FEASIBILITY OF FORMING CURRENCY UNION IN ASEAN-5 COUNTRIES

Transcription

1 FEASIBILITY OF FORMING CURRENCY UNION IN ASEAN-5 COUNTRIES ABSTRACTS TELISA AULIA FALIANTY Research Laboratory of Economics, Department of Economics Faculty of Economics, University of Indonesia Indonesia Keywords : 1.Optimum Currency Area 4.Endogeneity of OCA 2.Currency Union 5.Intra-industry trade 3.OCA Index The formation of currency union is the last step in sequence of policy initiatives toward regional economic integration. Currency union is an area where single currency is circulated. Since the Asian crisis in year 1997, a popular view among economist and policymakers is that developing countries with open capital account have a bipolar solution to the exchange rate dilemma they face: a free float or hard peg. One type of a hard peg is currency union. The success of Euro has also created initiatives for ASEAN countries to move to a closer financial integration. The increasing trend to move to a closer financial integration is supported by the implementation of ASEAN Free Trade Area (AFTA). The Asian Financial Crises in 1997 that gave a devastating impact on East Asian countries has provided a strong imputes for East Asia to push ahead with various initiatives for monetary cooperation in order to avoid the next crisis in the region. Several forms of financial cooperation have been implemented among ASEAN and ASEAN+3 countries in the spirit of crisis prevention. In 1976, ASEAN countries agreed on the implementation of ASEAN Swap Arrangement (ASA) although the facility had failed to prevent major financial crisis such as in This prompted further initiative to launch Bilateral Swap Arrangement (BSA) between ASEAN countries and three countries in Asia (Japan, Korea, China) totaling in US$ 3 billion. The objective is to provide short term financial facility and complementing other existing facility such ASA and financial aid from the IMF. Seven BSA agreements have been signed so far and another seven arrangements are under negotiation process. The objective of the paper is to find whether ASEAN is an optimum currency area (OCA) using the framework of OCA. There are three phases of the OCA theory: the pioneering phase, the cost benefit phase, and the endogeneity of OCA hypothesis. The pioneering phase is the main OCA properties, also called prerequisites, characteristics, or criteria. The properties of OCA are price and wage flexibility, mobility of factors of production, financial market integration, the degree of economic openness, the diversification in production and consumption, similarities of inflation rates, and fiscal integration. Cost benefit phase assesses the cost and benefit of forming currency union by constructing OCA index. The endogeneity of OCA criteria is formulated in the sense of the Lucas Critique. Currency union affects the underlying OCA criteria in such a way that they are more likely to be satisfied ex post as both monetary and trade integration. The result of the study finds that Singapore, Malaysia, and Thailand are an Optimum Currency Area, whether Indonesia and Philippines are not. The result is consistent for the two phase of OCA : properties of OCA phase and cost and benefit phase. The test on endogeneity of OCA shows that there is a weak evidence of endogeneity of OCA properties for ASEAN countries. The paper also produces the policy recommendation for ASEAN countries to enhance the convergence between them to move to a closer integration and to form currency union in the last step of their integration.

2 I. Introduction 1.1 Background Construction of a Currency Union is the last policy steps taken towards regional integration. Currency Union defines as an area where a single currency is commonly used. Debate related with pros and cons in adopting a common currency in ASEAN (Association of South East Asian Nations) countries has arisen rapidly since 1997 Asian Crisis and 1999 Euros establishment. One of the important factors that triggered the desire to use Currency Union is the increasing trade integration among ASEAN countries through ASEAN Free Trade Area (AFTA). Based on the facts above, the backgrounds that supported research on Optimum Currency Area (OCA) in ASEAN countries are: the Asian economic crisis; the capital flow openness in ASEAN countries; the successful establishment of Euros; AFTA development; the increasing of Intra-ASEAN Trade as well as Intra-Industry Trade in ASEAN countries; East Asian countries international trade development; and a stable trend in ASEAN countries cross rate. I.2 Research Objectives This research objectives are: a. To study the potential of Currency Union determination in ASEAN 5. b. To recommend an alternative policy that ASEAN 5 countries might use in the preparation stage of Currency Union implementation. c. To determine group of countries in ASEAN 5 that has an optimum condition to construct Currency Union. I.3 Research Hypothesis a. Not all ASEAN 5 countries follow the optimum condition that is necessary for one Currency Union. Currently, Currency Union can only be adopted in several ASEAN 5 countries. b. OCA Index among ASEAN 5 countries are relatively larger compare to OCA Index among EMU countries. c. Indicators of Currency Union Asymmetric Shocks and Wage Flexibility properties are endogenous. II. Literature Review Generally, research on Currency Union is divided into three main streams: first, probability of Currency Union formation using Optimum Currency Area (OCA) properties. Second, cost and benefit calculation of Currency Union construction using OCA Index. Finally is research on endogeneity of OCA indicators. This dissertation is a comprehensive study of the three main research streams on Currency Union construction in ASEAN countries. 2

