Received: 17 April 2020

| Revised: 14 December 2021

| Accepted: 17 December 2021

DOI: 10.1111/rode.12858

REGULAR ARTICLE

Exchange rate uncertainty and foreign

direct investment in Africa: Does financial
development matter?

Michael Effah Asamoah1 | Imhotep Paul Alagidede2 |

Frank Adu3

1
Accounting Department, University of
Ghana Business School, University of Abstract
Ghana, Accra, Ghana The orthodox view is that uncertainty deters invest-
2
Wits Business School, University of the ments and, by extension, private capital inflows. Paying
Witwatersrand, Johannesburg, South
Africa
specific attention to the volatility of the domestic ex-
3
African Center for Economic change rate, foreign direct investment (FDI), and fi-
Transformation, Accra, Ghana nancial development indicators, this study investigates
the impact of exchange rate uncertainty on FDI and
Correspondence
Michael Effah Asamoah, Accounting
whether financial development matters in such asso-
Department, University of Ghana ciation. We establish our empirical relationship with a
Business School, University of Ghana, system general methods of moments (GMM) two-­step
Accra, Ghana.
Emails: [email protected], robust estimator with orthogonal deviations. We found
[email protected] evidence supporting a nonlinear U-­shaped relationship
between uncertainty and FDI and that the impact of
uncertainty on FDI depends on varying levels of uncer-
tainty. We also document that uncertainty deters FDI
flows and that countries with a well-­functioning finan-
cial development can transform the adverse impact of
volatility on FDI. However, curbing the adverse effect
depends on the specific indicator and the threshold
value of financial development (financial institutions or
financial markets).

KEYWORDS
exchange rate, financial development, foreign direct investment,
marginal effects, nonlinearity, systems GMM, uncertainty

Frank Adu is a Transformation Fellow.

878 | © 2022 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/rode Rev Dev Econ. 2022;26:878–898.
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 879

JEL CLASSIFICATION
O2; O24

1 | I N T RO DU CT ION

The unpredictability of volatilities associated with capital flows adversely affects the quantum of
inward capital flows to destination countries. Macroeconomic uncertainties can lead to sudden
stops and reversal of capital flows resulting in a change in quantum and direction. Uncertainties
can hinder the attraction of private capital, which can adversely impact the growth of countries
that hugely depend on them. Shocks affecting capital flows may impede the directional move-
ment of the flows. According to a UNDP (2011) report, “A financial shock can result in the sud-
den reversal of capital flows and also in a sharp decline in inflows” (p. 86). The report further
admonished, “For this reason, policy measures to build a country's resilience to private capital-­
related shocks should focus on stabilizing the volatility associated with private capital flows”
(UNDP, 2011, p. 87). Such volatilities could result from the devaluation of the exchange rate,
inflationary pressures, and growth volatility. Lensink and Morrissey (2006) posit that countries
deemed political and economically volatile attract few foreign investors. Instabilities associated
with capital flows could also lead to distortions in the economic progress of countries (Forbes &
Warnock, 2012).
The theory of investment irreversibility primarily influences the adverse impact of volatili-
ties on investment and, by extension, capital flows. According to the proponents of investment
irreversibility and the “options to wait,” future uncertainty leads to a postponement of current
investment as economic agents will hold onto additional investment until the level of uncertainty
has been dispelled. Again, since the returns on investment are unpredictable during uncertainty,
investors will curtail additional financing. The unpredictability of returns indicates an inverse re-
lationship between investments and uncertainty (Dixit et al., 1994; Nickell, 1978; Pindyck, 1993).
In line with the aforementioned theoretical stance, some empirical studies have found a negative
relationship between uncertainty and private capital flows (Asamoah et al., 2016; Cavallari &
d'Addona, 2013; Jehan & Hamid, 2017). However, the empirical conclusions on the uncertainty-­
investment dynamics do not support a vigorous one-­sided conclusion of a negative relationship.
Some studies have found volatility to positively affect investment and capital flows (Corden,
1990; Daly & Vo, 2013). The positive association results from the theories of risk neutrality and
investment reversibility. Hartman (1972) and Abel (1983) posit that uncertainty causes the desire
to increase current investment, as the marginal product of capital and profits is expected to in-
crease with increasing uncertainty.
Considering these opposing theories, we contend that the association between capital flows
and volatility may not be strictly monotonic (positive or negative). With a prime focus on the
domestic exchange rate, the study seeks to reexamine the dynamics between macroeconomic
uncertainty and private capital flows in Africa by testing the two contrasting theories (investment
irreversibility and investment reversibility). Many existing studies take a linear relationship as
given, raising doubt on the validity of the conclusions from such studies. Therefore, the study
contends that there exists a U-­shaped relationship between exchange rate volatility and foreign
direct investment (FDI), where both theories could exist at extreme ends of the curve. First, the
study focuses on a more stringent nonlinear procedure proposed by Lind and Mehlum (2010) to
examine the two opposing theories (i.e., investment irreversibility and risk neutrality) (Table A1).
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
880 |    ASAMOAH et al.

Second, we examine the structural dynamics in host countries to shape the macroeconomic
volatilities–­capital flows relationship. Based on the investment irreversibility theory and the
empirical works of Asamoah et al. (2016) and Kyereboah-­Coleman and Agyire-­Tettey (2008),
exchange rate volatility deters FDI flows. However, on the notion of absorptive capacities
(Durham, 2004) and the perceived positive impact of financial development (FD) on capital
flows (Agbloyor et al., 2014; Dutta & Roy, 2011), we employ FD as the mediating variable in the
volatility–­FDI nexus. Lane (2015) and Guichard (2017) note that the adverse impact from capital
flows could be profound in the absence of weak institutions and FD. Consequently, the study fo-
cuses on the mediating role of FD in dealing with any adverse impact of exchange rate volatility
on FDI flows. Thus, this study is concerned with the impact of exchange rate volatility on FDI
and whether FD moderates the relationship. Regarding FD, the study employed recently devel-
oped indicators that overcome the limitations of single indices. These indicators account for the
complexities and multifaceted characteristics of FD regarding financial access, efficiency, and
depth. The study is the first to employ these indicators as mediating variables in the exchange
rate volatility–­FDI relationship.
Finally, the study fills a void in African literature. In separate studies, Caporale et al. (2015)
and Demir (2009) posit that empirical conclusions on the linkage between capital flows and
economic uncertainties are considerably very few, especially in the African context. Our study,
therefore, seeks to expand the scope of literature with a focus on the volatilities associated with
capital flows in Africa. The study thus differs from previous African studies in varied aspects.
First, Kodongo and Ojah (2012) study the linkages between the exchange rate and equity flows
in Africa; they did not consider the impact of exchange rate volatility and the mediating role
of FD. Even though Kyereboah-­Coleman and Agyire-­Tettey (2008) and Udoh and Egwaikhide
(2008) concentrated on exchange rate uncertainty and FDI inflows in two separate countries,
Ghana and Nigeria, both studies did not consider the effect of a moderating variable. Second,
these studies did not account for the nonlinear impact of exchange rate volatility. In a recent
study on exchange rate volatility and FDI flows in Africa, Asamoah et al. (2016) accounted for the
moderating effects of institutional quality and not FD. Again, their study failed to examine the
marginal effect of the moderating variable in the association between exchange rate uncertainty
and FDI inflows.
The study, therefore, seeks to address the following issues:

1. What is the nature of the relationship between FDIs and exchange rate volatility?
2. Can FD moderate the association between exchange rate volatility and inward FDI?
3. At what points or critical values can FD neutralize the potential adverse impact of exchange
rate volatility on FDI?

The rest of the study is structured as follows. Section 2 explains the data and estimation pro-
cedures. Section 3 focuses on the results from the econometric estimations. Section 4 concludes
the study and proffers policy directions.

