A Monetary Union in East Asia: What does the Common Cycles Approach Tell?

Transcription

1 A Moneary Union in Eas Asia: Wha does he Common Cycles Approach Tell? Sao, K. 1, D. Allen and Z.Y. Zhang * 1 Faculy of Economics, Yokohama Naional Universiy, Japan School of Accouning, Finance and Economics, Edih Cowan Universiy, Ausralia sao@ynu.ac.jp (Sao) and zhaoyong.zhang@ecu.edu.au (Zhang) Keywords: Moneary union; Coinegraion; Common feaure business cycle; Eas Asia EXTENDED ABSTRACT There is conroversy abou wheher a moneary union is feasible in he Eas Asian region. Amongs he crieria for esablishing a moneary union, mos of he exising sudies focus on he symmeric issue of fundamenal shocks and he exen of correlaions by applying he Blanchard and Quah (1989) srucural vecor auoregression (VAR) echnique, which includes he firsdifferenced variables in he model and examines only bilaeral relaionships. When forming a moneary union, he member counries need o renounce heir moneary policy auonomy. If shocks o respecive economies are symmeric, he cos of relinquishing he discreionary moneary policy is likely o be ouweighed by he benefis of esablishing a common currency. In conras, if shocks are asymmeric, i will be more cosly o give up he auonomous moneary policy and, hence, o esablish a moneary union. However, he shock symmery does no necessarily mean he comovemens of he real oupu variables (common business cycles) beween he counries concerned are presen. The presen paper employs he Johansen (1988) coinegraion es o check he long-run comovemens of real oupus and also conducs he Vahid and Engle (1993) common feaure es o deec he shor-erm common business cycles. The novely of his paper is wofold. Firs, whereas he srucural VAR approach considers shocks o correlaion bilaerally, we use a mulivariae VAR framework o allow for he relaionships wihin a specific group of counries. Second, we employ he coinegraion echnique o examine boh he longrun and he shor-run dynamics of linkage of he real variables o deermine he suiabiliy and coss of forming a moneary union in he region. We include in his sudy Japan and he Unied Saes in addiion o he nine Eas Asian economies including hree Asian NIEs (Korea, Taiwan and Hong Kong), ASEAN5 (Singapore, Malaysia, Indonesia, Thailand and he Philippines), and Mainland China o invesigae he co-movemens of he real oupu variables spanning a period from 1978Q1 o 006Q4. We firs perform he Johansen (1988) coinegraion es o check wheher a group of counries concerned shares common sochasic rend(s), and hen, conduc he Vahid and Engle (1993) common feaure es o explore he exisence of shor-erm common business cycles among he counries if he real oupu series are coinegraed. This will allow he assessmen of how he oupu variables among hese counries inerac in boh he shor-erm and long-erm wihin a mulivariae framework. The coinegraion resuls and he common feaure ess will ensure business cycle synchronizaion across he economies and deermine he effeciveness of a common moneary policy o a union-wide shock. Based on a mulivariae framework, his sudy will provide imporan implicaions for cos effeciveness in esablishing a regional moneary union. The resuls of he Vahid and Engle (1993) common feaure es indicae ha here exiss a linearly independen common feaure vecor, i.e., a linear independen combinaion of real oupu growh which has no correlaion wih he relevan pas. This leads o he conclusion ha besides he coinegraing relaionship of real oupus, he concerned counries share common shor-erm business cycles. In paricular, he resuls for he presence of one or wo common feaure vecor(s) indicae he exisence of synchronized common business cycles in wo groups: he firs one includes he Asian NIEs ha consiss of Korea, Hong Kong and Singapore, and he second one he ASEAN5 plus Japan. These economies would be he good candidaes for a moneary union as hey share boh long-run oupu co-movemens as well as synchronized common business cycles. However, he resuls show ha he Unied Saes and China are no suiable for a membership of he grouped economies, as do he ASEAN5 and Japan. 1007

