Optimum currency area : evidence in the West African Region?

(1)

Specialization: International Economics and Globalization 29th August 2014

Optimum Currency Area:

Evidence in the West African Region?

Chukwudumebi Egbuna (10651985)

dumebiegbuna@gmail.com

Supervisor:

Dr. Boe Thio

Word Count: 9175

This research uses the OCA index methodology to empirically determine the relationship between bilateral exchange rate variability and different criteria of the Optimum Currency Area (OCA) theory. This method, as implemented by Bayoumi and Eichengreen (1997), helps to determine the suitability of adopting a single currency amongst the fifteen ECOWAS countries. The OCA index is found to be very high amongst the 15 ECOWAS countries, most especially between the big countries like Nigeria, Ghana and Cote d’Ivoire, and the small countries like Guinea-Bissau and Liberia. This suggests that ECOWAS does not fulfill the necessary conditions for creating a monetary union and therefore, introducing a common currency would result in a low benefit to cost ratio for the region as a whole.

Keywords: Optimum Currency Area, Monetary Union, ECOWAS, OCA Index, exchange rate.

(2)

List of Abbreviations:

CFA – Communauté Financière Africaine (African Financial Community) ECOWAS – Economic Community of West African States

EMU – European Economic and Monetary Union GDP – Gross Domestic Product

GPPP – Generalized Purchasing Power Parity OCA – Optimum currency area

WAEMU – West African Economic and Monetary Union WAMI – West African Monetary Institute

WAMZ – West African Monetary Zone TSLS - Two-stage least squares regression

(4)

1. INTRODUCTION

Graph 1 shows the Gross Domestic Product (GDP) per capita growth in Sub-Saharan Africa. It is evident from this figure that although there was a significant increase in GDP per capita between 1960 and 2011, there have been fluctuations along the way. These fluctuations could be attributed to the financial crisis, but more so to the exchange rate instability, financial fragility and high inflation that often characterize the economic performance of Sub-Saharan Africa as acclaimed by many. The eradication of such problems could be a fundamental stepping-stone towards a better GDP per capita and therefore, economic growth. As Martin Feldstein (2008) states, joining an optimal currency area can and will lead to inflation and exchange rate discipline, therefore introducing a common currency into the region is perhaps one solution to the current underperformance.

Graph 1: GDP per capita growth (annual %). Source: World Bank Data

The Western region of Sub-Saharan Africa seems to acknowledge this line of reasoning, as the economic community created to promote economic integration in West Africa known as Economic Community of West African States (ECOWAS), has plans to introduce a common currency. This initiative comes years after the implementation of a common currency called the African Financial Community

-6 -4 -2 0 2 4 6 8 1960 1970 1980 1990 2000 2010

(5)

(CFA) franc1, which is used by eight2 of the fifteen3 countries that make up ECOWAS. These eight countries form a monetary union within ECOWAS called West African Economic and Monetary Union (WAEMU) and the success of its CFA Franc is the very reason why a single currency ECOWAS wide seems viable and beneficial. Alaba (2006) shows that committing to an economic and monetary union can induce economic coordination by improving trade, just like with the CFA zone. Furthermore, a previous study on ECOWAS Role in Achieving Regional Integration (Egbuna and Otten, 2013), investigated the major barriers to trade within West Africa and found that amongst other variables, trading activities between the eight WAEMU countries was largely induced by their common currency. For such reasons and more, economists are of the belief that introducing a common currency in the ECOWAS region-to be used by both the eight CFA franc countries, as well as the six that make up a second monetary zone within ECOWAS called West African Monetary Zone (WAMZ), and Cape Verde-would be rewarding in terms of increased trade and therefore, development.

Some economists however have doubts about the realization of this plan. Ngozi Okonjo-Iweala, the current Minister of Finance in Nigeria explains in an article written by Premium Times (2013), that introducing a single currency in the ECOWAS regions requires more than an act of law. “It requires action by the monetary authorities and by market participants in order to make it viable and sustainable” (Premium Times, 2013). Although having a common currency has its advantages such as currency convenience, price comparability, cross border investment, inflation discipline and exchange rate discipline, it also has its disadvantages. Introducing a common currency area would result in a loss of national monetary policy for the countries involved, as well as restricted expansionary fiscal policy. Moreover, not all regions are suitable to adopt a common currency, and this is the main issue to be considered within this research.

1_{NOTE: the CFA Franc officially refers to both Central African CFA Franc and West African CFA} Franc. However, for this paper, when I speak of the CFA Franc, I am referring solely to the West African CFA Franc.

2_{Made up of eight francophone West African Countries: Benin, Burkina Faso, Cote D’Ivoire,} Guinea Bissau, Mali, Niger, Senegal and Togo

(6)

Most literature on optimal currency theory focuses on the theoretical frameworks, indicating and explaining the different optimum currency area (OCA) criteria that indicate if the group of countries is better off adopting a common currency or not. Only a few of these literatures have attempted a quantitative analysis to test if a region/area can be deemed optimal according to the theory. Of those that do exist, not many have focused on Africa and even less on West Africa. Benassy-Quere and Coupet (2005), and Derun et al (2003) both tried to determine if the ECOWAS region can be an OCA using a correlation and cluster analysis and the macroeconomic modeling method respectively. The former showed that ECOWAS is unsuitable for a monetary union, while the latter found opposite results. As enlightening as both studies are, the authors use highly complicated methods involving detailed economic data, which are scarce in Africa and therefore had to be manipulated to fit the region, thus providing less reliable results. Moreover, the data used are from decades ago, which may produce different results today following the significant improvement in the economy.

Based on more recent data (2000-2012), the main objective of this paper is therefore to evaluate whether it would be beneficial to introduce a common currency ECOWAS wide, by answering the question, “Does ECOWAS constitute an optimum currency area (OCA)?” using a straightforward, less complicated quantitative

analysis, namely the OCA Index. This method aims to show if a monetary union is

feasible for a group of countries, by empirically testing for presence of OCA criteria within the region. The OCA index introduced by Bayoumi and Eichengreen (1997) uses various OCA criteria as independent variables to first determine the exchange rate variability in a region between country pairs. Thereafter, an OCA index is calculated based on the estimates of the regression for each country pair. The resulting figure is then used to determine which countries would constitute an optimal currency area.