2 | DATA AN D E ST IMAT ION TECHNIQUE

The study consists of 40 African countries1 for the period 1990–­2018. The choice of the study
period was influenced by two issues. First, the assertions of Gourinchas and Jeanne (2013)
on when most countries were deemed to be financially open. Gourinchas and Jeanne (2013)
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 881

contend that measures of financial openness show that from the mid-­1980s through the 1990s,
most countries had liberalized their economies and were open to external financial inflows.
Second, the Chinn and Ito (2008) measure of financial openness indicates that the levels of
financial openness for the samples in our study ranged between −1.917 and 2.347, indicating
a high level of financial openness among our sample. 2 Our dependent variable is FDI, which
is the decision by a foreign entity to acquire a lasting interest in another entity other than
one in its home country, where such interest is usually not less than a 10% stake. The interest
consists of accumulating equity capital, reinvestment of earnings, and other long-­ and short-­
term capital as shown in the balance of payments. FDI is net inflows scaled by gross domestic
product (GDP).

2.1 | Exchange rate volatility

Though Multi National Enterprises (MNEs) desire to invest abroad as a form of diversification,
the domestic exchange rate volatility may lead to increases in the cost of international business
transactions, a decrease in profits, and a decrease in the volumes of cross-­border capital flows.
We first estimate the real effective exchange rate (REER) using the purchasing price parity (PPP)
approach. According to the PPP, a country's REER is a function of its nominal exchange rate
(NER) relative to a foreign price level ratio to the national price levels. We proxy our foreign price
levels by the United States producers' price index and adjust the NER by the price differentials
between the United States and each of our sample countries. Following Elbadawi (1992), we
define the REER for each country as follows:

PPIUS
RERRi = NERi × (1)
CPIi

where NERi denotes the NER of a country, which is the value of the domestic currency needed to
exchange a unit of the U.S. dollar. PPIUS represents the producers' price index of the United States,
which is our proxy for foreign price levels. CPIi denotes the domestic price level, which is captured
by the consumer price index. It implies that a decrease (an increase) in the REER leads to a real ap-
preciation (depreciation) of the domestic currency. We obtain data on the United States producers’
price index from Federal Reserve Economic Data.
Following Bah and Amusa (2003) and McKenzie (1999), we employ the GARCH family mod-
els developed by Bollerslev (1986) to capture REER volatility. The GARCH procedures allow
current volatility to depend on its previous volatility. Using the GARCH process, we derive our
volatility as follows:

VOLt = 𝛿 + 𝜙VOLt−1 + 𝜀t (2)

where 𝜀t ≈ iid N(0, ht )

ht = 𝛽 + 𝜓𝜀2t−1 + 𝜒ht−1 (3)

Note that 𝛽 > 0; 𝜓 ≥ 0; 𝜒 ≥ 0.

From Equation 3, our conditional variance (ht ) is a function of the mean (𝛽) of the condi-
tional variance, information about the past volatility, which is the lag of the squared residual
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
882 |    ASAMOAH et al.

𝜀2t−1 (ARCH term), and the previous forecast error variance, 𝜒ht−1 (GARCH term). We first test
for stationarity to ensure that our REER is stationary to avoid the incidence of spurious re-
gression. Thus, we perform a unit root test by applying an augmented Dickey–­Fuller test
(Dickey & Fuller, 1979). Suppose the REER is not stationary in levels or integrated at order
zero. In that case, it must be differenced at the first level (integrated at order 1) or the second
level (integrated at order 2). Once it is deemed stationary, we estimate the mean-­variance
equation. The mean-­variance equation allows us to generate the mean-­variance series, which
captures the volatilities in our REER variable. A GARCH (1,1) shows the presence of a first-­
order moving average ARCH term and a first-­order autoregressive GARCH term. Based on
Equation 3, the ARCH term denotes 𝜓, while the GARCH term denotes 𝜒 . The ARCH term
captures current news on volatility, while the GARCH term captures the impact of previous
volatility on the current volatility.

We determine the evidence of volatility clustering and persistence by adding the sum of
the coefficients of the ARCH (𝜓) and GARCH (𝜒) terms. According to Enders (2015), volatility
shocks are deemed to be persistent when the sum of the coefficients is close to unity (𝜓 + 𝜒 ≈ 1).
That means the impact from volatility will linger over a long period. Significantly, we found the
sum of all ARCH and GARCH terms to be closer to unity. Again, we use the Ljung–­Box statistics
to confirm that our series does not suffer from autocorrelation up to a lagged value of 12. Also,
the ARCH LM statistics for heteroscedasticity test the null hypothesis that there is no ARCH
effect present in the residuals. The insignificance of the observation × R2 settles on the absence
of conditional heteroscedasticity. Thus, we derive our volatility variables with the correct speci-
fications and procedures.

2.2 | Financial development

The literature has primarily focused on either banking or stock market–­specific indicators as
proxies for FD while neglecting the impact of other equally essential indicators such as insur-
ance, pension funds, bonds, and mutual funds (Ito & Kawai, 2018). Equally important are the
additional roles of nonbank financial institutions such as venture capitals, microfinance institu-
tions, investment banks, credit unions, and savings and loan institutions. However, the concept
of FD is multidimensional and should not be confined to only traditional indicators. The cur-
rent proxies are also skewed mainly in terms of the quantity aspect of FD (size and depth) to
the neglect of the qualitative aspects of FD such as efficiency, liquidity, cost-­profit performance,
diversity, and the institutional environment, including legal systems (Hasan et al., 2009; Ito &
Kawai, 2018). We, therefore, employ a broad-­based index of FD that overcomes the limitations
of single indices while accounting for the complexities and multifaceted characteristics of FD.
Recently developed by the International Monetary Fund (IMF), the index considers financial
markets’ and institutions' development in terms of depth (liquidity and size of markets), effi-
ciency (low-­cost financial services amid sustained revenues, and capital market activities), and
access (i.e., the ability to access financial services) (Svirydzenka, 2016). The index is constructed
using data from various sources, including the IMF's financial access survey, the BIS debt securi-
ties database, the Dealogic corporate debt database, and the World Bank FinStats 2015. The index
has recently been deployed as the ultimate measure of FD (Hannan, 2017; Tchamyou et al., 2019).
For an in-­depth understanding of the impact of the development of the financial sector, we fur-
ther assess the conditional impact of FD from the two subindices (financial institutions and the
financial markets) making up the broad FD index. Therefore, Svirydzenka (2016) transcripts
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 883

that these two subindices focus on the development of financial markets and financial institu-
tions in terms of access, depth, and efficiency. While financial institutions focus on the standard
banking sector, financial markets are concerned with stock and debt market development. On
the relative importance of the new indicators, Svirydzenka (2016) is convinced that “the indices
are an improvement over the traditional measures of financial development. Conceptually, they
incorporate information on a broader range of financial development features for a wider array of
financial agents” (Svirydzenka, 2016, p. 20). The index has recently been employed in the capital
flows–­Africa literature (see Asamoah & Alagidede, 2021; Asamoah et al., 2021b).

2.3 | Controls

We employ relevant control variables as consistent with the capital flow literature. We measure
trade as the sum of imports and exports of goods and services scaled by GDP and financial open-
ness by the Chinn and Ito (2008) index of financial openness. We expect a positive impact of both
trade and financial openness as we control for both. We account for natural resources endow-
ment as the sum of natural gas, minerals, coal, forest, and oil rents, expressed as a percentage of
GDP. Though natural resource is one of the driving factors of FDI in the extractive sector, the
literature remains inconclusive on the directional effect. We measure human resources based
on the years of schooling and return to education. We anticipate a positive impact of human
capital on FDI inflows to Africa. We also control for the REER and expect a positive effect on
FDI inflows. All data except for human capital, financial openness, and FD were sourced from
the World Development Indicators of the World Bank. Data on FD are from the IMF. In con-
trast, human capital and financial openness are obtained from Penn World Tables 9.1 (Feenstra
et al., 2015) and the Chinn and Ito (2008) financial openness index, respectively.

2.4 | Regression model and data estimation procedure

Beginning with Equation 5, we first test the evidence of a linear association between exchange
rate volatility and private capital flows to examine whether exchange rate volatility deters FDI
inflows to Africa.

FDIit = 𝛼FDIit−1 + Σ𝛽 1 EXRVit + Σ𝛽 2 Xit + Ui + 𝜀t + 𝜆tit (4)

FDIit measures FDI for country i at time t. FDIit−1 is a lag of FDI testing for convergence and
reinforcing effects in a dynamic panel data setting. EXRVit indicates exchange rate volatility. Xit
denotes a set of controls in a standard capital flow model that includes trade openness, financial
openness, human capital, and natural resources. Ui , 𝜀t , and 𝜆tit signify country effects, a time-­
varying idiosyncratic shock with the standard iid assumption, and a model error term.