2 1. INTRODUCTION The feasibiliy of forming a moneary union and esablishing a regional (common) moneary uni in Eas Asia has been lively debaed in indusrial, governmenal and academic arena. There are a number of sudies on his issue which examine some of he precondiions for forming a moneary union. These include (i) he openness and goods marke inegraion; (ii) facor marke inegraion; (iii) similariy in economic srucure and symmery in (real) shocks; (iv) financial marke inegraion; and (v) policy coordinaion. Sudies of he symmeric naure of fundamenal (real) shocks emphasize ha shocks o he candidae economies mus be symmeric so ha he coss of relinquishing he discreionary moneary policy when forming a moneary union are likely o be ouweighed by he benefis of esablishing a common currency. In conras, if shocks are asymmeric, i will be more cosly o give up he auonomous moneary policy and, hence, o esablish a moneary union. Mos of he exising sudies use he Blanchard and Quah (1989) srucural vecor auoregression (VAR) echnique o idenify fundamenal shocks and conduc he correlaion analysis o deermine he symmery of he shocks (see Bayoumi and Eichengreen 1994, Bayoumi, Eichengree and Mauro 000, Zhang, Sao and McAleer 004, and Zhang and Sao 007). However, his approach has several weaknesses. Firs, a correlaion analysis of shocks idenified by he srucural VAR is inherenly a bivariae mehod, whereas i is obvious ha an analysis of OCA mus be based on a muli-counry framework. More specifically, he bivariae approach reveals jus counry-o-counry correlaions wihou aking ino accoun he relaionship wih oher possible parner counries. In his paper we will adop a muli-counry framework o assess he common business cycles. Second, i is imporan o disinguish beween he shor- and he long-run dynamics in consideraion of a moneary union (Beine, Candelon and Hecq 000). If real oupu variables are no coinegraed among he counries concerned, each oupu variable wonders randomly over ime, which leads o a differen growh pah for each counry. Since nominal exchange rae changes as well as oher macroeconomic policies have only ransiory effecs o sabilize he economy, he long-run economic divergence among he economies can be an obsacle for forming a moneary union. A commonaliy of business cycle phase is also an imminen concern o he counries paricipaing in he moneary union, even hough business cycle shocks end o originae from he demand side and o be relaively shor-lived. As long as hey face a well-synchronized business cycle, i will be less cosly for he counries o renounce he moneary policy auonomy. The srucural VAR approach generally employs a bivariae VAR model including he firs-differenced variables and imposes a resricion o allow only supply shock o affec he real oupu series in he long run, a resul of which would be a lack of disincion beween sochasic rends and common cycles. In his paper we adop an esimaion procedure based on he Johansen (1988) coinegraion es and he Vahid and Engle (1993) common feaure es, which would be more appropriae o disinguish beween he shor- and he long-run dynamics. Recenly, Cheung and Yuen (005) use he coinegraion echnique and he common business cycles approach o assess he level of inegraion among he hree Greaer China economies (he Mainland, Hong Kong and Taiwan). Sao and Zhang (006) apply a similar approach o he 9 Eas Asian economies plus Japan and he Unied Saes o assess 54 pairs of counries for he coinegraion and common cycle ess and o explore wheher hese economies share common business cycles. However, Sao and Zhang (006) sill employ a bivariae VAR of real oupu series for possible pairs of counries, namely, a counryo-counry analysis. In conras o he previous sudies, he