3 2.1 Optimum Currency Area (OCA) Properties The construction of Currency Union implies that members countries will loose its exchange rate as a tool to maintain equilibrium from asymmetric shocks. As a response several properties are needed to compensate the loss of exchange rate as a tool to maintain equilibrium. The properties needed for Currency Union construction are: a. The existence of positive shock correlation among countries (Mundell, 1961). b. Production factors mobility (Mundell, 1961). c. Wages and Price Flexibility (Friedman, 1963). d. Size and openness of the economy (McKinnon, 1963). e. Variety in production and consumption structure (Kenen, 1969). f. Degree of financial development (Stanoeva, 2001). g. Integration of factor productions market (Scitovsky and Ingram, 1962). h. Similar inflation rate (Fleming, 1971). i. Fiscal Integration (Kennen, 1969 and De Bandt & Mongelli, 2000). j. Political Integration (Cohen, 1993 and Mintz & Harberler, 1970). 2.2 Cost and Benefit of Currency Union Construction There are at least five benefits of Currency Union. First, benefit for micro economics efficiency development from an increasing use of money. The benefit specifically resulted from the use of a liquid single currency in a wider area as unit of account, medium of exchange, standard for deferred payment, and store of value. Second, benefit from an increase in macro economics stability (and growth) which is resulted from price stability development; more transparent and wider access to money market; an increase in external financing availability; a recover reputation for high inflation countries (by using low inflation countries as a benchmark); and reduced output and employment fluctuation among currency union members countries that occurs due to economics policies differences. Third, benefit from the positive external effect which was caused by the decreasing transactional cost as a result of a more widely use single currency; the existence of income from international seignoirage; a declining needs for foreign exchange reserves; and a more straightforward international coordination among countries. Fourth, benefit from protection. Memberships in currency union have made countries less vulnerable to currency manipulation by other countries outside its political block. However, it is still vulnerable to manipulation by the leader of its own political block. Fifth, will received privilege such as: an increase in access to foreign lending, and access to pooled reserves. On the other hand, there are also costs in implementing currency union. There are three costs of implementing currency union: First, the cost from deterioration in micro economics efficiency, for at least a while. Costs of exchanging to a new currency can varies and can include administrative cost, legal cost, psychological cost in using new numeraire, and other cost. Special circumstances especially applied when member countries incorrectly choose nominal exchange rate parity value. This will cause member country s competitiveness to be lower or higher compared to other member countries competitiveness. Second is a declining opportunity for countries to maintain their macroeconomic stability. OCA will decrease the amount of national government policy since it will later be the responsibility of Supra-national Central Bank. 3

4 Therefore, if one member country has a more rigid structure of nominal price and wages compare to other countries in the Currency Union, the inflation rate will be lower in the area and can cause a frictional unemployment. This situation will occurs until nominal rigidity decrease by structural reformation. Third is the loss from negative externalities. If one or more member countries in the Currency Union experienced budget deficit and have debt accumulation that can not be controlled, then Currency Union will suffer for externalities. Debt might be monetized, therefore will increase interest rate in the Currency Union. International confidence to Currency Union might be declining and thus speculation for Currency Union will arise. All members country in the union will suffer especially those who s previously have a stable currency. 2.3 Endogeneity of OCA Properties OCA properties might be an endogenous with respect to other variables. This is called the endogeneity of the Currency Union construction properties indicators. Therefore, OCA indicators are not an exogenous variable but it is a function of other variables determine in a system. Endogoneity of OCA might appear for these reasons: The existence of Common Currency can create relatively large push in trade among countries: trade openness and volume of intra-union trade will be larger under Common Currency than under Floating Exchange Rate Regime. As a result, the degree of trade openness and volume of intra-union trade will be endogenous with respect to exchange rate regime. An extensive application of Lucas Principle which states that economic structure can be influenced by changes in exchange rate regime argues that: a more credible common currency (compared with Floating Exchange rate Regime) will create more flexible wages in labor market and more flexible price in goods market. This does not mean that price and wages will be fully flexible when Pegged Exchange Rate Regime is implemented, however, this means that in such situation price and wages determination will adapt in order to avoid increases in unemployment (up to the level needed to maintain country s participation in the Currency Union). Frankel and Rose (1998) open an extensive debate on endogeneity of OCA properties. They argued that a tighter trade linkage can result in a more synchronize business cycle or increase in shock asymmetry. Frankel and Rose (1998) view, however, is different with the old view. The old view argued that international trade will increase specialization, thus, will create a more asymmetric shock. Formal model of OCA doesn t seem to give unique answer to this debate. Frankel and Rose (1998) argue that the importance of further analysis of international trade engagement should be emphasized, with differentiation on inter-industry trade and intra-industry trade. The logic behind the OCA properties endogeneity is that monetary integration will reduce trade cost which occur due to exchange rate fluctuation. Common currency among member countries will appear as a much more serious and durable commitment (McCallum, 1995). Common currency includes: increase in trade and financial integration; given used of Foreign Direct Investment (FDI) and long term development relationship; moreover an increase in political integration determination. This situation will then push business cycle among union countries to 4