This research is divided into six sections. The next section provides an in-depth review of the necessary criteria for the creation of a common currency by explaining the theory of OCA. It also provides a short but detailed description on different methodologies that can be used and how researchers have used them in the past. Section three highlights the current situation in West Africa in terms of some of the OCA criteria. Empirical methodology and data collection is explained fully in

(7)

section four and the results and analysis is outlined in section five. Finally, a conclusion of the research is given and possible steps that should be considered in the future.

2. LITERATURE REVIEW

In order to be able to assess if ECOWAS is considered an optimum currency area, it is necessary to identify what the conditions/criteria are in order to be deemed an optimum currency area. Mundell (1961) first introduced the theory of optimum currency area, which was subsequently contributed to by other economists like McKinnon (1963) and Kenen (1969). Mundell (1961) defines a currency area as a “domain within which exchange rates are fixed” (p. 657) and explains that a currency area is considered to be optimum when the benefits of adopting a single currency or a fixed exchange rate regime exceed the costs of abandoning the exchange rate as an instrument of adjustment in that area (Mundell, 1961). While the benefits of joining a monetary union and adopting a single currency relate to micro-level factors, the costs relate more to macro-level factors. Identifying these advantages within a region and comparing them to the costs involved will be ideal, however, it is only possible to do such an evaluation when a region has already adopted the single currency. Such a method may be ideal for the eight ECOWAS countries but not for the region as a whole.

Therefore, there is a need for indicators that can show when a group of countries will reap the benefits of a single currency more than the costs. This is what OCA properties/criteria identify. A currency area is said to be optimal when it fulfills “various OCA properties such as price and wage flexibility, financial integration, etc.” (Mongelli, 2008 p.2). The OCA properties as noted by Mongelli (2008), help lessen the need for nominal exchange rate adjustments within areas that share a common currency. These criteria do so by reducing the effects of some types of shocks or aiding their adjustments. A list of the various OCA properties as presented by Mongelli (2008) is summarized in Table 1.

(8)

Table 1: OCA Properties.

Source: Mongelli (2008), Box 2.1 “The OCA properties (I): the seminal contributions” pp. 2-3

a. Price and wage flexibility. Nominal prices and wages need to be flexible between the countries contemplating a single currency so as to lessen the problem of unemployment and/or inflation due to shocks.

b. Mobility of factors of production including labor. This enhances efficiency and welfare. Mundell (1961) explains that a fixed exchange rate system is better for areas where factors of production are mobile, as necessity to amend real factor prices and the nominal exchange rate between countries due to different shocks and disturbances would be lessened.

c. Financial market integration. As McKinnon (2004) argues, financial integration results in the diversification of income sources between countries that make use of a single currency. Such results can also help moderate the effects of asymmetric shocks. For example through income insurance, as a citizen of one country can acquire claims to interests, rental revenue and dividends from another country.

d. The degree of economic openness. Changes in international prices of goods that are imported is more likely to be transferred to the domestic cost of living in countries with high degree of openness. According to McKinnon (1963), this therefore reduces the exchange rate illusion most wage earners have.

e. The diversification in production and consumption. Trading partner countries that have highly diversified production and consumption portfolios, bear lesser costs as a result of relinquishing nominal exchange rate changes between them. They therefore find that creating a monetary union with a single currency is very beneficial. Kenen (1969) supports this motion as he explains that when a region is highly diversified in production, they have a greater chance of withstanding asymmetric shocks.

f. Similarities of inflation rates. As Fleming (1971) notes, low and similar inflation rates results in a stable terms of trade for the countries in question. It induces more balanced current account transactions as well as trade, and therefore also reduces the need for nominal exchange rate adjustments.

g. Political integration. Mintz (1970) and Haberler (1970) regard this as one of the most important conditions for a successful single currency area. Political integration boosts institutional linkages between the countries and helps favor cooperation between the countries on various economic policies.

(9)

The values for some of the criteria are not exactly quantifiable, which makes it difficult to determine the point at which a currency area is insufficient or sufficient; hence there is the need for a methodology that incorporates the theory and its criteria into a quantitative analysis. There are many of such methodologies, like the Generalized Purchasing Power Parity analysis (GPPP), the Correlation and Cluster Analysis, Macroeconomic Modeling and OCA index. Although different economic researchers have successfully used these methodologies, they all have their different limitations and advantages. In line with Adams (2005), these different methodologies are discussed below, highlighting the limitations and advantages as they pertain to this research.

2.1 Methodologies That Incorporate the OCA Theory

2.1.1 Generalized Purchasing Power Parity

This method as developed by Ender and Hum (1994), assesses the extent to which a group of countries exhibit integration of their real exchange rates. That is, by using cointegration analysis it “assesses the level of similarity in the movements of the real exchange rate relative to a central dominant country” (Adams, 2005 p.29). Breitenbach and Zerihn (2012) used the GPPP method to investigate the desirability of a policy of a monetary union in the Southern African Development Community (SADC) region, as well as the introduction of common currency. The results of the GPPP analysis proved positive in SADC, as there was evidence of cointegration of the real exchange rates of the countries involved. Mkenda (2001) also used the same analysis for African countries, but focused on three East African countries (Kenya, Tanzania and Uganda) and found similar results.

Although this method produced positive results in the above-mentioned studies, Adams (2005) suggests that the GPPP analysis “method alone cannot form the basis of worthwhile conclusions as the assumption that real exchange rate capture the economic fundamentals is too implausible” (p.30), as other factors like political interventions can affect the real exchange rates. Moreover, it captures just one aspect of the OCA theory.

(10)

2.1.2 The Correlation and Cluster Analysis

The cluster analysis as implemented by Atris and Zhang (2001), “examines the similarities and dissimilarities of economic structure which appear in the variables proposed to proxy OCA criteria” (p.6). Using a group of countries and different variables equivalent to OCA criteria (labor market flexibility, volatility in exchange rate, correlation in business cycles, etc.), the aim is to determine the group of countries within the data set that can be characterized as the core group, and therefore suitable for monetary union, due to the similar characteristics. Atris and Zhang (2001) used this within the European Monetary Union countries and found that it “consist in core group revolving around Germany” (p.13). Similarly, Benassy-Quere and Coupet (2005) used this analysis within the ECOWAS region, and the results showed that ECOWAS is unsuitable for a monetary union.

Although this methodology seems more valid than the GPPP analysis as it incorporates more variables, it still posses some problems. It is highly complicated method that requires detailed economic data like the labor market flexibility, which is hard to capture especially in within the African context.

