IUJ Research Institute Working Paper 03-2 International Development Series

Transcription

1 IUJ Research Insiue Working Paper 03-2 Inernaional Developmen Series Exchange Rae Volailiy and Expors from Eas Asian Counries o Japan and he U.S. SaangJoon Baak * Arif Al-Mahmood Souksavanh Vixahep Inernaional Univ. of Japan Yamao-machi Minami Uonuma-gun Niigaa, Japan sbaak@iuj.ac.jp Phone: Fax: Royal Holloway Universiy of London Egham, Surrey UK Asian Developmen Bank Lao Residen Mission Laos ABSTRACT The purpose of his paper is o invesigae he impac of exchange rae volailiy on expors in four Eas Asian counries (Hong Kong, Souh Korea, Singapore, and Thailand). Specifically, his paper aims o deermine wheher he bilaeral real exchange rae volailiy beween an Eas Asian counry and is rading parner negaively affecs he expors of he Eas Asian counry. Considering he dominan roles of he U.S. and Japan as rading parners of hose Eas Asian counries, his paper focuses on he monhly expor volumes of Eas Asian counries o he U.S. and Japan for he period from 990 o 200. Excep for he case of Hong Kong s expors o Japan, coinegraion ess and esimaions of error correcion models indicae exchange rae volailiy has negaive impacs on expors eiher in he shor run or in he long-run, or boh. On he oher hand, manufacuring producion indices of imporing counries and depreciaion of real bilaeral exchange raes urn ou, in general, o have posiive effecs on he expors of he Eas Asian counries examined. JEL Classificaion: C2, F, F3 Keywords: Exchange rae volailiy, Expor, Eas Asia, Coinegraion, Error correcion model * The corresponding auhor

2 . Inroducion The purpose of his paper is o invesigae he impac of exchange rae volailiy on expors in four Eas Asian counries (Hong Kong, Souh Korea, Singapore, and Thailand) where expors have been he major engine of economic growh. Even hough hese Eas Asian counries has implemened expor-oriened economic policies since heir early sages of developmen, he impac of exchange rae volailiy on expors, which has araced he ineress of researchers and policy makers since early 970s, has rarely been sudied for hose counries. One major reason of his neglec may be rooed in he facs ha he exchange raes of Eas Asian currencies agains he U.S. dollar had been relaively sable since hey had been implicily pegged o he U.S. dollar unil he 997 financial crisis and ha he U.S. has been he main expor marke of mos Eas Asian counries. As Eas Asian counries has moved o a floaing exchange rae sysem since he 997 financial crisis and as he share of non-us markes in he expors of Eas Asian counries has been increasing, however, he issue of he impac of exchange rae volailiy on expors has gained some aenions of researchers and policy makers in Eas Asia. Even hough Eas Asian counries manage o sabilize heir currency values agains he U.S. dollar, i does no mean heir currency values are also sable agains he currencies of oher major rading parners of heirs han he U.S., such as he Japanese yen. In fac, since he Japanese yen floaed more freely agains he U.S. dollar while oher Eas Asian currencies were effecively pegged o he U.S. dollar, he exchange raes of Eas Asian currencies agains he Japanese yen were relaively unsable. Therefore, he impac of exchange rae volailiy on expors is an issue o a counry whose exchange rae agains he U.S. dollar is managed quie sable bu where he U.S. is no he only dominan rading parner. As Kawai and Takagi (200) poin ou, his issue is especially imporan o he pos-crisis Eas Asia which is seeking a new regional exchange rae regime because See Tables - hrough -4. Excep for he case of Thailand, he share of he U.S. in he expors of he Eas Asian counries examined in his paper has declined for he las 5 years. According o 2

