Sander van Ruitenbeek
10092994 Supervisor:
dr. Hudab Al-Kobaisi
Is the Gulf Cooperation Council an
Optimum Currency Area?
Acknowledgements
I would like to thank dr. Hudab Al-Kobaisi for her help and support. Her enthusiasm and inspiration have been a great help in writing this thesis, which was not always easy.
Table of contents
Acknowledgements 2
1. Preface 4
2. Current literature 6
3.1 The Optimum Currency Area theory 9
3.2 The Optimum Currency-Area Index 10
3.3 Estimating the OCA-Index 12
4.1 The Optimum Currency Area-Index for the GCC member countries 15
4.2 Possible candidates: Morocco and Yemen 16
5. Limitations of the empirical research in this thesis 17
6. Conclusion 18 Reference list 21 Appendix one 23 Appendix two 23 Appendix three 24 Appendix four 24 Appendix five 25 Appendix six 26
1. Preface
On may 25 1981 the Arabic states bordering the Arab Gulf established a political and economic union called the Cooperation Council for the Arab States of the Gulf also know as the Gulf Cooperating Council (GCC). The Gulf Cooperation Council consists of the Kingdom of Saudi Arabia, the Sultanate of Oman, the State of Qatar, the State of Kuwait, the United Arab Emirates and the State of Bahrain. Recently Jordan and Morocco have requested to join as well (GCC, 2011). Ever since the establishment of the CCG important progress has been made toward economic and financial integration including the adoption of a unified common external tariff (Fasano and Schaechter, 2003).
In 2001 The GCC countries decided to create a monetary union, which should have been effected from 2010 onwards. If the Gulf Cooperation Council monetary union were to be established it would be the second largest supranational monetary union in the world. According to Fasano and Schaechter (2003), the creation of a monetary union represents an evolution to the integration efforts that started with the creation of the GCC. The first practical step towards the creation of a monetary union was taken during 2002 and early 2003 when all the involved countries officially pegged their currencies to the U.S Dollar (although before that time most of the Gulf countries had unofficially pegged their currencies to the U.S. Dollar). A second big step was taken in 2007, when the GCC countries agreed on the convergence criteria, which are rather similar to the EU’s Maastricht Criteria (Takagi, 2011).
However, the actual realization of a monetary union seems far away. In 2006 Oman opted not to join by 2010, saying it did not want to meet the spending curbs agreed with its neighbors (Al-Awsat , 2007). In 2007 the plan of forming a monetary union suffered another setback when Kuwait decided to de-peg from the U.S. Dollar. The main reason that moved Kuwait to de-peg was the high inflation caused by the weakness of the dollar (Zhou, 2007). But perhaps the biggest setback came in 2009 when the UAE withdraw as well because of a disagreement with Saudi Arabia over the location of the planned GCC central bank (GCC, 2011) and due to the problems in the Euro Area.
Despite these setbacks, there is still good hope that a monetary union will be formed among the members of the GCC. In fact, in 2009 the foreign ministers of Bahrain, Kuwait, Qatar and Saudi Arabia signed the Monetary Union Agreement and created the accompanying Basic Statute of the Monetary Council, a precursor to the GCC central
bank (Takagi, 2011).The Monetary Union Agreement explicitly declares that “A
monetary union shall be established”.
Forming a monetary Union is important to the GCC because it will likely expand markets and help promote competitiveness and economic diversification, thus
facilitating integration with the global economy (Fasano and Schaechter, 2003). Also it is hoped that closer economic and monetary integration among the GCC members could become a catalyst for a broader economic cooperation among the Middle Eastern countries. Buiter (2008) also mentions that the pursuit of price stability within the GCC does require abandoning the US dollar peg, currently all countries but Bahrain maintain this peg. Abandoning this peg is crucial for price stability because, as oil-wealth driven demand is rising and the dollar is weakening further, inflation in the gulf countries will shoot up (Buiter 2008). Forming a monetary union would allow these countries to abandon the US dollar peg without jeopardizing their mutual bilateral exchange rate stability.
As will be discussed in the next section, there is quite some work available in the current economic literature on the economics of evolving the GCC into a monetary union. However the current literature is mostly of qualitative nature and most of the work consists of comparative studies between the GCC monetary union and the Eurozone. Up until now there is very little to no quantitative work done on estimating whether or not the GCC is an optimum currency area. In this thesis research will be done of a more qualitative empirical nature. With the use of the Optimum Currency-Area Index developed by Bayoumi and Eichengreen (1997) an answer will be constructed to the question: To what economical extend, based on the Optimum Currency-Area Index, do the countries that form the Cooperation Council for the Arab States of the Gulf also form an optimum currency area?
This thesis will be structured on the following manner: Section two will discuss some of the more prominent work currently available on the subject. Section three will elaborate on the Optimum Currency-Area Index developed by Bayoumi and Eichengreen (1997). Section four will discuss the results obtained with the OCA-index and compare them with the OCA-indices calculated by Bayoumi and Eichengreen for the Euro area.