2.4.1 | Evidence of nonlinearity

We test for the existence of nonlinearities in Equation 4. The principal question is whether any
adverse effect of volatilities on FDI depends on the levels of volatilities or whether the relation-
ship decreases at the start of the interval and increases at the end of the interval, or vice versa. To
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
884 |    ASAMOAH et al.

test this hypothesis of nonlinearity, we follow Lind and Mehlum (2010) and test for evidence of
a U-­shaped or an inverted U-­shaped relationship, as Nier et al. (2014) recently employed in the
capital flows literature. We extend Equation 4 to include the quadratic term of the exchange rate
volatility. Equation 5 is specified as follows:

(5)
( )2
FDIit = 𝛼FDIit−1 + Σ𝛽 1 EXRVit + Σ𝛿 EXRVit + Σ𝛽 2 Xit + Ui + 𝜀t + 𝜆tit

where EXRV2 is a quadratic term for the domestic exchange rate volatility. The remaining variables
are as explained previously. Regarding the existence or otherwise of a nonlinear relationship, Lind
and Mehlum (2010) propose a set of three-­step procedures of regressing the variable to be explained
(FDI) on the explanatory variable of interest (exchange rate volatility).
The first is the sign and direction of the quadratic term. For the existence of a U-­shaped re-
lationship on a given interval of values, Lind and Mehlum (2010, p. 110) contend that “we need
to test whether the relationship is decreasing at low values within this interval and increasing at
high values within the interval.” Therefore, a U-­shaped relationship exists when the coefficient
of the linear term is negative and that of the quadratic term is positive and significant. In con-
trast, an inverted U-­shape exists when the coefficient of the linear term is positive and that of the
quadratic term is negative and significant. The quadratic term must be significant. The second
step notes that the slopes at the extreme ends of the data (minimum and maximum) must be
sufficiently steep. Lind and Mehlum (2010, p. 111) note that “the requirement for a U shape is
that the slope of the curve is negative at the start and positive at the end of a reasonably chosen
interval” of the explanatory variable of interest. Our study employs the observed data range as
our chosen interval, that is, EXRVmin and EXRVmax. Therefore, for a U-­shaped relationship, the
slope at the minimum data point must exhibit adverse and significant effects, while the slope
and the maximum data point should be positive and significant. The opposite holds under an
inverted U-­shaped relationship. Using the EXRV in the regression model (Equation 5), the joint
null hypothesis at the extreme ends of the data under an inverse U-­shaped relationship according
to Lind and Mehlum (2010) and further by Arcand et al. (2015) is stated in Equations 6 and 7:

(6)
( ) ( )
H0 : 𝛽 1 + 2δEXRVmin ≤ 0 U 𝛽 1 + 2δEXRVmax ≥ 0

(7)
( ) ( )
H1 : 𝛽 1 + 2δEXRVmin > 0 ∩ 𝛽 1 + 2δEXRVmax > 0

EXRVmin and EXRVmax are the minimum and maximum values of the exchange rate volatility,
respectively. The corresponding t-­statistics, which also corresponds to the rejection zone, can
also be estimated. Following Kuo et al. (2014), let 𝜃 1 represent the variance estimation of ̂𝛽 1 , 𝜃2
the estimated variance of 2̂
δ , and 𝜃 3 the covariance estimation of ̂𝛽 1 and 2̂
δ . The values ̂
𝛽 1 and
2̂
δ represent the corresponding estimated values of 𝛽 1 and 2δ. Therefore, the corresponding t-­
statistics at the minimum and maximum values of EXRV is stipulated in Equation 8:

̂
𝛽 1 + 2̂
δ(EXRV )
Ti = √ � � ; i = max or min point (8)
[𝜃 1 + 2𝜃 3 EXRVi + 𝜃 2 (EXRVi )2 ]

The procedure further suggests the estimation of the point of inflection of the quadratic term
at minimum and maximum values, which must lie within the range of the data set. To achieve
this, we take the partial derivative of Equation 5, which yields the threshold or turning point at
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 885

which the effect of volatilities on capital flows becomes nonmonotonic. The partial derivative is
presented in Equation 9:

d(FDI)
= 𝛽 1 + 2δEXRV (9)
d(EXRV )

At this point, any additional negative surges in volatilities will have no adverse impact on
FDIs. Such a relationship could be concave or convex. That point of inflection is achieved by
setting Equation 9 to zero and making EXRV the subject, as shown in Equation 10.

̂
𝛽1
𝛽 1 + 2δEXRV = 0; EXRV = (10)
̂
2δ

We test this condition at a 95% confidence interval of the turning point. There is evidence of
a nonlinearity once the confidence interval lies within the range of the data (Haans et al., 2016;
Kuo et al., 2014). Even though the relationship could be decreasing at the left and increasing at
the right within the minimum and maximum intervals of the data, the relationship may not be
U-­shaped. The third and final procedure is to perform the overall test of the presence of a U-­ or
inverse U-­shaped relationship. This is executed using the stata command (Utest). In our case,
the null hypothesis supports the presence of a U-­shaped relationship. The alternative hypothesis
tests the presence of an inverse U-­shaped relationship. Thus, a rejection of the null hypothesis
confirms the existence of an inverted U-­shaped relationship regarding the impact of exchange
rate volatility on FDI flows to Africa. Our hypothesis supports the existence of a significant non-
linear association between exchange rate volatility on FDI flows, which is U-­shaped.
Furthermore, Brambor et al. (2006) posit that we can also interpret Equation 9 by studying
the interpretation of the interaction models, where the effect of EXRV on the attraction of FDI is
dependent on varying levels of EXRV itself.

2.4.2 | FDI, macroeconomic volatility, and FD

Following Durham (2004), Alfaro et al. (2010), and Asamoah et al. (2021a), we assess the ability
of a host country's structural characteristics to deal with any form of volatilities associated with
FDI. We, therefore, determine the moderating effect of FD in the FDI-­exchange rate volatility
dynamics in two simple steps. The first is to include an interaction term of volatility and FD in
the linear equation in Equation 4. We thus estimate Equation 11:

FDIit = 𝛼FDIit−1 + Σ𝛽 1 EXRVit + Σ𝛽 3 Xit + 𝛽 4 FDit + 𝛽 5 (EXRV × FD)it + Ui + 𝜀t + 𝜆tit (11)

From Equation 11, while 𝛽 4 and 𝛽 1 examine the direct effect of FD and EXRV on FDI, our
variable of interest 𝛽 5 evaluates the effect of the volatility on FDI conditioned on varying levels of
FD. The assessment of the conditional varying points (marginal effects) indicates that the effect
of EXRV on FDI in the presence of FD is not static, as may be in the case of Equation 4, but the
impact of any change in FDI resulting from the volatility of the domestic exchange rate depends
on the different levels of FD.
Therefore, the marginal effect in our case is the partial derivative of Equation 11, where we
take the first derivative of FDI with respect to EXRV , which results in Equation 12.
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
886 |    ASAMOAH et al.

d(FDI)
= 𝛽 1 + 𝛽 5 FD (12)
d(EXRV )

Should we find both 𝛽 1 and 𝛽 5 to be positive, then fractional increases in volatilities will lead to
increases in FDI based on increasing FD values up to the point where FDI is optimized. However,
with the hindsight that volatilities deter capital inflows (Asamoah et al., 2016; Cavallari &
d'Addona, 2013) and that FD attracts FDI (Agbloyor et al., 2014; Asamoah & Alagidede, 2021;
Asamoah et al., 2021b), there is the probability of an adverse effect of EXRV and positive effect
of FD on FDI. In such a scenario, we seek to determine the percentile levels of FD necessary to
reduce any adverse impact of volatility on FDI and, if possible, to completely eradicate the nega-
tive impact. Brambor et al. (2006) further require that standard errors for the multiplicative term
be captured separately from the standard errors of the constitutive terms. Equation 13 shows the
standard errors for the interactive term:
( ) √
d (FDI) ( ) ( ) ( )
se = var ̂
𝛽 1 + FD2 var ̂
𝛽 5 + 2FDcov ̂𝛽1̂
𝛽5 (13)
d (EXRV )

From Equation 13, a negative covariance indicates the possibility of a significant marginal ef-
fect (𝛽 1 + 𝛽 5 FD) from FD, even if all other indicators are insignificant. Thus, the analysis of such
a multiplicative term equation needs further study. Given the continuous nature of the measures
of FD, we assess the marginal effect of EXRV on FDI at the different percentiles of FD (25th, 50th,
mean, 75th,, and 90th percentiles).

