Analysis of FDI Inflows into China from ASEAN-5 Countries: A Panel Cointegration Approach

Transcription

1 Journal of Economc Cooperaon and Developmen 35 3 (04) -8 Analyss of FDI Inflows no Chna from ASEA-5 Counres: Sallahuddn Hassan or Aznn Abu Bakar and Hussn Abdullah 3 hs sudy analyses he foregn drec nvesmen (FDI) nflows no Chna from ASEA 5 counres usng he panel conegraon approach. he FDI model has been ulzed n deermnng facors ha nfluence FDI nflows no Chna from ASEA 5 counres namely Malaysa haland he Phlppnes Indonesa and Sngapore. Varables such as rade openness (OPEESS) exchange rae of Chna relave o each of ndvdual ASEA 5 counres (RELEXC) fxed capal formaon (FCF) and gross domesc produc (GDP) are used for he perod of he emprcal resuls ndcae ha for mos counres OPEESS and GDP are sgnfcan varables n explanng he flow of FDI o Chna. Meanwhle FCF s only sgnfcan for Malaysa. Conversely RELEXC s no sascally sgnfcan for all counres. I s hoped ha hs fndng can be used by researchers and polcy makers n makng decson on mullaeral relaonshp beween Chna and ASEA 5 counres.. Inroducon Foregn drec nvesmen (FDI) plays an mporan role n he growh and developmen of no only he developed counres bu also n he developng counres. Besdes he capal ha brough n also nroduces new and modern echnology whch provdes marke opporunes and lnkages o expor. Counres are compeng wh each oher o offer a lucrave ncenve plans o arac FDI. & 3 Assocae Professor a College of Busness Unvers Uara Malaysa. Emal: dn636@uum.edu.my or 3 Emal: noraznn@uum.edu.my

2 Analyss of FDI Inflows no Chna from ASEA-5 Counres: he Asan regon has always been consdered as a pruden cenre for nvesmen especally from he Uned Saes of Amerca (USA) Japan Uned Kngdom (UK) and oher European counres. Globalzaon and negraon of economc acves across he world forced he governmen of he Asan counres o arac FDI whch laer on ranslaed no rapd growh n hese economes. Asan counres are mplemenng new plans and polces o arac more and more FDI whch wll brng n new nnovaon and auomaon based echnologes ha can rejuvenae he hos counry s exsng manufacurng base. Furhermore he Asan regon aracs FDI nflows as a resul of her abundan naural resources hghly sklled experenced and knowledgeversed labour and huge sze of domesc marke. Chna has been openng up s economy for more han weny years; however s accesson o World rade Organzaon (WO) on December 00 mples exensve consequences for s economy. Chna s openng up polcy has amed a promong expors whle proecng he domesc marke. hs was acheved hrough a dualsc rade regme whch has graned arff exempons on mpors of nermedae goods by expor orened ndusres and hrough a selecve polcy whch has channelled FDI no manufacurng producon argeed for expors or for mpor subsuon. As a resul FDI nflows have played a major par n he openng up of Chna s ndusry and s negraon no he nernaonal dvson of labour. he rapd expanson of s nernaonal rade and large capal nflows provde evdence of he ncreasng negraon of Chna n he world economy. Snce 980 Chna s share n nernaonal rade has rebled rsng from less han one percen o more han hree percen n 999. Durng he frs 6 monhs of 0 Chna has surpassed he USA as he world s larges recpen of global FDI wh a oal of USD59 bllon compared o FDI flowng o he USA oallng USD57.4 bllon he cooperaon and parnershp beween Chna and ASEA has long been esablshed hrough varous channels for aanng ceran goals. One of he major channels s FDI nflows. Accordng o Shu and Zeng (006) FDI nflows from ASEA 5 no Chna n 004 was abou ffy mes as much as was n 990 (see able ). Durng he cumulave amoun of Chna s acually ulzed FDI from ASEA 5 reached USD33.73 bllon whch exceeded he cumulave amoun of Chna s acually ulzed FDI from he UK France and Germany

3 Journal of Economc Cooperaon and Developmen 3 combned whch was USD7. bllon. Based on counry Sngapore recorded he hghes amoun of FDI nflows o Chna and hen followed by Malaysa. hs fac s n lne wh he argumen by Ellngsen e al. (006) ha Sngapore s one of he mos mporan ouward nvesors n he developed counres. Sngaporean drec nvesors were srongly encouraged by he governmen o reach beyond ASEA and were ncreasngly shfng her aenon o oher Asan hos counres parcularly Chna. Furhermore he growh of FDI nflows from ASEA 5 no Chna ncreased remendously n 99 and 993 bu showed a declnng rend afer he year 994. he wors FDI growh recorded n 999 ha due o Eas Asan fnancal crss n he mddle of 997. hs largely unforeseen crss and s afermah caused deeroraon n he macroeconomc fundamenals parcularly slower economc growh of haland Malaysa and Indonesa. As a resul hese counres reduced her FDI nflows no Chna. hs gves rse o he ssues of wheher he ASEA 5 counres are able o manan nvesmen n Chna and he emergng facors ha sgnfcanly affec he ably of ASEA 5 counres nves n Chna. able : Foregn Drec Invesmen n Chna by ASEA 5 Counres (Amoun conraced n USD mllon) Year Counry Malaysa Indonesa Sngapore haland he Phlppnes ASEA 5 Growh (%) Sources: Shu and Zeng (006).

