Journal of African Business
ISSN: 1522-8916 (Print) 1522-9076 (Online) Journal homepage: https://www.tandfonline.com/loi/wjab20
Interest Rate and Exchange Rate Exposure of
Portfolio Stock Returns: Does the Financial Crisis
Matter?
Richard Adjei Dwumfour & Naa Adokarley Addy
To cite this article: Richard Adjei Dwumfour & Naa Adokarley Addy (2019) Interest Rate and
Exchange Rate Exposure of Portfolio Stock Returns: Does the Financial Crisis Matter?, Journal of
African Business, 20:3, 339-357, DOI: 10.1080/15228916.2019.1583977
To link to this article:  https://doi.org/10.1080/15228916.2019.1583977
Published online: 01 Mar 2019.
Submit your article to this journal 
Article views: 91
View related articles 
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=wjab20
JOURNAL OF AFRICAN BUSINESS
2019, VOL. 20, NO. 3, 339–357
https://doi.org/10.1080/15228916.2019.1583977
Interest Rate and Exchange Rate Exposure of Portfolio Stock
Returns: Does the Financial Crisis Matter?
Richard Adjei Dwumfour and Naa Adokarley Addy
Department Of Finance, University of Ghana Business School, Legon, Ghana
ABSTRACT KEYWORDS
The study examines the impact of changes in interest rate and interest rate; exchange rate;
exchange rates and their unexpected changes on industry and size portfolio stock returns;
portfolio returns on the Ghana Stock Exchange (GSE) controlling emerging market; financialcrisis
for the 2007/2008 financial crisis. Three main exchange rates (FX)
namely, Ghana cedis (Gh¢)/US dollar, Gh¢/Great Britain Pounds
(GBP) and the Gh¢/Euro are used. We use OLS, GARCH and
ARIMA in our estimations. The study found that only depreciation
of the Gh¢/USD reduces the returns of financial stocks and large
firms. There is a direct positive impact of the financial crisis on the
returns due to investment shift from developed markets where
crisis occurs. Variations in the returns are mostly explained by the
market index returns (RM), which has a positive impact. However,
we find that the positive impact of RM on the portfolio returns
(finance, medium and large portfolios) is reduced during the
financial crisis. The results largely reveal that shocks to the condi-
tional variance are highly persistent and the response of volatility
decays at a slower rate.
1. Introduction
The liberalization of emerging financial markets results in the integration of developing
countries’ capital markets into global capital markets (Francis, Hasan, & Hunter, 2002).
Consequently, the global nature of capital markets increases the likelihood of financial
crisis in emerging markets if the same occurs in developed economies. This impact was
evidenced in the spillovers of the financial crisis that started in the US in 2007 and
threw the world economy into a state of serious turmoil and its consequent volatilities
in stock returns. Some empirical studies (Akhigbe, Madura, & Marciniak, 2012;
Kenourgios & Padhi, 2012) have been carried out to explore the impact of the 2007/
2008 crisis through different transmission channels including interest rates and foreign
exchange (FX) rates. Interest rates and FX rates are key variables in determining stock
returns, hence have received much attention in recent studies. These results have
remained mixed. Some studies (Joseph & Vezos, 2006) do not find a pronounced
degree of sensitivity of stock returns to interest rates and FX rates. Other studies like
those of Kolari, Moorman, and Sorescu (2008), Choi, Elyasiani, and Kopecky (1992),
CONTACT Richard Adjei Dwumfour radwumfour@st.ug.edu.gh Department Of Finance, University of Ghana
Business School, Legon, Ghana
© 2019 Informa UK Limited, trading as Taylor & Francis Group
340 R.A. DWUMFOUR AND N.A. ADDY
and Chamberlain, Howe, and Popper (1997) provide some evidence that interest rates
and exchange rates are priced into stock returns. Most of these studies have been for
developed markets. The issue, however, is that regardless of the ongoing deepening of
global financial integration, firms still retain some country-specific characteristics. As
Hirtle (1991) posits, the internationalization of financial markets is not complete.
Hence, the operations of firms in Ghana are more likely than not to differ from those
in other countries developed or developing. In this regard, studies on the interest rates
and exchange rates exposure of, for example, US firms, cannot accurately be used to
draw conclusions about non-US firms.
To the best of our knowledge, no study has been carried out to examine the impact of
the global financial crisis on portfolio stock returns on the Ghana Stock Exchange. Also, it
appears that previous studies carried out to explore the FX rates and interest rates
exposure of firms in Ghana have been limited. Those found in the literature that explored
foreign exchange exposure of listed firms in Ghana were those of Adjasi, Harvey, and
Agyapong (2008) and Salifu, Osei, and Adjasi (2007). Adjasi et al. (2008) uses the stock
market index to calculate the returns and they use trade weighted index of the exchange
rate using monthly data. Our study is different from these previous studies in five ways.
First, we estimate returns for portfolios grouped into industry (Finance, Manufacturing
and Retail) and size (Large, Medium and Small) to better understand the size and industry
differences. Analyzing industry and size exposure is important since industries differ in
terms of pass-through and mark-ups (Allayannis & Ihrig, 2001; Bodnar, Dumas, &
Marston, 2002), competitive structure (Griffin & Stulz, 2001; Marston, 2001) or industry
concentration (Bartram & Karolyi, 2006) and hence may face different levels of exposure.
Secondly, we use the three main exchange rates traded in Ghana and estimate them
individually rather than creating an index. This will help us understand how each of the
FX rates affects firms listed on the exchange.
Thirdly, the frequency of data (weekly) we use in the study is more suited for GARCH
estimation. Unlike Salifu et al. (2007), who estimate monthly data of the three exchange
rates using OLS estimation, this study uses GARCH (1, 1) estimation of weekly data. Due
to volatility clustering and the ARCH effects of high frequency data (weekly), the linear
(OLS) estimation methods produce bias and inconsistent results. Fourthly, we estimate
the impact of unexpected changes of interest rate and FX rates on portfolio stock returns.
This was used in similar studies like those of Fang and Loo (1994) and more recently by
El-Masry and Abdel-Salam (2007) who found that the unexpected changes in FX rates
have a significant cross-sectional effect on common stock returns.
Finally, the impact of the financial crisis on stock returns in Ghana needs to be
examined. As indicated by Ackah, Aryeetey, and Aryeetey (2009), Ghana, just like other
developing countries in the region, became increasingly vulnerable to the financial
crisis, as its current account, fiscal deficit, exchange rate, inflation and debt indicators
worsened during the crisis period. Thus, as globalization increases, with a high degree
of international capital mobility, portfolio flows can have a significant effect on domes-
tic asset prices (Sarno, Tsiakas, & Ulloa, 2016), especially during financial crisis
(Fratzscher, 2012). Hence, investors are likely to shift their investment into emerging
markets like Ghana, causing an impact on domestic stock returns. We therefore
examine the impact of the financial crisis on the portfolio stock returns in Ghana.