2.4.3 | Estimation procedure—­Systems GMM

We perform the aforementioned analysis using the system GMM estimation technique. We opt
for the two-­step GMM estimator because it is known to be asymptotically more effective and
vigorous for heteroskedasticity. We further use the Windmeijer (2005) finite-­sample correction
to obtain efficient standard errors and apply the small option to correct small sample bias to the
covariance matrix (Roodman, 2009). We use the forward orthogonal deviations to improve the
efficiency of our results due to the availability of gaps in our panel sample. To check the validity
of the system GMM estimator, we report the p-­values of two significance tests. The first is the
second-­order serial correlation [AR(2)], which tests whether the error terms are serially cor-
related. If there are serial correlations in the first order, it may not count, unlike in the second
order. The second is the Hansen J test, for overidentification restrictions on the validity of instru-
ments employed in the regression estimation. Hansen J tests the null hypothesis that overidenti-
fying restrictions in instruments do not correlate with the error term. The cogency of these tests
depends on the nature of the relationship between the sample size and the number of instru-
ment counts. Thus, to avoid instrument proliferation and model overfitness, we follow Roodman
(2009) and collapse our instrument matrix. The appropriate relationship is that the ratio (r) of
the sample size (n) to the number of instrument counts (i) should be higher than or equal to one
(i.e., r = n/I ≥ 1). In the GMM estimator, the first difference of the exogenous variables, the lags of
all endogenous variables, and lagged difference of the endogenous variables are used as standard
instruments in the dynamic panel estimation (Arellano & Bond, 1991). We treat all variables,
except for the indicator of volatility, the levels of the volatility variables, and FD as exogenous.
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 887

We use only internal instruments as we seek to maintain the assumption of fewer instruments
relative to the sample size (Roodman, 2009).

3 | R EG R E SSION R E SU LT S

We first determine the linear and nonlinearity relationship between FDI flows and REER volatil-
ity. We then determine the conditional effect of FD in the volatility–­FDI dynamics. We end the
discussion with our marginal effect analysis.

3.1 | Linear and nonlinear effects of exchange rate volatilities on

capital flows

From Table 1, the linear estimation in model 1 provides credence to the theoretical proposition
that uncertainty hurts investment. We found evidence at conventional significance levels that
EXRV tends to reduce FDI to Africa. The negative relationship between exchange rate volatility
and FDI is consistent with real options and investment irreversibility theories. These theories
suggest that the level of uncertainty allows investors to postpone current investments, causing a
decrease in current investment levels. Thus, economic agents are likely to invest less (Akkina &
Celibi, 2002; Dixit et al., 1994). From model 2, we show evidence of nonlinearity in the associa-
tion between uncertainty and investments. We found the coefficient of the linear exchange rate
volatility to be negative and the coefficient of the squared exchange rate volatility to be positive
and significant on FDI flows. The significant negative coefficients for the linear terms and the
subsequent significant positive coefficients of the quadratic terms indicate a nonlinear relation-
ship between volatility and FDI flows. The results indicate that the impact of volatilities on FDI
in Africa is dependent on varying levels of volatility. Model 2 shows the points of inflection at 95%
confidence interval. Reflecting on the maximum functions, the point of inflection for exchange
rate volatility is 23.408. There is evidence of a nonlinear relationship once the confidence inter-
val lies within the range of the data (Haans et al., 2016; Kuo et al., 2014; Lind & Mehlum, 2010).
The points of inflection imply that beyond the estimated values, increasing volatilities in the ex-
change rate will lead to increasing levels of FDI to Africa, conditioned on the current volatilities.
However, for evidence of a U-­shaped relationship, the inflection point must lie within the
confidence interval and with a negative slope at the minimum bounds and positive slope at the
maximum bounds, both significant at conventional levels of significance (Kuo et al., 2014; Lind
& Mehlum, 2010).
From model 2 of Table 1, the analysis exhibits a nonlinear and a U-­shaped relationship be-
tween the exchange rate volatility and FDI flows. This is because the linear exchange rate vola-
tility is significantly negative, while the squared exchange rate volatility is positively significant,
both at 5% levels. Furthermore, the U-­shaped relationship is supported by the negative slope
experience at the EXRVmin and positive slope at the EXRVmax and the point of inflection within
the data intervals. Finally, Table 1 presents the rejection of an inverted U-­shaped relationship
based on the overall test of the U-­shape. The nonlinear relationship exhibited by EXRV on FDI
suggests that the association is not strictly linear. Beyond the point of inflection, foreign investors
are mainly insensitive to exchange rate volatility. Investors could increase investments during
increasing macroeconomic volatilities, as benefits also tend to increase with the increasing vola-
tility of the domestic exchange rate. The results suggest that the investment irreversibility theory
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
888 |    ASAMOAH et al.

TABLE 1 FDI and exchange rate volatility—­Linear and nonlinear relationships

Linear model Quadratic model

(1) (2)

Dependent variables FDI FDI

Constant −3.895 (1.118)*** −4.465 (1.160)***
Lag dependent variable 0.087 (0.034)** 0.086 (0.030)***
Exchange rate 0.148 (0.069)** 0.196 (0.081)**
Exchange rate volatility −0.068 (0.01)** −0.117 (0.045)**
Exchange rate volatility2 0.003 (0.001)**
Financial openness 0.114 (0.055)** 0.093 (0.056)
Human capital 0.831 (0.391)** 0.76 (0.448)*
Natural resources 0.165 (0.041)*** 0.156 (0.048)***
Trade openness 0.769 (0.226)*** 0.881 (0.253)***
Diagnostics
Observations 740 740
Number of groups (n) 35 35
Number of instruments (i) 15 15
Instrument ratio (n/i) 2.33 2.33
AR(1): p-­value 0.003 0.003
AR(2): p-­value 0.573 0.558
Hansen J: p-­value 0.976 0.930
F 15.420 10.530
Probability > F 0.000 0.000
The slope at minimum bound −0.149 (0.056) ***
The slope at maximum bound 0.538 (0.004)
Inflection points 23.408
Overall test for the presence of an inverse 0.45(0.329)
U-­shape
[95% confidence interval] [−6.320, 26.593]
Notes: Values in parentheses represent Windmeijer (2005) robust standard corrected errors. AR(1) = test of first-­order
autocorrelation; AR(2) = test of second-­order autocorrelation; Hansen J = test of overidentifying restrictions. *, **, and ***
denote significance levels at 1%, 5%, and 10%, respectively. The slope represents the lower and upper bounds of EXRV.
Abbreviation: FDI, foreign direct investment.

dominates the initial relationship between exchange rate volatility and FDI flows, where the im-
pact is detrimental. This implies that until the point of inflection, increases in the volatility of the
domestic exchange rate harm the influx of FDI. However, the risk neutrality theory dominates
after the point of inflection, as increasing domestic exchange rate volatility leads to increases in
FDI flows.
Consistent with the requirements of interactive coefficients,3 we determine the marginal ef-
fect of EXRV on FDI at varying levels of volatility. We also use the marginal effect to validate our
assertion of a nonlinear relationship, where at lower levels of volatility, the effect of volatility
on capital flows is high, but the negative effect decreases at increasing levels of volatility. The
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 889

TABLE 2 Marginal effects of EXRV on FDI at varying levels of EXRV

25th 50th 75th 90th

percentile percentile Mean value percentile percentile
EXRV @ (0.088) (0.133) (0.142) (0.161) (0.265)
FDI −0.096*** −0.092*** −0.089*** −0.086*** −0.073***
(1.033) (1.046) (1.060) (1.070) (1.123)
Notes: Values in parentheses under FDI represent standard corrected errors. *, **, and *** denote significance levels at 1%, 5%,
and 10%, respectively.
Abbreviation: FDI, foreign direct investment.

marginal analysis will further indicate the critical points at which the negative impact of volatili-
ties on FDI flows will be positive. From Table 2, we observe that at lower levels of the EXRV (25th
percentile), the adverse impact of volatilities on FDI flows is high. However, at the upper levels
of volatility (90th percentile), the adverse impact of EXRV on FDI decreases. Thus, at a 1% signif-
icance level, the adverse impact of exchange rate uncertainty on the attraction of FDI is reduced
at increasing levels of volatility, from the 25th to the 90th percentiles of volatility. The results are
consistent with the conclusions of Nier et al. (2014) that the impact of volatility on capital flows
varies at increasing levels of volatility.

