EFFECT OF EXCHANGE RATE VOLATILITY ON THE GHANA STOCK EXCHANGE
Textonly Preview
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
EFFECT OF EXCHANGE RATE VOLATILITY ON THE
GHANA STOCK EXCHANGE
Charles Adjasi1
University of Ghana, Ghana.
Simon K. Harvey2,
University of Ghana, Ghana.
Daniel Agyapong3
University of Cape Coast, Ghana.
Email: [email protected]
ABSTRACT
The study looked at the relationship between Stock Markets and Foreign
Exchange market, and determined whether movements in exchange rates have an
effect on stock market in Ghana. The Exponential Generalised Autoregressive
Conditional Heteroskedascity (EGARCH) model was used in establishing the
relationship between exchange rate volatility and stock market volatility. It was
found that there is negative relationship between exchange rate volatility and
stock market returns – a depreciation in the local currency leads to an increase in
stock market returns in the long run. Where as in the short run it reduces stock
market returns. Additionally, there is volatility persistence in most of the
macroeconomic variables; current period’s rate has an effect on forecast variance
of future rate. It was also revealed that an increase (decrease) in trade deficit and
expectation in future rise in trade deficit will decrease (increase) stock market
volatility. In addition, the consumer price index has a strong relationship with
stock market volatility. This means that an increase in consumer price will lead to
a rise in stock market volatility. Finally, there is the presence of leverage effect and
volatility shocks in stock returns on the Ghana Stock Exchange.
1Contact details: Dr. Adjasi, Department of Finance, University of Ghana, Accra, Ghana .
2 Contact details: Department of Finance, University of Ghana, Accra, Ghana .
3 Contact details: School of Business, University of Cape Coast, Cape Coast, Ghana.
28
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
Keywords: volatility, leverage, exchange rate, returns, Ghana Stock Exchange
JEL Classification: N2, N27
I. INTRODUCTION AND STATEMENT OF THE PROBLEM
An economy’s financial position is susceptible to its foreign exchange
volatility. Foreign exchange market developments have cost implications for the
households, firms and the state. Benita and Lauterbach (2004) showed that
exchange rate volatility have real economic costs that affect price stability, firm
profitability and a country’s stability. Exchange rate volatility has implications for
the financial system of a country especially the stock market. However a survey of
the available literature reveals divergent views of researchers on the issue of
whether foreign exchange rate variability influences stock market volatility (Frank
and Young, 1972; Solnik, 1987; Taylor and Tonks, 1989). Three events – Asian
Currency Crises, the advent of floating exchange rate in the early 1970s and
financial market reforms in the early 1990s have prompted financial economist
into determining the link between these two markets (Mishra, 2004). Also, the
internationalization of capital markets has resulted in inflow of vast sums of funds
between countries and in the cross listing of equities. This has therefore made
investors and firms more interested in the volatility of exchange rate and its effect
on stock market volatility. Floating exchange rate appreciation reduces the
competitiveness of export markets; and has a negative effect on the domestic stock
market (Yucel and Kurt, 2003). But, for import dominated country, it may have
positive effect on the stock market by lowering input costs.
Ghana presents an example of a small open economy which engages in
international trade with several countries and hence susceptible to foreign
exchange rate volatility. Ghana’s stock market described as one of the emerging
markets currently was established in July 1989 as a private company limited by
guarantee under the Companies Code, 1963. It recorded its highest turnover of
equities in volume in 1997, with 125.63 million shares, from a volume of 1.8
million shares by the end of 1991. After wards, the volume have been falling
steadily from 125.63 million in 1997 to 91.45 million in 1998, 49.57 million in 1999
to 30.72 million in 2000. In 2001, the volume increased to 55.3 million, fell to 44.12
million in 2002, inched up to 96.33 million in 2003 and 104.35 million in 2004. The
AllShare Index, by the close of 2003, topped performance of stock markets in the
world with yield of 154.7 per cent (or 142.7 percent in dollar terms) (GSE Fact
Book,2005). After such a performance, the market Share Index has continued to
fluctuate with an occasional rise or dip.