2.1.3 OCA index

The OCA index introduced by Bayoumi and Eichengreen (1997) enables one to quantify the convergence of a currency area and therefore note its optimality. It is often described as a tool used for “assessing the possibilities of successful working countries in the single currency area” (Drastichova, 2011 p.121). The “OCA theory focuses on characteristics which make stable exchange rates and monetary unification more or less desirable” (Bayoumi and Eichengreen, 1997 p.762). The most importation of which (the asymmetry of business cycles, the differences in trade

structures, the share of bilateral trade in total trade and the relative size of countries)

are used in the OCA index to “predict which countries will be best able to support stable exchange rates” (Bayoumi and Eichengreen, 1997 p.762). “The variability of real and nominal exchange rates is itself the outcome of the choice of exchange rate regime as such should contain information about the decision of what arrangement to adopt” (Bayoumi and Eichengreen, 1997 p.762).

Bayoumi and Eichengreen (1997) applied this method to evaluate the optimality of the European monetary union and found that the European countries are

(11)

divided into three groups: those that exhibit high levels of readiness, those with potential and those without. Adam (2005) also adopted the OCA index to investigate the feasibility and effect of a single currency in Africa, focusing on the various regional economic communities present. The results were found to be inconsistent over time and so a concrete conclusion could not be made.

The OCA index so far seems like a more plausible and straight forward methodology within the context of ECOWAS, and although Adams (2005) suggested that the traditional OCA theory should be adjusted to fit African countries, it is possible that his results were inconsistent due to the fact that some countries were eliminated from his research for being present in more than one regional community. Since this research concentrates on just ECOWAS, more concrete conclusions may be reached. Nevertheless, the OCA index methodology is one that should be interpreted with caution due to possible endogenous nature of the variables.

The hypothesis of endogeneity of OCA, which will be discussed in depth in the following sub section, signifies Lucas critique, which is associated with the theory of rational expectations. “When assuming rational expectations, economics political authorizes are not able to eliminate the arising shocks. So it is not possible to make conclusions about the future from analysis of the historical data” (Drastichova, 2011 p.123).

2.2 Endogeneity of OCA theory

Frankel and Rose (1997) brought the hypothesis of endogeneity of OCA forth when researching the outcome of the different monetary unions that were introduced previously. They discuss how international trade patterns and international business cycle correlations are endogenous, and come up with the conclusion that when two countries have closer trade linkages, it is more or less due to their tightly correlated economic activities, most especially with intra-industry trade. Artis and Zhang (1999) and Melitz (2004) also support the notion of endogeneity of symmetry of shock and synchronization of outputs, which is in line with the correlation of business cycles. Other economists claim that other sources of endogeneity of OCA exist, like the endogeneity of financial integration (see Baele et al., 2004) and endogeneity of product and labour market flexibility (see Bertola and Boeri, 2004).

(12)

The main conclusion of the hypothesis expressed by many economists including Mongelli (2008) is “that monetary integration represents a removal of “borders” (very broadly intended to include also national monies) that contributes to the narrowing of distances and a change in the incentive structure of agents” (p.8). That is, as analyzed within the European Union by Frankel and Rose (1997), historical data may characterize some countries as poor candidates for the European Economic and Monetary Union (EMU). However, entry into the union may very well lead to better economic conditions for the countries in question. Therefore, “a country is more likely to satisfy the criteria for entry into a currency union ex post than ex ante” (p.23).

It has been declared by the endogeneity of OCA researchers that a single currency actually reduces the monetary costs involved. For instance, the euro has helped reduce costs involved directly and indirectly with trade, by “removing exchange rate risks and the cost of currency hedging… as well as reducing information costs” (Mongelli, 2008 p.10). McCallum (1995) states that progress in the different aspects of the OCA criteria theory is boosted when a group of countries form a monetary union with a single currency.

Overall, the endogeneity of OCA contradicts the OCA theory, which implies that for a group of countries to make up an optimal currency area, the several OCA criteria should be present ex ante rather than ex post. However, the OCA index can still be used when corrected for the problem of endogeneity as is done in the methodology of this research. The forthcoming section assesses the current situation in West Africa based on some of the quantifiable OCA criteria.

(13)

3. OCA CRITERIA WITHIN ECOWAS

This section assesses the current situation within ECOWAS based on some aspects of the OCA criteria. According to theory, a group of countries are said to qualify as an optimal currency area when benefits of adopting a single currency outweigh the costs. As previously noted, analyzing the situation for a group of countries that use different currencies would be quite hard, therefore there is a need for the OCA properties. It is interesting to see where some of the countries currently stand based on the OCA criteria.

3.1 Mobility of factors of production including labor

As explained by Mundell (1961), when a fixed exchange rate system is introduced for a group of countries between which the factors of production are mobile, altering factors like real prices and nominal exchange rate between them so as to respond to disturbances is not needed as much. ECOWAS has a policy set in place to promote trade and factor mobility. Following this policy, an ECOWAS passport, which allows for the free movement of people, goods and capital within the region, was introduced. This led to an increased flow of people, and although goods and services still remain somewhat stagnant, due to poor regional infrastructures and the “continued existence of non-tariff barriers and road blocks which add to the cost of moving goods “(WAMI, 2014), this can be seen as a stepping-stone towards achieving a currency union within West Africa.

3.2 Financial market integration

McKinnon (2004) argues that financial integration between countries that use a single currency can reduce the effects of asymmetric shocks. Ingram (1962) also supports this notion as he explains that a stable financial system can mitigate the need for exchange rate adjustments as it helps absorb asymmetric shocks through capital inflows. The countries within ECOWAS are far from integrated when it comes to their financial system, and therefore contradicts the OCA criteria theory.

Table 2 outlines the different exchange rate regimes and monetary policy frameworks adopted by each country. According to the table, each country is classified based on their actual, de facto, arrangements as identified by IMF staff,

(14)

which may differ from their officially announced arrangements. All WAEMU countries and Cape Verde fall under the Soft Pegs exchange rate arrangements, Nigeria, Guinea and Liberia fall under the Residual exchange rate arrangement, while the rest are classified under Floating exchange rate arrangements.

The overall state of the financial market in West Africa is such that shocks to the real economy could affect the countries very differently at any point in time. “The prevalence of asymmetric shocks may impede the conduct and effectiveness of monetary policy” within the region “(WAMI, 2014).