novely of his paper is wo-hold. Firs, he presen paper invesigaes wheher a group of Eas Asian counries share common business cycles as well as a common sochasic rend of real oupus by using a mulivariae VAR framework. Second, we aemp o invesigae sixy groups of counries o deec possible regional currency areas, which is far more comprehensive han he previous lieraure. We include in his sudy Japan and he Unied Saes in addiion o he nine Eas Asian economies including hree Asian NIEs (Korea, Taiwan and Hong Kong), ASEAN5 (Singapore, Malaysia, Indonesia, Thailand and he Philippines), and Mainland China o invesigae he co-movemens of he real oupu variables spanning a period from 1978Q1 o 006Q4. We firs perform he Johansen (1988) coinegraion es o check wheher a group of counries concerned share common sochasic rend(s), and hen, conduc he Vahid and Engle (1993) common feaure es o explore he exisence of shor-erm common business cycles among he counries if he real oupu series are coinegraed. This will allow for he assessmen of how he oupu variables among hese counries inerac in boh he shor-erm and long-erm wihin a mulivariae framework. The coinegraion resuls and he common feaure ess will ensure business cycle 1008

3 synchronizaion across he economies and deermine he effeciveness of a common moneary policy o a union-wide shock. Based on a mulivariae framework, his sudy will provide imporan implicaions for cos effeciveness in esablishing a regional moneary union. The res of he paper is srucured as follows. Secion discusses he analyical framework. Secion 3 describes he daa and presens he resuls of empirical examinaion. Finally, secion 4 concludes he paper.. ANALYTICAL FRAMEWORK To invesigae he exisence of a sable linear seady-sae relaionship beween he variables, we need o conduc uni-roo and coinegraion ess o deermine wheher a ime-series variable is saionary, and wheher here is a long-run (coinegraing) relaionship beween he variables if all he variables are found non-saionary (i.e., have uni roos). If all variables sudied are I(1) non-saionary, we proceed o he Johansen maximum likelihood (ML) mehod (Johansen, 1988; Johansen and Juselius, 1990) o es wheher hese variables are coinegraed. The Johansen approach allows esing of he long run relaionship beween variables in a mulivariae framework, and considers he error srucure of he daa processes and he ineracions in he deerminaion of he relevan economic variables. If he variables are coinegraed, he real oupu series have a common sochasic rend, implying synchronous long-run movemens of he real oupus among he economies. The Johansen coinegraion echnique is based on he maximum likelihood esimaion of he vecor error-correcion model. Le X be an ( n 1) vecor of I(1) variables. Then, i is possible o specify he following unresriced VAR involving up o k-lags of X : X = A1 X Ak X k + ε, (1) where A i is an ( n n ) marix of parameers and ε are a Gaussian error erm. The above equaion can be expressed as a vecor error-correcion form: Δ X = ΠX 1 + Γ1 ΔX Γk 1ΔX k+ 1 + ε () k k where Π = A i =1 i I n and Γi = i = i + 1 Ai. Our major ineres is in he marix Π = α β, where α represens he speed of adjusmen o disequilibrium, while β is a marix of long-run coefficiens such ha he erm β X represens 1 up o ( n 1) coinegraion relaionship in he mulivariae model. Thus, he es for coinegraion is o deermine how many r ( n 1) coinegraion vecors exis in β, which amouns o esing wheher Π = α β has reduced rank. We use in his paper he race saisic by which he null hypohesis ha here are a mos r coinegraing vecors ( 0 r n ) can be esed: n = + ln( 1 i r 1 λ = T ˆ λ ), (3) race where λˆ i s are he ( n r) smalles squared canonical correlaions of X wih respec o 1 Δ X correced for lagged differences and T is he sample size used for esimaion. Rejecion of his hypohesis suggess he exisence of he maximum r coinegraing vecors. To avoid he finie-sample bias oward