4 4 Analyss of FDI Inflows no Chna from ASEA-5 Counres: Gven he above scenaro we are encouraged o underake furher analyss o denfy facors ha sgnfcanly affec he FDI nflows no Chna from he seleced ASEA 5 counres. In hs arcle we posulae ha Chna s macroeconomc fundamenals can nfluence FDI nflows from he ASEA 5 counres. Furhermore hs sudy s also movaed by he fac ha here has been lle economerc modelng and evdences of how he macro fundamenals of Chna affec he nflows of capal from he ASEA 5 counres. hs paper res o fll hs gap by analyzng he lnk beween Chna and he ASEA 5 counres hrough FDI nflows acves. hs lnk has never been addressed aggressvely n he leraure parcularly afer WO accesson proocol. herefore hs sudy ams o provde new evdence of macro-deermnans of FDI nflows beween Chna and ASEA 5. Snce Chna has become very mporan n he world economc operaon and an aracve se for FDI (aughon 996) afer admsson n WO ASEA 5 counres would gan some sgnfcan benefs hrough FDI acves n Chna. he paper s organzed as follows. Secon revews he exsng leraure on deermnans of FDI nflows. Secon 3 descrbes he daa used and he mehodology of deermnng facors ha nfluence he level of FDI nflows no Chna. he emprcal resuls of he sudy are repored n Secon 4 and Secon 5 presens he concluson.. Leraure Revew Many sudes have been carred ou o analyze he deermnans of FDI and s effec o hos counres parcularly developng counres. As saed n many sudes FDI plays an mporan role n hese economes whch are generally lackng n erms of echnology as well as capal o fund he projecs (Borenzsen e al. 998; Esrn e al. 000; Buckley e al. 00). In addon Calvo e al. (996) and Moreno (000) saed ha capal flows from one counry o anoher also allow nvesors o dversfy her rsks. As he me progress and because of many facors he rend of FDI nflows s no jus from developed counres o developng or under developng counres bu he reverse or ouward FDI may exs. hs phenomenon also happens among ASEA 5 counres. Even hough

5 Journal of Economc Cooperaon and Developmen 5 hese counres are consdered as developng counres hey nend o nves n abroad n order o grab he opporunes and advanageous offered by he recpen counres. One of he specfc sudes relaed o FDI nflows n Chna and ASEA has done by Shu and Zeng (006). her descrpve analyses have ndcaed hree facors ha exer sgnfcan nfluence on FDI nflows beween Chna and ASEA. hese facors are new blaeral economc agreemens Chna s new mega economc zone and ASEA reforms and new foregn polcy. Blaeral agreemens such as Framework Agreemen on Comprehensve Economc Cooperaon (FACEC) beween ASEA and People's Republc of Chna sgned n 00 for nsance can srenghen economc relaons whch s a sraegc goal of boh sdes. A large varey of Chnese producs have been expored on a large scale no ASEA member counres. Followng he FACEC agreemen Chna provdes survval and growh opporunes for ASEA nvesors and raders. here are many emergng mcroeconomc and macroeconomc fundamenals ha nfluence a counry o nves abroad or o receve FDI nflows from abroad. However hs bref revew gves more emphasze o he effecs of macroeconomc fundamenals on FDI nflows. In addon o geographc locaon he mporan macroeconomc fundamenals of hos/recpen counres such as gross domesc produc (GDP) exchange raes rade openness and capal accumulaon are consdered as mporan deermnans. Openness of one locaon s one of he radonal varables o explan FDI movemens. I s defned as he rao of oal rade (expor plus mpor) o GDP. MEs engaged n expor orened nvesmen prefer o locae n a more open economy as ncreased mperfecons ha accompany rade proecon generally mply hgher ransacon coss assocaed wh exporng. A sudy done by he Organzaon for Economc Cooperaon and Developmen (000) fnd ha rade openness s one of he man deermnans of FDI n Chna. Chna has adoped he expor promoon sraegy where has mplemened economc reforms and open-door polces a he same me also promoes rade by concludng several blaeral rade arrangemens. hs can be seen n 990s where here has been a subsanal progress n reducng arff barrers.e. he average arff rae on mpors declned

6 6 Analyss of FDI Inflows no Chna from ASEA-5 Counres: from 4.9 percen n 99 o 7.6 percen n 997. Wh regard o hs varable he expeced mpac of he degree of openness on FDI can be mxed; can arac he foregn capal o he hos area and also can ncrease compeon beween he foregn and domesc frms on. Meanwhle Acemoglu and Zlbo (997) explan ha greaer openness n nernaonal ransacons helps drec fnancal flows from capal-abundan owards capal-scarce counres. Adhkary (0) has clamed ha he degree of rade openness s lkely o nfluence he flows of nernaonal capal n erms of rsk-reurn relaonshp. Exchange rae has greaer mpac on he flow of foregn capal no hos counry [see Froo and Jeremy (99) Pan and Van Welsum (003) Phlps e al. (008) Goschalk e al. (008) Baek and Okawa (00) yarko and keah-amponsah (0) Chaudary e al. (0) Ullah e al. (0) and gowan (0)]. Chaudary e al. (0) for nsance have suded he effecs of exchange rae on FDI n he Asan economes; wh Paksan Inda Sr Lanka and Bangladesh represenng Souh Asan. Malaysa Indonesa Sngapore and haland represenng he Souheas Asan regon; Chna Japan and Souh Korea represenng he counry n Eas Asa. Whle urkey Iran and Israel represenng Wes Asa. hey have used Auoregressve Dsrbued Lag (ARDL) conegraon and error correcon model (ECM) for performng emprcal analyss. hey found ha more han half of he seleced counres show no sgnfcan relaonshp beween FDI and exchange rae. Effec of shor-erm and long-erm exss n Paksan Inda Sr Lanka Souh Korea urkey and Israel. he resuls ndcaed here was no relaonshp beween FDI and exchange raes for Bangladesh Chna Malaysa Indonesa haland Sngapore and Iran. Furhermore economc growh of he hos counry s anoher man deermnan of FDI nflows. hs varable s used n many sudes such as oda and Yamamoo (995) Wang and Swan (995) Zhang ( ) Barhelemy and Demurger (000) We and Lu (00) Jordan (00) Al and Guo (005) and Le and Lu (005). hey agreed ha one of he man facors s he growh of he hos counry whch can be measured by he GDP GDP per capa gross naonal produc (GP) and GP per capa. hus he varable GDP s expeced o have a posve sgnfcan relaonshp wh FDI nflows. Borenszen e al. (998) for nsance have argued ha FDI nflows are posvely relaed