Table 2: De Facto Exchange Rate Arrangements and Anchors of Monetary Policy Sources: IMF staff; Annual Report on Exchange Arrangements and Exchange Restrictions 2013 (pp5-6)

EXCHANGE RATE REGIME

MONETARY POLICY FRAMEWORK EXCHANGE RATE ANCHOR. EURO MONETARY AGGREGATE TARGET INFLATION TARGETING FRAMEWORK CONVENTIONAL PEG Benin Burkina Faso Cape Verde Cote d’Ivoire Guinea-Bissau Mali Niger Senegal Togo OTHER MANAGED ARRANGEMENT Liberia Guinea Nigeria

FLOATING Gambia, The

Sierra Leone Ghana

3.3 Similarities in inflation rate

Different national inflation rates cause external imbalances, and this is the base of Fleming (1971) argument. Fleming (1971) claims that low and similar inflation rates between countries lead to a fairly stable terms of trade, and therefore, the need for nominal exchange rate adjustments is reduced. Using consumer prices from World Bank database as an indicator for inflation, Table 3, shows the mean and standard deviation for a period of seven years (2007-2013). The CFA countries experienced much lower inflation rate levels than the other countries in the region, ranging anywhere from 2.47% to 3.44% for the eight countries. A similar result was found in

(15)

Bayoumi and Ostry (1995), with the reason being the fixity of the parity of the CFA franc vis-á-vis the French franc. The same conclusion can be drawn from the inflation correlation of all ECOWAS countries. For the WAMZ countries, the correlations are small, and in some cases, negative as seen in Table 1 in the appendix. Such results generally are not in favor of a single currency within ECOWAS.

Table 3: Inflation consumer prices (2007-2013) Source: World Bank database, own calculations

Country Mean Standard deviation

Benin 3,449 2,752 Burkina Faso 2,770 3,870 Cape Verde 3,255 2,056 Cote d'Ivoire 2,812 2,010 Gambia, The 4,746 0,410 Ghana 12,387 3,962 Guinea 15,691 6,142 Guinea-Bissau 3,403 3,856 Liberia 9,820 4,101 Mali 3,120 3,247 Niger 2,634 3,966 Nigeria 10,536 2,767 Senegal 2,475 2,624 Sierra Leone 13,100 2,888 Togo 3,252 2,563

3.4 The diversification in production and consumption

As acclaimed by many economic researchers including Bayoumi and Ostry (1995), Sub-Saharan Africa is a continent that is highly specialized in the production and export of mostly primary/agricultural commodities. Such facts counter the OCA criteria theory as Kenen (1969) explains that forgoing nominal exchange rate changes between partner countries that are highly diversified, more than likely leads to reduced costs. This therefore implies that a group of specialized countries as in ECOWAS face a greater propensity for asymmetric shocks.

From the evaluation of these criteria within ECOWAS, one is tempted to conclude that ECOWAS is not in line with the theory of optimal currency area. However, it is important to note that the above evaluation and observation just help point out the current situation in West Africa, and therefore is not sufficient enough to indicate the

(16)

optimality of ECOWAS as a currency area. Therefore, the OCA index methodology is needed to help give an in-depth analysis.

4. DATA AND METHODOLOGY

4.1 Empirical Methodology

The focus of this research is to determine the suitability of adopting a single currency amongst the fifteen ECOWAS countries. This is done in two steps, as proposed by Bayoumi and Eichengreen (1997). Firstly, the relation between bilateral exchange rate volatility and different OCA criteria is determined using a two-stage least squares (TSLS) regression. This is an estimation method under the simultaneous equation method that incorporates the use of instruments, which help deal with the problem of endogeneity. Secondly, the OCA index is computed for the different ECOWAS country pairs using the coefficients from the estimation in the first step which a method called out-of-sample approach.

The first step adopts a regression where the bilateral exchange rate volatility is seen as the dependent variable and the different OCA criteria are seen as the independent variables. These variables are: the asymmetry of business cycles, the

differences in trade structures, the share of bilateral trade in total trade and the relative size of countries.

Bilateral exchange rate volatility (SD(eij)) is measured as the standard

deviation of the change-from year t to year t+1-in the logarithm of the end year bilateral exchange rate between countries i and j. Although Bayoumi and Eichengreen (1997), as well as many other studies that have used this method, employ the use of nominal exchange rate, this study uses the real exchange rate because eight of the countries in the data already use a single currency. The use of real exchange rate rather than nominal will better capture the variability between these countries and therefore determine if all fifteen countries can be deemed an OCA, by not looking at the eight countries as one.

(17)

The asymmetry of business cycles (BSCij) is measured as the standard

deviation of the difference in the change-from year t to year t+1-in logarithm of real output between country i and j. Difference in trade structures (DISSIMij) is measured

as the sum of the absolute differences in the shares of agricultural, mineral and manufacturing in total merchandise trade, the share of bilateral trade (TRADEij) is

measured as the average of the ratio of bilateral exports to domestic GDP for country i and j, and the relative size of the economies (SIZEij) is measured as the mean of the

logarithm of the two GDPs measured in constant US dollars. It is important to note that each variable is measured as an average over the given sample period.

The estimation model is therefore as follows:

SD(eij) = β0 + β1BSCij + β2DISSIMij + β3TRADEij + β4SIZEij + εij The intuition behind the estimation is that the more the variables-which represent the OCA criteria-are fulfilled among the different countries, the lower the variability of the exchange rates should be. That is, the more the group of countries are better suited to join a monetary union.

Due to lack of data, this research only incorporates three of the four independent variables, leaving out Difference in trade structures. Table 4 outlines exactly how each variable is derived, and the revised equation is therefore:

(18)

Table 4:

Calculation of the variables

VARIABLE CALCULATION

SD(eij) D ( o )

The asymmetry of business cycles “BSCij” D ( )

The share of bilateral trade “TRADEij” ∑

The relative size of the economies “SIZEij” ∑

 eij is the end year bilateral real exchange rate between countries i and j.

 Yi and Yj are the logarithms of real output of country i and country j respectively.

 exijt and exjit are exports from country i to country j in year t and vice versa, respectively.

 yit and yit is the GDP of country i to country j in year t, respectively.

In line with the economic theory of this methodology, the bilateral exchange rate volatility is expected to be positively dependent on the asymmetry of business cycles (β1) in the sense that, as the standard deviation of the difference in the logarithm of real output between two countries increases, so does the bilateral exchange rate volatility between them. The bilateral exchange rate volatility is expected to be negatively dependent on the share of bilateral trade (β2), and positively dependent on the size on the economy (β3) indicating that the larger the country size, the greater the exchange rate variability.