5 synchronize. The paradigm shows that a group of countries using single currency might develop into an OCA ex-post, in such they did not developed into OCA ex-ante (Rose, 2000; Frankel and Rose, 1998, 2000). III. Research Methodology 3.1 Data This research used data from ASEAN-5 countries: Malaysia, Indonesia, Singapore, the Philippines; and Thailand. Source of the data varies from international organizations and each ASEAN-5 country publications (Table 1). Table 1. Data Source No. Publication Title Publisher 1 International Financial Statistics International Monetary Fund 2 Direction of Trade Statistics International Monetary Fund 3 Key Indicators Asian Development Bank 4 Trade Statistics: COMTRADE United Nations 5 Labor Statistics International Labor Organizations 6 Macroeconomic Indicators Each country Statistics Bureau Source: Authors summary. Data used in this research is an annual data sample, generally from However, for the purpose of OCA Index measurement and comparison this research used an annual data sample from Empirical Study on the Initial Stage of the OCA Theory: Pair wise Method Existence of positive shock correlation among member countries In order to measure shock correlation among member countries a dynamic analysis using Structural Vector Auto regression (SVAR) and a static analysis were used. a.1 Dynamic Analysis: SVAR SVAR is a development of Vector Auto regression (VAR) model which was developed by Christopher Sims (1980). Sims argued that if a simultaneous relation among variables exists, then all variables should be treated equally. Thus, there is no such thing as endogenous or exogenous variable. Just like any other model VAR also has its limitation, which does not depend much on the theory during model construction. Therefore it s often called unstructured model. Furthermore, on its development, SVAR was created as a bridge between economic theory and time-series analysis. The purpose of SVAR includes the determination of dynamic response of variables from various disturbances or shock in the economy. However, SVAR main purpose is to interpret fluctuation in the business cycle and to identify various effects from economic policy. To apply SVAR method there are several procedure that needed to be used: first, user must determine the stationerity of each variables (variables are either stationer I(0) or non-stationer I(1)). This step is important to determine whether Reduced Form on Level (2) or Level (3) is needed. The next step is to estimate VAR Reduced Form using Ordinary Least Squares (OLS) and ensures that the numbers of lag used will guarantee 5

6 the non existence of serial correlation on the residual. In order to have the proper numbers of lag based on VAR, statistical tests needs to be conducted, especially to avoid the exclusion restriction, which can caused over parameterized and thus loss in the degree of freedom which is important for estimation purposes. In order to examine the shock correlation among member country, decomposition of macroeconomic indicators fluctuation into and onto the shocks is needed. For this purposes, Bivariate Structural VAR (proposed by Blanchard and Quah (1989)) will later on be used. Macroeconomic Theory on Aggregate Demand (AD) and Aggregate Supply (AS) will be the base of Bivariate Structural VAR used. Aggregate demand curve has a negative slope in the short run and the long run, while aggregate supply curve has a positive slope in the short run and vertical in the long run. Shocks in a simple model show the shifting of aggregate demand and aggregate supply curve from equilibrium. Supply shocks related to the shifting of aggregate supply curve has a short run and long run effect on output and price. Demand shocks also have effect on output and price in the short run. However, it does not have long run effect on output and price because in the long run aggregate supply curve is vertical. Decomposition with Bivariate Structural VAR will allow us to identified supply shocks and demand shocks from output and price changes that can be observed. In this research, supply and demand disturbance identification will be done for each ASEAN-5 country. Later, after finding demand and supply shocks, a shock correlation will be determined. If the result shows a positive shock correlation, then symmetric shocks exist. While, if there is no positive shock correlation, then asymmetric shocks exist. This process is done to test research hypothesis that there are several pair of ASEAN member countries that have symmetric shocks, but there also those who have asymmetric shocks. a.2 Dynamic Analysis: Hodrick-Prescott Filter Hodrick-Prescott Filter is a smoothing methods which is commonly used by macroeconomics expert to get an estimated value of long term smooth variable with trend from a set of macroeconomic time-series data. This method was first used by Hodrick and Prescott in the early 1980s to analyze US business cycle after the World War II. Business Cycle symbolized with X is defined as:x = lny t lny. X variable represents deviation on GDP trend, which later on will describe the presence of business cycle. b. Static Analysis Static analysis shows that symmetric shocks can be measure with real shock measurement and monetary shock measurement as follows: t t = GDP GDP i j SDI ij SDln SDln t 1 t 1 GDPi GDPj Where: i = country i; j = country j; GDP = GDP in constant price; and SD ij = real shock. The amount of real shock on period t is measured as an absolute value of differences in the standard of deviation of log ration GDP in country i and j in period t and t-1. 6