However, empirical evidence on the influence of foreign exchange market
volatility on stock market is largely inconsistent. These have been in the contest of
developed economies. Mishra (2004) admitted no theoretical consensus on the
interaction between stock prices and exchange rate. Solnik (2000) on the other
hand posits that there is a negative correlation between stock market and local
currency.
29
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
The openness of a country’s economy is recognised as a cause of volatility
of its market. Ghana presents a classic example of an open economy which
engages in international trade transaction. Moreover, with advert of globalisation,
developing economies are becoming more integrated into developed economies
with the results of increasing flow of imports and exports. Ghana is not an
exception. A cursory examination of foreign exchange rate history in Ghana shows
some considerable level of volatility. Therefore, it would be interesting to explore
the effect of its foreign exchange volatility on its stock market. Again, much work
on the effect of the exchange rate volatility in the developing country like Ghana
has not been done. Thus, therefore the study intended look at the effect of foreign
exchange movements and stock market volatility in Ghana.
A. Research Objectives
The study determined the following:
1. Whether exchange volatility has effect on stock market volatility in Ghana,
2. If other macroeconomic variables affect stock market volatility in Ghana.
B. Research Hypothesis
H0: Exchange rate volatility has no impact on stock market volatility
H1: Exchange rate volatility has an impact on stock market volatility
H0: Macroeconomic variables have no effect on stock market volatility
H1: Macroeconomic variables have an effect on stock market volatility
II. LITERATURE REVIEW
Two portfolio models explain the interaction between exchange rate and stock
market volatility. First, the “FlowOriented” model (Dornbusch and Fischer, 1980
and Gavin, 1989) – in which exchange rate movement affects output levels of firms
and also the trade balance of an economy. Share price movements on the stock
market also affect aggregate demand through wealth, liquidity effects and
indirectly the exchange rate. Specifically a reduction in stock prices reduces wealth
of local investors and further reduces liquidity in the economy. The reduction in
liquidity also reduces interest rates which in turn induce capital outflows and in
turn causes currency depreciation. The second is the “StockOriented” model
(Branson, 1983 and Frankel, 1983). In the case of the “StockOriented” model the
stock market exchange rate link is explained through a country’s capital accounts.
In this model the exchange rate equates demand and supply for assets (bonds and
stocks). Therefore expectations of relative currency movements have a significant
impact on price movements of financially held assets. Thus stock price movements
may influence or be influenced by exchange rate movements. That is, if the cedi
for a example depreciates against a foreign currency (the British pound), it will
increase returns on the foreign currency (the pound). Such events will motivate
investors to move funds from domestic assets (stocks) towards pound assets,
depressing stock prices. Thus a depreciating currency has a negative impact on
stock market returns (Adjasi and Biekpe, 2005).
30
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
Officer (1973) showed that aggregate stock volatility, volatility of money
growth and industrial production increased during the period of depression. But
stock volatility was at similar levels before and after the depression. However,
Black (1976) and Christie (1982) discovered that stock market volatility can
partially be explained by financial leverage. This is a contrary finding to that of
Officer. Also French et al. (1987) and Schwert (1989) measured market volatility as
the variance of monthly returns of market index. They discovered that the market
volatility changes over time. Also they were of the view that the value of corporate
equity depends of the health of the economy, so a change in the level of
uncertainty about future macroeconomic conditions would cause a proportional
change in stock return volatility. But as French et al. fail to find a direct positive
relation between expected return and volatility, Schwert also failed to explain
much of the change in market volatility over time using macroeconomic variables.
In a related study, Schwert (1990) analyzed the behaviour of stock return volatility
around stock market crashes and discovered that stock market volatility jumps
dramatically during the crash and returns to low precrash levels quickly.
In related studies, Officer (1973) explained the drop in stock market
volatility in the 1960s with a reduced variability in industrial production. Schwert
(1989) and Hamilton and Lin (1996) discovered that stock market volatility is
increases in times of recession and Glosten et al. (1993) find interest rates to be an
important factor in explaining stock market volatility.