In the light of these discussions, this study answers the following research questions:
JOURNAL OF AFRICAN BUSINESS 341
(1) Do changes in interest rate and exchange rates and their unexpected changes
affect portfolio stock returns in Ghana?
(2) Did the 2007/2008 financial crisis have any impact on portfolio stock returns in
Ghana?
The rest of the paper is classified as follows. Section 2 reviews some related literature
on the subject matter. Section 3 explains the data. Section 4 describes the methodology
used. Section 5 describes the preliminary test on the data. Section 6 presents the results
and section 7 concludes the study.
2. Review of literature
Interest rate and exchange rate risks may vary across countries. Foreign exchange rate
exposure, interest rate risks and market risks are important financial and economic
factors that influence the value of common stocks. Below we provide a brief review of
related studies.
Abor (2005) examined how Ghanaian firms involved in international trade manage
their foreign exchange risk. The results show that one of the main problems that firms
in Ghana face is the frequent appreciation of foreign currencies against the local
currency, which tends to affect the prices of their final products sold locally. Adjasi
et al. (2008) find exchange rate to have a negative relationship with stock returns on
the Ghana Stock Exchange. In the US, Jorion (1990) finds a significant impact of
foreign exchange rate risk on stock prices only for 5.2% of the analyzed 287 US
multinationals. Choi et al. (1992) also provide empirical evidence that both interest
rates and foreign exchange rates are priced into the stock market for US banks.
According to Fraser, Madura, and Weigand. (2002), this interest rate risk extends
even to banks that are reliant on non-interest income since a high interest rate slows
down economic growth and reduces banks’ non-interest income from initial public
offerings and acquisitions of related activities. Choi and Prasad (1995) find that only
14.9% of the individual firms in the United States show a significant foreign exchange
rate exposure. Kolari et al. (2008) show that exchange rate risk measured by con-
temporaneous exchange rate changes is priced into the US stock market. However, Du
and Hu (2012) provide contrary findings explaining that the results of Kolari et al.
(2008) may have been spurious because their exchange rate risk factor had a strong
correlation with the size factor, and that their exchange rate sensitivity portfolios have
a strong factor structure.
On the Japanese market, Koch and Saporoschenko (2001) carried out an empirical
study on the sensitivity of individual and portfolio stock returns for Japanese horizontal
keiretsu financial firms to unanticipated changes in market returns, interest rates,
exchange rate changes, and nominal interest rate spread changes. Results indicate that
the stock returns of keiretsu financial firms often exhibit significant negative responses
to interest rate increases while they find that the firms have higher average market risk
but insignificant exposure to exchange rate changes. Using monthly Japanese data for
the period 1991–2005, Hartmann and Pierdzioch (2007) examined the link between
exchange rate movements and stock returns. They found that exchange rate movements
per se do not help to explain stock returns. They however provide evidence of in-sample
342 R.A. DWUMFOUR AND N.A. ADDY
predictability if one accounts for the interventions of the Japanese monetary authorities
in the foreign exchange market.
In a broader study, Bartram and Bodnar (2012) examined the importance of exchange rate
exposure in the return generating process for a large sample of non-financial firms from 37
countries and found that the effect of exchange rate exposure on stock returns is conditional
and show evidence of a significant return impact to firm-level currency exposures when
conditions of the exchange rate change. Their study shows that the realized return to exposure
is directly related to the size and sign of the exchange rate change, suggesting fluctuations in
exchange rates as a source of time-variation in currency return premia.
This study seeks to add to the body of knowledge of the literature by providing empirical
evidence from an emerging market perspective. Through this we understand the behavior
of stocks with regard to interest rates and FX rates changes. We compare the study’s results
with previous evidence in other markets to understand their consistency or otherwise.
2.1. Observations during the crisis
Here we plot the time series graph of changes in the interest rate and the exchange rates
used in the study. These are shown in Figure 1, 2, 3 and 4. From these figures the FX rates
show more volatilities than the interest rate. One observation is that GH¢/USD shows
a relative tranquility from 2005 to the end of 2007 after which volatility increased. It can be
also be seen with the FX rates that higher volatilities were experienced during the crisis
period and even periods after the crisis especially for the GBP and EURO.
3. Data
The data for this study involved firms listed on the Ghana Stock Exchange (GSE). The
selection of the firms listed on the GSE was based on the fact that these firms provided
enough data for the study period. Those companies which listed before 2005 met the data
Figure 1. Trend of Changes in Interest Rate(91-Day).
JOURNAL OF AFRICAN BUSINESS 343
Figure 2. Trend of Changes in GHȼ/USD Exchange Rate.
Figure 3. Trend of Changes in GHȼ/GBP Exchange Rate.
requirement criteria and a total of 24 firms were selected. These firms were classified into
portfolios based on their industry and their sizes (market capitalization). The formation of
portfolios provides an efficient way for condensing a substantial amount of information
about firm stock return behavior and it has the advantage of smoothing out the noisiness in
the data, due to transitory shocks to individual firms (Elyasiani & Mansur, 1998). These
shocks may, otherwise, distort the results significantly. The disadvantage of this approach,
however, is that it masks the dissimilarities among firms within each portfolio. The
portfolio returns were generated using equal weights. In all, three portfolios of industry,
namely, finance (6), manufacturing (12), and retail (6) were obtained. Three portfolios
based on size were also obtained and grouped into small (8), medium (8) and large (8) cap
firms. We employed a dummy variable to control for the 2007/2008 global financial crisis.
344 R.A. DWUMFOUR AND N.A. ADDY
Figure 4. Trend of Changes in GHȼ/EURO Exchange Rate.
We define the pre-crisis period from January 2005 toMarch 2007, the crisis period from
April 2007 to December 2008, and the post-crisis period from January 2009 to
December 2012. Weekly share prices (which we use to form the portfolios) and Ghana
Stock Exchange Composite Index (GSE-CI) information for the period were obtained from
the GSE. The currencies and the exchange rates chosen for the study were those widely used
in international transactions in Ghana. The exchange rates were the Ghana cedi to US
dollar (Gh¢/USD), the Ghana cedi to EURO (Gh¢/EURO) and the Ghana cedi to
UK pound (Gh¢/GBP). These were the end mid-point interbank weekly rates obtained
from the Bank of Ghana. The exchange rates are defined as the number of local currency
which could purchase a unit of foreign currency. Weekly data were also obtained from the
Bank of Ghana on the 91-day Treasury bill rates. We chose weekly data because this is the
best compromise between maximizing observations and minimizing the day-to-day fluc-
tuations that have less direct economic relevance (Chen & So, 2002).
4. Methodology
Since the ARCH and Generalized ARCH (GARCH) models proposed by Engle (1982)
and Bollerslev (1986), respectively, many researchers have applied these conditional
models and their extensions or modifications to revisit the CAPM and APT models. We
estimate both the OLS and GARCH models.
4.1. OLS model
In this study, the OLS model adopted is that of Salifu et al. (2007) with some addition of
variables (91-day Treasury bill rate, and a dummy variable to control for the crisis period).