7 The amount of monetary shock (SDM ij ) on period t is measured as an absolute value of differences in the standard of deviations of money supply log ratio in country i and j during period t and t-1. t t = M M i j SDMij SD ln SD ln t 1 t 1 M i M j Where: i = country i; j = country j; M = Money Supply (M 0 ); and SDM ij = monetary shock Size and Openness of the Economy The Size of Economy can be measure using the following formula: t GDPi SIZEij = Average t GDPj Here, SIZE ij = relative size of country i with respect to country j. Relative size of one country can be measure as the mean from GDP (in constant price) comparison between country i and country j. Thus, a smaller SIZE means a more stable exchange rate among countries. In addition, a more open economy, in terms of international trade, will create more country willingness to adopt Fixed Exchange Rate Regime. On the other hand, the Openness of Economy can be measure using the following formula: t, j t, i Exp + i Exp j TRij = Average t NGDPi Where: TR ij = the degree of openness of country i relative to country j which was resulted from export and import goods (on current prices) mean comparison between the two region; NGDP = GDP on current prices for period t. In addition to the static analysis above, openness of the economy will also be measured using methods suggested by Feldstein and Horioka (1980). This method is based on the regression model below: I S = β0 + β1 + ε Y Y Where: I Y is mean of Gross Domestic Investment Rate; and S is the mean of Gross Y National Saving Rate. According to Feldstein and Horioka (1980), if β 1 is equal to 1 then the economy can be state as a closed economy; in the contrary if β 1 is different from 1 then it is an open economy. 7

8 3.2.3 Degree of Financial Development According to economic literature, there are several ways to measure financial development. Here, two of them will be discussed. Most commonly used measurement is the ratio of monetary aggregate to GDP from each country. Therefore, the following formula can measure relative degree of financial development from ASEAN-5 member countries : t M i T NGDP i FD ij = Average T T M M i j + T T NGDPi NGDPj Where FD ij = degree of financial development relative to country i and country j; M = money supply (Mo)for country I and country j; NGDP = GDP (current price) from country I and country j. As an alternative, financial development can also be measured by the fraction of credit for local private companies which are divided by country s total domestic credit. t PC i FC = i Average t TDCi Where FC i = financial credit in country I; PC i = credit (on current price) given to local private companies in country i for t period; TDC i = total domestic credit (on current price) given in country i for t period. The degree of financial development can be measured by the ratio of total commercial bank asset to GDP Similarity in Inflation Rate Similarity in the inflation rate among countries can be observed by comparing inflation rate data among ASEAN-5 countries. Methods that can be used to determine similarity includes: calculating for inflation correlation among countries; and calculating for mean and standard of deviation for each member countries. 3.3 Empirical Study on the Initial Stage of the OCA Theory: Clustering Method Cluster analysis is one of the analyses in the multivariate techniques that aim to grouped research objects based on its characteristics. The result from clustering analysis shows that there is one cluster that shows homogeneity and thus shows differences with other cluster. With Cluster analysis, a group of research objects can be divided into several groups based on its similarity. The object used in this research is work field. A group of work field will be divided based on a particular characteristic it owns. Cluster Variate is a group of variables describing characteristics used to compare objects in cluster analysis. Cluster analysis is the only multivariate technique that does not estimate variate empirically, in stead it uses variate that has been specified by researcher. Cluster analysis focuses on objects comparison based on their variate (not comparing the variate it self).so the most important step on Cluster Analysis is the variate definition. Cluster Analysis is 8

9 commonly used for taxonomy on Biology, psychological classification based on personality, market segmentation analysis, individual groups, etc. 3.4 Optimum Currency Area (OCA) Index To avoid difficulties in interpreting the mixed results of currency union properties indicators, an index from Optimum Currency Area need to be made. According to Bayoumi and Eichengreen (1997), OCA Index can be defined as the prediction value of exchange rate variability. Variability of an exchange rate can be measure by using standard of deviation from changes of bilateral exchange rate between country i and j (in logarithm form), symbolized by SD(e ij ). A low value of OCA index means higher benefit (compare to cost) in constructing currency union. On the other hand, a high value of OCA index means lower benefit (compare to its cost) in constructing currency union. The reason behind these arguments is that countries with high asymmetric shocks and high trade linkage tend to have more stable exchange rate. Based on the arguments above, the following equation will be estimated: SD(e) ij = βo + β 1 SD (ΔY i -ΔY j ) + β 2 TRADE ij + β 3 SIZE ij + β 4 DISSIM ij Variables included in equation (i) are measured in the form of mean from all sample period. Sample period that is used to count OCA Index in this research range from 1983 to 1992, following Bayoumi and Eichengreen (1997) research period which has been used for the purpose of comparison. Note from equation (i): 1. SD(e) ij shows the standard of deviation from changes of bilateral exchange rate between country i and j (in logarithm form). 2. SD (ΔY i -ΔY j ) shows standard of deviation from differences between real output logarithm of country i and j. 3. TRADE ij shows the mean of bilateral export to Domestic GDP ratio between country i and country j. 4. SIZE ij shows the mean of GDP ratio between country i and country j (in logarithm form). 5. DISSIM ij shows the amount of absolute differences per share from agricultural sector, mineral sector, and manufacturing sector on trade between country i and j. DISSIM ij is the difference in export composition between two trading partners country. 3.5 Empirical Study on the Endogeneity of Optimum Currency Area Properties Asymmetric Shocks as an Endogenous Variable In order to test the endogeneity of asymmetric shocks there are three steps that need to be taken. The first step is shock identification, the second step is creating a time varying shock correlation and the third step is connecting the shock asymmetric with trade indicators (trade intensity and intra-industry trade) Changes in Wages as an Endogenous Variable Wages are treated as endogenous variable which is affected by Changes in monetary policy (changes in exchange rate regime), GDP (since contract wages will response to changes in GDP), 9