3 he impac of exchange rae volailiy on expors should be examined o consruc an opimal exchange rae scheme. Also, i should be considered by he local moneary auhoriies when hey se he weighs of differen foreign currencies in he deerminaion of he values of heir own currencies. Agains his background, he presen paper aims o deermine wheher he bilaeral exchange rae volailiy beween an Eas Asian counry and is rading parner negaively affecs he expors of he Eas Asian counry. Considering he dominan roles of he U.S. and Japan as rading parners of Eas Asian counries, his paper focuses he expors from Eas Asian counries o he U.S. and o Japan for he period from 990 o 200. In fac, numerous sudies, heoreically and empirically, have aemped o find he naure of he relaionship beween exchange rae volailiy and expors, and repored boh posiive and negaive relaionships. In addiion, some have repored no significan relaionship. 2 However, as menioned earlier, his issue was rarely invesigaed regarding he expors of Eas Asian counries. I should be, however, noed ha his paper disinguishes from he previous lieraure no only by he geographical focus of he sudy bu also by he empirical research ools. Mos of empirical research examining ime series daa in his area invesigaed quarerly daa of he oal expor volumes of one or more counries. 3 In conras, he presen paper invesigaes monhly daa of bilaeral expor volumes, which is expeced o yield more accurae resuls as Baum, Caglayan and Ozkan (200) and Dell Ariccia (999) argue. Following Arize, Osang and Sloje (2000), Chowdhury (993) and Hassan and Tufe (998) among ohers, he long-run relaionship beween exchange rae volailiy and expors is examined by performing coinegraion ess, and he shor run impacs of exchange rae volailiy on expors is examined by esimaing error- Nakamura and Masuzaki (997) and Takagi (996), Japan share in he expors of he whole Eas Asian counries became close o he U.S. in mid 990s. 2 See De Grauwe (988) and Secru and Uppal (2000, Ch. 6) for heoreical examples showing an ambiguous relaionship and Bacchea and Wincoop (2000) for a heoreical example showing no relaionship. The empirical researches of Arize, Osang and Sloje (2000), Chowdhurry (993), Kim and Lee (996) and Peree and Seinherr (989) repor a negaive relaionship while Bahmani-Oskooee and Payesech (993) and Hooper and Kohlhagen (978) repor an insignifican relaionship. 3 See, for example, Arize, Osang and Sloje (999, 2000), Chowdhurry (993), Hassan and Tufe (998), and Kim and Lee (996). 3

4 correcion models. Along wih exchange rae volailiy, manufacuring producion indices and real bilaeral exchange raes are also employed as explanaory variables of real expor volumes. Exchange rae volailiy is measured by compuing he monhly sandard deviaions of daily real bilaeral exchange raes. Since daily exchange raes are nominal and price indices are no available on a daily basis, monhly price indices were convered ino daily price indices using he mehod of Quadraic-Mach Average available in he sofware package, E-Views 4, o compue daily real bilaeral exchange raes. In he case of expors o he U.S., preliminary empirical es resuls indicae a negaive long-run relaionship beween expors and exchange rae volailiy in Souh Korea and Singapore, and no long-run relaionship in Hong Kong and Thailand. However, negaive shor-run impacs of exchange rae volailiy on expors are deeced in Hong Kong and Souh Korea. In he case of he expors o Japan, empirical sudies indicae a negaive longrun relaionship beween expors and exchange rae volailiy in Souh Korea and Thailand, and a posiive long-run relaionship in Hong Kong and Singapore. In conras, negaive shor-run impacs of exchange rae volailiy on expors are deeced in all he counries examined, excep for Hong Kong. On he oher hand, manufacuring producion indices of imporing counries and depreciaion of real bilaeral exchange raes urn ou, in general, o have posiive effecs on expor volumes of he Eas Asian counries examined. 2. Descripion of he model and daa 2.. The coinegraion equaion This paper invesigaes he long-run relaionship beween exchange rae volailiy and expors by performing coinegraion ess and he shor run impacs of volailiy on expors by esimaing error-correcion models as in Arize, Osang and Sloje (999, 2000), Chowdhury (993) and Hassan and Tufe (998). 4

5 Following he ypical specificaion of oher papers, he long-run equilibrium relaion beween expors and oher economic variables is examined in his paper by he following equaion: X = 0 + ξ. i + ξ 2 p + ξ 3 ξ σ + ε () where X denoes real expors from an Eas Asian counry o eiher he U.S. or Japan, p he real bilaeral exchange rae reflecing he price compeiiveness, i he manufacuring producion index of he imporing counry, σ he exchange rae volailiy, and ε a disurbance erm. All variables are in naural logarihm. In his equaion, i is used as a proxy for economic aciviy in he imporing counry because monhly daa for GDP are no available. I is expeced ha he higher he economic aciviy in he imporing counry, he higher he demand for expors. Therefore, he value for ξ is expeced o be posiive. Since a higher real exchange rae implies a lower relaive price, he value for ξ 2 is also expeced o be posiive. Exchange rae volailiy is measured by compuing he monhly sandard deviaions of daily real bilaeral exchange raes. Since daily exchange raes are nominal and price indices are no available on a daily basis, monhly price indices were convered ino daily price indices using he mehod of Quadraic-Mach Average available in he sofware package, E-Views 4, o compue daily real bilaeral exchange raes. The following subsecion shows more specifically how he daa for he variables were compued The variables Real expors (X i ) 5