Again, we group the stocks into industry and size using weekly data. In addition, we use the
GARCH (1, 1) model instead of the OLS. As stated earlier, most empirical studies employ
(linear) OLS methods to estimate the sensitivity of stock returns to interest rate and FX rate
changes.We initially follow this approach to compare our results with earlier results although
JOURNAL OF AFRICAN BUSINESS 345
this comparison is limited due to differences in the time periods of those studies and their data
frequencies. The suitability of the OLS estimation is tested with the ARCH test.
The OLS model is stated below:
rt ¼ β0 þ β1TBRt þ β2USDt þ β3GBPt þ β4EUROt þ β5RMt þ β6CRISISt þ μt (1)
where: rt denotes the capital gain or loss (portfolio return) of the portfolio for the
period t. This variable is assumed to capture all financial and economic factors that are
specific to the portfolio. TBRt denotes the change in the short term risk-free interest
rate (91-day Treasury bill rate) at time t. USDt is the change in the cedi per US dollar.
GBPt is the change of the cedi per Great Britain Pound (GBP) and EUROt is the change
of the cedi per Euro. RMt is the return on the market (GSE-CI) which is considered to
reflect economy-wide factors at time t. β1  β5 represents the sensitivity of the portfo-
lio returns to changes in interest rate, USD, GBP, EURO and RM, respectively. For
example, concerning the FX rates, a positive β2 indicates that a portfolio returns
appreciates with a depreciation of the local currency while a negative β2 indicates
that a portfolio return depreciates with a depreciation of the local currency. NB: the
change is calculated as for example: USD ¼ USD1USD0t USD , where USD1 is current exchange0
rate of Gh¢/USD and USD0 is previous exchange rate of Gh¢/USD. This is done for all
the FX rates, TBR and the RM. CRISIS is a dummy variable taking the value of 1 during
the crisis period (5 April 2007 to 24 December 2008) and 0 in other periods (pre-crisis
period–7 January 2005 to 30 March 2007 and post-crisis period 2 January 2009 to
28 De
cember 2012.). μt reflects the stochastic disturbance term assumed to be iid,
ut,N 0; 2t . Eq. (1) is estimated for each of the portfolios.
The above model states that returns are a function of contemporaneous changes in
interest rate, exchange rate and a market index with portfolio-specific intercept and slope
coefficients. The market variable is intended to capture the influence of the general market
on individual portfolio stock returns. The estimated interest rate and exchange rate
coefficients will provide a measure of the effect of interest rate and exchange rates changes
on the portfolio returns given its relation to the market index.
4.1.1. Testing for ARCH effects
Since we have argued for the case of estimation based on conditional variance, it is
useful to test for ARCH effects in the conditional variance of μt. The test is
performed on one lag of the OLS residuals of Eq. (1). Thus we test the simple ARCH
(1) specification for 2 2 2t which is given by t ¼ ρ0 þ ρ1μt1 . The null hypothesis of no
ARCH effects requires that ρ1 ¼ 0 against the alternative that ρ1
0. If the residuals in
Eq. (1) contain autocorrelation or heteroscedasticity it is likely that the null hypothesis
will be rejected. We use the LM test of order 1 for ARCH.
4.2. The GARCH model
In the empirical analysis of financial data, GARCH (1, 1) models have often been found
to appropriately account for conditional heteroscedasticity (Palm, 1996). The Generalized
346 R.A. DWUMFOUR AND N.A. ADDY
Autoregressive Conditional Heteroscedasticity (GARCH) process, first introduced by
Bollerslev (1986), is estimated next. The GARCH (1, 1) process is specified as follows1:
rt ¼ λ0 þ λ1TBRt þ λ2USDt þ λ3GBPt þ λ4EUROt þ λ5RMt þ λ6CRISISt þ εt (2:1)

 
ε 2t,N 0; σt
σ2 ¼ θ 2 2t 0 þ θ1εt1 þ Φσt1 (2:2)
where other variables are as defined before, the variance equation (2.2) includes the
long-term average volatility θ0, news about volatility from the previous period which is
defined as an ARCH term and the previous period’s forecast variance which is defined
as the GARCH term. The GARCH specification requires that in the conditional
variance equation, parameters θ0, θ1and Φ should be positive for a non-negativity
condition and the sum of θ1 and Φ should be less than one to secure the covariance
stationarity of the conditional variance.
4.2.1 Generating unexpected changes using ARIMA
We then estimate whether unexpected changes in interest rate, FX rates and market
index returns have an impact on the portfolio stock returns. We use ARIMA to extract
the unexpected changes in the interest rate, FX rates and market return variables. Fang
and Loo (1994) and, more recently, El-Masry and Abdel-Salam (2007) used the ARIMA
residuals as a proxy of unexpected changes in exchange rates and found that they have
a significant cross-sectional effect on common stock returns. Following these authors,
we use a two-step procedure to identify the unexpected changes in these variables. The
first step is to find an ARIMA model. The Akaike Information Criterion (AIC) and the
Bayesian Information Criterion (BIC) indicate ARIMA (1, 0, 0) to be appropriate for
these factors. The residuals are then defined as the unanticipated changes in interest
rate, FX rates and market returns. The second step is to substitute these residuals for the
interest rate, FX rates, and market return variables in model 2.
5. Preliminary tests
5.1. Descriptive statistics and unit root test
Table 1 reports the descriptive statistics, normality test and the unit root test (ADF) for
the returns of portfolios, the market, interest rate and FX rates. The table shows much
volatility (standard deviation) for the EURO and GBP compared to the USD. It can be
seen from the Shapiro-Wilk test that none of the variables is normally distributed at 1%
significance level. Unit root tests are important in examining the stationarity of time
series. In general, a stationary series is considered to have a constant mean and variance
and time invariant covariance. To test for stationarity we applied the well-known
Augmented Dickey-Fuller (ADF) tests which are now standard in time series applica-
tions. The tests were applied separately with a constant term only, a constant term and
1Several GARCH models including EGARCH, GARCH-M and different lag values of GARCH were estimated for all series
and the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) indicated GARCH (1, 1) to be
the appropriate parameterization for all series.
JOURNAL OF AFRICAN BUSINESS 347
trend. Even though only the results for the constant term are shown, the results for the
constant term and trend also yielded similar results; thus, from Table 1 we reject the
null hypothesis of unit root. Hence, all the variables used are stationary at levels since
all the test statistics were greater than the 1% critical values. Thus, overall, the OLS does
not appear very reliable due to violations of certain important OLS classical
assumptions.
We proceed to describe the data into pre-crisis, crisis and post-crisis to understand
the trend of the data in these sub-periods. From Table 1, it can be seen that all
portfolios generated negative returns (Mean) prior to the crisis. In particular, with
the industry portfolios, financial stocks had the maximum loss of 0.44% and the highest
risk of 1.68% while with the size portfolios, medium firms had the highest loss of 0.31%
even though their risk was 1.13% lower than that of large firms of 1.25% which was the
highest. During the crisis period, all the portfolios had positive returns. This suggests
the high risk-high return theory. Only the manufacturing, medium and large firms
sustained their positive returns even after the crisis period.