10 and. Unemployment Rate, since the contract wages will be revises as a response to changes in the level of unemployment. Regression on wages will be represent by the following model: [ ] [ ] [ ] logw logw = a+ b logy log Y + D* b logy logy + bu + b logw logw + ε t t 1 1 t k t k 1 2 t k t k 1 3 t k 4 t k t 1 t Where k is time lag operator, w is wages, Y is GDP, U is unemployment rate, and D is dummy variable (value 1 and 0) for the exchange rate regime. If the Dummy Variable is significant, then changes in exchange rate regime will affect changes in wages. Regression of the model will be conducted for each ASEAN-5 countries. The regression method being used is Generalized Least Squares (GLS) Method. In this case, GLS method will be used to correct violations in Ordinary Least Squares (OLS) assumptions. GLS method is actually OLS method for variables that have been transformed and thus satisfy the OLS assumptions. IV. Research Result and Analysis 4.1 Research Result and Analysis of Optimum Currency Area (OCA) Properties in ASEAN-5 Countries: Pair wise Method Based on the research result on OCA properties for calculation which is conducted by this research, a simple calculation was conducted by adding up OCA properties that has been completed by pairs of ASEAN-5 countries. Pairs of countries or country taken are they whom ranked in the big 3 or the big 4 on an OCA property. These pairs of countries or country will then be scored 1. Next, all score will be add up for each pairs of countries or country. The results in detail are in the appendix. Summary of the result for all ASEAN-5 countries is as follow: Table 2. Summary of Research Result on ASEAN-5 OCA Properties Accomplishment Sample Negara Indonesia Malaysia Singapura Filipina Thailand Indonesia Malaysia Singapura 3 20 Filipina 5 Thailand Sample Negara Indonesia Malaysia Singapura Filipina Thailand Indonesia Malaysia Singapura 3 18 Filipina 3 Thailand Source: Authors calculation Table 2 shows that for sample period , pair of countries that are mostly to fulfill OCA properties are: First Position: Malaysia-Thailand, Second Position: Singapore-Thailand, and Third Position : Singapore-Malaysia. While for sample period , pair of countries that 10

11 mostly fulfill OCA properties are: First Position: Malaysia-Thailand, Second Position: Singapore- Thailand, and Third Position: Singapore-Malaysia This results shows that currently there are three countries in ASEAN-5 that are tightly correlated and thus possible to construct a Currency Union: Singapore, Malaysia, and Thailand. Differences in the sample period which include crisis period ( ) and before crisis period ( ) did not change the results on countries that are optimum to construct Currency Union. 4.2 Research Result and Analysis of Optimum Currency Area (OCA) Properties in ASEAN-5 Countries: Clustering Method To avoid transitivity problem in pair wise method a multivariate method can be use. A suitable multivariate method for this research is clustering method. An often used clustering method is Hierarchical Clustering Method. Using this method, cluster output shows that there are two clusters (group) that exist: Group 1 : Indonesia and the Philippines Group 2 : Malaysia, Singapore, and Thailand. This result shows that clustering method is consistent with the result from pair wise Method. Thus this result strengthens the summary for countries that are optimum for currency union construction, which are Singapore, Malaysia, and Thailand. The result also strengthen the summary that Indonesia and the Philippines are located outside (or fall behind ) OCA cluster in ASEAN Calculation Result of OCA Index In this research the OCA Index calculation are using the same sample data period and the same estimation equation as Bayoumi and Eichengreen (1996). Sample data used by Bayoumi and Eichengreen (1996) range from for EMU countries. OCA Index calculation for ASEAN-5 countries is represented in the table below: Table 3. OCA Index of ASEAN-5 Countries Pair of Country OCA index Indonesia Malaysia Indonesia Singapura Indonesia Filipina Indonesia Thailand Malaysia Singapore Malaysia Filipina Malaysia Thailand Singapore Filipina Singapore Thailand Filipina Thailand Source: Authors calculation Pair of countries that are on the bottom three for OCA Index are: (1)Malaysia and Singapore (2)Singapore and Thailand, (3)Malaysia and Thailand. This result is consistent with the result from OCA properties research. Thus, Singapore, Malaysia and Thailand can be justified as countries 11