over-rejecion of he no coinegraion hypohesis, we employ he small sample correcion of he race es provided by CATS in RATS, Version, which is based on Johansen (000, 00). Once a coinegraing relaionship is found in real oupus among he economies, our nex ineres is o es wheher hey share common shor-erm oupu flucuaions. If he economies face an asynchronous business cycle, a common moneary policy would be ineffecive in responding o he asymmeric shocks across he economies. I would herefore be very cosly for hese economies o form a moneary union. In conras, if he economies share common business cycles, a common moneary policy would be desirable. Thus, i will be less cosly for hem o renounce an auonomous moneary policy and o form a moneary union among he economies. The es for a common business cycle will feaure a es for a serial correlaion common feaure in he difference of he variables. Engle and Kozicki (1993) devise he es for a serial correlaion common feaure for saionary variables based on he wo-sage leas square regression using he lagged value of all variables as he insrumens. If here exiss a linear combinaion of variables ha eliminaes all correlaion wih he pas and is no correlaed wih he pas informaion se, we hen conclude ha he se of variables shares a common cycle. Vahid and Engle (1993) exend he Engle and Kozicki es o propose he es procedure for common serial correlaion cycles given he i 1009

4 presence of coinegraion. I is o find a sample canonical correlaion beween Δ X and W ( p) ( ΔX 1,..., ΔX p, Z 1 ) where Z is 1 he error-correcion erm. Under he null hypohesis ha here exis a leas s linearly independen common feaure vecors, he es saisic is given by: s j = 1 j ) C( p, s) = ( T p 1) ln(1 λ, (4) where λ (j = 1,, s) is he sh smalles squared j canonical correlaions beween Δ X and W ( p). Under he null hypohesis, he saisic C ( p, s) has a χ disribuion wih ( s + snp + sr sn ) degrees of freedom, where n is he number of endogenous variables, p is he lag order of he differenced variables in he error-correcion model, and r is he number of coinegraing vecors.. 3. EMPIRICAL RESULTS 3.1 The Daa We use he quarerly series of real GDP for coinegraion analysis of real oupus among he concerned economies. All daa are expressed in naural logarihms and seasonally adjused using he Census X-1 mehod. The eleven economies aken up in his paper include he hree Asian NIEs (Korea, Taiwan and Hong Kong), ASEAN5 (Singapore, Malaysia, Indonesia, Thailand and he Philippines), China, Japan and he Unied Saes. The sample period covers 1978Q1 hrough 006Q4 for all economies. The daa on real GDP is obained from Abeysinghe and Gulasekaran (004), he CEIC Asia Daabase, and he web sies of he Japanese METI (Minisry of Economy, Trade and Indusry) and he FRB (Federal Reserve Board). We firs check he saionariy of he real GDP series using he ADF (Augmened Dickey-Fuller) es and DFGLS es (Dickey-Fuller es wih GLS derending) proposed by Ellio, Rohenberg and Sock (1996). The es saisics show ha for he levels of all he series, he null hypohesis ha a uni roo exiss canno be rejeced. The uni roo ess of he firs difference of he variables rejec he null hypohesis. These findings sugges ha each series conains one uni roo and is hus I(1) process (he resuls are no repored in he paper bu available upon reques). Then we proceed o he coinegraion analysis in he nex secion. 