Section five will discuss the possibility of the entrance of Yemen and Morocco. Finally results will be concluded in section six.
2. Current literature
Buiter (2008) wrote a paper on the subject, called: ’Economic, Political, and
Institutional Prerequisites for Monetary Union Among the Members of the Gulf Cooperation Council’’, which he had presented at the seminar “Preparing for GCC Currency Union’’. In order to form a conclusion on whether the GCC should form a
monetary union or not, Buiter uses both technical economic arguments, such as the stabilizing effect of a currency area, the extend to which output shocks are symmetric and financial stability considerations. As well as political economy considerations among which are the issues concerning the surrender of national sovereignty and the accountability of monetary policy makers to the legitimate political authorities.
Buiter (2008) argues that there is an economic case for the GCC Monetary Union, but not an overwhelming one. Especially because of the striking lack of economic
integration, free movements of goods, services and capital among the GCC members is still not achieved. The implementation of a common market would solve this problem. According to Buiter: “The economic case for monetary union is mainly based on the small
size of all GCC members other than Saudi Arabia, and their high degree of openness” (p.
579). Notice however that these days effort is being made to improve the economic integration, which would enforce the economic case for a monetary union. According to Gulf officials the customs union, which was officially launched in 2003, will be fully functioning by 2014 (Dokoupil, 2012). (The fact that it is still not fully functioning today is mainly caused by a disagreement on how to divide customs revenues).
From a political perspective Buiter’s arguments against a monetary union seem to be overwhelming. Buiter: “I believe that monetary union among GCC members will only
occur as part of a broad and broadly based movement towards far-reaching political integration. And there is little evidence of that as yet” (p. 611). Main reason for Buiter to
come to this conclusion is the absence of effective supranational political institutions, which implies that there could not be an effective political accountability of the GCC central bank.
Abu-Bader and Abu-Qarn (2006) take a different approach in examining the feasibility of a monetary union among the GCC members, in their paper “On the
Optimality of a GCC Monetary Union: Structural VAR, Common Trends, and Common Cycles Evidence”, Abu-Bader and Abu-Qarn explore the symmetry of external shocks that the
six countries are subject to. They also investigate the degree of synchronization of the long run economic activity and of the short run business cycles (2006). The method they use to do this is developed by Blanchard and Quah (1989), this is a procedure for
identifying both demand and supply shocks and is based on the aggregate demand and supply framework (AD-AS). Abu-Bader and Abu-Qarn also use the Johansen (1988) cointegration tests to find whether there exists a long run relationship of real GDP among member countries. Finally they test if the countries show a long run
synchronization of GDP1 and inflation rate.
Abu-Bader and Abu-Qarn (2006) conclude that the correlation of the shocks is significantly positive, indicating the low cost of forming a monetary union. However, looking at these shocks in more detail reveals that merely the demand shocks are symmetric. The supply shocks show no significant positive correlation and hence are asymmetric. They conclude that, since supply shocks are caused by external factors and not by domestic policy, the GCC lack significant positive correlation and thus there is no evidence of the readiness for the monetary union.
Secondly the authors conclude, based on the Johansen co-integration tests2, that
the economic activities of the GCC members are not linked and it will require effort to enhance their compatibility.
Finally Abu-Bader and Abu-Qarn argue that the GCC countries lack long run synchronization, since long run growth rates vary greatly. Also the economies of the UAE, Qatar and Kuwait are characterized by highly volatile economic activities
compared with the other GCC members. Looking at inflation the authors conclude that, however volatile, all GCC members have low inflation rates and hence exhibit a great deal of convergence.
1 Gross Domestic Product, i.e. the total market value of all goods and services produced
in a given year.
2 The Johansen test is a test for cointegration that allows for more than one
cointegrating relationship. A Cointegration is a statistical property of time series variables, two or more time series are cointegrated if they share a common drift.
Considering all three methods, the authors conclude that there is no evidence of the GCC countries being ready to form a monetary union.
In their paper Sturm and Siegfried (2005) look at some key economic and institutional aspects of the GCC monetary union. Looking at the intra-GCC trade they conclude that the GCC lacks economic integration. However according to the authors not only the legal and regulatory barriers are to be blamed for this lack in economic
integration, they also point out that similar factor endowments causes economic
integration to stay behind. On the other hand Sturm and Siegfried mention that the level of monetary convergence is remarkably high.
The authors also point out that in recent decades the GCC member countries have not had to adjust exchange rates to deal with asymmetric shocks, implying that the potential macroeconomic costs of forming a monetary union are limited. At the same time, the low level of economic integration combined with the relatively low exchange rate risk, lowers the economic benefits of sharing a currency. However financially speaking, Sturm and Siegfried see significant gains that would come with creating a monetary union. They argue that a single currency could promote the development of a more liquid and deeper financial market. Finally they mention that forming a monetary union would serve as a catalyst in the creation of a broad, multilateral, stabilizing macroeconomic framework.