5.2. Correlation matrix
To address the problem of multicollinearity between the independent variables – espe-
cially the problem of multicollinearity between the three exchange rates, a correlation
matrix of these weekly returns was examined as seen in Table 2. As indicated by Kennedy
(2008), correlation coefficients of below 0.70 indicate that weaker relationships exist
among the independent variables, hence the avoidance of any potential multicollinearity
problems in the regression estimates.
6. Results
6.1. OLS estimations
As indicated earlier, we follow previous studies using the OLS to compare and to test
the underlying reasons for using the GARCH (1, 1) model in our estimation. Thus, here
we test the ARCH effect in the OLS residuals. Looking at the results in Table 3, the
models hardly provide a goodness of fit with low Adjusted R square values. Also, few of
the portfolios are exposed to the interest rate and exchange rate variables as well as the
market returns. Our results for the direct effect of interest rate and foreign exchange
exposure applied to the portfolios are weak, but consistent with those documented
previously in the literature. ARCH effects are found for five out of six regression
estimates. ARCH effects are very strong for the majority of the regression estimates.
Failure to account for these effects would affect the reliability of the linear estimates.
This finding would also appear to question the reliability of prior studies. Therefore,
GARCH-type models would appear to be more appropriate for estimating such data.
6.2. GARCH (1, 1) estimations with interest rate and exchange rates changes
In Table 4, we estimate the GARCH (1, 1) model. The results of the mean equation show that
interest rate has no effect on any of the portfolios. Thus, changes in Government of Ghana
Treasury bill rate have no effect on the returns of any of the portfolios. This is inconsistent
348 R.A. DWUMFOUR AND N.A. ADDY
Table 1. Descriptive Statistics.
Obs. Mean Std. Dev. Min Max Shapiro-Wilk test (Z) ADF
Financial 417 −0.0001 0.0212 −0.1300 0.0800 8.3010*** −17.7950***
Manufacturing 417 0.0013 0.0190 −0.0900 0.0900 8.9500*** −24.9420***
Retail 417 0.0001 0.0154 −0.1300 0.0700 9.1300*** −17.374***
Small 417 −0.0007 0.0137 −0.0600 0.0800 8.6120*** −21.6560***
Medium 417 0.0006 0.0269 −0.1400 0.1400 9.1300*** −24.7790***
Large 417 0.0024 0.0169 −0.1000 0.0800 8.1810*** −16.4270***
RM 417 −0.0012 0.0474 −0.8649 0.1520 12.7180*** −19.457***
TBR 417 0.0009 0.0232 −0.1451 0.2367 10.6990*** −15.187 ***
USD 417 0.0018 0.0044 −0.0087 0.0313 10.2550*** −11.813***
EURO 417 0.0018 0.0133 −0.0676 0.0871 6.6340*** −19.493***
GBP 417 0.0014 0.0134 −0.0605 0.0591 4.8860*** −20.451***
Descriptive statistics-Sub periods (pre-crisis, crisis and post-crisis)
Variable Obs. Mean Std. Dev. Min Max
Pre-crisis (Jan. 2005 – Mar. 2007)
Finance 117 −0.0044 0.0168 −0.0800 0.0400
Manufacturing 117 −0.0013 0.0077 −0.0400 0.0200
Retail 117 −0.0010 0.0095 −0.0300 0.0600
Small 117 −0.0012 0.0084 −0.0300 0.0300
Medium 117 −0.0031 0.0113 −0.0600 0.0200
Large 117 −0.0012 0.0125 −0.0500 0.0600
RM 117 −0.0024 0.0100 −0.0692 0.0128
TBR 117 −0.0046 0.0138 −0.0683 0.0329
USD 117 0.0002 0.0012 −0.0027 0.0032
EURO 117 0.0001 0.0096 −0.0313 0.0242
GBP 117 0.0004 0.0098 −0.0233 0.0231
Crisis (Apr. 2007 – Dec. 2008)
Finance 91 0.0058 0.0119 −0.0300 0.0500
Manufacturing 91 0.0044 0.0112 −0.0200 0.0600
Retail 91 0.0029 0.0072 −0.0100 0.0300
Small 91 0.0022 0.0127 −0.0200 0.0800
Medium 91 0.0042 0.0097 −0.0200 0.0600
Large 91 0.0075 0.0117 −0.0400 0.0400
RM 91 0.0080 0.0158 −0.0148 0.0872
TBR 91 0.0105 0.0343 −0.1451 0.2367
USD 91 0.0030 0.0041 −0.0070 0.0169
EURO 91 0.0039 0.0175 −0.0676 0.0871
GBP 91 0.0001 0.0166 −0.0605 0.0374
Post-crisis (Jan. 2009 – Dec. 2012)
Finance 209 −0.0003 0.0255 −0.1300 0.0800
Manufacturing 209 0.0013 0.0251 −0.0900 0.0900
Retail 209 −0.0005 0.0199 −0.1300 0.0700
Small 209 −0.0018 0.0161 −0.0600 0.0600
Medium 209 0.0012 0.0364 −0.1400 0.1400
Large 209 0.0022 0.0202 −0.1000 0.0800
RM 209 −0.0046 0.0654 −0.8649 0.1520
TBR 209 −0.0001 0.0200 −0.0525 0.1098
USD 209 0.0021 0.0054 −0.0087 0.0313
EURO 209 0.0018 0.0129 −0.0318 0.0413
GBP 209 0.0025 0.0135 −0.0342 0.0591
Note.Computations based research data. Finance: portfolio of financial stocks, Manufacturing: portfolio of manufactur-
ing stocks, Retail: portfolio of retail services stocks, Small: portfolio of small value stocks, Medium: portfolio of
medium value stocks, Large: portfolio of large value stocks, RM: return on the market, TBR: Change in the government
of Ghana 91-day Treasury bill rate, USD: change in the number of Ghana cedis per unit of USD, EURO: change in
number of Ghana cedis per unit of Euros, GBP: change in number of Ghana cedis per unit of Great Britain Pounds.
ADF: Augmented Dickey-Fuller.
*** indicates significance at 1%
JOURNAL OF AFRICAN BUSINESS 349
Table 2. Correlation Matrix.
RM TBR USD EURO GBP CRISIS
RM 1
TBR 0.0441 1
USD −0.0201 0.3039*** 1
EURO 0.0188 0.0596 0.1161** 1
GBP −0.0105 0.0461 0.0761 0.5276*** 1
CRISIS 0.1035** 0.2189*** 0.1426*** 0.0829* −0.0508 1
Note. Computations from research data. RM: return on the market, TBR: Change in the government of Ghana 91-day
Treasury bill rate, USD: change in the number of Ghana cedis per unit of USD, EURO: change in number of Ghana
cedis per unit of Euros, GBP: change in number of Ghana cedis per unit of Great Britain Pounds.