In Mao and Kao (1990) exporting firms’ stock values were seen to be more
sensitive to changes in foreign exchange rates. Their findings also revealed
another topical issue of the relationship between stock prices at the macro and
micro levels. Although theories suggest causal relationship between exchange rate
and stock prices, existing evidence indicates a weak link between them at a micro
level. On the macro level, Ma & Kao (1990) found that a currency appreciation
negatively affects the domestic stock market for an exportdominant country and
positively affects the domestic stock market for an importdominant country,
which seems to be consistent with goods market theory. Meanwhile, Khoo (1994)
estimated mining companies’ economic exposure by using exchange rates, interest
rates and price of oil and discovered that, the sensitivity of stock returns to
exchange rate movement and proportion of stock returns explained by exchange
rate movement are small. Domely and Sheehy (1996) also found a
contemporaneous relation between the foreign exchange rate and the market
value of large exporters in their study.
Adjasi and Biekpe (2005) investigated the relationship between stock
market returns and exchange rate movements in seven African countries.
Cointegration tests showed that in the longrun exchange rate depreciation leads
to increases in stock market prices in some of the countries, and in the shortrun,
exchange rate depreciations reduce stock market returns. In Mishra (2004) it was
identified that there is no Granger’s causality between the exchange rate and stock
return. The study of Mishra (2004) indicated that stock return, exchange rate
31
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
return, the demand for money and interest rate are related to each other though no
consistent relationship exist between them. Further more, forecast error variance
decomposition evidenced that exchange rate return affects the demand for money;
interest rate causes exchange rate to change; exchange rate affects the stock return;
demand for money affects stock return; interest rate affects the stock return, and
demand for money affects the interest rate.
Engle and Rangel (2005) also examine the link between the unconditional
volatility and a number of macroeconomic variables. Bercker and Clement (2005)
extended the SPLINE GARCH model proposed by Engle and Rangel (2005) when
they modelled stock market volatility conditional on macroeconomic conditions.
They incorporate macroeconomic information directly into the estimation of such
GARCH models. It was demonstrated that forecasts of macroeconomic variables
can be easily incorporated into volatility forecasts for share index returns. Thus
their model can lead to significantly different forecasts than traditional GARCH
type volatility models.
Among the few studies on emerging markets includes; Mishra (2004),
Chortareas et al (2000); and Koutmoa et al (1993). Studies like Smith (1992), Solnik
(1987), Aggarwal (1981), Frank and Young (1972), Phylaktis and Ravazzolo (2000),
Granger et al. (2000), Abdalla and Murinde (1997), and Apte (2001) have found a
significant positive relationship between stock prices and exchange rates while
others, such as Soenen and Hennigar (1998), Ajayi and Mougoue (1996), Mao and
Kao (1990) have reported a significant negative relationship between the two
variables. On the other hand, some studies, such as Bartov and Bodnar (1994),
Frank and Young (1972), found very weak or no relationship between stock prices
and exchange rates. On the issue of causation, most of the studies had mixed
results (Morley and Pentecost (2000); BahmaniOskooee and Sohrabian (1992);
Ibrahim (2000); Kanas (2000).
It is also evident that the standard Granger causality method has been the
most predominant model used in most studies. Even though some studies have
linked foreign exchange markets to stock markets in some emerging markets, the
researcher did not come across any of such study on the Ghana Stock Exchange.
Some works have been done on the issue of stock market volatility on the Ghana
Stock Exchange, but it has basically be on how stock market returns vary and not
linked with foreign exchange rate movement.