Sturm and Siegfried (2005) suggest that in order to create a sustainable and credible monetary union, the GCC members should establish a supranational monetary institution. This institution will be accountable for the design of monetary policies that will encompass the entire GCC region instead of coordinating the individual national policies.
Fasano and Schaechter (2003) use a more qualitative approach in their work as well, which lead them to an overall positive view with respect to forming a monetary union among the GCC members. Their main conclusion is very similar to that of Sturm and Siegfried (2005), which is that forming a monetary union, if combined with the right macroeconomic policies, will improve efficiency of the financial markets, increase price transparency, bring down transaction costs and help allocate resources on an efficient manner.
3.1 The Optimum Currency Area theory
The Optimum Currency Area (OCA) theory explores the criteria as well as the costs and benefits of forming a monetary union. The foundations of the OCA theory as we know it today were laid over 50 years ago, by Robert Mundell (1961) in his seminal paper: “A Theory of Optimum Currency Areas.” In this paper Mundell (1961) attempted to figure out what the appropriate domain of a common currency area is. The theory that Mundell developed in this paper focuses on a subset of considerations relevant to the decision of forming a monetary union. His theory emphasizes on asymmetric shocks3,
labor mobility and the transactions value of a single currency (Bayoumi and
Eichengreen, 1997). Let us consider the following example in order to illustrate why these factors are important in the consideration of forming a monetary union. Suppose there are two countries, A and B. Assume that both A and B are initially at full
employment and that wages and prices cannot be reduced in the short run without creating unemployment. Now consider an asymmetric shock, more specifically consider a shift in demand for products of county A towards county B. This shift in demand will cause unemployment to rise in A and create inflationary pressure in B. If these countries both have their own floating currencies the price of the currency of B will go up and of A will go down. Which will make the products of B relatively more expensive and the products of A relatively less expensive. This in turn will cause demand for the products of country B to go down and for country A to go up. In other words this exchange rate mechanism will automatically bring down the unemployment in A and inflationary pressure in B so it will act as a stabilizer.
Now consider the scenario in which these two countries have a common
currency. The shift in demand for products of county A towards county B will again raise unemployment in A and cause inflationary pressure in B. However since both countries now share the same currency the exchange rate mechanism is unable to act as a
stabilizer. In other words having a common currency calls for a different adjustment mechanism to restore equilibrium. Mundell (1961) stresses labor mobility as
adjustment mechanism to restore equilibrium. To illustrate how a high degree of labor mobility will act as an adjustment mechanism lets consider the example once more. The
3 An asymmetric shock occurs when an economic supply or demand shock is different
from one region to another, or when the shocks do not move in tandem.
shift of demand for products of county A towards county B will raise unemployment in A and cause inflationary pressure in B, however because the factor labor is highly mobile unemployed workers of A will migrate to B. This in turn will bring down unemployment in A and inflationary pressure in B. Now both countries are back at their full
employment equilibrium, so there will be no need for either country to have their own currency.
This highly simplified example illustrates clearly how asymmetric shocks affect a common currency area and how equilibrium can be restored, without hurting either countries economy, when labor is highly mobile. Mundell (1961) also emphasizes price and wage flexibilities as adjustment mechanisms.
After McKinnon (1963) and Kenen (1969) contributed to the OCA theory by adding, respectively, the importance of the degree of openness and the notion of product diversification, the theory has advanced only minimally (Bayoumi and Eichengreen, 1997). And although the intuition behind the OCA theory is clear and straightforward, it remains difficult to move from theory to empirical work and policy analysis. This
difficulty remains because the OCA theory does not specify any sort of guidelines and needs to be operationalized before it can be used for empirical work and policy analysis (Bayoumi and Eichengreen, 1997).
Buiter (2008) points out two other key shortcomings of the OCA theory. First of all the theory is highly ‘old-Keynesian’, resulting in a great overestimation of the power of monetary policy to affect real economic performance. The second key shortcoming is that the OCA theory has been designed for a world without endogenous capital
movements. In such a world the exchange rate acts as a stabilizer, just like in the earlier example of the two countries. However the modern reality is that we live in a world with a lot of capital movements, these movements have a great impact on exchange rates. In reality exchange rates are far from acting as stabilizers, in fact they can amplify shocks and destabilize the real economy (Buiter, 2008).