6 In order o ensure consisency in daa 4, expors of he Eas Asian economies under consideraion are convered from US dollar ino he respecive local currency uni (LCU) using corresponding nominal exchange raes, since he expor uni value index is based on domesic currency 5. Real expors of counry i are defined as follow: EX i X = ln 00 i, i=, 2, 3, 4 EXUVi where X i denoes real expors of counry i in domesic currency in naural logarihm scale, EX i is monhly nominal expors of counry i in domesic currency, EXUV i denoes he index of expor uni value of counry i and he index symbolizes he ime. Indusrial producion index (i ) As menioned in he previous secion, lack of monhly daa for income or GDP of he imporing counries leads o he applicaion of he indusrial producion index as a proxy variable for he economic condiion of he imporing counry. Indusrial producion indices are commonly used as a proxy for income in lieraure, for example Baum, Calagyan and Ozkan (2002). The variable i is he naural logarihm of he indusrial producion index of an imporing counry in ime. Real bilaeral exchange rae ( p ) Bilaeral rade beween wo counries depends upon, among oher hings, exchange raes and he relaive price level of he wo parners. Hence, he following definiion of real exchange raes for counry i capures boh effecs relaed o he price of currencies, and of goods and services. 4 Variables, which were no seasonally pre-adjused, were adjused for seasonaliy prior o aking logarihm by applying he mehod Census X2 available in he sofware package Eviews 4. 5 See IFS documens, such as IFS yearbook 200, for deailed explanaion abou he uni value index for expors. 6

7 CPI j p = ln i Ei, i=, 2, 3, 4 CPI i where p i symbolizes real monhly exchange rae in naural logarihm scale; E i is he nominal monhly exchange rae; CPI i and CPI j denoe he monhly consumer price index of an exporing counry i and an imporing counry j, respecively; and symbolizes he ime index. Real exchange rae volailiy ( σ ) Alhough here exis numerous measures for exchange rae risks, he presen sudy applies sandard deviaion of exchange raes, since his measure is common in lieraure, for insance Akhar and Hilon (984) and Baum e al. (2002). The monhly real exchange rae volailiy is defined as he naural logarihm of he sandard deviaion of daily real exchange raes (RERij) wihin one monh. σ 2 ( RER RER ), n σ ln ij = ik i i=, 2, 3, 4 n k = where RER ik is he daily real exchange rae of counry i in normal scale; RER i denoes he monhly average of daily real exchange raes in normal scale and k is he index of he days in a monh, on which exchange rae daa are available. RER ik is defined as he produc of counry i s daily nominal exchange rae and he raio of he daily CPI of he imporing counry over he daily CPI of he exporing counry. As illusraed above, he compuaion of daily real exchange raes requires daily daa for he consumer price index. Hence, monhly CPI was used o compue daily CPI for he six economies involved because of lack of daily CPI daa. Derived from he mehodology applied in Baum e al. (2002), he frequency conversion from low frequency (monhly) o high frequency (daily) was conduced by applying he mehod Quadraic-mach Average available in he sofware package Eviews4. Daa Sources 7

8 The monhly daa sars from January 990 and ends a November 200. Consumer Price Indices (CPI) have been colleced from he Inernaional Financial Saisics (IFS) of he Inernaional Moneary Fund (IMF). The daa for expors from each Eas Asian counry o Japan and o he U.S. have been obained from he Direcion of Trade Saisics (DOTS) of he IMF. The daa for he indusrial producion index of Japan have been colleced from he Minisry of Economy, Trade and Indusry (METI) of Japan, while he daa for he indusrial producion index of he U.S. have been colleced from he Federal Reserve Board (FRB) of he U.S. Daily exchange rae daa have also been colleced from he FRB of he U.S The error-correcion model Afer observing he resuls of coinegraion ess, he following dynamic error correcion (EC) model is consruced and esimaed o see he shor-run impacs of exchange rae volailiy on expors: X = k n n n n 0 + λec + αi X i + α 2i pi + α 3i ii + α 4i σ i + i= 0 i= o i= 0 i= 0 u --- (2) If he variables in equaion () are no coinegraed, he error correcion erm,, will be eliminaed from equaion (2). EC 3. Empirical es resuls 3.. Uni Roo ess As preparaion for coinegraion ess, he presence of uni roos in he variables included in equaion () are examined using he augmened Dickey-Fuller (ADF) 8