METHODOLOGY AND ANALYSIS
A. Data Collection
The data obtained was mainly from secondary sources including the IMF
Direction of Trade Statistics Yearbook and the Ghana Statistical Service where
data on volumes and values of exports and imports were obtained, the Bank of
Ghana quarterly bulletins, where the data on treasury bill rates, money supply,
foreign exchange rates and inflation rates and trade deficit was obtained; the
32
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
Ghana Stock Exchange quarterly publications, where the data on stock indices was
obtained. Nominal figures were used for the work for the study. The data was
tested for stationarity or the order of integration of the data series in order to
eliminate spurious regression results. This test used was the Augmented Dickey
Fuller method. Yt, CPI, TBR, St were not stationary, but became stationary on the
first difference.
B. Model Specification and Estimation
The Exponential Generalised Autoregressive Conditional Heteroskedascity
(EGARCH) was used. This is specifically designed to model and forecast
conditional variance especially in financial assets. This is most often preferred to
the GARCH model in studying financial markets. As identified by Koulakiotis et
al (2006) the GARCH is relatively weaker than the EGARCH in studying financial
markets phenomenon. The weaknesses of the GARCH includes (i) it assumes that
there is a negative correlation between current returns and future volatility; (ii) it
imposes parameter restrictions that are often violated by estimated coefficients
which may unduly restrict the dynamics of the conditional variance process; and
(iii) it is difficulty to interpret whether shocks to conditional variance persist or not
in the GARCH. This is because the usual norms measuring persistence often do
not agree. The model is stated with the mean and variance equation equations,
similar to Koulakiotis et al (2006) and Adjasi (2004). The mean and variance
equations are stated in equations (1) and (2) respectively:
Y ? ? ? ? S
? ? ? MS ? ? TBR ? ? DT ? ? CPI ? ? MS ?
t
0
1
t
2
t
3
t
4
t
5
t
6
t 1
?
? TBR ? ? DT ? ? CPI ? ? ..............................................................(1)
7
t 1
?
8
t 1
?
9
t 1
?
t
?
?
log(
2
? ) ? ? ? ? log( 2
? ) +
t 1
? ? +
t 1
? ? + ? ….……………..…. (2)
t
t 1
?
?
?
t
t 1
?
t 1
?
Where:
log( 2
? ) = log of conditional variance of stock market returns
t
?t = the error term
? = vector of coefficient
?S  changes in exchange rate at time t
t
? =exchange rate volatility
t
? = leverage effect
Y = stock market returns
t
33
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
The macroeconomic variables included in the mean equation are value of
money supply in the market measured monthly (MS), treasury bill rate (TBR ),
trade deficit?DT ? and inflation measured by?CPI ? . ?S was determined by using
t
the Trade Weighted Index (TWI) model proposed by White (1997). The major (or
core) Trade Weighted Index (TWI) is an index measure of the value (January
1995=100) of the cedi relative to the currencies of Ghana's top three trading
currencies the euro, the pound and the dollar.
The model is specified as:
wit
? e ?
it
S ? ?
sf ………………………………………………(3)
t
*
?
?
e
? i0 ?
Where:
S ? TWI = The trade weighted index
t
? = The multiplication sign
e ? The number of foreign currency units for trading partner i per the Cedi at
it
time t
e = The number of foreign currency units for trading partner i per the Cedi in the
i0
base period
w ? The trade weight for the currency of trading partner i at time t;
it
sf = a scale factor which ensures that the exchange rate index does not change on
a monthly reweighting solely as the result of change in currency weights
S ? S
Change in exchange rate was obtained by
t
t 1
S
?
? ?
..................................(4)
t
St 1
?
Where: S ? Current month’s trade weighted index
t
S
? Previous month’s trade weighted index
t ?1
C. Statioarity Test
A stationary test was first carried out on the variables. In applying the Augmented
Dickey – Fuller (ADF) test to the variables, DTt, CPIt and MSt were found to be
stationary at levels. The critical values at 1% and 5% were 3.4804 and 2.8832
respectively. However, the ADF test statistic of Yt, St, and TBRt were less than the
critical value at 1% and 5% respectively, and hence not stationary. However, these
variables attained stationarity after the first differencing. Therefore regression
could be run without any spurious results. This is indicated in table 1.1.