The estimation of the above regression can also be done using a simple OLS regression. However, as noted, endogeneity may be present, following the “Endogeneity of the OCA theory”, with some of the independent variables; namely

TRADEij and BSCij, being endogenous to the dependent variable. Frankel and Rose

(1997) came up with the conclusion that “a closer trade linkage between two countries is strongly and consistently associated with more tightly correlated economic activity between the two countries” (p.19), most especially with intra-industry trade, which is more likely present within the fifteen ECOWAS countries. Therefore, a simple OLS regression could result in an invalid interpretation of the estimation. Introducing instruments to the equation and using an IV regression could deal with the bias of endogeneity. Such instruments include:

(19)

 Distanceij: the logarithm of bilateral distance between countries i and j.

 Borderij: A dummy variable, which takes a value of 1 if the two countries share a border, and 0 otherwise.

 Languageij: A dummy variable, which takes a value of 1 if the two countries speak the same language, and 0 otherwise.

These three instruments affect both trade and the asymmetry of business cycles between two countries, as if the pair of countries share the same language and/or are in very close proximity to one another, trade is more likely to occur. This will also lead to more synchronized business cycles between them. This set of instruments was also adopted by other economists doing similar research, including Bayoumi and Eichengreen (1998).

4.2 Data

The OCA index is obtained form a cross sectional analysis-due to the nature of the generated variables-using an annual dataset for the fifteen ECOWAS countries mentioned earlier, over a thirteen year period (2000-2012). This sample period is used because it captures the recent times in which most of these countries have shown significant amount of improvement in their various economies when compared to the time period used by other economists. This analysis takes into account the relationship between each country, and given the bilateral nature of the data, the relationship between country i and j is the same as the relationship between country j and i. Also, as explained earlier, each variable is measured as an average over the sample period of thirteen years. For example, each country pair (e.g. Benin-Nigeria) has a single value for each variable, and is counted once since Benin to Nigeria = Nigeria to Benin. The whole nature of the data therefore leads to matrix with 105 observations , one for each pair of countries.

Data for calculating Bilateral exchange rate volatility SD(eij) is obtained from

the IMF’s International Statistics (IFS). The asymmetry of business cycles (BSCij) and

the relative size of the economies (SIZEij), are both calculated using data form the

World Bank database, while the share of bilateral trade (TRADEij), is calculated

(20)

5. RESULTS AND ANALYSIS

5.1 Empirical Analysis

The results of estimation of equation (1) by OLS are shown in column one of Table 5 and are in line with the initial assumptions. The signs for the coefficients of all three independent variables are as predicted and all but TRADEij are significant on a 10% level or lower. The insignificance of trade can be caused by the fact that the data showed signs of no trade between some of the countries on an average of ten out of the thirteen years in the sample period. This is not due to lack of data but rather that some countries may import from other countries but do not offer any products that are worth exporting in return. The equation explains 20% of the variation in bilateral real exchange rate, which is a reasonable r-squared and can therefore still support the notion of the OCA theory.

According to the estimation, a one percentage point increase in asymmetry of business cycles will increase the bilateral exchange rate variability between the two countries by approximately 82%, while a one percent increase in the size of the countries will increase the bilateral exchange rate variability by 1%, ceteris paribus. However, a one percent increase of trade linkages between the countries leads to a 123.7% decrease in bilateral exchange rate viability. Such effects can however be misleading as it is noted that the coefficients of the OLS estimation may be biased due to endogeneity, hence the need for instruments to resolve such issues. Tests were done to support the assumption of the likelihood of TRADEij and BSCij being endogenous. A test was done both separately and together for both independent variables. The results of the test produce a p-value of 0.7706 and 0.0000 for TRADEij and BSCij respectively. The low p-value for BSCij means that the null assumption of exogeneity of the variable can be rejected; therefore indicating that asymmetry of business cycles is indeed endogenous. On the other hand, the high p-value for trade indicates that trade is actually exogenous. This results leads to the instrumentation of BSCij only in this research.

The results of the IV regression estimates are presented in column two of Table 5. Although TRADEij is still insignificant as in the OLS estimation, the other two independent variables are more so significant on a 0.1% level. The coefficients also have a larger impact on bilateral exchange rate when compared to the coefficients

(21)

of the OLS estimation. A one percent increase in asymmetry of business cycles will increase the bilateral exchange rate variability between the two countries by 416%, while a one percent increase in the size of the countries will increase the bilateral exchange rate variability by 3.2%, ceteris paribus. However, a one percent increase of trade linkages between the countries still leads to a 98% decrease in bilateral exchange rate viability.

Table 5:

Variability of bilateral exchange rates, OLS and IV estimations Sources: Own estimations based on data

OLS IV

The asymmetry of business cycles “BSCij”

The share of bilateral trade “TRADEij” The relative size of the economies

“SIZEij” Constant 0.821* (0.13) 4.166* (1.12) -1.237 (1.07) -0.981 (1.53) 0.010** (0.00) 0.032* (0.01) -0.325*** (0.18) -1.509* (0.51) R-squared N 0.204 . 105 105

*** p<0.10, ** p<0.05, * p<0.01. The Standard errors are reported in parenthesize.

The large impact of asymmetry of business cycles and its relevance over trade in this estimation is not unlikely as other studied found similar conclusions. Eva Spanikova (2006) found similar results in her discussion of the possible entrance of Slovakia into the European Union. “The results suggest that the factors that exercise pressure on the (nominal) exchange rate (especially variability of output) are more important than the factors that normally tend to stabilize (such as trade linkages)” (Eva Spanikova, 2006 p.52). Bayoumi and Eichengreen (1998) also found that variability of output has a large positive effect of 209% on the variability of actual exchange rate when testing the optimality in industrial countries.

As stated above, the results of the IV regression also show that the asymmetry of business cycles has the largest impact on exchange rate variability while trade

(22)

between some countries in the data set. The variables are found to be jointly significantly different from zero, which suggests that they do explain some variation in bilateral exchange rate. The regression therefore shows a strong support of the empirical implication of the OCA theory and therefore can be used to forecast and predict the OCA index, which is the next step in the methodology of the research.