12 which are currently more prepared compare to other country in ASEAN-5 to construct Currency Union. Generally it can be concluded that the mean of OCA Index in ASEAN-5 countries are higher then the mean of OCA Index in EMU countries. This means that benefit from currency union construction in ASEAN-5 countries is not as big as benefit for its construction in EMU countries. Smaller OCA Index means higher benefit from the currency union construction (compared with the cost). This result is appropriate with the research hypothesis. Thus, the first research hypothesis can not be rejected. 4.4 Research Result on Endogeneity of the two OCA properties (Asymmetric Shocks and Changes in Wages) Asymmetric Shock as Endogenous Variable The research result here shows that endogeneity exists on the asymmetric shocks in ASEAN- 5 countries. A low and decreasing asymmetric shock, as one of the requirement for currency union, was significantly affected by the member s country trade patterns, especially intra industry trade. An increase in trade intensity will not necessarily affect the asymmetric shocks. However, an increase in trade intensity which is followed by an increase in intra-industry trade will create a decline in asymmetric shocks especially from the demand side Changes in Wages as Endogenous Variable Based on Lucas Principle (1976) it can be hypothesized that changes in monetary policy regime will change the degree of wage flexibility. The existence of Supply Shock and Demand Shock in the regional level will be stabilized through a counter-cyclical monetary policy or through exchange rate fluctuation. However, these can not be done in a currency union. Therefore, in currency union, the adjustment weight on shock in regional level will be burden on wages. In the model wages act as dependent variable which is affected by monetary policy regime. It can be hypothesized that wages will be more pro-cyclical to GDP in monetary policy regime with fixed exchange rate compare to monetary regime with floating exchange rate. There is an assumption stated that changes in monetary policy regime interact with GDP. This assumption is shown on the slope of Dummy Variable with respect to GDP parameter. The existence of time lag, which differentiates the left hand side and the right hand side of the wage changes model equations stated on Research Methodology section, will help eliminate simultaneous equations biasness. Changes on ASEAN-5 countries exchange rate regime are as follow: Table 4. Exchange Rate Regime changes in ASEAN-5 countries Country Indonesia : Fixed/Peg to US Floating : Managed floating 1997 : Multiple Currency Basket Malaysia Managed floating Fixed/peg Singapore Managed floating Managed float band widening Filipina Managed floating Floating Thailand : Peg to US Floating : Multiple currency basket 12

13 Source: Authors Summary In doing regression with changes in wage as endogenous variable, Dummy Variable is needed. In this research Dummy Variable is define as follow: 1 = Floating (Free) Exchange Rate Regime 0 = More Fixed Exchange Rate Regime Dummy Variable varies among countries and depend on changes in exchange rate regime in each country. Table 5: Dummy Variable Explanation for Each Member Countries Countries Research Dummy=0 Dummy=1 Period Indonesia Malaysia Singapore Philippines Thailand Research Period for Malaysia, the Philippines, and Thailand started from However, for Indonesia and Singapore, due to the data limitation problem, the research period was different. The main problem was the data on wages. Wages data being used was for wages in manufacturing sector. For Indonesia, the Central Bureau of Statistics (BPS) started to have wages data on manufacturing sector since Therefore, the research period for Indonesia started from 1989 until For Singapore, the available data for wages on manufacturing sector started from 1976 thus the research period for Singapore started from 1977 until Hypothesis on wages as an endogenous variable are as follow: Dummy coefficient is negative and significant for five ASEAN countries (Indonesia, Malaysia, Singapore, the Philippines, and Thailand). Changes on wages are treated as endogenous variable which affected by lag of wages it self, lag of unemployment rate, production index or GDP, and by dummy variable for exchange rate regime. The Dummy Variable assumed to interact with the production index or GDP. The regression result for changes on wages with GLS in each ASEAN-5 country is represented in the following table: Table 6 : Summary of Regression Result For ASEAN-5 Countries For Changes in Wages as endogenous variable Countries Constant Lag and Wages Lag and unemployment rate Production Index (GDP) Dummy R 2 Adjusted R 2 Indonesia (0.089) (0.060) (0.713) (0.378) (0.658) Malaysia (0.879) (0.216) (0.029) (0.753) (0.026)** Singapore (0.045) (0.964) (0.001) (0.054) (0.015)** Philippines

14 (0.003) (0.933) (0.093) (0.932) (0.098)* Thailand (0.404) (0.153) (0.100) (0.036) (0.407) Notes : ( ) shows prob-value or significance * significant on α=10% ** significant on α=5% The regression result shows that dummy coefficient for Indonesia, Malaysia, Singapore, the Philippines, and Thailand are negative. The results are consistent with the early hypothesis on the sign of the dummy coefficient. However, only dummy coefficient for Malaysia, Singapore, and Thailand that are significant at α = 10 per cent. For Indonesia and Thailand, the dummy coefficient are not significant at α = 10 per cent. A negative dummy coefficient sign means that wages become more procyclical on a more fixed exchange rate regime. Therefore, a negative dummy coefficient supports the endogeneity of wages hypothesis. In this research it is realized that the regression result contains several limitations which are: The research period limitation, which mainly caused by the use of several variables with low frequency data range, such as data on wages and the unemployment rate. Limited sample period for Dummy Variable value 1 (Indonesia, Singapore, the Philippines, and Thailand) and Dummy Variable value 0 (Malaysia). As a result, the effects of different regime to wages were not too clear or conclusive. A less fit model to explain wage behavior shown by the low value of R square. However, in spite of these limitations, the regression results have at least describe and early picture for endogeneity in wages. Thus, the result might provide a weak evidence for the existence of monetary policy regime influence on wages. So, this can be a scientific consideration for the analysis of currency union construction in ASEAN. The regression model that was used is based on equation Changes in wages as endogenous variable was influenced by the wage lag itself, lag from unemployment rate, production index or GDP, and from dummy variable for exchange rate regime. The dummy variable is assumed to interact with the production index or GDP. Regression result for changes in wages with GLS in each ASEAN-5 member countries can be seen in the following table: The regression result shows that dummy coefficient for Indonesia, Malaysia, Singapore, Philippines, and Thailand is negative. This result is similar to the initial hypothesis regarding the signs for dummy coefficients. However, only dummy coefficients for Malaysia, Singapore, and Philippines that are significant at α=10%. For Indonesia and Thailand, their dummy coefficients are not significant at α=10%. A negative dummy coefficient V. Conclusion, Policy Recommendation, Limitations, and Future Research 5.1 Conclusion This research concludes that: 1. Based on study on currency union construction properties and calculation of OCA Index, there are three countries on ASEAN-5 that currently had a bigger chance to construct 14