Table 1.1 ADF Unit Root Test of Variables
34
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
Variables
Levels
First Difference I(1)
St
0.814175
3.328392
Yt
1.003925
6.531121
MSt
4.283642

TBRt
1.042575
4.574844
CPIt
4.283642

DTt
4.284977

The ADF critical values at 1% and 5% are 3.4804 and 2.8832 respectively
Not stationary
Stationary at 5%
Stationary at both 1% and 5%
III. RESULTS FROM MODELLING VOLATILITY
The main model of the study was EGARCH (1,1). A variant of the in mean
specification used the conditional standard deviation in place of conditional
variance. In order to determine the nature of exchange rate and other
macroeconomic variables’ volatility, a GARCH (1,1) was employed to estimate the
conditional variance of these variables. Then the volatility of the exchange rate
and other macroeconomic variables were introduced in the conditional variance of
the stock market returns equation using an EGARCH (1,1). The results are
illustrated in the tables 1.2 and 1.3.
The results of an EGARCH (1,1) estimation of Stock Market Volatility,
Exchange Rate and other Macroeconomic variables are depicted in table 1.2 below.
Even though the original model contains money supply ( MS ), it had to be
t
eliminated due to its perfect positive correlation with the consumer price index
(CPIt). The results shown in table 1.2 indicated that there is a positive relationship
between consumer price index and stock market volatility.
Table 1.2: Effect of Exchange Rate, Macroeconomic Variables on
Stock Market
Dependent Variable: LY
t
Method: ML – ARCH
Date: 07/05/07 Time: 19:48
Sample(adjusted): 1995:03 2006:06
Included observations: 136 after adjusting endpoints
Convergence not achieved after 100 iterations
Coefficient
Std. Error
zStatistic
Prob.
C
0.001272
0.000117
10.88564
0.0000
LCPI
4.158309
0.747694
5.561513
0.0000
LTBR
0.113066
0.013520
8.362907
0.0000
LDT
5.177410
0.779522
6.641772
0.0000
L?t
0.188043
0.023179
8.112585
0.0000
LCPI(t1)
1.224610
1.239101
0.988305
0.3230
LTBR(t1)
0.003417
0.014638
0.233457
0.8154
LDT(t1)
1.348548
1.312910
1.027145
0.3044
L?(t1)
0.114453
0.026404
4.334680
0.0000
35
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
LY(t1)
1.035695
0.003402
304.4638
0.0000
Variance Equation
C
6.803226
1.788158
3.804600
0.0001
RES/SQR[GARCH](1)
0.832085
0.253089
3.287714
0.0010
RES/SQR[GARCH](1)
0.028974
0.134708
0.215084
0.8297
EGARCH(1)
0.720664
0.087792
8.208763
0.0000
LCPI
44.00593
7359.075
0.005980
0.9952
LTBR
245.7033
41.69450
5.892944
0.0000
LDT
46.84151
7860.877
0.005959
0.9952
L?t
161.5657
140.9899
1.145938
0.2518
LCPI(t1)
28.24004
10957.69
0.002577
0.9979
LTBR(t1)
49.03193
57.07838
0.859028
0.3903
LDT(t1)
22.88789
11717.94
0.001953
0.9984
L? (t1)
157.7215
177.6653
0.887745
0.3747
LY(t1)
34.71869
27.35591
1.269148
0.2044
Rsquared
0.986625
Mean dependent var
0.013255
Adjusted Rsquared
0.984021
S.D. dependent var
0.003792
S.E. of regression
0.000479
Akaike info criterion
13.89782
Sum squared resid
2.60E05
Schwarz criterion
13.40524
Log likelihood
968.0518
Fstatistic
378.8832
DurbinWatson stat
1.876963
Prob(Fstatistic)
0.000000
RES/SQR [GARCH] (1) = (?) = Leverage effect
The results of an EGARCH (1, 1) estimation of Stock Market Volatility,
Exchange Rate Volatility are depicted in table 1.3
Table 1.3: Stock Market Returns and Exchange Rate Volatility
Dependent Variable: LYt
Method: ML – ARCH
Date: 07/05/07 Time: 19:54
Sample(adjusted): 1995:03 2006:06
Included observations: 136 after adjusting endpoints
Convergence achieved after 58 iterations
Coefficient
Std. Error
zStatistic
Prob.