***, **, * indicate significance at 1%, 5% and 10% respectively.
with the results of Adjasi et al. (2008) who found that interest rate affects the stock market
index. Again, unlike Salifu et al. (2007) who use OLS andmonthly data on the GSE and show
that the financial stocks did not show any risk exposure to any of the international currencies,
the results of this study show that changes in the US dollar rate had a significant negative
impact in the estimations specifically for financial and large stocks portfolios. Both GBP
and EURO had no impact on any of the portfolio returns. Thus, even though the descriptive
statistics show higher volatilities of GBP (1.34%) and Euro (1.33%) than theUSD, it seems the
financial and large portfolios are more engaged in foreign transactions in the USD, hence
their exposure. Thus, the rationale for the negative relationship with the USD and the
financial stocks can probably be explained by the size of foreign currency denominated assets
and liabilities in the balance sheets of these firms. Themovements in exchange rates can affect
a bank balance sheet directly by generating translation gains or losses based on the net foreign
positions. When foreign currency denominated liabilities exceed foreign currency denomi-
nated assets, the depreciation of the Ghana cedis may lead to damage in the bank balance
sheet, and the weakening of bank equity may result in a decline in bank stock returns. This is
consistent with the results of Kasman, Vardar, and Tunc (2011), who also find the US dollar
to negatively affect the financial stocks. This is also consistent with the results of El-Masry and
Abdel-Salam (2007), who also find exchange rate exposure to have a more significant impact
on stock returns of the large firms compared with the small and medium-sized companies.
The results also show that changes in the market returns significantly and positively
affect four out of the six portfolios returns. Thus, compared with the interest rate and
the exchange rates, the market index returns explain most of the variations in the
portfolio returns. This is consistent with the results of Joseph and Vezos (2006) and
Kasman et al. (2011), who find market returns to explain most variations of stock
returns in their samples. In addition, we find that generally the 2007/2008 financial
crisis had a positive impact on the portfolio stock returns. This supports the classical
positive risk-return relationship as espoused by Merton (1973, 1980)), contrary to the
normal expectation of a negative risk-return relationship during crisis periods. This
result may be because the GSE is not well integrated into the global stock market and
hence, during the 2007/2008 financial crisis, investors shifted their investment from
developed markets to invest in an emerging economy like Ghana, causing increases in
expected returns of the stocks. This presents emerging stock markets as good alter-
natives to invest in during financial crises that happen in developed markets. This
seems to affirm the contribution of Sarno et al. (2016) and Fratzscher (2012) who
conclude that, during the recent global financial crisis, the influence of push factors
350 R.A. DWUMFOUR AND N.A. ADDY
Table 3. OLS Estimation.
β0 β1 β2 β3 β4 β5 β6 R
2 ARCH (1)
Finance −0.0005 (0.0012) 0.0413 (0.0460) −0.6668 *** (0.2378) 0.0781 (0.0885) −0.0710 (0.0893) 0.0924 *** (0.0212) 0.0073 *** (0.0025) 0.0732 8.6560***
Manufacturing 0.0004 (0.0011) 0.0121 (0.0430) 0.0722 (0.2219) 0.0953 (0.0826) −0.1423 * (0.0833) 0.0169 (0.0198) 0.0041 * (0.0023) 0.0024 7.4430 ***
Retail −0.0005 (0.0008) 0.0131 (0.0347) −0.1904 (0.1791) 0.0239 (0.0667) −0.0290 (0.0672) −0.0352 ** (0.0160) 0.0042 ** (0.0019) 0.0092 0.0010
Small −0.0016** (0.0008) 0.0140 (0.0306) 0.0913 (0.1579) 0.0935 (0.0587) −0.0738 (0.0593) 0.0348 ** (0.0141) 0.0034** (0.0017) 0.0212 3.2340*
Medium −0.0003 (0.0016) 0.0254 (0.0611) −0.0761 (0.3155) 0.0880 (0.1174) −0.0632 (0.1154) −0.0218 (0.0282) 0.0049 (0.0033) −0.0064 6.5680**
Large 0.0023** (0.0010) 0.0262 (0.0369) −0.6248 *** (0.1904) 0.0199 (0.0709) −0.1090 (0.0715) 0.0583*** (0.0170) 0.0067*** (0.0020) 0.0725 14.9940***
No. of significant 2/6 0/6 2/6 0/6 1/6 4/6 5/6 5/6
cases
Note. Computations from research data. Standard errors in parenthesis. *** indicates significance at 1%, ** significance at 5% and * significance at 10%. Finance: portfolio of financial stocks,
Manufacturing: portfolio of manufacturing stocks, Retail: portfolio of retail services stocks, Small: portfolio of small value stocks, Medium: portfolio of medium value stocks, Large: portfolio of large
value stocks, RM: return on the market, TBR: Change in the government of Ghana 91-day Treasury bill rate, USD: change in the number of Ghana cedis per unit of USD, EURO: change in number
of Ghana cedis per unit of Euros, GBP: change in number of Ghana cedis per unit of Great Britain Pounds.
Estimation for Model 1: rt ¼ β0 þ β1TBRt þ β2USDt þ β3GBPt þ β4EUROt þ β5RMt þ β6CRISISt þ μt
JOURNAL OF AFRICAN BUSINESS 351
Table 4. GARCH (1,1) results with changes in interest rate and exchange rates.