7 Journal of Economc Cooperaon and Developmen 7 o per capa GDP growh provded he hos counry has a hghly educaed workforce. Meanwhle Chowdhury and Mavroas (003) fnd undreconal causaly runnng from growh o FDI n he case of Chle bu fnd bdreconal causaly for haland and Malaysa. he earler growh models by Harrod (939) and Domar (946) explaned ha capal formaon rases he sandard of lvng whch n urn resuls n hgher growh. hen Solow (956) argues ha capal formaon ncreases labor producvy n a dynamc process of nvesmen growh and fnally enhancng economc growh. hese evdences recognze he role of capal formaon n economc growh and hen creae a bass for aracng FDI nflows no a parcular counry. Analycally capal formaon can mprove domesc nvesmen hrough posve spllovers and by creang complemenary ndusres. hus capal formaon has a dynamc effec on FDI nflows. hs s suppored by Adhkary (0) where he menoned ha he level of capal formaon s lkely o nfluence FDI and economc growh. eo-classcal growh model posulaes ha developng economes ha have a lower nal level of capal sock end o have hgher margnal rae of reurns (producvy) and growh raes f adequae capal sock s njeced. In oher words n a capal shorage economy he margnal producvy of nvesmen s ncreased n he shor-run when addonal capal s njeced n he form of long-erm nvesmen lke FDI and hs ncreased producvy nfluences economc growh n he long-run. A he macro level by and large prevous leraure suggess ha macroeconomc deermnans conrbue o FDI nflows hrough eher drec or ndrec FDI. Drec FDI s a ype of nvesmen where nvesmen acves are carred ou by he foregn-owned frm such as paren (headquarer) drecly. Meanwhle ndrec FDI s a ype of FDI by foregn afflae. Invesmen acves are carred ou by frm whch hemselves are afflaes of foregn Mulnaonal Enerprses (MEs) such as a regonal headquarer.

8 8 Analyss of FDI Inflows no Chna from ASEA-5 Counres: 3. Mehodology 3. Daa Secondary daa are used n he sudy. Due o shor me spans of me seres daa we have pooled cross secon and me seres daa o form a balanced panel daa. Usng he shor me span of me seres daa may yeld unrelable resuls. he balanced panel consss of annual daa for FDI nflows no Chna from fve seleced ASEA 5 counres namely Malaysa haland he Phlppnes Indonesa and Sngapore for he perod of and daa of each varable s measured n US dollars. he use of panel daa allows us o sudy boh he changes n he ndependen varables of a sngle ASEA 5 counry over me and he varaon n hese varables of many ASEA 5 counres a a gven pon n me. he daa are gahered and verfed from varous sources.e. Inernaonal Fnancal Sascs by Inernaonal Moneary Fund (IMF) Drecon of rade Sascs World Developmen Indcaors and World Deb ables. 3. Model Specfcaon Based on prevous leraure FDI nflows are manly deermned by rade openness (OPEESS) exchange rae (RELEXC) fxed capal formaon (FCF) and GDP. I s descrbed by he basc funcon as Equaon []. Furhermore he emprcal model form for hs funcon s gven by Equaon []. FDI f ( OPEESS RELEXC FCF GDP) [] FDI 0 OPEESS RELEXC 3FCF 4GDP [] In Equaon [] FDI represens he oal FDI nflows no Chna from ASEA 5 counres OPEESS represens he level of rade openness and s measured by he rade-nensy rao whch s he share of expor and mpor n GDP RELEXC s he exchange rae of Chna relave o each of he ndvdual ASEA 5 counres FCF s he oal of fxed capal formaon n Chna and GDP s he gross domesc produc of Chna. he choce of hese varables reles on he daa accessbly and he value of daa n dollar Amerca. he β 0 s a

9 Journal of Economc Cooperaon and Developmen 9 consan erm and β o β 4 are esmaed parameers n he model and s a cross secon daa for counres referred o and s a me seres daa and ε s an error erm. Boh he coeffcens β 3 and β 4 are expeced o carry a posve sgn. 3.3 Esmaon Procedure Esmaon procedure of he relaonshp beween FDI nflows no Chna and ndependen varables n Equaon [] nvolves wo man seps conegraon analyss and coeffcen esmaon Conegraon Analyss We have employed panel conegraon approach o perform conegraon analyss. hs mehod s used o nvesgae he exsence of he long run conegraon among he varables. In order o nvesgae he possbly of panel conegraon s frs necessary o deermne he exsence of un roos n he daa seres. For hs sudy we have chosen he Im Pesaran and Shn (IPS hereafer) (997) whch s based on he well-known Dckey Fuller procedure. IPS proposed a es for he presence of un roos n panels ha combnes nformaon from he me seres dmenson wh ha from he cross secon dmenson such ha fewer me observaons are requred for he es o have power. he use of hs mehod s movaed by ha hs s commonly used and superor es power (Chou and Lee 003). Many economc researchers such as Lee e al. (997) Sarans and Seward (999) Canzoner e al. (999) and Chou and Lee (003) have also appled hs mehod n her analyss of he long run relaonshps n panel daa. IPS begns by specfyng a separae ADF regresson for each crosssecon wh ndvdual effecs and no me rend: y y p y j j [3] j where =... and =... IPS use separae un roo ess for he cross secon uns. her es s based on he Augmened Dckey Fuller (ADF) sascs averaged across groups. Afer esmang he separae ADF regressons he