3.2 | FDIs, exchange rate volatility, and FD

We present the results of the association between EXRV and FDI conditioned on FD. We explore
three indices of FD. We first determine the direct effects of both FD and EXRV in the presence of
other control variables. Table 3 presents the results. Models 1–­3 study the unconditional associa-
tion between exchange rate volatility and FDI flows in the presence of the three indices of FD.
In contrast, models 4–­6 present the association between EXRV and FDI flows conditioned on the
levels of each FD index.
From models 1 to 3, we observe a significant direct negative relationship between exchange
rate volatility and FDI flows at conventional levels of significance, in the presence of all three
measures of FD. In particular, a 1% increase in the levels of the domestic EXRV leads to a 0.08%
decrease in FDI flows in the presence of the overall FD index, 0.09% in the presence of financial
institutions, and 0.37% in the presence of financial markets. The significant adverse relationship
supports the investment irreversibility theory (Akkina & Celibi, 2002; Dixit et al., 1994) and the
risk-­averse nature of many investors (Bénassy-­Quéré et al., 2001). These theories predict that
where macroeconomic uncertainty abounds, in this case, as captured by the domestic exchange
rate, there will be a decrease in volumes of investments (FDI). Therefore, the volatility of the
domestic exchange rate increases the risk borne by foreign investors as it leads to a decrease in
projected returns on investments. The instability of the local currency affects the investment de-
cision of MNEs by restraining further investments due to the rate of the unpredictability of the
exchange rate between the local and foreign currencies. The volatility increases the anticipated
cost of production and decreases the value of assets of MNEs.
The impact of volatility is much stronger in cases of profit repatriation by MNEs as much of
the volatile domestic currency may be needed for a few of the foreign currencies. The overall
effect is that the volatility of the domestic exchange rate automatically triggers “the option to
wait” on future investments leading to the observed negative relationship between FDI and the
domestic EXRV. Our results support earlier studies that have found domestic EXRV to discourage
 TABLE 3 Foreign direct investments, exchange rate volatility, and financial development
| 890

Dependent variable FDI (1) FDI (2) FDI (3) FDI (4) FDI (5) FDI (6)
  

Constant −3.368 (2.325) −1.498 (0.854) −1.974 (1.481) −0.584 (1.839) −0.858 (1.700) −1.640 (2.715)
Lag FDI 0.091*** (0.042) 0.103** (0.046) 0.097** (0.037) 0.097** (0.036) 0.104*** (0.033) 0.070 (0.045)
Exchange rate volatility (ERV) −0.083** (0.037) −0.088** (0.043) −0.369*** (0.114) −0.107** (0.048) −0.120** (0.048) −0.277** (0.128)
Financial development index (FD) 1.392** (0.596) 1.256** (0.514)
Financial institutions index (FIN) 1.031** (0.505) 1.125** (0.509)
Financial market index (FMK) 2.744** (1.105) −0.390* (0.218)
FD × ERV 0.129** (0.063)
FIN × ERV 0.145** (0.071)
FMK × ERV 0.249**
Exchange rate 0.066 (0.105) 0.150 (0.087) 0.001 (0.062) 0.043 (0.062) 0.010 (0.065) 0.109 (0.075)
Financial openness 0.006 (0.133) 0.020 (0.160) 0.169 (0.158) 0.072 (0.086) 0.081 (0.079) 0.083 (0.112)
Human resources −0.324 (0.824) −0.015 (0.715) −0.133 (0.629) −0.396 (0.590) −0.284 (0.474) −0.020 (0.964)
Natural resources 0.242*** (0.075) 0.202*** (0.062) 0.195*** (0.054) −0.220*** (0.055) 0.198*** (0.052) 0.205*** (0.067)
Trade 0.758* (0.420) 0.770* (0.380) 0.785** (0.362) 0.829** (0.365) 0.750** (0.325) 0.910** (0.391)
Diagnostics
Observations 751 755 779 755 755 722
Number of groups (n) 35 35 35 35 35 33
Number of instruments (i) 28 28 31 34 27 21
Instrument ratio (n/i) 1.25 1.25 1.13 1.03 1.30 1.57
Critical points 0.829 0.828 1.12
AR(1): p-­value 0.008 0.009 0.0.02 0.007 0.006 0.006
AR(2): p-­value 0.219 0.201 0.239 0.255 0.220 0.589
Hansen J: p-­value 0.187 0.092 0.382 0.475 0.301 0.550
F 6.010 7.710 19.580 5.790 10.770 10.170
Probability > F 0.000 0.000 0.000 0.000 0.000 0.000
Notes: Values in parentheses represent Windmeijer (2005) robust standard corrected errors. FD × ERV is the interaction between the FD index and ERV. FIN × ERV is the interaction between
financial institutions index and ERV. FMK × ERV is the interaction between the financial market index and ERV.
ASAMOAH et al.

Abbreviation: FDI, foreign direct investment.

14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 891

FDI inflows (Asamoah et al., 2016; Balaban et al., 2019; Cavallari & d'Addona, 2013; Dal Bianco
& Loan, 2017; Jehan & Hamid, 2017). However, not all studies are in tandem with our results
as it contradicts studies that have found no significant impact of exchange rate volatility on FDI
inflows (Abbot et al., 2012; Chowdhury & Wheeler, 2015; Dhakal et al., 2010). It also differs from
those studies that have found a positive impact of EXRV on FDI inflows (Bénassy-­Quéré et al.,
2001; Urata & Kawai, 2000).
Regarding the direct linkages between FDI and FD, our expectation was in line with Alfaro
et al. (2004), that countries with developed financial markets benefit both directly and indirectly
from the growth enhancement of FDI. From models 1 to 3, we observe a significant positive
association between FDI and all FD indices. The significance of the positive relationship lies in
the fact we employ a new indicator of FD not previously employed in the volatility–­FDI nexus,
especially in the context of Africa.
The positive correlation shows that foreign investors are willing to partake in the domestic
market through borrowing from banks, undertaking insurance agreements, pensions, and gen-
erating capital from the domestic market. Developed financial markets provide liquidity to both
domestic and foreign investors. This also helps foreign investors to borrow in the local currency
and avoid issues of exchange rate uncertainty, where borrowed funds are denominated in foreign
currency. Investors become confident that a developed financial market absorbs moral hazard
issues, encourages savings and resource mobilization, and reduces borrowing costs. The over-
whelming conclusion is that FD in terms of bank and stock market access, efficiency, and depth
is a crucial decision foreign investors consider in their decision to invest overseas. Our results
support the conclusion of Agbloyor et al. (2014) that advanced banking systems can attract more
FDI inflows, while stock market development may also attract FDI inflows. Similarly, Jehan and
Hamid (2017) found that the bank and private sector credit attract FDI flows. Though Soumaré
and Tchana Tchana (2015) found bidirectional causality between FDI and financial market de-
velopment, they contend the impact differs based on whether the analysis is focused on bank or
stock market development. Further support is the studies of Ezeoha and Cattaneo (2012) on FDI
flows to sub-­Saharan Africa.
We turn to the indirect association between EXRV and FDI conditioned on the domestic fi-
nancial sector. We interact each of the three FD indicators with exchange rate volatility and as-
sess their combined impact on FDI flows. Based on our estimation results, we expected that
conditioned on FD, the indirect impact of EXRV on FDI will be positive. Models 4–­6 illustrate the
results of our indirect analysis. Based on just the coefficient of the interactive terms, we observe
a positive and significant relationship with FDI in all models. We can infer from the results that
FD is vital in curtailing the potential adverse impact of EXRV on FDI flows. The results further
suggest that curbing the exchange rate volatility damming impact on FDI is an increasing func-
tion of FD. Given the negative coefficient of volatilities and the subsequent positive coefficient
of the interactive terms, we can confidently say that improving the level of FD will significantly
reduce the adverse impact of exchange rate volatility on FDI flows. Therefore, one can allude that
dealing with the impact of volatility on FDI can be attained at increasing levels of FD. Another
plausible explanation is that as FD improves, there will be a gradual decline in the adverse im-
pact of exchange rate volatility on the attraction and retention of FDI flows. It also holds that as
the quality of FD declines or in the absence of FD, the increasing volatility of the exchange rate
leads to a decrease in FDI volumes. We can allude from the coefficient of the interaction term
that FD can mitigate the negative impact of exchange rate volatility on FDI flows. Practically, for-
eign investors can still invest in an environment of exchange rate uncertainty once the domestic
financial market is developed. Therefore, it can be said that though instability of the domestic
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
892 |    ASAMOAH et al.

exchange rate may abound, with compliments from the financial sector, countries can still in-
crease the volumes of FDI flows.