λ0 λ1 λ2 λ3 λ4 λ5 λ6 θ0 θ1 Φ
Finance −0.0009 (0.0008) 0.0005 (0.0340) −0.5240*** (0.1514) 0.0900 (0.0560) −0.0481 (0.0617) 0.3559*** (0.0231) 0.0057*** (0.0018) 0.0001*** (0.0000) 0.3776*** (0.0833) 0.2764*** (0.0738)
Manufacturing −0.0005 (0.0006) 0.0089 (0.0294) 0.1677 (0.1319) −0.0120 (0.0432) −0.0590 (0.0483) 0.0818*** (0.0113) 0.0020* (0.0010) 0.0000*** (0.0000) 0.0759*** (0.0114) 0.9023*** (0.0114)
Retail −0.0009 (0.0007) 0.0131 (0.0145) 0.0377 (0.1174) 0.0066 (0.0352) −0.0447 (0.0353) −0.0026 (0.0168) 0.0024** (0.0012) 0.0000** (0.0000) 0.0644*** (0.0101) 0.9052*** (0.0120)
Small −0.0014 (0.0010) 0.0277 (0.0330) 0.0578 (0.1753) 0.0548 (0.0530) −0.0421 (0.0672) 0.0297 (0.0404) 0.0050*** (0.0017) 0.0000** (0.0000) 0.0406*** (0.0111) 0.9150*** (0.0213)
Medium −0.0011 (0.0008) 0.0494 (0.0356) −0.1078 (0.1189) 0.0580 (0.0488) −0.0709 (0.0505) 0.1148*** (0.0150) 0.0035*** (0.0012) 0.0000** (0.0000) 0.1035*** (0.0158) 0.8753*** (0.0138)
Large 0.0016*** (0.0006) −0.0022 (0.0218) −0.1842* (0.1003) 0.0179 (0.0424) −0.0481 (0.0431) 0.4383*** (0.0234) 0.0039*** (0.0011) 0.0001*** (0.0000) 0.3625*** (0.0979) 0.2081*** (0.0680)
No. of significant. 1/6 0/6 2/6 0/6 0/6 4/6 6/6 6/6 6/6 6/6
cases
Note. Computations from research data. Standard errors in parenthesis. *** indicates significance at 1%, ** significance at 5% and * significance at 10%. Finance: portfolio of financial stocks,
Manufacturing: portfolio of manufacturing stocks, Retail: portfolio of retail services stocks, Small: portfolio of small value stocks, Medium: portfolio of medium value stocks, Large: portfolio of large
value stocks, RM: return on the market, TBR: Change in the government of Ghana 91-day Treasury bill rate, USD: change in the number of Ghana cedis per unit of USD, EURO: change in number of
Ghana cedis per unit of Euros, GBP: change in number of Ghana cedis per unit of Great Britain Pounds
Estimation for Model: rt ¼ λ0 þ λ1TBRt þ λ2USDt þ λ3GBPt þ λ4EUROt þ λ5RMt þ λ6CRISISt þ εt ð2:1Þ, σ2t ¼ θ0 þ θ ε21 t1 þ Φσ2t1 (2.2)
352 R.A. DWUMFOUR AND N.A. ADDY
(mostly explained by global economic forces) increases in explaining international
portfolio flows. This may not be surprising as Cenedese and Mallucci (2016) find that
international equity outflows are influenced by cash-flow news not real interest rate
shocks and exchange rate shocks.
In the conditional variance equation, the intercept term is significantly positive for
all the portfolios. This indicates that there is a significant time-invariant component
in the return generating process. Further, both the ARCH and GARCH parameters
satisfy the non-negativity condition. Also, the GARCH parameter is significantly
greater than the ARCH parameter for all portfolios except for financial and large
stocks. This implies that the volatility of each stock return is more sensitive to its own
lagged values than it is to new surprises. However, the financial and large stocks are
more sensitive to new surprises. Again, the sum of the ARCH and GARCH para-
meters for most of the regressions are close to unity, indicating that these returns have
highly persistent effects and the response of volatility decays at a slower rate. In
particular, market returns are more persistent than interest rate and exchange rate
changes in explaining the conditional stock returns
6.3. GARCH (1, 1) estimations with unexpected changes in interest rate and
exchange rates
Table 5 shows the impact of unexpected changes of interest rate and exchange rates on
the portfolio returns. Here, in the conditional mean equation, the market returns again
are seen to explain more of the conditional stock return changes than unexpected
changes in interest rate and exchange rates. These results are similar to the earlier
estimations for changes in these rates except that here the unexpected changes in the
Gh¢/USD had no impact on the portfolio returns. The crisis period also positively
affects all the portfolio returns. In the conditional variance equation, similar results are
obtained as in the previous estimation.
6.4. GARCH (1, 1) estimations with interactions with CRISIS
Here, knowing that the crisis period did have an impact in almost all the portfolios, we
proceed to identify the main transmission channel of this impact by interacting the
crisis dummy variables with our other variables. We estimate using Model 2. The results
are presented in Table 6. The results indicate that the major transmission channel of the
financial crisis was through the market index returns not the interest rate and FX rates.
As indicated earlier, this seems to affirm the contribution of Sarno et al. (2016) and
Fratzscher (2012) who conclude that during the financial crisis, the influence of push
factors (mostly explained by global economic forces) increases in explaining interna-
tional portfolio flows, and the results of Cenedese and Mallucci (2016) who find that
international equity flows are not influenced by real interest rate shocks or exchange
rate shocks.
The interaction term (RM*CRISIS) was significantly negative for financial, medium
and large portfolios. Given that the unconditional effects and conditional effects
corresponding to all other inclusive variables (USD, GBP, EURO, TBR and their
interactions with CRISIS) are not jointly significant, the computation of net effects
JOURNAL OF AFRICAN BUSINESS 353
Table 5. GARCH (1,1) sector results with unexpected changes in interest rate, exchange rates and market returns.
λ0 λ1 λ2 λ3 λ4 λ5 λ6 θ0 θ1 Φ
Finance −0.0017** (0.0008) −0.0089 (0.0377) −0.2126 (0.2175) 0.0797 (0.0607) −0.0438 (0.0643) 0.3566*** (0.0239) 0.0046** (0.0019) 0.0001*** (0.0000) 0.3264*** (0.0736) 0.3213*** (0.0742)
Manufacturing −0.0005 (0.0006) 0.0083 (0.0339) 0.1379 (0.1488) −0.0127 (0.0424) −0.0560 (0.0486) 0.0797*** (0.0112) 0.0021** (0.0010) 0.0000*** (0.0000) 0.0772*** (0.0116) 0.9006*** (0.0116)
Retail −0.0009 (0.0007) 0.0050 (0.0128) 0.0771 (0.1285) 0.0056 (0.0363) −0.0435 (0.0356) −0.0028 (0.0168) 0.0025** (0.0012) 0.0000** (0.0000) 0.0656*** (0.0102) 0.9037*** (0.0121)
Small −0.0013 (0.0009) 0.0323 (0.0316) 0.0742 (0.2238) 0.0577 (0.0546) −0.0459 (0.0692) 0.0302 (0.0425) 0.0051*** (0.0016) 0.0000** (0.0000) 0.0414*** (0.0113) 0.9138*** (0.0212)
Medium −0.0014** (0.0007) 0.0375 (0.0414) −0.0211 (0.1354) 0.0564 (0.0488) −0.0695 (0.0494) 0.1180*** (0.0152) 0.0036*** (0.0012) 0.0000** (0.0000) 0.1010*** (0.0154) 0.8776*** (0.0136)
Large 0.0007 (0.0005) −0.0111 (0.0232) −0.1020 (0.1361) 0.0106 (0.0431) −0.0485 (0.0429) 0.4388*** (0.0237) 0.0039*** (0.0011) 0.0001*** (0.0000) 0.3683*** (0.0992) 0.2176*** (0.0686)
No. of significant 2/6 0/6 0/6 0/6 0/6 4/6 6/6 6/6 6/6 6/6
cases
Note. Computations from research data. Standard errors in parenthesis. *** indicates significance at 1%, ** significance at 5% and * significance at 10%. Finance: portfolio of financial stocks,
Manufacturing: portfolio of manufacturing stocks, Retail: portfolio of retail services stocks, Small: portfolio of small value stocks, Medium: portfolio of medium value stocks, Large: portfolio of large
value stocks, RM: return on the market, TBR: Change in the government of Ghana 91-day Treasury bill rate, USD: change in the number of Ghana cedis per unit of USD, EURO: change in number of
Ghana cedis per unit of Euros, GBP: change in number of Ghana cedis per unit of Great Britain Pounds
Estimation for Model: rt ¼ λ0 þ λ1TBRt þ λ2USDt þ λ GBP þ λ EURO þ λ RM þ λ CRISIS þ ε (2.1), σ2 ¼ θ þ θ ε2 þ Φσ23 t 4 t 5 t 6 t t t 0 1 t1 t1 (2.2)
354 R.A. DWUMFOUR AND N.A. ADDY
Table 6. GARCH (1,1) results with interaction with CRISIS.