10 0 Analyss of FDI Inflows no Chna from ASEA-5 Counres: average of he sascs for p from he ndvdual ADF regressons p ) : ( p [4] he bar s hen sandardzed and s shown ha he sandardzed sasc converges o he sandard normal dsrbuon as and. IPS (997) showed ha es has beer performance when and are small. hey proposed a cross seconally demeaned verson of boh es o be used n he case where he errors n dfferen regressons conan a common me-specfc componen. he nex sep s o es for he exsence of a long run conegraon beween FDI and he ndependen varables usng panel conegraon ess suggesed by Pedron ( ). Pedron (999) deermned ha he ess are approprae o esmae resduals from a conegraon regresson afer normalzng he panel sascs wh correcon erms. he es procedures proposed by Pedron make use of esmaed resdual from he hypoheszed long run regresson of he followng form: y x x... MxM e [5] for =..; =.; m =. M where s he number of observaons over me number of crossseconal uns n he panel and M number of regressors. In hs se up s he member specfc nercep or fxed effecs parameer whch vares across ndvdual cross-seconal uns. he same s rue of he slope coeffcens and member specfc me effecs. Pedron ( ) proposes he heerogeneous panel and heerogeneous group mean panel es sascs o es for panel conegraon. He defnes wo ses of sascs. he frs se of hree sascs Z v Z and Z s based on poolng he resduals along he whn dmenson of he panel. he sascs are as follows 3 / Z v L e [6]

11 Journal of Economc Cooperaon and Developmen e e L e L Z [7] I e e e L e L Z ~ / [8] where e s he resdual vecor of he OLS esmaon of Equaon [5] and where he oher erms are properly defned n Pedron. he second se of sascs s based on poolng he resduals along he beween dmenson of he panel. I allows for a heerogeneous auocorrelaon parameer across members. he sascs are as follows: e e e Z ~ [9] e e e Z ~ / [0] hese sascs compue he group mean of he ndvdual convenonal me seres sascs. he asympoc dsrbuon of each of hose fve sascs can be expressed n he followng form: ) ( 0 v X [] where X s he correspondng from of he es sascs whle and v are he mean and varance of each es respecvely. hey are gven n able n Pedron (999). Under he alernave hypohess Panel v sascs dverges o posve nfny. herefore s a one sded es were large posve values rejec he null of no conegraon. he remanng sascs dverge o negave nfny whch means ha large negave values rejec he null.

12 Analyss of FDI Inflows no Chna from ASEA-5 Counres: 3.3. Coeffcens Esmaon For coeffcen esmaon we have adoped he Fully Modfed Ordnary Leas Squares (FMOLS) procedure based on Chrsopoulos and sonas (004). In order o oban asympocally effcen conssen esmaes n panel seres non exogeney and seral correlaon problems are ackled by employng FMOLS nroduced by Pedron (996). Snce he explanaory varables are conegraed wh a me rend and hus a long-run equlbrum relaonshp exss among hese varables hrough he panel un roo es and panel conegraon es we proceed o esmae he Equaon [] by he mehod FMOLS for heerogenous conegraed panels. hs mehodology allows conssen and effcen esmaon of conegraon vecor and also addresses he problem of non saonary regressors as well as he problem of smulaney bases. I s well known ha Ordnary Leas Squares (OLS) esmaon yelds based resuls because he regressors are endogenously deermned n he I() case. he sarng pon OLS as n he followng conegraed sysem for panel daa: y x e [] x x where s he saonary wh covarance marx e. he esmaor wll be conssen when he error process e sasfes he assumpon of conegraon beween y and x. he lmng dsrbuon of OLS esmaor depends upon nusance parameers. Followng Phllps and Hansen (990) a sem paramerc correcon can be made o he OLS esmaor ha elmnaes he second order bas caused by he fac ha he regressors are endogenous. Pedron ( ) follows he same prncple n he panel daa conex and allows for he heerogeney n he shor run dynamcs and he fxed effecs. FMOLS Pedron s esmaor s consruced as follow: FM ( x x ) x x e [3] 0 0 e e

13 Journal of Economc Cooperaon and Developmen 3 where he covarance marx can be decomposed as 0 0 where s he conemporaneous covarance marx and s a 0 weghed sum of auocovarances. Also denoes an approprae esmaor of. 0 In hs sudy we employed panel group FMOLS es from Pedron ( ). An mporan advanage of he panel group esmaors s ha he form n whch he daa s pooled allows for greaer flexbly n he presence of heerogeney of he conegrang vecors. es sascs consruced from he panel group esmaors are desgned o es he null hypohess H0 : 0 for all agans he alernave hypohess H A : 0 so ha he values for β are no consraned o be he same under he alernave hypohess. Clearly hs s an mporan advanage for applcaons such as he presen one because here s no reason o beleve ha f he conegrang slopes are no equal o one whch hey necessarly ake on some oher arbrary common value. Anoher advanage of he panel group esmaors s ha he pon esmaes have a more useful nerpreaon n he even ha he rue conegrang vecors are heerogeneous. Specfcally pon esmaes for he panel group esmaor can be nerpreed as he mean value for he conegrang vecors (Pedron 00). 4. Emprcal Resuls able presens he resuls of he IPS panel un roo es a level ndcang ha all varables are I() n he consan of he panel un roo regresson. hese resuls clearly show ha he null hypohess of a panel un roo n he level of he seres canno be rejeced a varous lag lenghs. We assume ha here s no me rend. herefore we es for saonary allowng for a consan plus me rend. In he absence of a consan plus me rend agan we found ha he null hypohess of havng panel un roo s generally rejeced n all seres a level form and varous lag lenghs. We can conclude ha mos of he varables are non-saonary n wh and whou me rend specfcaons a level by applyng he IPS es whch s also appled for heerogeneous panel o es he seres for he presence of a un roo.