SQR(GARCH)
1.043849
0.167307
6.239134
0.0000
C
0.000747
4.27E05
17.47705
0.0000
L?t
0.105751
0.054258
1.949034
0.0513
L? (t1)
0.051070
0.061475
0.830741
0.4061
LY(t1)
1.005999
0.004253
236.5350
0.0000
Variance Equation
C
11.11815
0.364042
30.54088
0.0000
RES/SQR[GARCH](
1.415530
0.226496
6.249698
0.0000
1)
RES/SQR[GARCH](1)
1.453967
0.208849
6.961827
0.0000
EGARCH(1)
0.286132
0.012553
22.79435
0.0000
L?t
0.431048
71.77992
0.006005
0.9952
L? (t1)
0.028142
135.2342
0.000208
0.9998
LY(t1)
2.418404
27.78157
0.087051
0.9306
36
African Journal of Accounting, Economics, Finance and Banking Research Vol. 3. No. 3. 2008.
Charles Adjasi, Simon K. Harvey and Daniel Agyapong
Rsquared
0.829828 Mean dependent var
0.013255
Adjusted Rsquared
0.814732 S.D. dependent var
0.003792
S.E. of regression
0.001632 Akaike info criterion
11.43701
Sum squared resid
0.000330 Schwarz criterion
11.18001
Log likelihood
789.7169 Fstatistic
54.97039
DurbinWatson stat
1.640142 Prob(Fstatistic)
0.000000
RES/SQR [GARCH] (1) = (?) = leverage effect
IV. INTERPRETATION OF RESULTS
Using the EGARCH (1,1), the results shown in table 1.2 indicated that there
is a positive relationship between consumer price index and stock market
volatility. The value of consumer price index CPIt is positive; meaning that an
increase in consumer price index will lead to a rise in stock market volatility.
When prices in the domestic economy are uncertain, the volatility of nominal asset
returns should reflect consumer price index volatility (Schwert, 1989). Similarly,
Skousen (2006) submitted that in principle, the stock market should do well under
conditions of strong economic growth with relatively stable price levels. Under
conditions of unstable prices, analysts do not think strong job creation and
economic growth are sustainable, and this may create uncertainties in the stock
market leading to variability in price. This means that stock markets are not likely
to perform well in periods of unstable prices.
The finding in this study is consistent with Erb et al (1995; 1997). They
found out that market volatility is high when inflation (changes in prices) risk is
high, which is typically the case in many emerging countries. The positive
coefficients of LCPIt and its lag suggest volatility clustering; that is if stock market
volatility was high in the previous period due to high price levels, it is likely to be
high in the current period if level of price levels remains high. More over, the
values were statistically significant, indicating a strong relationship between stock
market volatility and consumer price index.
The coefficient of the lag of LY is positive and statistically significant. This
t
indicates that the past period volatility of stock market returns affect the current
period. This is consistent with the ARCH model Engle (1982) postulates that
volatility in the current period is related to its value in the previous period.
The coefficient of LTBRt is negative and statistically significant indicating
that higher volatility in Treasury bill rate dampens stock market activities. This
means an increase in treasury bill rate volatility will lead to a fall in stock market
volatility. Under conditions of attractive Treasury bill rates, analysts believe that
investors may shift their funds from stocks into treasury bills that may affect stock
market activities and vice versa.
The coefficient of LDT is negative and statistically significant indicating
that that higher volatility in trade deficit dampens stock market activities. This
means an increase in trade deficit volatility will lead to a fall in stock market
volatility. Conditions of high trade deficit create uncertainty about the general
health of an economy and its stock markets activities. For instance, high trade
37