3.2.1 | Marginal effect analysis of the effect of exchange rate volatility on

FDI at levels of FD

As Brambor et al. (2006) indicated on interaction models, it is impossible to determine if the

independent variable impacts the dependent variable by merely focusing on the significance or
otherwise of the coefficient of the interactive term. However, they contend that the effect of a
change in the independent variable on the dependent variable relies on varying values of the
conditional variable. Therefore, we determine the marginal effect of exchange rate uncertainty
on FDI at diverse values of all three FD indices. In this case, the marginal effect will indicate the
threshold value of FD that can neutralize the negative impact of exchange rate volatility on FDI.
We conduct the marginal analysis at the 25th, 50th, mean, 75th,, and 90th percentiles of FD.
We estimate the marginal analysis by applying Equation 12, where we evaluate the impact of
exchange rate volatility on FDI flows. From Table 4, model A shows that the adverse impact on
exchange volatility decreases as the overall FD index increases at the 1% significance level. As the
level of FD increases by 1% from the 25th to the 90th percentiles, the negative effect of exchange
rate volatility on FDI decreases from −0.096% to −0.073%. However, our observation shows the
impact of the exchange rate on FDI remains negative even at the highest (90th) percentile of
our FD index. We must, therefore, determine the critical point or threshold level of the FD that
completely eradicates the adverse impact. To achieve that optimal point, we set Equation 12 to
zero and make FD the subject.
Regarding the FD index, the critical point is ~0.829.4 We can make two critical observations
at this point. Given that the overall FD index ranges between 0 and 1, it suggests that there are
periods in the development of the financial sector where the adverse impact of exchange rate
volatility on FDI can be wholly eradicated. However, using the minimum (0) and maximum

T A B L E 4 Marginal effects of exchange rate volatility on foreign direct investment at varying levels of
financial development indicators

Model FD index Model FIN Model FMK

@ A FDI index @ B FDI index @ C FDI
25th percentile −0.096*** 25th percentile −0.097*** 25th percentile −0.277***
(0.088) (1.033) (0.161) (0.822) (0.002) (0.981)
50th percentile −0.092*** 50th percentile −0.091** 50th percentile −0.274
(0.133) (1.046) (0.203) (0.844) (0.011) (0.983)
Mean value −0.089*** Mean value −0.086*** Mean value −0.265**
(0.142) (1.060) (0.234) (0.859) (0.048) (0.128)
75th percentile −0.086*** 75th percentile −0.080*** 75th percentile −0.266***
(0.161) (1.070) (0.274) (0.879) (0.041) (0.989)
90th percentile −0.073*** 90th percentile −0.060*** 90th percentile −0.236**
(0.265) (1.123) (0.411) (0.948) (0.166) (0.989)
Notes: FD is the overall financial development index; FIN is the financial institutions index; FMK is the financial market index;
FDI is foreign direct investment. Values in parentheses under FDI represent standard corrected errors. *, **, and *** denote
significance levels at 1%, 5%, and 10%, respectively.
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 893

(0.627) values of our data,5 the apparent analogue is that given the current state of Africa's FD,
we can only infer that financial sector development minimizes the exchange rate volatility–­FDI
negative relationship but cannot eradicate the adverse impact. The same observation holds for
models B and C of Table 4, where a substantial increase in the financial institutions and market
indices leads to a decrease in the adverse impact of exchange rate volatility on FDI, at 1% and 5%
significance levels, respectively. A 1% development in financial institutions from the 25th to the
90th percentile leads to a decrease in the adverse impact of EXRV on FDI from −0.097 to −0.060.
However, the effect cannot be defused outrightly. The critical point for the financial institutions'
index is ~0.828, which also lies within the ranges of the financial institutions' index but outside
the ranges (0 and 0.738) of our data. Again, financial institutions’ development regarding access,
depth, and efficiency can reduce the exchange rate volatility–­FDI antagonistic relationship only
at the current state on institutions’ development but not up to the point of complete eradication.
Regarding financial markets, we see marginal decreases in the adverse effect of exchange
rate volatility on FDI flows from the 25th percentile (−0.277) to the 90th percentile (−0.236).
The critical or tuning point regarding financial markets is 1.12, which falls outside the index
values of 0 and 1. The critical value shows that it is not plausible for financial market develop-
ment in Africa to eradicate the negative effect of exchange rate volatility on FDI at its current
state, even though financial market development can marginally reduce the adverse effect. The
results on financial markets only confirm the assertions by Svirydzenka (2016) that financial
market development in Africa is fragile when compared to other regions in the world. For in-
stance, while the average financial market development for America, much of Europe, Asia,
and Australia was above 0.596, and much of South America ranged between 0.392 and 0.592,
the average financial market development for Africa was less than 0.046 (Svirydzenka, 2016).
The observation is that financial institutions' development can better neutralize the adverse
impact of volatility faster than financial markets and the overall development index. Thus, the
slow growth of FD in Africa is a function of the weak, inactive, ill-­liquid, and inefficiency of
most stock markets on the continent.

4 | CO N C LUSION S AN D POLICY RECOMMENDATIONS

The study aimed to determine the linear and nonlinear impact of exchange rate uncertainty
on FDI in Africa and whether FD moderates such association. From the linear system GMM
estimation, we found evidence that countries with unstable exchange rates are often associated
with low FDI inflows. The result from the quadratic analysis shows evidence of a nonlinear rela-
tionship between exchange rate volatility and FDI and that there exists a U-­shaped relationship
between the two. The nonlinear relationship confirms that both theories of investment irrevers-
ibility and reversibility have dominant influences on FDI flows. While the former dominates at
lower levels of volatility, the latter dominates at higher levels of uncertainty. This implies that
although countries may be affected by decreases in FDI flows to initial volatility beyond the point
of inflection, increases in uncertainty boost investments due to higher anticipated returns. The
results further show that previously mixed findings on the exchange rate volatility–­FDI nexus
could result from the inadvertent failure to determine a nonlinear association. We further ascer-
tain that in as much as exchange rate volatility significantly decreases FDI flows, countries with
strong FD can attract high inflows of FDI as the adverse impact of volatility on FDI will be nulli-
fied in the robust financial sector. From the marginal effect analysis, we show that increasing the
levels of FD, be it the overall sector, financial institutions, or financial markets, will mitigate the
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
894 |    ASAMOAH et al.

adverse effect of exchange rate uncertainty on FDI flows, with the most significant impact from
financial institutions’ development.
The findings of the study provide several policy directions. First is the call for methods to stabilize
the exchange rate of most economies in Africa. Issues of exchange rate misalignment are a funda-
mental problem because of the underlying economic regimes of many African countries, such as im-
port dependence and the dollarization of many economies. Efficient management of the economy is
critical. It will improve the general economic fundamentals, reduce economic agents' willingness to
hedge their wealth in foreign currency, lead to a halt in the scramble for foreign currency, and curb
the volatility associated with the domestic exchange rate. Clear-­cut policies regarding the use and
demand of a foreign currency in the domestic retail environment should be firm in dwarfing the
appetite for foreign currency in the domestic space. Growth in manufacturing and industries will
reduce the dependence on foreign goods; reduce imports and the scramble for foreign currency; and,
ultimately, reduce fluctuations in the exchange rate. Governments should focus on the growth of
local industries or encourage foreign entities to establish production plants within. Policies aimed at
the stabilization of the exchange rate should be in sync with the overall development of the financial
sector. Strengthening financial institutions is a quicker method of curbing the volatilities associated
with FDI inflows than the financial market channel. The development of the banking sector can
drive the overall development of the financial sector in eradicating volatilities associated with capital
flows. Further studies could focus on other forms of capital flows such as portfolio equity and private
and public debt flows, as well as other macroeconomic variables such as growth, inflation, and inter-
est rate uncertainty. Regarding nonlinearity, further studies can also consider the application of the
sample splitting method, as well as dynamic thresholds that deal with endogeneity.