14 4 Analyss of FDI Inflows no Chna from ASEA-5 Counres: Varable able : Panel Un Roo es Im Pesaran and Shn (IPS) Level Frs Order Dfference Consan Consan + rend Consan Consan + rend FDI ** -.703** (0.04) (0.48) (0.003) (0.044) OPEESS * ** (0.93) (0.87) (0.000) (0.00) RELEXC * -3.33** (0.586) (0.8) (0.000) (0.005) FCF ** -.549*** (0.899) (0.899) (0.07) (0.06) GDP * -.887** (0.7) (0.683) (0.00) (0.05) oe: * and ** ndcae rejecon of he null hypohess of no conegraon a one percen and fve percen levels of sgnfcance he resuls of he panel un roo ess confrm ha he varables are non saonary a level. able also presens he resuls of he ess a frs dfference for IPS es n consan and consan plus me rend. We can see ha for all seres he null hypohess of un roo es s rejeced a boh one percen and fve percen levels of sgnfcance. Hence based on IPS es here srong evdence ha all he seres are n fac negraed of orders one. We can conclude ha he resuls of panel un roo es (IPS es) repored n able suppor he hypohess of a un roo n all varables across counres as well as he hypohess of zero order negraon n frs dfferences. A mos of one percen sgnfcance level we found ha all ess sascs n boh wh and whou rends sgnfcanly confrm ha all seres srongly rejec he un roo null. Gven he resul of IPS es s possble o apply panel conegraon mehod n order o es for he exsence of he sable long run relaon among he varables. he nex sep s o es wheher he varables are conegraed usng Pedron s ( ). hs s o nvesgae wheher long run seady sae or conegraon exs among he varables and o confrm wha Coeux and Olver (000) sae ha he panel conegraon ess have much hgher esng power han convenonal conegraon es. Snce he varables are found o be negraed n he same order I () we connue wh he panel conegraon ess proposed by Pedron (999

15 Journal of Economc Cooperaon and Developmen ). Conegraons are carred ou for consan and consan plus me rend and he summary of he resuls of conegraons analyses are presened n able 3. In consan level we found ha hree ou of seven sascs rejec null by hypohess of no conegraon a he fve percen level of sgnfcance for he adf sasc and group ρ sasc whle he group adf s sgnfcan a one percen level. he resuls of he panel conegraon ess n he model wh consan level show ha ndependen varables do hold conegraon n he long run for a group of ASEA 5 counres wh respec o FDI. In he panel conegraon es for our model wh consan plus rend level he resuls ndcae ha four ou of seven sascs rejec he null hypohess of non conegraon a he one percen and fve percen level of sgnfcance. I s shown ha ndependen varables do hold conegraon n he long run for a group of ASEA 5 counres wh respec o FDI. However snce all he sascs conclude n favour of conegraon and hs combned wh he fac ha he accordng o Pedron (999) he panel non paramerc ( sasc) and paramerc (adf sasc) sascs are more relable n consan plus me rend we conclude ha here s a long run conegraon among our varables n ASEA 5 counres. able 3: he Pedron Panel Conegraon es es Consan rend Consan + rend Panel v Sasc Panel ρ Sasc ** Panel Sasc: (non paramerc) Panel Sasc (adf): (paramerc) -.98** -.37** Group ρ Sasc -.97** -.654** Group Sasc: (non paramerc) Group Sasc (adf): (paramerc) -3.7* * oe: All sascs are from Pedron s procedure (999) where he adjused values can be compared o he (0) dsrbuon. he Pedron (004) sascs are one-sded ess wh a crcal value of.64 (k <.64 mples rejecon of he null) excep he v sasc ha has a crcal value of.64 (k >.64 suggess rejecon of he null). * and ** ndcae rejecon of he null hypohess of no conegraon a one percen and fve percen levels of sgnfcance respecvely.

16 6 Analyss of FDI Inflows no Chna from ASEA-5 Counres: Once we have confrmed ha here s a presence of a long run relaonshp beween FDI nflows no Chna and ndependen varables we have done coeffcens esmaon. he resuls of esmaon are shown n able 4. able 4: FMOLS Regresson Resuls FDI Inflows Varables from OPEESS RELEXC FCF GDP Indonesa 5.55* (3.86) 0.4 ( 0.60).7* ( 3.8) 30.85* (4.7) Malaysa 6.33* (.54) 0.5 (0.30).9** (.7).5 (.0) Phlppnes.58 (.03) 0.49 (0.7).30 (0.45). (0.89) Sngapore 4.67* ( 3.) 0.66 (.0) 0.37 (0.45) 3.97* (4.95) haland ** 5.74** (0.64) ( 0.34) (.09) (.59) Panel Group.89** (.7) 0. ( 0.66) 4. (.7) 8.86* (6.08) oe: he null hypohess for he rao s H : 0. 0 Fgures n parenheses are sascs. * and ** ndcae rejecon of he null hypohess a one percen and fve percen levels of sgnfcance respecvely. As shown by able 4 he sgn of esmaed coeffcen for OPEESS are correc and sascally sgnfcan a one percen level for Indonesa and Malaysa respecvely. One un ncrease n rade-nensy rao caused as much as USD5.5 and USD6.63 of FDI nflows no Chna from Indonesa and Malaysa. Hgher degree of openness economy of Chna means he resrcon o rade s low or fallng hus faclang more nernaonal exchange of capal beween Chna and hese wo counres. For he case of he Phlppnes and haland even hough he esmaed coeffcens show posve relaonshps wh he value of coeffcens are.58 and 0.65 respecvely bu hey are no sgnfcan. In conras OPEESS has negave effec on FDI nflows no Chna from Sngapore snce he esmaed coeffcen s negave and sascally sgnfcan. I means ha a hgher degree of openness mples a lower level of FDI nflows. hs can be explaned by lookng a he ype of nvesmens made by he counres. If s horzonal