5.2 OCA index for ECOWAS countries

As explained earlier, calculating the OCA index is the second step in the “out-of-sample approach” applied by Bayoumi and Eichengreen (1997). It is the predicted/forecasted value of “exchange rate variability in a given period adjusted for exchange rate variability based on the OCA criteria.” (Eva Spanikova, 2006 p.60). The values of the IV estimation from the pervious section are used to calculate the so-called OCA index. Bayoumi and Eichengreen (1997) explain that long-term variability of bilateral exchange rates should indicate the level of fulfillment of the OCA criteria. That is, a low variability of bilateral exchange rates indicates that the countries fulfill the OCA criteria. All in all, the lower the value of the OCA index, the less the cost of adopting a single currency, which therefore indicates that the countries are substantial candidates for a currency union. The OCA index formula according to the estimations is as follows:

OCA index = -1.509 + 4.166*BSCij – 0.981*TRADEij + 0.032* SIZEij (2) The OCA index is calculated using the derived independent variables averages for the sample period, for each country pair. Judging from most research that concluded with signs of optimality within the different countries involved (Eva Spanikova, 2006 and Bayoumi and Eichengreen, 1997), a low OCA index ranges anywhere from zero to about 0.05. With respect to the CFA countries, one would expect relatively low indexes that fall within the aforementioned range, indicating low variability of their respective bilateral exchange rates, since they currently use a single currency and are part of a monetary union WAEMU. However the results indicate the opposite.

Table 6 presents the results for these eight countries. Cote d’Ivoire is considered the biggest country within the group based on its real GDP value compared to the rest, but its relationship within the group in terms of the OCA index

(23)

is high. That is, it has the highest predicted volatility relationship between itself and the rest of the CFA countries, with the highest value being between itself and Guinea-Bissau at 0.177. According to the theory, this suggests that a monetary union between Cote d’Ivoire and the remaining seven CFA countries would not be beneficial. Such results could be due to the size of Cote d’Ivoire in comparison to the other countries; with the difference being so huge that the cost of a monetary union would weigh more on Cote d’Ivoire. Especially as, for instance, Guinea-Bissau is considered as one of the weak countries within the group based on it GDP value. One the other hand, OCA index between Benin and the remaining six CFA countries (excluding Cote d’Ivoire) is low, with the lowest value being between Benin and Burkina Faso at 0.02.

In general, the indexes are too high to indicate any sign of an optimal currency area amongst the CFA countries. This result is in line with Benassy-Querre and Coupet (2003) who, as previously mentioned, used correlation and cluster analysis methodology to assess the adequacy of monetary arrangements in Sub-Saharan Africa. Their result shows that the CFA franc zone cannot be viewed as an optimal currency area as the countries did not belong to the same clusters. They however, suggested the inclusion of Gambia, Ghana and Sierra Leone into the WAEMU monetary arrangement. The result of this research supports the inclusion of Gambia, but not Ghana and Sierra Leone, as the OCA indexes between these two countries and the rest of WAEMU are too high as seen in Table 2 of the appendix. However, Guinea may be a better match has it has one of the lowest index in the matrix, at 0.006 between itself and Burkina Faso.

(24)

Table 6: OCA Index between CFA Franc Countries Source: Own calculations based on data

OCA Index Benin Burkina Faso

Cote d'Ivoire

Guinea-Bissau Mali Niger

Sierra Leone Burkina Faso 0,020 Cote d'Ivoire 0,100 0,075 Guinea-Bissau 0,046 0,107 0,177 Mali 0,045 0,065 0,166 0,058 Niger 0,047 0,007 0,095 0,105 0,092 Senegal 0,018 0,035 0,118 0,083 0,043 0,088 Togo -0,011 -0,025 0,066 0,089 0,070 0,028 0,149

Furthermore, Zhao and Kim (2009) also came to a similar conclusion that the CFA franc zone does not appear to form an optimum currency area and that “the monetary union may have been costly to maintain unless the member countries are compensated with some other benefits” (p.1882). Table 7 shows the OCA indexes calculated for the WAMZ countries. The results also indicate that the introduction of a single currency to be used by these six countries would incur more cost than expected benefits. The high-predicted exchange rate volatility is most evident between Nigeria and Liberia at 0.822, as well as between Liberia and the rest of the WAMZ countries. The relationship between Nigeria and Liberia is the same as that between Cote d’Ivoire and Guinea-Bissau, with Nigeria being the biggest country in terms of Real GDP within the WAMZ countries and ECOWAS as a whole, and Liberia being the smallest.

(25)

Table 7: OCA Index between WAMZ Countries Source: Own calculations based on data

OCA Index Gambia, The Ghana Guinea Liberia Nigeria

Ghana 0,120

Guinea 0,039 0,042

Liberia 0,552 0,589 0,509

Nigeria 0,213 0,346 0,301 0,802

Sierra Leone 0,199 0,216 0,140 0,556 0,388

As noted, the overall relationship between each WAMZ country pair is far worse than that of the CFA countries as the lowest exchange variability is between Gambia and Guinea at 0.039. The none optimal nature of the WAMZ monetary institution is in line with Benassy-Querre and Coupet (2003). Their analysis concludes, that creating a separate WAMZ monetary union is not advised and neither is the inclusion of Nigeria in the CFA zone.

Although the results show that a monetary union around Nigeria is not efficient, Nigeria is still one of the biggest countries in West Africa, and thus can be used as a benchmark for all the other fourteen countries in the research. Taking Nigeria as a benchmark, an analysis can be done to show possible signs of convergence, and hence improvements, over the years towards achieving a singe currency.

Table 3 in the appendix shows forecasts of the dependent variable, i.e. the OCA index, vis-á-vis Nigeria in 2000-2007 and 2008-2012. The sample period is divided into these two time periods to see what impact the financial crisis has played towards the achievement of a single currency within ECOWAS. The first period represents a period before the financial crisis, and the second represents the period

during and after the financial crisis. It is expected that the financial crisis would affect

the OCA index negatively, in the sense that the second time period would have higher predicted exchange rate volatility. However the result is quite the opposite. For all fourteen relationships, the OCA index is lower during and after the financial crisis than before, more so with Liberia. Graph 2 below shows that before the financial

(26)

crisis, Liberia was the only outlier within the group with an OCA index of 1.041, but there is sign of convergence as the OCA index drops drastically to 0.029 during the second time period.