ACKNOWLEDGMENTS
The authors are grateful to the coeditor (Prof. Andy McKay) and two anonymous reviewers for their
helpful comments. The authors thank Tracy Dede Asiedu-­Dicker for proofreading the article.

CONFLICT OF I NTEREST
There are no issues of conflicts of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available in the supplementary material of
this article.

ORCID
Michael Effah Asamoah https://orcid.org/0000-0001-6702-7804
Frank Adu https://orcid.org/0000-0003-0072-882X

ENDNOTES
1
These countries are Algeria, Angola, Benin, Botswana, Burkina Faso, Burundi, Cape Verde, Cameroun CAR,
Chad, Congo DR., Congo Republic, Côte d’Ivoire, Egypt, Equatorial Guinea, Ethiopia, Gabon, Gambia, Ghana,
Guinea-­Bissau, Kenya, Lesotho, Madagascar, Malawi, Mali, Mauritania, Mauritius, Morocco, Niger, Nigeria,
Rwanda, Senegal, Seychelles, South Africa, Sudan, Tanzania, Togo, Tunisia, Uganda, Zambia.
2
According to Chinn and Ito (2008), the index ranges between −2.6 (very closed financially) and 2.6 (very open
financially).
3
Brambor et al. (2006) and Nier et al. (2014).
4 𝛽̂1
𝛽 1 + 𝛽 5 FD = 0; FD = = 0.107/0.129 = 0.829.
𝛽̂5
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 895

5
Per the summary statistics (see Table A1), the minimum and maximum values are FD index [0, 0.627], financial
institutions [0, 0.738], and financial markets [0, 0.586].

REFERENCES
Abbott, A., Cushman, D. O., & De Vita, G. (2012). Exchange rate regimes and foreign direct invest-
ment flows to developing countries. Review of International Economics, 20(1), 95–­107. https://doi.
org/10.1111/j.1467-­9396.2011.01010.x
Abel, A. B. (1983). Optimal investment under uncertainty. The American Economic Review, 73(1), 228.
Agbloyor, E. K., Abor, J. Y., Adjasi, C. K. D., & Yawson, A. (2014). Private capital flows and economic growth in
Africa: The role of domestic financial markets. Journal of International Financial Markets, Institutions, and
Money, 30, 137–­152. https://doi.org/10.1016/j.intfin.2014.02.003
Akkina, K. R., & Celebi, M. A. (2002). The determinants of private fixed investment and the relationship between
public and private capital accumulation in Turkey. The Pakistan Development Review, 243–­254.
Alfaro, L., Chanda, A., Kalemli-­Ozcan, S., & Sayek, S.(2004). FDI and Economic Growth: the role of local financial
markets. Journal of International Economics,, 64(1), 89–­112.
Alfaro, L., Chanda, A., Kalemli-­Ozcan, S., & Sayek, S. (2010). Does foreign direct investment promote growth?
Exploring the role of financial markets on linkages. Journal of Development Economics, 91(2), 242–­256.
https://doi.org/10.1016/j.jdeve​co.2009.09.004
Arcand, J. L., Berkes, E., & Panizza, U. (2015). Too much finance? Journal of Economic Growth, 20(2), 105–­148.
https://doi.org/10.1007/s1088​7-­015-­9115-­2
Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an
application to employment equations. The Review of Economic Studies, 58(2), 277–­297. https://doi.
org/10.2307/2297968
Asamoah, M. E., Adjasi, C. K., & Alhassan, A. L. (2016). Macroeconomic uncertainty, foreign direct investment,
and institutional quality: Evidence from sub-­Saharan Africa. Economic Systems, 40(4), 612–­621. https://doi.
org/10.1016/j.ecosys.2016.02.010
Asamoah, M. E., & Alagidede, I. P. (2021). Foreign direct investment, real sector growth and financial develop-
ment. International Journal of Finance & Economics, 1–­18. https://doi.org/10.1002/ijfe.2440
Asamoah, M. E., Alagidede, I. P., & Adu, F. (2021a). Private capital flows, real sector growth and institutional qual-
ity in Africa. Journal of International Development, 33(1), 171–­207. https://doi.org/10.1002/jid.3518
Asamoah, M. E., Alagidede, I. P., & Adu, F. (2021b). Financial development, portfolio investments and the real
economy in Africa. Economic Systems, 45(4), 1–­18. https://doi.org/10.1016/j.ecosys.2021.100872
Bah, I., & Amusa, H. A. (2003). Real exchange rate volatility and foreign trade: Evidence from South Africa's ex-
ports to the United States. African Finance Journal, 5(2), 1–­20.
Balaban, S., Živkov, D., & Milenković, I. (2019). Impact of an unexplained component of real exchange rate volatil-
ity on FDI: Evidence from transition countries. Economic Systems, 43(3-­4), 100719. https://doi.org/10.1016/j.
ecosys.2019.100719
Bénassy-­Quéré, A., Fontagné, L., & Lahrèche-­Révil, A. (2001). Exchange-­rate strategies in the competition for
attracting foreign direct investment. Journal of the Japanese and International Economies, 15(2), 178–­198.
https://doi.org/10.1006/jjie.2001.0472
Bollerslev, T. (1986). Generalized autoregressive conditional heteroskedasticity. Journal of Econometrics, 31(3),
307–­327.
Brambor, T., Clark, W. R., & Golder, M. (2006). Understanding interaction models: Improving empirical analyses.
Political Analysis, 14(1), 63–­82. https://doi.org/10.1093/pan/mpi014
Caporale, G. M., Ali, F. M., & Spagnolo, N. (2015). Exchange rate uncertainty and international portfolio flows: A
multivariate GARCH-­in-­mean approach. Journal of International Money and Finance, 54, 70–­92. https://doi.
org/10.1016/j.jimon​fin.2015.02.020
Cavallari, L., & d'Addona, S. (2013). Nominal and real volatility as determinants of FDI. Applied Economics, 45(18),
2603–­2610. https://doi.org/10.1080/00036​846.2012.674206
Chinn, M. D., & Ito, H. (2008). A new measure of financial openness. Journal of Comparative Policy Analysis,
10(3), 309–­322. https://doi.org/10.1080/13876​98080​2231123
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
896 |    ASAMOAH et al.