17 Journal of Economc Cooperaon and Developmen 7 (marke-seekng) nvesmens rade resrcons (means less openness) would have a posve effec on FDI nflows. hs s due o he arff jumpng hypohess whch argues ha foregn frms ha seek o serve local markes may decde o se up subsdares n he hos marke as s dffcul o mpor her producs o he counry. he oher way around mulnaonal frms engagng n expor-orened nvesmens wll prefer o locae n a more open economy snce ncreased mperfecons ha accompany rade proecon mply hgher ransacon coss assocaed wh exporng. Furhermore all coeffcens of RELEXC are no sascally sgnfcan. herefore here s no sascally sgnfcan long run relaonshp beween RELEXC and FDI nflows no Chna from ASEA 5. hs fndng suggess ha he changes n relave exchange raes do no nfluence sgnfcanly FDI nflows no Chna from ASEA 5 whch s conras o he explanaon provded by Xu and Wang (007). Our resul shows ha he movemen of exchange raes plays no role n explanng he level of FDI nflows no Chna. hs can be jusfed wh producon flexbly argumen (De Meza & van der Ploeg 987; Azenman 99; Phllps e al. 008). Accordng o hs argumen foregn nvesors are rsk neural and hey comm o domesc and foregn capacy ex ane afer he realzaon of nomnal or real shocks. hs resul s conssen wh Baley and avlas (99) and Goldberg and Kolsad (995). Goldberg and Kolsad for nsance argued ha f nvesors are classfed as rsk neural here s no sgnfcan relaonshp beween exchange rae volaly and he allocaon of producon facles beween local and foregn markes. However hs resul s also conras o he common argumen of economc heory and avalable leraure on exchange raes and FDI ha based on rsk-averson argumen mperfec capal marke hypohess and rreversble nvesmen decson (Dx & Pndyck 994). Economc heory saes ha as a deprecaed exchange rae lowers he cos of producon and hgher compeveness of counry. herefore nvesmen acves parcularly FDI nflows no he hos counres wll ncrease. Furhermore rsk-averson argumen and mperfec capal marke hypohess heorecally clam ha he level or exchange rae may negavely nfluence FDI. Usng mperfec capal marke consderaon sudy of Froo and Sen (99) provde he evdence of negave relaonshp beween exchange raes and FDI. A weaker hos

18 8 Analyss of FDI Inflows no Chna from ASEA-5 Counres: counry currency nward FDI ends o ncrease whn an mperfec capal marke model as deprecaon makes hos counry asses less expensve relave o asses n he home counry. Walsh and Yu (00) have also saed he smlar argumen. A weaker real exchange rae mgh be expeced o ncrease vercal FDI as frms ake advanage of relavely low prces n hos markes o purchase facles or f producon s re-expored o ncrease home-counry profs on goods sen o a hrd marke. Meanwhle depends on he expecaons of fuure profably Campa (993) predced ha an apprecaon of he hos currency wll ncrease FDI no he hos counry. Usng daa on Japanese acqusons n he USA Blongen (997) has concluded dfferen argumen from Froo and Sen. Oher numercal sudes by Cushman (988) and Dewener (995) have suppored Campa and Blongen s vews. Ineresngly he posve coeffcen of FCF and s sascally sgnfcan a fve percen level mples ha ncrease USD mllon of FCF caused ncremen of FDI nflows no Chna from Malaysa as much as USD.9 mllon. hus hs fndng s n lne wh he heory ha explans counres wh hgh capal formaon normally have more access o capal and hey end o become home counry and do nvesmen abroad. On he oher hand FCF has negave effec on FDI nflows from Indonesa and haland snce s coeffcens values are negave and sascally sgnfcan a one percen and fve percen level. he resul suppors he argumen of Adhkary (0). He argued ha he level of capal formaon s lkely o nfluence FDI and economc growh as well. eo-classcal growh model posulaes ha developng economes ha have a lower nal level of capal sock end o have hgher margnal rae of reurns (producvy) and growh raes f adequae capal sock s njeced As expeced we found ha he coeffcens of GDP are posve and sascally sgnfcan a one percen level only. herefore hese resuls sugges ha GDP of Chna has posve effec on FDI nflows from Indonesa Sngapore and haland. he resuls show ha ncrease USD mllon Chna s GDP wll cause ncrease FDI nflows from Indonesa Sngapore and haland as much as USD30.85 mllon USD3.97 mllon and USD5.74 mllon respecvely. We can conclude ha he hgh rae of economc growh n Chna wll lead o ncrease he nflows of FDI nflows from hese counres. FDI nflows from