3.2 The Optimum Currency-Area Index
In their paper “Ever Closer to Heaven? An Optimum- Currency-Area Index for
the core implications of the theory of optimum currency area. Bayoumi and Eichengreen use the OCA theory as the foundation of their index despite its shortcomings, discussed in section 3.1, because there is simply no better alternative. Or as the two authors put it: “Like it or not, the theory of optimum currency areas remains the workhorse for analyses
of monetary unification’’ (p. 761)
As discussed in the previous section the OCA theory focuses on characteristics that make stable exchange rates and monetary unification more or less desirable. The most important of these characteristics are asymmetric disturbances to output, trade linkages, the usefulness of money for transaction, the mobility of labor and the extent of automatic stabilizers (Bayoumi and Eichengreen, 1997). Although the last two
characteristics are clearly important for behavior across regions within a country, Bayoumi and Eichengreen found that in their sample period these characteristics have not played a significant role in responding to shocks that are felt asymmetrically across countries. For that reason they have focused their empirical work on capturing the first three characteristics. Bayoumi and Eichengreen call their operationalization of the OCA theory the Optimum Currency-Area Index and is stated as follows:
SD(eij) = α + ß1 SD(Δyi - Δyj) + ß2*DISSIMij + ß3*TRADEij + ß4*SIZEij
Where SD(eij) is the standard deviation of the change in the logarithm of the end-year bilateral exchange rate between countries i and j. SD(Δ yi- Δyj) is the standard deviation of the difference in the logarithm of real output between i and j. DISSIMij is the sum of the absolute differences in the shares of agricultural, mineral, and manufacturing trade in total merchandize trade. TRADEij is the mean of the ratio of bilateral exports to domestic GDP for the two countries. And SIZEij is the mean of the logarithm of the two countries GDP measured in dollars. In each case, the independent variables are
measured as averages over the sample period (Bayoumi and Eichengreen, 1997). At first glance this Index looks somewhat complicated, however a closer look at the chosen variables reveals that this OCA-Index is as intelligible as the OCA theory discussed in section 3.1. The first independent variable, SD(Δ yi- Δyj), is essentially a measure of asymmetric output disturbances. Countries with symmetric business cycles will have their national outputs moving together, thus will have a small value for this measure. Bayoumi and Eichengreen have chosen to measure the output disturbances as
the standard deviation (of the log) of real output disturbances. They have chosen to use the standard deviation because it is not per se the correlation of output shocks but rather the variance which determines whether or not a single monetary policy will be appropriate. For example if the output of country A is highly correlated to the output of country B, but the output of country A is much more volatile than that of country B, then the monetary policy of B will still be inappropriate for A (Alesina et al, 2002). The
second independent variable, DISSIMij, captures the dissimilarities of the export
structure. This variable is added because countries with similar compositions of exports have similar business cycles. The variable ‘’Trade’’, which measures bilateral trade, is added because a common currency is more beneficial for countries that intensively trade with each other. Finally, a variable which measures size is added because small countries should benefit the most from the unit of account, means of payment and store of value services provided by a common currency (Bayoumi and Eichengreen, 1997).
3.3 Estimating the OCA-Index
In this thesis a slightly alternated version of the OCA-Index is used, one in which the original DISSIM variable is replaced with a variable which measures the absolute dissimilarities in the rate of oil exports to total exports (were oil exports are defined as: the exports of fuels, mineral oils and products of their distillation; bituminous
substances; mineral waxes), this variable is called OILDIS. The DISSIM variable, which measures the absolute differences in the shares of agricultural, mineral and
manufacturing trade in total merchandize trade variable, is replaced due to data unavailability. The variable OILDIS is proper substitute because the involved countries have got export structures which are characterized by a high level of oil exports.
In order to determine to what extend the variables in the OCA-Index influence bilateral exchange rate movements the coefficients, the betas, need to be estimated. This can be done relatively easy by running an OLS regression on data sets of all variables. However, estimating the coefficients using data of the GCC makes no sense since all the GCC member countries but Kuwait have a hard currency peg with the US Dollar (Fasano and Schaechter, 2003). This implies that the standard deviation of the bilateral exchange rates between these countries is near zero. Attempting to estimate the coefficients of the independent variables would then result in all coefficients to be highly insignificant. One
might argue that trying to estimate the OCA-Index for countries that already maintain a highly stable bilateral exchange rate over an extensive period of time is useless. Why try to estimate whether or not countries could share a single currency if they have already proven to maintain a stable bilateral exchange rate? Because it is not the ability to
artificially maintain a stable bilateral exchange rate, that should matter in the decision of forming a monetary union or not. A far more important issue is whether or not the countries could maintain a stable exchange rate on a more natural manner, i.e. without central bank interference. So estimating the OCA-Index does make sense, because the OCA-Index shows if the economies are aligned enough for the countries to maintain a natural, fixed, bilateral exchange rate. However, as discussed earlier, estimating the coefficients is a problem due to the currency pegs.
One possible solution to this problem would be to simply use the coefficients calculated by Bayoumi and Eichengreen on the European countries, but it is very likely that the impact of the independent variables on the standard deviation of bilateral exchange rates differ from region to region (Stein and Allen, 1995). Using data of the European countries for the GCC member countries would than make no sense. For this reason the best possible solution is to estimate the coefficients of the OCA-Index using data of similar countries.
However finding similar countries that do not have a currency peg or similar countries with different anchor currencies is not an easy task. For instance, a country like Azerbaijan seems ideal at first glance with its relatively small size and high level of oil exports, but their recent introduction of a new currency causes the standard
deviation of bilateral exchange rates to be un-pure. Countries like Iran and Iraq on the other hand are facing a lot of political unrest causing the exchange rate movements to be unstable as well.