9 ess. Tables <2-> and <2-2> presen he augmened Dickey-Fuller es saisics for he firs differences all he four variables in equaion (). The lengh of he lags included in he ess were deermined by he Akaike infomaion crierion. The ADF saisics for he levels of all he series were below he criical values implying he presence of uni roos. However, he saisics obained from he firs differences of he variables rejec he null hypohesis of a uni roo a he five percen significance level Coinegraion ess and Error correcion model Johansen (988,99) coinegraion ess were applied o es for he presence of a long-run equilibrium relaionship in he variables in equaion (). The resuls of coinegraion ess are presened in Tables <3-> and <3-2>, where r denoes he number of coinegraing vecors. The es saisics imply he presence of one coinegraing relaionship for all he four counries examined.. The esimaed coefficiens for he long-run relaionship are presened in Tables <4-> and <4-2> and he esimaed coefficiens for he error correced models are presened in Tables <5-> and <5-2>. In all counries, he level of economic aciviy measured by he manufacuring producion index urns ou o posiively affec expors o Japan and expors o he U.S. boh in he long run and in he shor run. In conras, he impac of exchange rae volailiy urns ou o be a lile bi ambiguous as in oher lieraure. In he case of expors o he U.S., preliminary empirical es resuls indicae a negaive long-run relaionship beween expors and exchange rae volailiy in Souh Korea and Singapore, and no long-run relaionship in Hong Kong and Thailand. However, negaive shor-run impacs of exchange rae volailiy on expors are deeced in Hong Kong and Souh Korea. In he case of he expors o Japan, empirical sudies indicae a negaive longrun relaionship beween expors and exchange rae volailiy in Souh Korea and Thailand, and a posiive long-run relaionship in Hong Kong and Singapore. In conras, negaive shor-run impacs of exchange rae volailiy on expors are deeced in all he counries examined, excep for Hong Kong. 9

10 References Arize, Augusine C., Thomas. Osang and Daniel J. Sloje (2000), Exchange-rae volailiy in Lain America and is impac on foreign rade, manuscrip, Texas A&M Universiy. Arize, Augusine C., Thomas. Osang and Daniel J. Sloje (2000), Exchange-rae volailiy and foreign rade: evidence from hireen LDC s, Journal of Business & Economic Saisics, vol. 8, no., pp Baak, S. J. (200), Japanese yen and Eas-Asian currencies: before and afer he Asian fianacial crisis, Inernaional Developmen Series 7, Inernaional Universiy of Japan. Bacchea, Philippe and Eric van Wincoop (2000), Does exchange-rae sabiliy increase rade and welfare? The American Economic Review, vol. 90, no. 5, pp Bahmani-Oskooee, Mohsen and Sayeed Payeseh (993), Does exchange rae volailiy deer rade volume of LDCs? Journal of Economic Developmen, vol. 8, pp Baum, Chrisopher F., Musafa Caglayan and Neslihan Ozkan (200), Exchange rae effecs on he volume of rade flows: an empirical analysis employing highfrequency daa, manuscrip, Boson College. Chowdhury, Abdur R. (993), Does exchange rae volailiy depress rade flows? Evidence from Error-Correcion Models, The Review of Economics and Saisics, vol. 75, no. 4, pp Côé, Agahe (994), Exchange rae volailiy and rade: a survey, Working Paper 94-5, Bank of Canada. De Grauwe, Paul (988), Exchange rae variabiliy and slowdown in growh of inernaional rade, IMF Saff Papers, vol. 35, pp Dell Ariccia, Giovanni (999), Exchange rae flucuaions and rade flows: evidence from he European Union, IMF Saff Papers, vol. 46, no. 3, pp Dickey, D. A., and W. A. Fuller (979), Disribuion of he esimaors for auoregressive ime series wih a uni roo, Journal of he American Saisical Associaion, vol. 74, pp