Graph 2: Movements in OCA indexes over time Source: World Bank Data

The overall declining rate of the OCA index despite the financial crisis could be an indication of the improved economic situation within the region compared to pervious years. Although the OCA indexes are still quite high and hence suggest that an optimal currency area would not be the right move for ECOWAS at the moment, the sign of convergence could mean that some time in the near future the West African economy would reap more benefits than costs with the use of a single currency. The delay with the introduction of a single currency within ECOWAS is perhaps the right decision as for now. Accordingly, WAMI (2014) concludes that although “considerable progress has been made within the region, the level of macroeconomic and convergence and legal and institutional preparedness required for the successful launch of the union by January 1, 2015 were still inadequate” (p. 103). They claim that the whole monetary union project cannot be advantageous without stronger political commitment and public understanding hence the reason for the further postponement till January 1, 2020. It would therefore be valid to analyze beyond the present years, so as to predict and forecast if 2020 will be the right time.

Liberia 0 0,2 0,4 0,6 0,8 1 1,2 0 0,2 0,4 0,6 0,8 1 1,2 2008 -2012 2000-2007

(27)

6. CONCLUSION

The main objective of this research has been to evaluate whether it would be beneficial to introduce a common currency ECOWAS wide based on the more recent developments within the region, by answering the question, “Does ECOWAS constitute an optimum currency area (OCA)?” To do so, the theory of optimum currency areas is operationalized in other to construct an OCA index, which helps answer the proposed research question. Low OCA index values indicate high benefit to cost ratios for monetary integration for a given pair of countries and high values result in the opposite.

Many economic researchers have found this methodology useful in the European context, and very few have applied it within Africa. Adam (2005) was unable to draw a concrete conclusion using the OCA index methodology, as his results were found to be inconsistent over time, suggesting that the method be adjusted to fit the African countries. Other economists that tried to answer a similar research question adopted different methodologies and came up with similar results (see Benassy-Querre and Coupet (2003), Zhao and Kim (2009), Breitenbach and Zerihn (2012) and Mkenda (2001))

The results of this research were expected to be different due to better economic conditions within the region compared to the past years during which most of these researches were conducted. However, similar conclusions to past researches were drawn. The OCA index was found to be very high between more than half of the 105 country pairs. Most especially between the big countries in the matrix like Nigeria, Ghana and Cote d’Ivoire, and the small countries like Guinea-Bissau and Liberia, suggesting that the creation of monetary union made of the fifteen ECOWAS countries would result in a low benefit to cost ratio.

Dividing the region into two separate monetary unions isn’t advisable either. The results show that the CFA currency area is not optimal and is probably very much dependent on outside benefits to help offset the cost, while the WAMZ zone would definitely incur more cost. However, it is interesting to observe that the OCA index is declining despite the financial crisis, and note that this could be an indication of the improved economic situation within the region compared to pervious years.

(28)

Although the OCA indexes are still quite high and hence suggest that an optimal currency area would not be the right move for ECOWAS at the moment, the sign of convergence could mean more positive results some time in future. According to the WAMI 2014 report, as January 1, 2020 is the current set date for the introduction of the single currency in West Africa, it would be valid to analyze beyond the present years, so as to predict and forecast if this is the valid time.

(29)

References

Adams, P. D. (2005). Optimal Currency Areas: Theory and Evidence for an African

Single Currency. University of Manchester.

Alaba, O.B. (2006). EU-ECOWAS EPA: Regional Integration, Trade Facilitation

and Development in West Africa

Artis, M. J., & Zhang, W. (2001). Core and periphery in EMU: A cluster analysis. Economic Issues Journal Articles, 6(2), 47-58.

Artis, M. J., & Zhang, W. (1999). Further evidence on the international business cycle and the ERM: is there a European business cycle?. Oxford Economic

Papers, 51(1), 120-132.

Baele, L., A. Ferrando, P. Hördahl, E. Krylova and C. Monnet (2004), “Measuring financial integration in the euro area”, ECB Occasional Paper No 14.

Bayoumi, T. (1994). A formal model of optimum currency areas. Staff

Papers-International Monetary Fund, 537-554.

Bayoumi, T., &Eichengreen, B. (1997). Ever closer to heaven? An

optimum-currency-area index for European countries. European economic

review, 41(3), 761-770.

Bayoumi, T., & Eichengreen, B. (1998). Exchange rate volatility and intervention: implications of the theory of optimum currency areas. Journal of International

Economics, 45(2), 191-209.

Bayoumi, T., & Ostry, J. D., (1995),“Macroeconomic Shocks and Trade Flows

Within Sub-Saharan Africa: Implications for Optimum Currency

Arrangements”(Vol. 142). IMF Working Paper No. WP/95.

Bénassy‐Quéré, A., & Coupet, M. (2005). On the Adequacy of Monetary Arrangements in Sub‐Saharan Africa. The World Economy, 28(3), 349-373. Bertola, G. and Boeri, T. (2004) ‘Product Market Integrations, Institutions and the

(30)

Breitenbach, M. C., & Zerihun, .M. F. (2012) Exchange rate policy options and Currency Union in SADC–a GPPP Approach to assessing OCA criteria.

Derun, X., Mason, P. and Patillo, C. (2003) ‘West African Currency Unions: Rationale and Sustainability’ CESifo Economic Studies 49(3): 381-413

Drastichova, K. (2011). The Assessment of Convergence in the EU Using the Optimum Currency Area Index. In 13th International Conference on Finance and Banking: Lessons learned from the Financial Crisis (pp. 121-135).

Egbuna, C. & Otten, S. (2013) ECOWAS Role in Achieving Regional Integration.

Enders, W., & Hum, S. (1994). Theory and Tests of Generalized Purchasing‐Power Parity: Common Trends and Real Exchange Rates In the Pacific Rim. Review

of International Economics, 2(2), 179-190.

Feldstein, M. (2008). Optimal Currency Areas.

Fleming. J.M. (1971). ‘On Exchange Rate Unification’. Economic Journal, Vol. 81, pp. 467–88.

Frankel, J. A., & Rose, A. K. (1998). The endogeneity of the optimum currency area criteria. The Economic Journal, 108(449), 1009-1025.

Haberler, G. (1970), “The International Monetary System: Some recent Developments and Discussions”, in Approaches to Greater Flexibility in Exchange Rates, edited by George Halm, Princeton University Press, pp. 115-23.

Ingram, J.C. (1962) Regional Payments Mechanisms: The Case of Puerto Rico (Chapel Hill: University of North Carolina Press)

IMF (2003). Annual Report on Exchange Arrangements and Exchange Restrictions 2013.