After examining a lot of possible countries, the following six countries were selected: Algeria, Egypt, Tunisia, Kuwait, Yemen and the United Arab Emirates. The regression was done using of twelve years of data (2000-2011) of these countries. The twelve years were divided into four sub periods and pairs were formed between all countries (except for the pair Kuwait and the UAE before the first de-pegged from the US Dollar). Due to some missing data and the relatively small amount of countries used, the total number of observations was limited to 56.
Because of the limited amount of observation the first regression, including all variables, resulted in very insignificant coefficients and a R-squared equal to 0.2676 (appendix six). In order to achieve more significant results given the number of observations, the variables Size and OILDIS are replaced by dummy variables
(DummyOil and DummySize). Where the variable dummyOil equals zero if the average rate of oil exports to total exports of a pair of countries in a given period is smaller than 71% and equals one otherwise. The dummy Size equals zero in a given period if the mean of the logarithm of GDP of a country pair is smaller than 10.95 in that same period and equals one otherwise. With these alternations the regression improved surprisingly well, all variables now differ from zero at the five percent significance level and the R-squared increased to 0.3415 (appendix one). Also the low correlation between the independent variables (appendix two) implies that multicollinearity is probably not a problem. The alternated OCA-Index with the estimated coefficients looks as follows (with p-values in parentheses):
SD(eij) = .0408353 - .5312304*SD(Δyi - Δyj) - 1.086203 *TRADEij (0.000) (0.004) (0.016)
- .0097369*DummyOil - .0101685*DummySize (0.040) (0.035)
The constant is, as expected, positive, which implies that if all other variables equal zero there will still be a standard deviation of the bilateral exchange rate. This makes perfect sense since the four independent variables could not possibly explain all bilateral exchange rate variance. However, the negative sign of the first independent variable is unreasonable according to the Optimum Currency Area theory, as discussed in section 3.1 (asymmetric output shocks could not lower bilateral exchange rate
movements). The negative coefficient of TRADE was expected, since a low SD of bilateral exchange rates positively affects bilateral trade (De Grauwe, 1988). The negative sings on the oil dummy could also be explained with common sense, if both countries have a high share of oil export to total export, than both have similar export structures and are both affected by shocks in the same sectors. Finally, the negative sign for the dummy Size implies that larger countries have a more stable bilateral exchange rate.
4.1 The Optimum Currency Area-Index for the GCC member countries
Bayoumi and Eichengreen (1997) divide their results on the Eurozone into three different groups. Group one consisted of prime candidates for the EMU (at that time the Eurozone was not formed yet), these candidates were Austria, Belgium, the Netherlands Ireland and Switzerland. These countries all had an OCA-Index vis-à-vis Germany below 0.025 (Bayoumi and Eichengreen, 1997). The countries that were converging to the EMU, such as Sweden, Italy, Portugal and Spain, formed group two. Group two countries had an average OCA-Index vis-à-vis Germany of 0.06 and were expected to converge to 0.05. Finally, group three consists of countries that showed little convergence and all had OCA-Indices greater than 0.07, this group consisted of countries as the United Kingdom, Denmark and France. Note that, surprisingly, France is placed in group three while Ireland is placed in group one. Bayoumi and Eichengreen (1997) argue that this caused by the high level of importance of the variables measuring bilateral trade and relative size in the OCA-Index. Thus the high index for France reflects its large and, for European standards, closed nature of its economy, while for Ireland the low index is explained by its small and relatively open economy.
The three groups of Bayoumi and Eichengreen (1997) will serve as a benchmark for the analysis of the GCC members. Countries with OCA-Index values below 0.025 will be considered as potentially good monetary union members, all country pairs with OCA-Indices above 0.025 will be considered as countries that are not yet ready to form a monetary union. Note that in the model estimated in this thesis, see section 3.3, the OCA-Index cannot fall below .0408353 (the constant) because all variables show a negative sing. This is probably due to the relative low amount of data used to estimate the coefficients, causing the results to be somewhat imprecise. Or perhaps, however less likely, the countries used to estimate the coefficients are all good candidates to form a monetary union together causing the maximum OCA-Index to be on the low side.
In order to estimate the Optimum Currency Area Index, the six GCC countries need to be divided into pairs. The following six pairs were randomly chosen: Bahrain and Oman, Oman and Saudi Arabia, Saudi Arabia and Qatar, Qatar and Kuwait, Kuwait and the UAE, UAE and Saudi Arabia. Appendix three summarizes the OCA-Indices for these country pairs for the period 2010-2011. Looking at the index numbers of these country pairs it becomes clear that all of them fit in the first group of Bayoumi and
Eichengreen’s (1997) three groups. All of them have index numbers well below 0.025, which would make them all prime candidates for forming a monetary union. Taking a closer look at the index numbers reveals that Oman and Saudi Arabia is the most eligible pair of countries to form a monetary union. This could be explained by the relatively high level of bilateral trade between these nations, the same holds for Bahrain and Oman. The fact that according to the index numbers Oman and Bahrain are so well suited to form a monetary union with the other GCC members is probably because they both do not rely completely on oil and gas exports but have got other export products as well (appendix four), causing a relatively high level of bilateral trade. It is highly likely that these other exports are the factors driving the bilateral trade in the GCC region, since one could imagine that bilateral trade in oil and gas products is near, if not completely, zero.