3.. Resuls of Coinegraion Tess We employ he Johansen coinegraion es o es wheher he I(1) oupu series for he economies concerned move ogeher in he long-run. We firs esimae vecor auoregressions (VAR) wih four lags and hen conduc he lag reducion ess based on he χ -disribued Likelihood Raio (LR) ess. Once he common lag lengh is deermined, we perform he es for reduced rank. Doornik e al. (1999) propose o include he impulse dummies o allow for he ouliers so ha he VAR residuals may be normally disribued. We include impulse dummies in a VAR model since he Johansen coinegraion es is very sensiive o he assumpion ha errors are independenly normal. The inclusion of impulse dummy variables is necessary in his sudy given ha our sample includes he currency crisis period in Following Doornik e al. (1999), we iniially make a preliminary VAR esimaion wihou dummies o invesigae he hisogram of he sandardized residuals. Then, in he presence of exreme ouliers, we include he impulse dummies and reconduc he VAR esimaion. In paricular, we aemped much closer inspecion of he esimaed residuals han he visual invesigaion of he residual graph. If we deec large residuals wih absolue values larger han he hreshold (.576), we included impulse dummies and re-esimae a VAR. The dummies are included when he following economies are in he VAR (he daes of dummies are lised in parenhesis): Korea (1980Q4, 1988Q1), Taiwan (1999Q, 003Q), Hong Kong (1984Q4-1985Q, 003Q), Singapore (1985Q, 003Q), Malaysia (1984Q4, 1998Q1), Indonesia (1993Q1, 1998Q1, 1998Q), Thailand (1980Q, 1997Q4), he Philippines (1979Q4, 1984Q3, 1987Q4), China (1986Q1, 1989Q1, 1989Q3), Japan (1993Q) and he Unied Saes (1981Q, 1981Q4, 198Q1). In conducing he VAR esimaion, we ried o scruinize he exisence of large residuals carefully and o include as small a number of dummies as possible. Hence, all he dummies above were no used a he same ime for esimaion. The resuls for coinegraion rank ess are repored in Table 1. Due o he space limiaion, we repor if he coinegraing relaionship exiss or no only (he deailed es resuls are available upon reques). The resuls show ha, wih he exclusion of Japan and he Unied Saes, he hypohesis of no coinegraing relaionships is rejeced in 8 ou of 0 cases (groups). Among ohers, a leas one is observed in mos Eas Asian groups and Norheas Asia/NIEs groups. In conras, no coinegraing relaionships are found in ASEAN groups. When Japan is included, he resuls indicae ha here are 7 ou of 0 cases (groups) ha show a leas one coinegraing relaionship and mosly wihin he 1010

5 Japan and ASEAN groups. In conras, he inclusion of Japan reduces he number of coinegraing relaionship in Norheas Asia/NIEs groups. If he Unied Saes is included, he number of coinegraing relaionship improves subsanially: I is found ha 15 ou of 0 groups share he long-run oupu co-movemens. Thus, he inclusion of Japan or he Unied Saes in a group considerably changes he paern and he number of possible combinaions of counries ha exhibi coinegraing relaionship of real oupus. We have also conduced he significance es of he coinegraing vecors, wih he null hypohesis se as zero for he coefficien. If we canno rejec he zero resricion on one of β coefficiens, i means ha he variable corresponding o zero-resriced coefficiens will be excluded from he coinegraing relaionship. The resuls are no repored bu are available upon reques. I is found ha only 6 ou of 5 groups rejec he zeroresricions in all coefficiens of heir coinegraing vecors; oherwise, a leas one coefficien in longrun coinegraing vecors canno rejec he zeroresricion. As a resul, he es for common business cycles is conduced for jus 6 groups of counries Resuls of Common Business Cycle Tess Once he long-erm real oupu co-movemen is found, he nex sep is o examine wheher he group of counries concerned share he synchronous business cycles. We conduc he Vahid and Engle (1993) procedure o es for he common serial correlaion of he business cycles in he presence of coinegraing relaionship. The es resuls are repored in Table. If he null hypohesis of, say, s = 1 is no rejeced, i means ha here exiss a linearly independen common feaure vecor, i.e., we have found a linear independen combinaion of real oupu growh which has no correlaion wih he relevan pas. Then, we can say ha besides he coinegraing relaionship of real oupus, he concerned counries share common shor-erm business cycles. As i can be seen from Table, he null hypohesis of one common