15 currency union: Singapore, Malaysia, and Thailand. The conclusions on these three optimum countries are equal in both Pairwise and Multivariate methods. 2. The mean of OCA Index for ASEAN-5 countries are bigger compared to the mean of OCA Index for EMU countries. This results shows that ASEAN-5 currently were not too ready to construct a currency union. However, the value of OCA Index in ASEAN-3: Singapore, Malaysia, and Thailand are already low. Thus, compared to the other countries on ASEAN- 5, these countries are currently more ready to construct currency union. 3. In the research on ASEAN-5 countries, evidence on endogeneity of asymmetric shocks, as currency union construction property, was found. Increase in intra-industry trade can decline the asymmetric shocks among member countries. Research also found weak evidence for endogeneity in wages, as one of the currency union construction property. Wages tend to be more pro-cyclical in a more fixed exchange rate regime. 4. With the evidence found on endogeneity of the asymmetric shocks, then an expectation on currency union construction on ASEAN-5 countries arise. ASEAN-5 countries need to do several co ordinations on their economic policy, in order to increase the economics convergence. A convergence economics can create a more synchronized business cycle which in the end will be able to reduced asymmetric shocks. One of the policies that are needed to be taken is to support the intra-industry trade increase among ASEAN-5 member countries. 5. Convergence pattern in ASEAN-5 countries with respect to leader candidate country, which is Singapore, varies; depend on different benchmark country as well as different research period. However, the most convergence country is Malaysia. On the contrary, country that is most divergence is the Philippines. These results strengthen an estimation that stated Indonesia and the Philippines as fall behind countries, which are not ready to enter the currency union. 5.2 Policy Recommendation 1. Based on the research result on this dissertation, which shows that currently ASEAN-5 member countries are not completely optimum for currency union construction and does not fulfilled all OCA properties, policies to support for currency union construction in ASEAN-5 countries are needed. If ASEAN-5 countries are serious and want to take further step on constructing a currency union, then policies leading to OCA properties fulfillment are needed. Such policies are: (1) Employment Policy More efforts need to be taken to increase wage flexibility in ASEAN-5 countries. Wage flexibility is a very important instrument to confront asymmetric shocks. More efforts also needed to be taken to increase labor mobility and capital mobility among ASEAN-5 countries. High labor and capital mobility are essentially needed to tackle asymmetric shocks problem and to reduce the need for nominal exchange rate adjustment. An important note is needed to be taken for the employment policy in this dissertation, which is the speed of the policy application varies in each country and depend on economic and social condition of each country. For example: With its current situation, Indonesia will find more obstacles that can slow down the applications of the recommended employment policy. However, in the long run (based on OCA theory), if 15

16 a currency union constructed then wages shall be more flexible, and labor as well as capital shall have high mobility. (2) Monetary and Financial Policy Developing each country financial market, for examples by developing obligation and stock market which then will create financial market structure that is not only dominate by the banking sector. Maintain the level of inflation rate so that it is not much different among countries, especially with country where its inflation rate (low and stable inflation rate) is being used as benchmark for others. In ASEAN-5, country with low and stable inflation rate is Singapore. Effort to control inflation is especially important for Indonesia and the Philippines, which structurally have higher inflation rate compare to other countries. Increased External Sustainability in order to maintain the level of reserves and Net Foreign Assets. Members of ASEAN countries have to maintain reserves and NFA on the relatively same level. Countries with relatively similar NFA and reserves position can apply the same monetary policy to reduce the negative impact of asymmetric shock. (3) International Trade Policy An increase in cooperation on trade among ASEAN-5 countries is needed to be made, especially intra-industry trade, in order to decrease the asymmetric shock. Increase relative economic openness among ASEAN-5 countries. Relative economic openness among member countries is closely related with increase in intra-industry trade in ASEAN. Efforts to increase intra-industry trade includes: increasing Foreign Direct Investment (FDI) as a main channel to strengthen intra-industry, since FDI can strengthen production network; increase the use of technology, increase trade liberalization, make a standardization on port and custom, improve structure of logistic cost and supply chain to strengthen the production network; and increase competitiveness among member countries (for example in improving port infrastructure and other infrastructure relevant with international trade), as well as increase entrepreneurship as one of the requirement for innovation process. (4) Fiscal Policy Strengthen the fiscal transfer mechanism as an instrument to reduced exchange rate adjustment need. Among countries that shared one fiscal transfer system, transfer of fund for other member that faced negative shocks is allowed. Each country member has to maintain the level of their budget deficit since one the OCA property is low level of budget deficits. (5) Real Sector Policy Increase in investment flexibility on each country. Increases in investment flexibility are needed to facilitate increase on intra-industry trade. Increase in investment flexibility is closely related with efforts to increase FDI. Increase in investment flexibility can be done by reducing the amount of restriction (barrier) on investment procedure on each country, so that it will attract investors to invest their money. 2. Singapore, Malaysia, and Thailand can immediately prepare them selves to construct a currency union among them since they are relatively ready, in terms of their economy, 16