It is true that the United Arab Emirates has an even lower oil export to total export ratio than Oman and Bahrain and for that reason should have low index numbers as well (caused by a high level of bilateral trade). However this low ratio also implies that the dissimilarities of export structures are also much higher between the UAE and the other GCC members, causing the OCA-Index to increase. Whereas the dissimilarities of the export structure between Oman/Bahrain and the other GCC members is much lower, i.e. the dummy specifying the level of oil exports to total exports equals one for Oman and Bahrain and zero for the UAE, which causes the OCA-Index of the UAE to be higher.
4.2 Possible candidates: Morocco and Yemen
Recently Yemen and Morocco have requested a Gulf Cooperation Council membership (GCC, 2011), with the use of the OCA-Index estimated is section 3.3 it is a small step to calculate how well suited Yemen and Morocco are to join the GCC monetary union as well. The OCA-Index for these two countries will be calculated vis-à-vis Saudi Arabia, the latter has been selected to forms pairs with Yemen and Morocco because it is by far the largest economy in the GCC.
Estimating the OCA-Index, using exactly the same method as in the previous section, yields the following results: Yemen and Saudi Arabia have got an OCA-Index of -0,014833627; Morocco and Saudi Arabia have got an OCA-Index of 0,023180703. The
negative index for the pair Yemen and Saudi Arabia is irrational according to OCA theory, it implies that this pair of countries has got a negative standard deviation of bilateral exchange rate. It is probably caused by the fact that Yemen is the only country in the GCC that experienced a relatively large economic downturn in the period
2010/2011. This results in a high value for the first independent variable, which
measures asymmetric GDP shocks, causing the index to drop significantly. Looking at the OCA-Index for the pair Morocco and Saudi Arabia one could conclude that it is relatively high with respect to the other index numbers (appendix three), this was also expected. Morocco has not got a very large economy, but more importantly the distance between Morocco and the Gulf countries is quite large, causing bilateral exports to be low, which in turn causes the OCA-Index to be high. Also the dissimilarities in exports of Morocco and the GCC member countries is highly significant, oil exports in Morocco make up roughly three to four per cent of total exports which is significantly lower than the same figure for the Gulf countries, causing the Index to be higher.
All in All, Yemen seems to be a good candidate to join a possible monetary union, it has got a similar structure of exports as most of the other Gulf countries and it has relatively high trade linkages (appendix five). But with the OCA-Index as it is, with a negative coefficients for the variable measuring output shocks it is not possible to account for asymmetric shocks the way they should be accounted for, causing the verdict to be somewhat unreliable. But more on that in the following section.
The case for Morocco is a bit different, its export structure far from similar to that of the Gulf countries. Also the trade linkages between Morocco and the Gulf countries are low. And, despite the fact that in the model used in this thesis asymmetric output shocks have a positive effect on the OCA-Index, it has an OCA-Index vis-à-vis Saudi Arabia of approximately 0.023 which is only just in the first group of the benchmark discussed earlier. One could conclude that, if the sign of the first independent variable were positive, as it should be, Morocco probably would not be in the first group and thus would not make an ideal monetary union member for the Gulf countries.
5. Limitations of the empirical research in this thesis
Before discussing the conclusion it is important to elaborate on the limitations of the research method on which that conclusion is based. As discussed earlier, the
coefficient on the variable SD(Δyi - Δyj), which measures asymmetric output shocks, has got a negative sign. This could not be explained by the optimum currency area theory, in fact according to this theory it should have a positive sign (see section 3.1). The most logical explanation for this anomaly is the relatively little amount of data used to estimate the coefficients, especially when keeping in mind that Bayoumi and
Eichengreen (1997) were able to produce a positive sign using the same method applied to a larger set of data (n=210 vs n=56).
The second problem with the research done in this thesis is data unreliability, this holds especially true for the data on bilateral trade. Data on bilateral trade is extracted from the United Nations Comtrade database and although this is a reliable institution it has some problems with its data. For instance, the total amount of exports of country A to country B in a certain year is almost in every single case not equal to the total amount of imports of country B from country A, in the UN Comtrade database. Also not all data was available, causing the number of observations to be lower than it would have been if all data were available.
Thirdly, in order to obtain a significant result given the limited amount of data two variables had to be replaced with dummy variables. This causes the research to be less specific, i.e. instead of accounting for the marginal effect of an independent variable to the dependent variable a dummy variable can only take two values and thus is less specific.
Despite these problems the Optimum Currency Area-Index is still a very powerful, straightforward and easy to interpret method, as discussed earlier.