feaure vecor (s = 1) is no rejeced even a he 10 percen significance level bu he null of wo or more common feaure vecors is rejeced for he group A09 (Korea, Hong Kong and Singapore) and B19 (Japan, Singapore, Malaysia and Thailand). The null hypoheses of one and wo common feaure vecors (s = 1 and s = ) are no rejeced a he 10 percen level for he group B15 (Japan, Singapore, Malaysia, Thailand and he Philippines) and B17 (Japan, Singapore, Malaysia, Indonesia and Thailand). However, he group of A10 (Taiwan, Hong Kong and Singapore) rejeced he null hypohesis ha here exiss common feaure vecor(s) a leas a he 10 percen significance level. This holds rue even if he Unied Saes is included in he group A10, as he group C10 rejecs he hypohesis ha here exiss one common feaure vecor. 4. CONCLUSION This sudy adops a mulivariae coinegraion approach o es for synchronized common business cycles in order o assess he feasibiliy of a moneary union in he Eas Asian region. The resuls sugges ha some Asian NIEs (Korea, Hong Kong and Singapore) and ASEAN5 plus Japan should be a poenial candidae group o form a moneary union as hey share boh long-run oupu co-movemens as well as synchronous business cycles. Ineresingly, he Unied Saes and Mainland China are excluded from he candidaes groups. Furhermore, ASEAN counries canno form a candidae group unless Japan is included as a member counry, which has imporan implicaions for he role of Japan o esablish a regional moneary union. However, here are some limiaions o our analysis which will be addressed in he fuure work. Firs, we need o conduc a robus es o check he sensiiviy of he resuls for he formaion of oher possible groups as we have only sixy groups of counries. Second, our analysis migh have no fully refleced he impacs of he regional inegraion process and he emerging Chinese economy. Third, our analysis focuses on he synchronizaion of business cycles among he regional counries. However, counries may have differen iniial response o shocks and have sared o reac symmerically o shocks wih one or wo period lag(s). Such asymmeric response a he iniial sage and he synchronous reacion in laer periods are no couned in he curren sudy, bu reward a consideraion in furher analysis of a regional moneary union. 5. ACKNOWLEDGMENTS The auhors would like o hank Colin McKenzie, Kenaro Kawasaki, Mica Panić, Masanaga Kumakura, Tze Haw Chan and Craig Parsons for heir insighful commens on he earlier version of he paper. The second and he hird auhors wish o acknowledge he Faculy Sraegic Research Gran of Edih Cowan Universiy. The sudy is financially suppored by Japan Sociey for he Promoion of Science hrough he Gran-in-Aid for Scienific Research (B),

6 6. REFERENCES Abeysinghe, T. and R. Gulasekaran, 004, Quarerly Real GDP Esimaes for China and ASEAN4 wih a Forecas Evaluaion, Journal of Forecasing, 3, pp Bayoumi, T. and B. Eichengreen, 1994, One Money or Many? Analyzing he Prospecs for Moneary Unificaion in Various Pars of he World, Princeon Sudies in Inernaional Finance, No.16. Bayoumi, T., B. Eichengreen and P. Mauro, 000, On Regional Moneary Arrangemens for ASEAN, Journal of he Japanese and Inernaional Economies, 14, pp Beine, M., B. Candelon and A. Hecq, 000, Assessing a Perfec European Opimum Currency Area: A Common Cycles Approach, Empirica, 7, pp Blanchard, O.J. and D. Quah, 1989, The Dynamic Effecs of Aggregae Demand and Supply Disurbances, American Economic Review, 79, pp Cheung, Y.