17 compare to other countries in ASEAN. As for Indonesia and the Philippines, if they want to joint Singapore, Malaysia, and Thailand, they have to give a lot of efforts to harmonize their economy with the other three. By improving their economic performance the two countries can be expected to reduce their OCA Index and increase the benefit from Optimum Currency Area. The economic performance improvement here is consistent with recommendation number Learning from EMU successful experience in constructing currency union, if ASEAN countries want to have the same result, then they need to prepare the followings: Found a candidate for leader country that has a well economic performance and strong bargaining position. ASEAN have to focus more on economic factor as the base of the cooperation and excavate internal factors that can strengthen the needs for the economic cooperation. Strengthen ASEAN institution by having a more often regular meeting which is not only attended by the ministers but also by the head of the state. In addition, ASEAN also need to established several institutions such as: ASEAN institute, ASEAN Central Bank, and ASEAN council; that can help the preparation for currency union implementation. Do an advanced study on all the criteria that need to be fulfill by currency union members. These criteria can be partly learned from the convergence criteria in EMU. However, ASEAN needs its own criteria for its trademark. ASEAN need to realize that currency union construction does bring its own cost and benefit. There is a risk in constructing a currency union. Although constructing currency union in ASEAN is still an aspiration, early step can be made by strengthening Free Trade Area (FTA) and increase intra-industry trade. Increase in dominant intraindustry trade will automatically create the needs for currency union. This then will create a stronger improvement towards currency union construction. 5.3 Limitation of the Study and Future Research The research on this dissertation contains several limitations: 1. Sample country in this research is limited to ASEAN-5. For further study it is expected that the sample country can be broaden into East Asian countries, with the inclusion of highly influential countries such Japan, China, and Korea. The roles of Japan, China, and Korea on the construction of single currency in East Asia are very important. Thus East Asian countries roles in ASEAN integration development are also important. It is showed by the role of Japan, China and Korea on CMI, as well as the East Asian trade pattern. Therefore further study on the possibility of Single Currency determination on East Asian will be an interesting discussion in the future. 2. The data used in the research is yearly data. The reason for using this data has already been mention on the Research Methodology section. However, using yearly data contain weakness. This weakness appears since yearly data is a period s average data from monthly or quarterly 17

18 data, or, an end of period s data every December or else the end of a quarter. As a result the usage of yearly data will affect the research result. However, for comparison, conclusion from literature study on Optimum Currency Area in ASEAN, especially countries that fulfill currency union construction properties, in general shows the same results with this research. Although minor differences still exist. 3. The model that was used in this research is still conventional panel model. Further study can be made with the application of dynamic panel model such as general Method of Moment (GMM). 4. Convergence study with half-life model to see when or how long it will takes for ASEAN-5 countries to converge; or else when or how long it will take for ASEAN-5 countries to converge and reach specific convergence criteria (such as convergence criteria in Maastricht Treaty). REFERENCES AFTA Reader. Vol. 4. Jakarta : ASEAN Secretariat, September Ananta, Aris and Arifin, Evi N, ed. International Migration in Southeast Asia. Institute of Southeast Asian Studies, Babetskii, Ian. EU Enlargement and Endogeneity of Some OCA Criteria : Evidence From the CEEs, Czech National Bank Research, Januari Babetskii, Ian, Boone, Laurence, and Maurel, Mathilde. Exchange Rate Regimes and Shocks Asymmetry: The Case of the Accession Countries, in Journal of Comparative Economics 32, Barrel, Ray, and Choy, Amanda. Economic Integration and Opennes in Europe and East Asia, paper to be presented at the ASEM Symposium on Multilateral and Regional Economic Relations, Tokyo, 25 Maret Bayoumi, Tamim and Mauro, Paulo. The Suitability of ASEAN for a Regional Currency Arrangements, IMF Working Paper, Desember Bayoumi, Tamim and Eichengreen, Barry. Ever Closer to Heaven? An Optimum-Currency-Area Index for European Countries, European Economic Review 41, Blanchard, O. and Quah, D. Dynamic Effects of Aggregate Demand and Supply, American Economic Review 79, Boone, Laurence. Symmetry and Asymmetry of Supply and Demand Shocks in The European Union : A Dynamic Analysis, CEPII Working Paper, No.97/03. Cecchetti, S.G. Legal Structure, Financial Structure, and the Monetary Policy-Transmission Mechanism, The Monetary Transmission Process, Recent Development and Lessons for Europe, edited by the Deutsche Bundesbank,