Finance Manufacturing Retail Small Medium Large
Mean equation
Constant −0.0009 (0.0008) −0.0005 (0.0006) −0.0010 (0.0007) −0.0013 (0.0009) −0.0012 (0.0008) 0.0016*** (0.0006)
TBR −0.0008 (0.04870) −0.0237 (0.03662) −0.0231 (0.0365) 0.0069 (0.0560) 0.0577 (0.0440) −0.0189 (0.0316)
TBR*CRISIS 0.0081 (0.0652) 0.08266* (0.0434) 0.0485 (0.0410) 0.0604 (0.0649) −0.0195 (0.0778) 0.1022* (0.0496)
USD −0.5184*** (0.1668) −0.0275 (0.1761) −0.0059 (0.1869) −0.0827 (0.1917) −0.0195 (0.1436) −0.2433** (0.1107)
USD*CRISIS 0.0052 (0.5142) 0.7908*** (0.2831) 0.1899 (0.2637) 0.6771 (0.5563) 0.0672 (0.2608) 0.2725 (0.2674)
GBP 0.1477** (0.0626) −0.0334 (0.0557) −0.0070 (0.0529) 0.0892 (0.0580) 0.0787 (0.0627) 0.0297 (0.0496)
GBP*CRISIS −0.1361 (0.0717) 0.1054 (0.1007) 0.0322 (0.0779) −0.1056 (0.2014) −0.036 (0.0983) −0.0235 (0.0921)
EURO −0.1312* (0.0717) −0.0459 (0.0163) 0.0335 (0.0588) −0.0597 (0.0830) −0.0101 (0.0880) −0.0877* (0.0534)
EURO*CRISIS 0.1833 (0.1681) −0.0622 (0.0919) −0.1406* (0.0767) 0.0670 (0.1583) −0.1004 (0.1102) 0.0699 (0.0872)
RM 0.3826*** (0.02630 0.0655*** (0.0113) −0.0051 (0.0184) 0.0271 (0.0358) 0.1802*** (0.0191) 0.4394*** (0.0253)
RM*CRISIS −0.1607* (0.0946) 0.0432 (0.0596) 0.0016 (0.0458) 0.0038 (0.1468) −0.1635*** (0.0502) −0.3380*** (0.0567)
CRISIS 0.0059** (0.0023) 0.0006 (0.0011) 0.0025* (0.0013) 0.0025 (0.0024) 0.0043*** (0.0013) 0.0036*** (0.0013)
Net effects 0.2219 n.a. n.a. n.a. 0.0167 0.1014
Variance equation
Constant 0.0000*** (0.0000) 0.0000*** (0.0000) 0.0000** (0.0000) 0.0000** (0.0000) 0.0000** (0.0000) 0.0001*** (0.0000)
Arch(1) 0.3626*** (0.0801) 0.0789*** (0.0118) 0.0643*** (0.0103) 0.0520*** (0.0144) 0.0985*** (0.0147) 0.3625*** (0.1001)
Garch(1) 0.2626*** (0.0757) 0.8989*** (0.0116) 0.9057*** (0.0122) 0.8879*** (0.0276) 0.8783*** (0.0131) 0.2075*** (0.0692)
Note. Computations from research data. Standard errors in parenthesis. *** indicates significance at 1%, ** significance at 5% and * significance at 10%. Finance: portfolio of financial stocks,
Manufacturing: portfolio of manufacturing stocks, Retail: portfolio of retail services stocks, Small: portfolio of small value stocks, Medium: portfolio of medium value stocks, Large: portfolio
of large value stocks, RM: return on the market, TBR: change in the government of Ghana 91-day Treasury bill rate, USD: change in the number of Ghana cedis per unit of USD, EURO:
change in number of Ghana cedis per unit of Euros, GBP: change in number of Ghana cedis per unit of Great Britain Pounds.
Net effects calculated for RM*CRISIS as their conditional and unconditional effects are significant. n.a: not applicable because at least one estimated coefficient needed for the computation of
net effects is not significant.
JOURNAL OF AFRICAN BUSINESS 355
is not feasible for them. We therefore compute the net effects for the ‘RM*CRISIS’
relationship to evaluate the complementary role of financial crisis in the effect of RM
on the financial, medium and large portfolios. For instance, in the second column of
Table 6, the net effect obtained from the interaction between RM and CRISIS is 0.2219
([−0.1607 × 1] + [0.3826]), where: 0.3826 is the unconditional effect of RM; 1 is the
value when crisis occurs and −0.1607 is the conditional effect from the interaction
between RM and CRISIS. The results show that all the net effects of CRISIS on RM for
financial, medium and large portfolios remain positive (columns 2, 6 and 7).
Thus, we find that the net effect of financial crisis at the macro level enhances the
positive impact of the market index returns (RM) on financial, medium and large
portfolios even though the impact on these portfolios is lower than the impact of the
unconditional effect of RM. This shows that in periods of financial crises, higher
volatility of the economy-wide factors leads to a decrease in the positive impact of
RM on these portfolio returns.
7. Conclusion and recommendations
It is observed that few studies have been undertaken to explore the impact of interest rate
and exchange rates and their unanticipated changes on stock returns in Ghana controlling
for market returns (as well as unanticipated market returns) and the financial crisis.
Especially after the 2007/2008 financial crisis, no study has been carried out in Ghana to
test the impact of the crisis on stock returns. This study fills the gap. The results show that
depreciation of the Ghana cedis against the US dollar reduces the returns of the financial
stocks and large cap firms. Interest rate, GBP and EURO had no impact on any of the
portfolios. The study also shows that changes in market returns explain most of the
variations of the portfolio returns. The results, however, show that in the period of financial
crisis, higher volatility of the economy-wide factors leads to a decrease in the positive
impact of RM on these portfolio returns. The return generating process of all the portfolios
is affected by their own previous information and new surprises, even though their own
previous information exerts more variations in almost all the portfolios except for financial
and large stock portfolios, where new surprises explain most of their return variations. The
results for the unanticipated changes in interest rate, exchange rates and market returns
were similar to the results in their level changes.