11 Dickey, D. A. and W. A. Fuller (98), Likelihood raio saisics for auoregressive ime series wih a uni roo, Journal of he American Saisical Associaion, vol. 74, pp Frankel, J. A. and S. J. Wei (994), Yen bloc or dollar bloc? Exchange rae policies of he Eas Asian economies, Io, T and A. O. Krueger eds., Macroeconomic linkage: savings, exchange raes, and capial flows, NBER-Eas Asia seminar on economics, vol 3, Naional Bureau of Economic Research, Cambridge. Hassan, M. Kabir and David R.Tufe (998), Exchange rae volailiy and aggregae expor growh in Bangladesh, Applied Economics, vol. 30, pp Hooper, Peer and Seven Kohlhagen (978), The effec of exchange rae uncerainy on he prices and volume of inernaional rade, Journal of Inernaional Economics, vol. 8, pp Io, T., E. Ogawa and Y. N. Sasaki (998), How did he dollar peg fail in Asia? Journal of Japanese and Inernaional Economies, vol. 2, no. 4, pp Johansen, Søren (988), Saisical analysis of coinegraion vecors, Journal of Economic Dynamics and Conrol, vol.2, pp Johansen, Søren (99), "Esimaion and hypohesis esing of coinegraion vecors in Gaussian Vecor Auoregressive Models," Economerica, Vol. 59, pp Kawai, Masahiro and Shinji Takagi (200), Proposed sraegy for a regional exchange rae arrangemen in pos-crisis Eas Asia, unpublished manuscrip, World Bank. Kwan, C. H. (995), Economics of he yen zone: A perspecive on currency union in Asia, Nihon Keizai shimbunsha (wrien in Japanese). Nakamura, Yoichi and Izumi Masuzaki (997), Economic Inerdependence: Japan, Asia, and he World, Journal of Asian Economics, vol. 8, no. 2, pp Ohno, K. (999), Exchange rae managemen in developing Asia: Reassessmen of he pre-crisis sof dollar zone, ADBI working paper. Asian Developmen Bank Insiue, Tokyo. Secru, Pie and Uppal, Raman (2000), Exchange Rae Volailiy, Trade, Capial Flows under Alernaive Exchange Rae Regimes. Cambridge: Cambridge Universiy Press. Chaper 6. Takagi, S. (996), The Yen and is Asian neighbors, : cooperaion or compeiion, NBER working paper 5720, Naional Bureau of Economic Research, Cambridge.

12 Year <Table -> Expors of Hong Kong DOTS Toal (in million USD) World Share of Share of Share Expors o he Expors o Expors U.S. Japan (in percen) (in percen) China (in percen) , , , , , , , , , , , , , , Source: IMF, Direcion of Trade Saisics, Yearbook (various issues) of o <TABLE -2> EXPORTS OF KOREA DOTS World Share of Share of Share of Toal Expors o he Expors o Expors o Year (in million USD U.S. (in percen) Japan (in percen) China (in percen) , N.A , N.A , N.A , N.A , N.A. 99 7, , , , , , , , , Source: IMF, Direcion of Trade Saisics, Yearbook (various issues) Noe: N.A. denoes no available 2

13 <TABLE -3> EXPORTS OF SINGAPORE DOTS World Share of Share of Share of Toal Expors o he Expors o Expors o Year (in million USD) U.S. (in percen) Japan (in percen) China (in percen) , , , , , , , , , , , , , , Source: IMF, Direcion of Trade Saisics, Yearbook (various issues) <TABLE -4> EXPORTS OF THAILAND DOTS World Share of Share of Share Toal Expors o he Expors o Expors Year (in million USD) U.S. (in percen) Japan (in percen) China (in percen) 986 8, , , , , , , , , , , , , , Source: IMF, Direcion of Trade Saisics, Yearbook (various issues) of o 3

14 <Table 2-> ADF Uni Roo Tes for Expors o US Economy/ Variable ADF Tes Observaions Lags Counry Firs Difference Saisic y x Hong Kong x x y x Korea x x y x Singapore x x y x Thailand x x Noes: ) Lags denoes he included augmenaion lags in uni roo es. 2) ADF is he augmened Dickey-Fuller es. 3) The ADF regression includes only he inercep. 4) The Mckinnon criical value for rejecion of hypohesis of a uni roo a, 5 and 0 percen level is approximaely 3.48, 2.88 and 2.57, respecively. 5) The number of lags was deermined based on Akaike info crierion and he F- es (he F-es was conduced from 2 lags downward. The larger number of lags is seleced if he F- es for 2lags and he minimum Akaike consan rejecs he null hypohesis favoring he shorer lags). <Table 2-2> ADF Uni Roo Tes for Expors o Japan Economy/ Variable Counry Firs Difference y x Hong Kong x 2 x 3 y x Korea x 2 x 3 y x Singapore x 2 x 3 y x Thailand x 2 x 3 Refer o he noes under <Table 2-> Observaions Lags ADF Tes Saisic * * *

