljeblom, E. and Stenius, M. (1997). Macroeconomic volatility and
stock market volatility: empirical evidence on Finnish data.
Applied financial economics, 7,419-426.
Nelson, D. B. (1991). Conditional hetroskedasticity in asset returns: a
new approach. Econometrica, 59, 347-370.
Officer, R. R. (1973). The Variability of the market factor of New York
stock exchange. Journal of business, 46, 434-453.
Pagan, A. R. and Schwert, G. W. (1990). Alternative models for
conditional stock volatilities, Journal of econometrics, 45, 267-
90.
Pesaran, H.M. and Shin, Y. (1999). Autoregressive distributed lag
modeling approach to cointegration analysis. Econometrics and
economic theory in the 20th. Century: The Ragnar Frisch
centennial symposium. Cambridge university press, Cambridge.

PAKISTAN BUSINESS REVIEW JAN 2015 742

Research Relationship between Stock Market Volatility . . .

Pesaran, H.M. Shin, Y. and Smith, R. J. (2001). Bounds testing

approaches to the analysis of level relationships. Journal of
applied econometrics, 16, 289-326.
Rafique, A. and Rehman, K, U. (2011). Comparing the volatility
clustering of different frequencies of stock returns in an
emerging market: A case study of Pakistan. Journal of
economics and behavioral studies, 3, 332-336.
Saryal, F, S. (2007). Does inflation have an impact on conditional
stock market volatility? Evidence from Turkey and Canada.
International research journal of finance and economics, 13,
123-133
Varder, G., Aksoy, G. and Can, E. (2008). Effects of interest and exchange
rate on volatility and return of sector price indices at Istanbul
Stock Exchange. European journal of economics, finance and
administrative sciences. 11, 126-135.
Wang, X. (2010). The relationship between stock market volatility
and macroeconomic volatility: Evidence from China.
International research journal of finance and economics, 5,
149-160.
Zafar, N., Urooj, S. F. and Durrani, T. K. (2008). Interest rate volatility
and stock return and volatility. European journal of
economics, finance and administrative sciences. 14, 135-140.

743 PAKISTAN BUSINESS REVIEW JAN 2015