-W. and J. Yuen, 005, The Suiabiliy of a Greaer China Currency Union, Pacific Economic Review, 10(1), pp De Grauwe, P., 005, Economics of Moneary Union, Sixh Ediion, Oxford and New York: Oxford Universiy Press. Dennis, J.G., H. Hansen, S. Johansen and K. Juselius, 006, CATS in RATS, Coinegraion Analysis of Time Series, Version, Evanson, Illinois: Esima. Doornik, J.A., D.F. Hendry and B. Nielsen, 1999, Inference in Coinegraed Models: UK M1 Revisied, in M. McAleer and L. Oxley, eds., Pracical Issues in Coinegraion Analysis, Oxford: Blackwell Publishers. Ellio, G., T. J. Rohenberg and J. H, Sock, 1996, Efficien Tess for an Auoregressive Uni Roo, Economerica, 64, pp Engle, R.F. and S. Kozicki, 1993, Tesing for Common Feaures, Journal of Business and Economic Saisics, 11, pp Johansen, S., 1988, Saisical Analysis of Coinegraion Vecors, Journal of Economic Dynamics and Conrol, 1, pp Johansen, S., 000, A Barle Correcion Facor for Tess on he Coinegraing Relaions, Economeric Theory, 16, pp Johansen, S. 00, A Small Sample Correcion of he Tes for Coinegraing Rank in he Vecor Auoregressive Model, Economerica, 70, pp Johansen, S. and K. Juselius, 1990, Maximum Likelihood Esimaion and Inference on Coinegraion Wih Applicaion o he Demand for Money, Oxford Bullein of Economics and Saisics, 5, pp Kawai, M., 1987, Opimum Currency Areas, in J. Eawell, M. Milgae, and P. Newman, eds., The New Palgrave: A Dicionary of Economics, London: Macmillan Press, Ld., pp Kawasaki, K. and E. Ogawa, 006, Wha Should he Weighs of he Three Major Currencies be in a Common Currency Baske in Eas Asia? Asian Economic Journal, 0(1), pp Maddala, G.S. and I.-M. Kim, 1998, Uni Roos, Coinegraion, and Srucural Change, Cambridge: Cambridge Universiy Press. Ogawa, E. and K. Kawasaki, 006, Adoping a Common Currency Baske Arrangemen ino he ASEAN plus Three, paper presened a he 17h NBER Annual Eas Asian Seminar on Economics, Hawaii, June -4. Rubin, J. and N. Thygesen, 1996, Moneary Union and he Ousiders: A Coinegraion Codependence Analysis of Business Cycles in Europe, Economie Appliquée, XLIX(3), pp Sao, K., 007, Eas Asian Moneary Inegraion: An Empirical Assessmen of he Opimum Currency Area Crieria, in H. Misuo, ed., New Developmens of he Exchange Rae Regimes in Developing Counries, UK: Palgrave Macmillan, pp Sao, K. and Z.Y. Zhang, 006, Real Oupu Comovemens in Eas Asia: Any Evidence for a Moneary Union? The World Economy, 9(1), pp Tavlas, G.S., 1993, The New Theory of Opimum Currency Areas, The World Economy, 16(6), pp Vahid, F. and R.F. Engle, 1993, Common Trends and Common Cycles, Journal of Applied Economerics, 8, pp Zhang, Z.Y. and K. Sao, 007, Whiher A Currency Union in Greaer China? Open Economies Review, forhcoming. Zhang, Z.Y., K. Sao and M. McAleer, 004, Is a Moneary Union Feasible for Eas Asia? Applied Economics, 36(10), pp

7 Table 1. Tess for Coinegraion Rank Groups (A) Eas Asia (B) Including (C) Including (Counry Name) Only Japan USA a) Eas Asia (EA) (01) EA9 Yes Yes Yes (0) EA8 Yes Yes Yes b) Norheas Asia/NIEs (03) NIEs4 (Kr,Tw,Hk,Sg) + Ch No No Yes (04) NIEs3 (Kr,Tw,Hk) + Ch Yes No Yes (05) Greaer China (Tw,HK,Ch) Yes No Yes (06) NIEs4 (Kr,Tw,Hk,Sg) Yes No Yes (07) NIEs3 (Kr,Tw,Hk) Yes No Yes (08) Kr, Tw, Sg No No Yes (09) Kr, Hk, Sg Yes Yes Yes (10) Tw, Hk, Sg Yes No Yes Table 1. Tess for Coinegraion Rank (con'd) Groups (A) Eas Asia (B) Including (C) Including (Counry Name) Only Japan USA c) ASEAN (11) ASEAN5 + Ch No No Yes (1) ASEAN5 (Sg,My,Id,Th,Ph) No Yes Yes (13) ASEAN4 (My,Id,Th,Ph) No Yes Yes (14) Sg, Id, Th, Ph No No No (15) Sg, My, Th, Ph No Yes Yes (16) Sg, My, Id, Ph No No No (17) Sg, My, Id, Th No Yes No (18) Sg, My, Id No No No (19) Sg, My, Th No Yes No (0) My, Id, Th No Yes Yes Table : Resuls of Common Feaure Tess Group (Counry Name) Null Hypohesis Degrees of Freedom Squared Canonical Correlaion Common Feaure Sa. C(p,s ) Criical Value (10% level) Criical Value (5% level) (A09) Kr, Hk, Sg s = s = * s = * (A10) Tw, Hk, Sg s = # s = * s = * (B15) Jp, Sg, My, Th, Ph s = s = s = # s = * s = * (B17) Jp, Sg, My, Id, Th s = s = s = # s = * s = * (B19) Jp, Sg, My, Th s = s = # s = * s = * (C10) US, Tw, Hk, Sg s = # s = * s = * s = * Noe: "s" denoes he number of common feaure vecors. An aserisk (*) and a sharp (#) denoe ha he null hypohesis is rejeced a he 5 percen and 10 percen significance level, respecively. 1013