Chowdhury, A. R., & Wheeler, M. (2015). The impact of output and exchange rate volatility on fixed private
investment: Evidence from selected G7 countries. Applied Economics, 47(25), 2628–­2641. https://doi.
org/10.1080/00036​846.2015.1008761
Corden, W. M. (1990). Strategic trade policy: How new? How sensible? (Vol. 396). World Bank Publications.
Dal Bianco, S., & Loan, N. (2017). FDI inflows, price, and exchange rate volatility: New empirical evidence
from Latin America. International Journal of Financial Studies, 5(1), 1–­17. https://doi.org/10.3390/ijfs5​
010006
Daly, K., & Vo, X. V. (2013). The determinants of home bias puzzle in equity portfolio investment in Australia.
International Review of Financial Analysis, 27, 34–­42. https://doi.org/10.1016/j.irfa.2012.05.005
Demir, F. (2009). Macroeconomic uncertainty and private investment in Argentina, Mexico, and Turkey. Applied
Economics Letters, 16(6), 567–­571. https://doi.org/10.1080/13504​85070​1206528
Dhakal, D., Nag, R., Pradhan, G., & Upadhyaya, K. P. (2010). Exchange rate volatility and foreign direct invest-
ment: Evidence from East Asian countries. International Business & Economics Research Journal (IBER), 9(7),
121–­128. https://doi.org/10.19030/​iber.v9i7.603
Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root.
Journal of the American Statistical Association, 74(366a), 427–­431.
Dixit, A. K., Dixit, R. K., & Pindyck, R. S. (1994). Investment under uncertainty. Princeton University Press.
Durham, J. B. (2004). Absorptive capacity and the effects of foreign direct investment and foreign equity portfolio
investment on economic growth. European Economic Review, 48, 285–­306.
Dutta, N., & Roy, S. (2011). Foreign direct investment, financial development, and political risks. The Journal of
Developing Areas, 44(2), 303–­327. https://doi.org/10.1353/jda.0.0106
Elbadawi, I. (1992). Terms of trade, commercial policy and the black market for foreign exchange: an empirical
model of real exchange rate determination. Journal of African Finance and Economic Development, 1(2),
1–­26.
Enders, W. (2015). Applied econometric time series (4th. ed.). New York, United States: Wiley.
Ezeoha, A. E., & Cattaneo, N. (2012). FDI flows to sub-­Saharan Africa: The impact of finance, institutions, and
natural resource endowment. Comparative Economic Studies, 54(3), 597–­632. https://doi.org/10.1057/
ces.2012.18
Feenstra, R. C., Inklaar, R., & Timmer, M. P. (2015). The next generation of the Penn World Table. American
Economic Review, 105(10), 3150–­3182. https://doi.org/10.1257/aer.20130954
Forbes, K. J., & Warnock, F. E. (2012). Capital flow waves: Surges, stops, flight, and retrenchment. Journal of
International Economics, 88(2), 235–­251. https://doi.org/10.1016/j.jinte​co.2012.03.006
Gourinchas, P. O., & Jeanne, O. (2013). Capital flows to developing countries: The allocation puzzle. Review of
Economic Studies, 80(4), 1484–­1515.
Guichard, S. (2017). Findings of the recent literature on international capital flows: Implications and suggestions for
further research (OECD Economics Department Working Papers, 1410).
Haans, R. F., Pieters, C., & He, Z. L. (2016). Thinking about U: Theorizing and testing U-­ and inverted U-­shaped
relationships in strategy research. Strategic Management Journal, 37(7), 1177–­1195. https://doi.org/10.1002/
smj.2399
Hannan, S. A. (2017). The drivers of capital flows in emerging markets post-­global financial crisis. Journal of
International Commerce, Economics and Policy, 8(02), 1–­26.
Hartman, R. (1972). The effects of price and cost uncertainty on investment. Journal of Economic Theory, 5(2),
258–­266. https://doi.org/10.1016/0022-­0531(72)90105​-­6
Hasan, I., Koetter, M., & Wedow, M. (2009). Regional growth and finance in Europe: Is there a quality ef-
fect of bank efficiency? Journal of Banking & Finance, 33(8), 1446–­1453. https://doi.org/10.1016/j.jbank​
fin.2009.02.018
Ito, H., & Kawai, M. (2018). Quantity and quality measures of financial development: Implications for macro-
economic performance. Policy Research Institute, Ministry of Finance, Japan, Public Policy Review, 14(4),
1–­32.
Jehan, Z., & Hamid, A. (2017). Exchange rate volatility and capital inflows: The role of financial development.
Portuguese Economic Journal, 16(3), 189–­203.
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ASAMOAH et al.    | 897

Kodongo, O., & Ojah, K. (2012). The dynamic relation between foreign exchange rates and international port-
folio flows: Evidence from Africa's capital markets. International Review of Economics & Finance, 24,
71–­87.
Kuo, H. C., Lin, D., Lien, D., Wang, L. H., & Yeh, L. J. (2014). Is there an inverse U-­shaped relationship between
pay and performance? The North American Journal of Economics and Finance, 28, 347–­357. https://doi.
org/10.1016/j.najef.2014.03.007
Kyereboah-­Coleman, A., & Agyire-­Tettey, K. F. (2008). Effect of exchange-­rate volatility on foreign direct in-
vestment in sub-­Saharan Africa. The case of Ghana. The Journal of Risk Finance, 9(1), 52–­70. https://doi.
org/10.1108/15265​94081​0842410
Lane, P. R. (2015). International financial flows and the Eurozone crisis. The Eurozone Crisis, 129, 129–­133.
Lensink, R., & Morrissey, O. (2006). Foreign direct investment: Flows, volatility, and the impact on growth. Review
of International Economics, 14(3), 478–­493. https://doi.org/10.1111/j.1467-­9396.2006.00632.x
Lind, J. T., & Mehlum, H. (2010). With or without U? The appropriate test for a U-­shaped relationship. Oxford
Bulletin of Economics and Statistics, 72(1), 109–­118. https://doi.org/10.1111/j.1468-­0084.2009.00569.x
McKenzie, M. D. (1999). The impact of exchange rate volatility on international trade flows. Journal of Economic
Surveys, 13(1), 71–­106.
Nickell, S. J. (1978). The investment decisions of firms. Cambridge University Press Archive.
Nier, E., Sedik, T. S., & Mondino, T. (2014). Gross private capital flows to emerging markets: Can the global financial
cycle be tamed? (No. 14-­196). International Monetary Fund.
Pindyck, R. S. (1993). A note on competitive investment under uncertainty. The American Economic Review, 83(1),
273–­277.
Roodman, D. (2009). How to do xtabond2: An introduction to difference and system GMM in Stata. The Stata
Journal, 9(1), 86–­136. https://doi.org/10.1177/15368​67X09​00900106
Soumaré, I., & Tchana Tchana, F. (2015). Causality between FDI and financial market development: Evidence
from emerging markets. The World Bank Economic Review, 29(Suppl. 1), S205–­S216. https://doi.org/10.1093/
wber/lhv015
Svirydzenka, K. (2016). Introducing a new broad-­based index of financial development (International Monetary
Fund Working Paper, WP/16/5). https://doi.org/10.5089/97815​13583​709.001
Tchamyou, V. S., Erreygers, G., & Cassimon, D. (2019). Inequality, ICT, and financial access in Africa. Technological
Forecasting and Social Change, 139, 169–­184. https://doi.org/10.1016/j.techf​ore.2018.11.004
Udoh, E., & Egwaikhide, F. O. (2008). Exchange rate volatility, inflation uncertainty, and foreign direct investment
in Nigeria. Botswana Journal of Economics, 5(7), 14–­31.
UNDP (2011). TOWARDS HUMAN RESILIENCE: Sustaining MDG progress in an age of economic uncertainty.
United Nations Development Programme.
Urata, S., & Kawai, H. (2000). The determinants of the location of foreign direct investment by Japanese small and
medium-­sized enterprises. Small Business Economics, 15(2), 79–­103.
Windmeijer, F. (2005). A finite sample correction for the variance of linear efficient two-­step GMM estimators.
Journal of Econometrics, 126(1), 25–­51. https://doi.org/10.1016/j.jecon​om.2004.02.005

How to cite this article: Asamoah, M. E., Alagidede, I. P., & Adu, F. (2022). Exchange rate
uncertainty and foreign direct investment in Africa: Does financial development matter?
Review of Development Economics, 26, 878–­898. https://doi.org/10.1111/rode.12858
 14679361, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rode.12858 by State Bank of Pakistan, Wiley Online Library on [23/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
898 |    ASAMOAH et al.

APPENDIX

TABLE A1 Summary statistics

Standard Coefficient
Variable Mean Median deviation of variance Minimum Maximum N
Foreign direct 3.605 1.996 8.052 2.234 −8.589 161.824 1,119
investment
Real exchange rate 5.324 6.153 2.255 0.424 0.432 14.610 1,143
Exchange rate 4.130 4.268 1.523 0.369 −9.605 8.777 1,076
volatility
Financial 0.142 0.113 0.095 0.669 0 0.627 1,120
development
index
Financial institutions 0.234 0.203 0.121 0.517 0 0.738 1,120
index
Financial market 0.048 0.011 0.090 1.85 0 0.586 1,120
index
Financial openness −0.648 −1.210 1.177 −1.816 −1.917 2.347 1,111
Human capital 1.693 1.614 0.410 0.242 1.030 2.885 980
Natural resources 12.531 8.750 12.136 0.968 0.001 84.240 1,116
Trade openness 67.801 59.643 32.338 0.477 11.087 225.023 1,073