15 Economy/ Counry <TABLE 3-> JOHANSEN CO-INTEGRATION TESTS FOR EXPORTS TO JAPAN Trace Saisics H 0 : H A : r = 0 r r r 2 r 2 r 3 r 3 r = 4 r = 0 r = Maximum Eigenvalue r r = 2 r 2 r = 3 r 3 r = 4 Hong Kong Singapore Souh Korea Thailand * 5.482* 70.33* * * 3.379* * * Criical Values Hong Kong (5%) (%) Singapore (5%) (%) Souh Korea (5%) (%) Thailand (5%) (%) Noes: ) r denoes he number of co-inegraing vecors. 2) The aserisks (*) and (**) indicae he rejecion of he null hypohesis a he % and 5% significance level, respecively. Economy/ Counry < TABLE 3-2> JOHANSEN CO-INTEGRATION TEST FOR EXPORTS TO THE U.S. Trace Saisics H 0 : H A : r = 0 r r r 2 r 2 r 3 r 3 r = 4 r = 0 r = Maximum Eigenvalue r r = 2 r 2 r = 3 r 3 r = 4 Hong Kong Singapore Souh Korea Thailand 7.44* * * 7.590* ** ** ** 34.83* 34.97** 38.9* Criical Values Hong Kong (5%) (%) Singapore (5%) (%) Souh Korea (5%) (%) Thailand (5%) (%) Refer o he noes under <Table 3->

16 <Table 4-> Esimaes of he conegraing vecors for expors o Japan Counry Normalized coinegraing equaion Consan i Hong Kong (.498) Singapore (0.864) Souh Korea (7.384) (3.254) Thailand (5.249) (.33) Noes : () Numbers in he parenheses are sandard errors. p (.29) (0.823) 8.78 (.396) 3.93 (0.433) σ (0.96).25 (0.72) (0.346) (0.090) <Table 4-2> Esimaes of he conegraing vecors for expors o he U.S. Counry Normalized coinegraing equaion i Hong Kong (7.98) 3.57 (.263) Singapore (0.50) (0.2) Souh Korea (0.004) (.26) Thailand (0.025) (6.547) Noes : () Numbers in he parenheses are sandard errors. p (.333) (0.8).278 (0.286) (.544) σ (0.225) (0.028) (0.057) (0.92) 6

17 <Table 5-> Regression Resuls for Error-Correcion Models for Expor o Japan Variables Hong Kong Singapore Souh Korea Thailand C 0.4*** ** (0.040) (0.006) (0.02) -0.00*** 0.000** (0.000) (0.000) EC *** 0.09** 0.08 (0.09) (0.09) (0.009) (0.09) *** -0.47*** -0.60*** *** (0.065) (0.08) (0.00) (0.097) *** *** *** 2 (0.09) (0.2) (0.09) ** ** 3 (0.0) (0.092) (0.23) -.224*** -0.59* ** 4 (0.35) (0.092) (0.3) (0.09) -.09*** -0.79** ** 5 (0.44) (0.085) (0.04) (0.089) ** 6 (0.230) 0.53* 0.67* 7 (0.080) (0.09) (0.252) (0.070) (0.09) (0.248) *** 2 (0.076).099***.386*** 0.625** (0.334) (0.333) (0.288) 2.835*** (0.967) ** ** 2 (0.997) (0.385) (0.306) * ** 3 (0.367) (0.394) (0.328) 0.806** ** 4 (0.406) (0.300) 2.973*** ** 5 (0.900) (0.374) (0.42) 0.626* 0.859*** 6 (0.363) (0.302) 0.823*** 7 (0.304) 2.64*** 9 (0.953) 0.823** -0.97*** 2 (0.37) (0.304) * 3 (0.306) -.403** 0.574** 0.660*** 0.258* (0.623) (0.277) (0.92) (0.40).533** *** 0.44*** 2 (0.644) (0.298) (0.49) 0.439** 3 (0.97) 7