6. Conclusion
In this thesis an answer was sought for the question: To what economical extend, based on the Optimum Currency-Area Index, do the countries that form the Cooperation Council for the Arab States of the Gulf also form an optimum currency area?
From the current literature review the conclusion can be drawn that the GCC member countries are not yet ready to form a currency area, main reasons for that conclusion are the lack of economic integration, long run synchronization and the absence of effective supranational political institutions. Sturm and Siegfried (2005) also point out that the potential gains of forming a monetary union are limited due to similar
factor endowments, which causes intra-area trade to stay behind. On the other hand they argue that a single currency could promote the development of a more liquid and deeper financial market and would serve as a catalyst in the creation of a broad, multilateral, stabilizing macroeconomic framework.
Nearly all authors conclude that, although the GCC is not yet ready to form a monetary union, there is definitely a good chance of success for a sustainable and credible monetary union if combined with the right macroeconomic policies. Thus the creation of a supranational institution that will be accountable for the design of
monetary policies is a must.
The empirical research used in this thesis, based on the OCA-Index developed by Bayoumi and Eichengreen (1997) has not been applied to the Gulf Cooperation Council members before. One possible reason for that is the great deal of effort that needs to be done in order to estimate the coefficients of the index, since these coefficients need to be estimated on a different, but similar, group of countries. The results from this analysis are in favor of monetary unification. In fact the OCA-Indices of all country pairs
investigated in this thesis are well below the benchmark value.
Applying the same method to the possible new candidates for the GCC, Yemen and Morocco, shows that Yemen, with its similar export structure and relatively high trade linkages with the other Gulf countries, is probably a proper addition to the to be formed monetary union. The case for Morocco is less positive, its trade with the Gulf countries is hardly significant and its export structure is not nearly similar to that of most Gulf countries. With a relatively high index number, approximately equal to 0.023, it is still within the benchmark. However this index number is most likely
underestimated, given the fact that the sign of the first independent variable of the model created in this thesis is incorrect according to the OCA-theory. For that reason the conclusion is drawn that Morocco is not an ideal candidate for joining the to be formed monetary union.
Finally I would like to note that the Optimum Currency Area-Index has never been applied to this specific group of countries before and for that reason is subjected to certain flaws. Using a larger set of data to estimate the coefficients can solve most of these flaws. The Optimum Currency Area-Index is an easy to interpret and powerful tool, however more work needs to be done in order to make it easier to apply and to test its relevance further. This can be achieved by creating a more general form of the
OCA-Index, one in which the coefficients are standard and do not need to be estimated every time a different group of countries is being investigated.
Reference list
Abu-Bader, S., and Abu-Qarn, A.S., 2006, “On the Optimality of a GCC Monetary Union:
Structural VAR, Common Trends, and Common Cycles Evidence.” The World
Economy 31,(2008), 5, 612-630.
Al-Awsat, A., 2007, “Gulf Poised to Delay Monetary Union, Official Says.” Reuters, 26, June, 2013. http://www.aawsat.net/2007/10/article55260998
Alesina, A., S. Ardagna, R. Perotti, and F. Schiantarelli, 2002, “Fiscal Policy, Profits, and
Investment.” American Economic Review, 92, (2002), 571-589.
Bayoumi, T., and Eichengreen, B., 1997. “Ever Closer to Heaven? An Optimum-
Currency-Area Index for European Countries.” European Economic Review, 41(3-5): 761–
770.
Blanchard, O.J., Quah, D., 1989, “The Dynamic Effects of Aggregate Demands and Supply
Disturbances.” American Economic Review, 79, (1989), 655-673.
Buiter, W.H., 2008, ‘’Economic, Political, and Institutional Prerequisites for Monetary
Union Among the Members of the Gulf Cooperation Council.’’ Open Economies
Review (2008) 19: 579-612.
GCC, 2011, “Morocco and Jordan ask to join GCC.” The National, July, 1, 2013.
http://www.thenational.ae/news/world/middle-east/morocco-and-jordan-ask-to-join-gcc
Dokoupil, M., 2012, “With customs union, Gulf edges toward closer economic ties.” Reuters, June, 21, 2013. http://www.reuters.com/article/2012/07/23/us-gulf-integration-idUSBRE86M0F920120723
Fasano, U., Schaechter, A., 2003, ‘’Monetary Union Among Member Countries of the Gulf
De Grauwe, P., 1988, “Exchange rate variability and the slowdown in growth of
international trade.” Staff Papers – International Monetary Fund, (1988), 35(1), pp.63-84.
Johansen, E., 2002, “On the Common Currency for the GCC Countries.” IMF Policy Discussion Papers, (2002), 02/12.
Kenen, Peter (1969), "The Theory of Optimum Currency Areas: An Eclectic View," in Robert A. Mundell and Alexander K. Swoboda (eds), Monetary Problems of the International Economy, Chicago: University of Chicago Press, pp.41-60.
McKinnon, Ronald. 1963. “Optimum Currency Areas.” American Economic Review, 53(4): 509-517.