19 Journal of Economc Cooperaon and Developmen 9 Indonesa are he larges conrbuon among hem. he resul obaned from hs varable s also smlar o sudes done by sa (994) and Schneder and Frey (985). Boh sudes ndcae ha real GDP per capa as a proxy of marke sze has a srong posve relaonshp wh FDI. However should be noed ha whle marke sze s sll a sgnfcan varable s mporance has been slowly decreasng recenly. hs resul s also echoed by (UCAD 996; 97) whch found ha FDI n developng counres are slowly shfng from resource and marke seekng o more (vercal) effcency seekng. 5. Polcy Implcaon and Concluson hs sudy examnes he deermnans of FDI nflows no Chna from ASEA 5 usng he panel conegraon approach. he un roo es (IPS) s used o confrm he saonary of all varables before performng conegraon es. Afer confrmng ha all varables are non saonary a level he panel conegraon approach s appled. Usng Pedron s approach he long run conegraon es s performed o nvesgae he exsence of he long run conegraon among he varables. hs paper provdes sascal evdence ha he long run deermnans of FDI nflows no Chna may dffer among ASEA 5 counres. Resuls obaned ndcae he presence of he long run relaonshp beween OPEESS and FDI nflows from Indonesa Malaysa and Sngapore. However here s no sascally sgnfcan long run conegraon beween RELEXC and FDI nflows from ASEA 5 counres. For he case of Malaysa he coeffcen of FCF s sascally sgnfcan wh he value of.9 and herefore we can conclude ha here s sll a long run conegraon beween FCF and FDI. he resuls sugges ha he macroeconomc polcy framework s mporan. Chna can ndeed usefully underake polces o foser he volume of FDI nflow from he ASEA 5. In specfc we found ha GDP and FCF are crucal as par of fuure polcy o furher arac new FDI o nflow o Chna. he low mpac of GDP and FCF does sugges ha Chna s ye o embark serously n hs ssue. hs s because hese ndependen varables could also serve as a cos of dong busness and mprovemen would surely be able o reverse he nflows no Chna. here s room for Chna o mprove s exchange rae GDP and fxed capal formaon as aracon for FDI o nflows no Chna. hs could

20 0 Analyss of FDI Inflows no Chna from ASEA-5 Counres: be explaned by he fac ha alhough he ASEA as a whole s no as populous as Chna he purchasng power of he ASEA s acually hgher han Chna and herefore could serve as a good avenue for hgh qualy produc of mulnaonal corporaons. Fnally he mpac of rade openness s posve and sgnfcan mplyng ha mananng he level of openness could be anoher good polcy. Due o posve developmen afer accesson of Chna n WO he ASEA 5 counres need o explore he possble sraeges for accelerang he flow of FDI o Chna. Snce FDI s vewed as one of he major smul of cooperaon some areas o he cooperaon among hem should be denfed and acvaed such as economc and nvesmen cooperaon rade parnershp and echnology developmen exchanges. Economc and nvesmen cooperaon for nsance wll conrbue excellen prospecs and beer opporunes n a wde range of felds for promong neracon ghenng relaonshp and achevng muual benefs n rade nvesmen and ndusral and commercal felds beween Chna and ASEA 5 counres as well as proecng he neress of all counres. he mplemenaon of he ASEA-Chna Free rade Agreemen n 00 for example has undoubedly boosed blaeral rade and nvesmen relaons beween Chna and ASEA 5 counres. Alhough Chna s accesson no WO wll have an mporan mpac on he ASEA counres especally when Chna s FDI nake and global share have ncreased n a perod when global FDI nflows have conraced Chna s rse should be nerpreed as wn opporuny. hs s because whle FDI furher promoes Chna s expor also expands s domesc marke where he ASEA counres can drecly benef by recordng surgng expors o Chna. A he same me Chna s cos compeveness forces he ASEA counres o move up he value chan by shfng from low echnology manufacurng o hgher value added pushng for economcs as well as ndusral ransformaon. I would ake several years before Chna expors FDI n sgnfcan amouns. he ASEA counres should ake hs opporuny o acvely pursue nsead of encumber economc cooperaon and free rade wh Chna. In shor he emprcal resuls confrm ha nflows of FDI no Chna by he ASEA 5 counres are sgnfcanly nfluenced by all seleced varables excep exchange rae. herefore he macroeconomc fundamenal changes are mporan facors ha arac he ASEA 5 o

21 Journal of Economc Cooperaon and Developmen nves capal n Chna. he ASEA 5 counres have also an mporan sake n encouragng FDI nflows n Chna. For susanng and enhancng good cooperaon beween Chna and he ASEA 5 counres he fuure flow of FDI from hese counres need o be mananed hrough muual benef agreemen and effecve foregn nvesmen polces.

22 Analyss of FDI Inflows no Chna from ASEA-5 Counres: References Acemoglu D. and Zlbo F. (997). Was promeheus unbound by chance? Rsk dversfcaon and growh. Journal of Polcal Economy 05(4) Adhkary B.K. (0). FDI rade openness capal formaon and economc growh n Bangladesh: A lnkage analyss. Inernaonal Journal of Busness and Managemen 6() 6 8. Azenman J. (99). Exchange rae flexbly volaly and paerns of domesc and foregn drec nvesmen. Inernaonal Moneary Fund Saff Paper 39(4) Al S. and Guo W. (005). Deermnans of FDI n Chna. Journal of Global Busness and echnology () 33. Baek. I M. and Okawa. (00). Foregn exchange raes and Japanese foregn drec nvesmen n Asa Journal of Economcs and Busness Baley M.J. and avlas G.S. (99). Exchange rae volaly and drec nvesmen Journal of he Amercan Academy of Polcal and Socal Scence Barhelemy J.C. and Demurger S. (000). Foregn drec nvesmen and economc growh: heory and applcaon o Chna Revew of Developmen Economcs 4() Blongen B.A. (997). Frm specfc asses and he lnk beween exchange raes and foregn drec nvesmen. Amercan Economc Revew Borenszen E. De Gregoro J. and Lee J.W. (998). How does foregn drec nvesmen affec economc growh? Journal of Inernaonal Economcs Bosworh B.M. Collns P.S. and Renhar C.M. (999). Capal flows o developng economes: Implcaons for savngs and nvesmen. Brookngs Papers on Economc Acvy ()