18 (0.80) (0.39) (0.644) (0.58).527** *** 7 (0.650) (0.289) 0.358** 8 (0.55) ** 9 (0.80) (0.62).293** 0.340* 0.337** (0.63) (0.77) (0.42) -0.73** * 2 (0.296) (0.85) (0.) 0.602** 3 (0.280) *** 4 (0.52) (0.35) 0.367** 5 (0.48) σ 0.099** -0.2*** ** (0.05) (0.025) (0.07) σ * -0.04*** (0.040) (0.024) (0.08) σ *** ** (0.022) (0.08) σ *** (0.038) (0.09) (0.07) σ *** * (0.037) (0.08) (0.07) σ ** ** *** (0.05) (0.02) (0.00) σ (0.035) σ * (0.0) σ 9 (0.04) σ (0.033) (0.06) σ ** (0.09) σ * (0.09) σ (0.09) σ ** (0.03) (0.07) σ ** ** (0.03) (0.03) σ (0.0) Adjused R 2 DW *** (0.03) *** (0.03) *** (0.03) *** (0.03) *** (0.03) ** (0.0)** -0.07** (0.008)

19 Noes: Figures in parenheses are sandard errors <Table 5-2> Regression Resuls for Error-Correcion Models for Expor o he U.S. Variables Hong Kong Singapore Souh Korea Thailand C 0.350*** (0.086) 0.047*** (0.07) ** *** (0.00) -0.00*** (0.000) 0.000*** (0.000) 0.000** (0.000) EC *** (0.077) 0.07 (0.090) EC *** (0.00) EC *** (0.058) *** *** *** *** (0.5) (0.2) (0.085) (0.09) * *** *** *** 2 (0.52) (0.090) (0.095) (0.089) ** *** 3 (0.69) (0.092) -.70*** -0.54*** 4 (0.7) (0.096) -.379*** ** -0.50*** (0.87) (0.094) (0.06) (0.093) *** ** *** (0.203) (0.4) (0.) (0.093) ** -0.79* 7 (0.0) (0.09) ** *** 8 (0.249) (0.08) (0.090) *** ** *** 9 (0.3) (0.07) (0.093) -.202*** *** *** -0.96** 0 (0.340) (0.08) (0.088) (0.078) -0.72** -0.23** *** (0.345) (0.05) (0.087) -0.82** ** *** 2 (0.33) (0.0) (0.086) -0.92** *** (0.347) (0.09) (0.076) (0.095) * -0.24* 4 (0.294) (0.3) ** (0.224) (0.09) 8.778** 2.743** 2.73** 2.785** (3.433) (.258) (.62) (.85) *** (.3) (.243) ** 5.047*** 3 (.55) (.230) 2.02* 4 (.209) 9

20 .905*** 6.30*** 2.40** 5 (3.474) (.264) (.53) *** 6 (3.345) (.206) ** 7 (.28) 0.940*** 2.67* 8 (3.636) (.25) (3.402) 8.227** -3.06*** 0 (3.575) (.45) (.47) 0.060** (3.97) (.26) 2.548*** *** (2.93) (0.392) (0.49) (0.7) 0.372* 0.607*** 2 (0.93) (0.69) (3.77) 0.444** 4 (0.206) * 0.94*** 6 (3.03) (0.20) *** 7 (0.207) (0.99) * (3.90) (0.426) (0.27) (0.62) * 9 (3.29) (0.209) ** (3.282) (0.66) (0.59) * * (3.285) (0.42) (0.56) (0.49) (0.52) ** 3 (0.385) (0.48) -.23*** (0.379) (0.56) σ *** (0.009) σ *** (0.028) (0.009) σ *** ** 0.00 (0.030) (0.0) (0.007) * ** σ 3 (0.036) σ 4-0.*** (0.038) -0.07* (0.009) (0.03) *** (0.05) 20

21 σ ** ** -0.04*** 0.022** (0.037) (0.009) (0.05) (0.009) σ *** *** 0.039*** (0.034) (0.06) (0.00) σ *** * 0.08* (0.027) (0.00) (0.00) (0.009) σ ** 0.024** (0.024) (0.0) (0.00) σ *** (0.00) σ (0.00) σ 0.08* (0.00) σ ** 0.023** (0.00) (0.0) σ *** * (0.02) (0.008) (0.02) σ *** 0.029** (0.0) (0.02) σ *** (0.009) Adjused R DW Noes: Figures in parenheses are sandard errors 2