The study recommends that financial firms and large stock firms should manage their
exchange rate exposure to the US dollar well. A look at their dollar denominated balance
sheet items would help them control any mismatch. Stakeholders should be more con-
cerned about the betas of firms to the market returns. All exposures to the market returns
were significantly positive showing that most of the returns on the GSE move in tandem
with the market changes. Particular attention should be paid to financial and large stocks as
new surprises influence most of its return generating process than previous information
about these stocks.
Disclosure statement
No potential conflict of interest was reported by the authors.
356 R.A. DWUMFOUR AND N.A. ADDY
ORCID
Richard Adjei Dwumfour http://orcid.org/0000-0002-0408-9159
References
Abor, J. (2005). Managing foreign exchange risk among Ghanaian firms. Journal of Risk Finance,
6(4), 306–318.
Ackah, C. G., Aryeetey, E. B., & Aryeetey, E. (2009). Global financial crisis: Discussion paper 5:
Ghana. London: Overseas Development Institute and Institute of Statistical, Social and
Economic Research.
Adjasi, C., Harvey, S. K., & Agyapong, D. (2008). Effect of exchange rate volatility on the Ghana
Stock exchange. African Journal of Accounting, Economics, Finance and Banking Research, 3
(3), 28–47.
Akhigbe, A., Madura, J., & Marciniak, M. (2012). Bank capital and exposure to the financial
crisis”. Journal of Economics and Business, 64, 377–392.
Allayannis, G., & Ihrig, J. (2001). Exposure and markups. Review of Financial Studies, 14(3),
805–835.
Bartram, S. M., & Bodnar, G. M. (2012). Crossing the lines: The conditional relation between
exchange rate exposure and stock returns in emerging and developed markets. Journal of
International Money and Finance, 31, 766–792.
Bartram, S. M., & Karolyi, G. A. (2006). The impact of the introduction of the euro on foreign
exchange rate risk exposures. Journal of Empirical Finance, 13(4–5), 519–549.
Bodnar, G. M., Dumas, B., & Marston, R. C. (2002). Pass-through and exposure. Journal of
Finance, 57, 199–231.
Bollerslev, T. (1986). Generalized autoregressive conditional heteroscedasticity. Journal of
Econometrics, 31, 307–327.
Cenedese, G., & Mallucci, E. (2016). What moves international stock and bond markets? Journal
of International Money and Finance, 60, 94–113.
Chamberlain, S., Howe, J., & Popper, H. (1997). The exchange rate exposure of US and Japanese
banking institutions. Journal of Banking and Finance, 21, 871–892.
Chen, C. C., & So, R. W. (2002). Exchange rate variability and the riskiness of US multinational
firms: Evidence from the Asian financial turmoil. Journal of Multinational Financial
Management, 12(4), 411–428.
Choi, J. J., Elyasiani, E., & Kopecky, K. J. (1992). The sensitivity of bank stock returns to market:
Interest and exchange risks. Journal of Banking and Finance, 16, 983–1004.
Choi, J. J., & Prasad, A. M. (1995). Exchange risk sensitivity and its determinants: A firm and
industry analysis of US multinationals. Financial Management, 24, 77–88.
Du, D., & Hu, O. (2012). Exchange rate risk in the US stock market. Journal of International
Financial Markets, Institutions and Money, 22, 137–150.
El-Masry, A., & Abdel-Salam, O. (2007). Exchange rate exposure: Do size and foreign operations
matter? Managerial Finance, 33(9), 741–765.
Elyasiani, E., & Mansur, I. (1998). Sensitivity of the bank stock returns distribution to changes in
the level and volatility of interest rate: A GARCH-M model. Journal of Banking and Finance,
22, 535–563.
Engle, R. F. (1982). Autoregressive conditional heteroscedasticity with estimates of the variance
of United Kingdom Inflation. Econometrica, 50, 987–1007.
Fang, H., & Loo, J. C. H. (1994). Dollar value and stock returns. International Review of
Economics and Finance, 3(2), 221–231.
Francis, B. B., Hasan, I., & Hunter, D. M. (2002). Emerging market liberalization and the impact
on uncovered interest rate parity. Journal of International Money and Finance, 21(6), 931–956.
Fraser, D. R., Madura, J., & Weigand, R. A. (2002). Sources of bank interest rate risk. Financial
Review, 37, 351–368.
JOURNAL OF AFRICAN BUSINESS 357
Fratzscher, M. (2012). Capital flows, push versus pull factors and the global financial crisis.
Journal of International Economics, 88(2), 341–356.
Griffin, J. M., & Stulz, R. M. (2001). International competition and exchange rate shocks: A
cross-country industry analysis of stock returns. Review of Financial Studies, 14, 215–241.
Hartmann, D., & Pierdzioch, C. (2007). Exchange rates, interventions, and the predictability of
stock returns in Japan. Journal of Multinational Financial Management, 17, 155–172.
Hirtle, B. (1991). Factors affecting the competitiveness of internationally active financial
institutions. FRBNY Quarterly Review, 16(1), 38–51.
Jorion, P. (1990). The exchange rate exposure of US multinationals. Journal of Business, 63,
331–346.
Joseph, N. L., & Vezos, V. (2006). The sensitivity of US banks’ stock returns to interest rate and
exchange rate changes. Managerial Finance, 32(2), 182–199.
Kasman, S., Vardar, G., & Tunç, G. (2011). The impact of interest rate and exchange rate
volatility on banks’ stock returns and volatility: Evidence from Turkey. Economic Modelling,
28, 1328–1334.
Kennedy, P. (2008). A guide to econometrics (6th ed.). Oxford, UK: Blackwell Publishing.
Kenourgios, D., & Padhi, P. (2012). Emerging markets and financial crises: Regional, global or
isolated shocks?”. Journal of Multinational Financial Management, 22, 24–38.
Koch, T. W., & Saporoschenko, A. (2001). The effect of market returns, interest rates, and
exchange rates on the stock returns of Japanese horizontal keiretsu financial firms. Journal of
Multinational Financial Management, 11, 165–182.
Kolari, J., Moorman, T., & Sorescu, S. M. (2008). Foreign exchange risk and the cross-section of
stock returns. Journal of International Money and Finance, 27, 1074–1097.
Marston, R. (2001). The effects of industrial structure on economic exposure. Journal of
International Money and Finance, 20, 149–164.
Merton, R. C. (1973). An intertemporal capital asset pricing model. Econometrica, 41, 867–887.
Merton, R. C. (1980). On estimating the expected return on the market. Journal of Financial
Economics, 8, 323–361.
Palm, F. C. (1996). GARCH models of volatility. In G. Maddala & C. Rao (Eds.), Handbook of
statistics: Statistical methods in finance (Vol. 14, pp. 209–240). Amsterdam: Elsevier.
Salifu, Z., Osei, K. A., & Adjasi, C. K. D. (2007). Foreign exchange risk exposure of listed
companies in Ghana. The Journal of Risk Finance, 8(4), 380–393.
Sarno, L., Tsiakas, I., & Ulloa, B. (2016). What drives international portfolio flows? Journal of
International Money and Finance, 60, 53–72.