Mundell, Robert. 1961. “A Theory of Optimum Currency Areas.” American Economic Review, 51(4): 657-665.
Pantin, T., 2009, “UAE monetary union withdrawal to stay.” The National, June, 23, 2013. http://www.thenational.ae/business/economy/uae-monetary-union-w
ithdrawal-to-stay
Stein, L., and Allen, PR., 1995, “Fundamental Determinants of Exchange Rates.” Oxford: Oxford University Press.
Sturm, M., and Siegfried, N., 2005, “Regional Monetary Integration in the Member States
of the Gulf Cooperation Council.” Occasional Paper series 31, European Central
Bank.
Takagi, S., 2012, “Establishing Monetary Union in the Gulf Cooperation Council: What
Lessons for Regional Cooperation?” ADBI Working Paper Series, July, 8, 2013.
http://www.adbi.org/files/2012.10.19.wp390.monetary.union.gulf.cooperation.c ouncil.pdf
Zhou, W., 2007, “Kuwait unhooks dinar and dollar, signalling a possible trend.” Marketwatch, June, 24, 2013. http://www.marketwatch.com/story/kuwait-u nhooks-dinar-and-dollar-signaling-a-possible-trend
Appendix 1:
Linear regression Number of obs = 56 F( 4, 51) = 7.24 Prob > F = 0.0001 R-squared = 0.3415 Root MSE = .01604 --- | Robust
SDeij | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---+--- SDlogyilogyj | -.5312304 .1757883 -3.02 0.004 -.88414 -.1783208 TRADE | -1.086203 .4337137 -2.50 0.016 -1.95692 -.2154866 DummyOil | -.0097369 .0046219 -2.11 0.040 -.0190157 -.000458 DummySize| | -.0101685 .0046882 -2.17 0.035 -.0195804 -.0007566 _cons | .0408353 .0043922 9.30 0.000 .0320176 .0496531 Appendix 2: Correlation Matrix
| SDeij SDlogy~j TRADE DummyOil DummyS~e SDeij | 1.0000
SDlogyilogyj | -0.4414 1.0000
TRADE | -0.3242 0.2520 1.0000
DummyOil | -0.1869 0.0197 -0.1901 1.0000
Appendix 3:
OCA-Index for the years 2010/2011 Bahrain and Oman 0,006407027
Oman and Saudi Arabia 0,003989402 Saudi Arabia and Qatar 0,009543954 Qatar and Kuwait 0,020635445 Kuwait and UAE 0,016693836 UAE and Saudi Arabia 0,011625778
Appendix 4:
Share of oil* export in total export
UAE Kuwait Qatar Oman Bahrain Saudi Arabia
2008 0,504615052 0,902996983 0,909233739 0,774610376 0,663903096 0,895444304
2009 0,36952857 0,94617498 0,891055979 0,676430614 0,647178744 0,846149231
2010 0,373312933 0,92134742 0,900706685 0,695083848 0,717625504 0,857075991
2011 0,414109161 0,979825805 0,917513282 0,737073585 0,77435 0,870923188
* Used definition of oil export: Exports of fuels. Mineral oils and products of their distillation; bituminous substances; mineral waxes
Appendix 5:
The OCA-Indices and the products of the individual variables
OCA-INDEX Constant SD(Δyi - Δyj) TRADEij DummyOil DummySize
2010/2011
0,0408353 -0,5312304 -1,086203 -0,0097369 -0,0101685
Bahrain and Oman 0,006407027 0,0408353 -0,006606559 -0,018084814 -0,0097369 0
Oman and Saudi Arabia 0,003989402 0,0408353 -0,000955088 -0,01598541 -0,0097369 -0,0101685
Saudi Arabia and Qatar 0,009543954 0,0408353 -0,000241841 -0,011144105 -0,0097369 -0,0101685
Qatar and Kuwait 0,020635445 0,0408353 -0,004322335 -0,00614062 -0,0097369 0
Kuwait and UAE 0,016693836 0,0408353 -0,001887438 -0,012085526 0 -0,0101685
UAE and Saudi Arabia 0,011625778 0,0408353 -0,002193056 -0,016847966 0 -0,0101685
Yemen and Saudi Arabia -0,014833627 0,0408353 -0,033455931 -0,012476095 -0,0097369 0
Appendix 6:
Linear regression Number of obs = 56 F( 4, 51) = 5.44 Prob > F = 0.0010 R-squared = 0.2676 Root MSE = .01691 --- | Robust
SDeij | Coef. Std. Err. t P>|t| [95% Conf. Interval]
--- SDlogyilogyj | -.5491212 .1769568 -3.10 0.003 -.9043768 -.1938655 TRADE | -1.05982 .5157063 -2.06 0.045 -2.095144 -.0244966 SIZE | -.0078393 .0103859 -0.75 0.454 -.0286897 .0130112 OILEXDIS | .0071554 .008649 0.83 0.412 -.0102082 .024519 _cons | .1166659 .1131678 1.03 0.307 -.1105279 .3438597