Kenen, P. (1969), “The Optimum Currency Area: An Eclectic View”, In Mundell and Swoboda, (eds.), Monetary Problems of the International Economy, Chicago: University of Chicago Press.

(31)

McCallum, J. (1995), ‘National Borders Matter: Canada – US Regional Trade Patterns’. American Economic Review, Vol. 85, No. 3, pp. 615–23.

McKinnon, R. I. (1963). Optimum currency areas. The American Economic Review, 717-725.

McKinnon, R. I. (2004). Optimum currency areas and key currencies: Mundell I versus Mundell II. JCMS: Journal of Common Market Studies, 42(4), 689-715.

Melitz, J. (2004), “Risk Sharing and EMU”, Journal of Common Market Studies. Special Issue Nov. 2004; 42(4): 815-40.

Mintz, N.N. (1970), “Monetary Union and Economic Integration”, The Bulletin, New York University.

Mkenda, B. K. (2001). Is East Africa an optimum currency area?. Göteberg

University Department of Economics Working Paper, 41.

Mongelli, F. P. (2008). European economic and monetary integration, and the

optimum currency area theory (No. 302). Directorate General Economic and

Monetary Affairs (DG ECFIN), European Commission.

Mundell, R. A. (1961). A theory of optimum currency areas. The American Economic

Review Vol. 51, 657-665.

Premium Times (2013).West Africa Must Learn From Eurozone on Single Currency – Okonjo-Iweala Premium Times. Retrieved from:

http://www.premiumtimesng.com/business/115864-west-africa-must-learn-from-eurozone-on-single-currency-okonjo-iweala.html

Spaniková. E. (2006). Is Slovakia making headway towards constituting an OCA with the EMU?

World Data Bank (n.d.). World development Indicators. Retrieved from:

http://databank.worldbank.org/data/views/variableSelection/selectvariables.as px?source=world-development-indicators#s_r

WAMI (2004). A study on The State of Preparedness of WAMZ Member States for Monetary Union. WAMI-IMAO/CC.A/2014.03

(32)

Zhao, X., & Kim, Y. (2009). Is the CFA Franc zone an optimum currency area?. World Development, 37(12), 1877-1886.

(33)

Appendix

Table 1: Correlation of Inflation values within ECOWAS Source: World Bank database; own calculations

Benin Burkina Faso Cape Verde Cote d’Ivoire The

Gambia Ghana Guinea

Guinea-Bissau Liberia Mali Niger Nigeria Senegal

Sierra-Leone Togo Benin 1,00 Burkina-Faso 0,86 1,00 Cape Verde 0,55 0,67 1,00 Cote d’Ivoire 0,46 0,72 0,84 1,00 The Gambia -0,37 -0,39 0,20 -0,10 1,00 Ghana 0,14 0,44 -0,04 0,06 -0,14 1,00 Guinea 0,15 0,07 0,75 0,49 0,46 -0,63 1,00 Guinea-Bissau 0,60 0,68 0,98 0,85 0,14 -0,06 0,73 1,00 Liberia 0,04 0,02 -0,10 0,03 0,32 0,05 -0,14 -0,15 1,00 Mali 0,95 0,94 0,69 0,59 -0,21 0,30 0,20 0,70 0,10 1,00 Niger 0,65 0,88 0,76 0,89 -0,28 0,37 0,22 0,81 -0,15 0,75 1,00 Nigeria 0,47 0,29 -0,16 0,05 -0,20 0,14 -0,36 -0,04 0,60 0,36 0,18 1,00 Senegal 0,32 0,39 0,93 0,63 0,37 -0,20 0,87 0,89 -0,30 0,45 0,52 -0,44 1,00 Sierra-Leone 0,33 0,18 0,50 0,51 0,35 -0,45 0,60 0,60 0,39 0,31 0,33 0,48 0,40 1,00 Togo 0,76 0,96 0,68 0,79 -0,28 0,51 0,06 0,70 0,10 0,87 0,94 0,36 0,38 0,29 1,00

(34)

Table 2: OCA Index for all ECOWAS Countries Source: own calculations.

OCA Index Benin

Burkina Faso Cape Verde Cote d'Ivoire Gambia,

The Ghana Guinea

Guinea-Bissau Liberia Mali Niger Nigeria Senegal

Sierra Leone Togo Benin 1,000 Burkina Faso 0,020 1,000 Cape Verde 0,040 0,107 1,000 Cote d'Ivoire 0,100 0,075 0,163 1,000 Gambia, The 0,016 0,045 0,040 0,057 1,000 Ghana 0,067 0,081 0,117 0,165 0,120 1,000 Guinea -0,046 0,006 0,039 0,099 0,039 0,042 1,000 Guinea-Bissau 0,046 0,107 0,031 0,177 0,067 0,098 0,025 1,000 Liberia 0,533 0,587 0,497 0,584 0,552 0,589 0,509 0,404 1,000 Mali 0,045 0,065 0,103 0,166 0,055 0,145 0,050 0,058 0,572 1,000 Niger 0,047 0,007 0,138 0,095 0,056 0,100 0,033 0,105 0,568 0,092 1,000 Nigeria 0,306 0,325 0,262 0,315 0,213 0,346 0,301 0,302 0,802 0,353 0,336 1,000 Senegal 0,018 0,035 0,060 0,118 0,005 0,108 0,024 0,083 0,611 0,043 0,088 0,289 1,000 Sierra Leone 0,173 0,185 0,158 0,232 0,199 0,216 0,140 0,228 0,556 0,261 0,202 0,388 0,222 1,000 Togo -0,011 -0,025 0,063 0,066 0,012 -0,005 0,008 0,089 0,576 0,070 0,028 0,261 0,021 0,149 1,000

(35)

Table 3: OCA Index Between Nigeria and the rest Source: own calculations.

Period Benin Burkina Faso Cape Verde Cote d'Ivoire Gambia,

The Ghana Guinea

Guinea-Bissau Liberia Mali Niger Senegal

Sierra

Leone Togo 2000-2007 0,379 0,392 0,329 0,352 0,251 0,374 0,368 0,380 1,041 0,438 0,395 0,355 0,478 0,315 2008-2012 0,104 0,136 0,091 0,200 0,121 0,240 0,132 0,176 0,029 0,137 0,182 0,081 0,186 0,056

Optimum currency area : evidence in the West African Region?