23 Journal of Economc Cooperaon and Developmen 3 Buckley P.J. Clegg J. Forsans. and Relly K.. (00). Increasng he sze of he counry: Regonal economc negraon and foregn drec nvesmen n a globalsed world economy. Managemen Inernaonal Revew 4(3) Calvo G.A. Lederman L. and Renhar C.M. (996). Inflows of capal o developng counres n he 990s. Journal of Economc Perspecve 0() Campa J.M. (993). Enry by foregn frms n he Uned Saes under exchange rae uncerany. Revew of Economcs and Sascs Canzoner M. Cumby R. and Dba B. (999). Relave labour producvy and he real exchange raes n he long run: Evdence for a panel of OECD counres. Journal of Inernaonal Economcs Chaudhary G.M. Al Shah S.Z. and Mahmood Bagram M.M. (0). Do exchange rae volaly effecs foregn drec nvesmen? Evdence from seleced Asan economes. Journal of Basc and Appled Scenfc Research (4) Chou W.L. and Lee D.S-Y. (003). Conegraon analyss of aud prcng model: A panel un roo es approach. Journal of Busness Fnance and Accounng 30(7 & 8) Chowdhury A. and Mavroas G. (003). FDI and growh: Wha causes wha? Paper presened a he Wder Conference Sepember. Chrsopoulos D.K. and sonas E.G. (004). Fnancal developmen and economc growh: Evdence from panel un roo and conegraon ess. Journal of Developmen Economcs Coeux M. and Olver S. (000). he savng reenon coeffcen n he long run and n he shor run: Evdence from panel daa. Journal of Inernaonal Money and Fnance Cushman D. (988). Exchange rae uncerany and foregn drec nvesmen n he Uned Saes. Welwrschaflches Archv

24 4 Analyss of FDI Inflows no Chna from ASEA-5 Counres: De Meza D. and van der Ploeg F. (987). Producon flexbly as a move for mulnaonaly. Journal of Indusral Economcs 35(3) Dewener K. (995). Do exchange rae changes drve foregn drec nvesmen? Journal of Busness 68(3) Dx A. and Pndyck R. (994). Invesmen under uncerany Prnceon Unversy Press. Domar E.D. (946). Capal expanson rae of growh and employmen. Economerca Ellngsen G. Lkumahuwa W. and unnenkamp P. (006). Ouward FDI by Sngapore: A dfferen anmal? ransnaonal Corporaon. 5() 40. Esrn S. Rche X. and Brada J.C. (000). Foregn drec nvesmen n Cenral Easern Europe: Case sudes of frms n ranson he mcroeconomcs of ranson economes. M. E. Sharpe Inc. ew York USA Froo K.A. and Jeremy C.S. (99). Exchange raes and foregn drec nvesmen: An mperfec capal markes approach. Quarerly Journal of Economcs Goldberg L.S. and Kolsad C.D. (995). Foregn drec nvesmen exchange rae varably and demand uncerany. Inernaonal Economc Revew Goschalk S. and Hall S. (008). Foregn drec nvesmen and exchange rae uncerany n Souh-Eas Asa. Inernaonal Journal of Fnance and Economcs Harrod R F. (939). An essay n dynamc heory. he Economc Journal Im K.S. Pesaran M.H. and Shn Y. (997). esng for un roos n heerogeneous panels. Manuscrps Deparmen of Appled Economcs Unversy of Cambrdge.

25 Journal of Economc Cooperaon and Developmen 5 Im K.S. Pesaran M.H. and Shn Y. (003). esng for un roos n heerogeneous panels. Journal of Economcs Jordan S. (00). A VAR approach o economcs of FDI n Chna. Appled Economcs 34(7) Lee K. Pesaran H.M. and Smh R. (997). Growh and convergence n a mul counry emprcal sochasc Solow model. Journal of Appled Economercs Lee H.H. and an H.B. (006). echnology ransfer FDI and growh n he ASEA regon. Journal of he Asa Pacfc Economy (4) Levn A. and Ln C.F. (99). Un roo es n panel daa: asympoc and fne sample properes. Dscusson Paper o Calforna USA: Unversy of Calforna a San Dego. Levn A. and Ln C.F. (993). Un roo es n panel daa: new resuls. Dscusson Paper o Calforna USA: Unversy of Calforna a San Dego. L X. and Lu X. (005). Foregn drec nvesmen and economc growh: An ncreasngly endogenous relaonshp World Developmen 33(3) Lloyd P.J. (996). he role of foregn nvesmen n he success of Asan ndusralzaon. Journal of Asan Economcs aughon B. (996). Chna's emergence and prospecs as a radng naon. Brookngs Papers on Economc Acvy gowan. A. (0). RMB exchange rae volaly and s mpac on FDI n emergng marke economes: he case of Zamba Inernaonal Journal of Busness and Socal Scence 3(9) 9 5. yarko P.A. and keah-amponsah E. (0). Effec of exchange rae regms on FDI nflow n Ghana Inernaonal Journal of Economcs and Fnance 3(3)

26 6 Analyss of FDI Inflows no Chna from ASEA-5 Counres: Pan. and Van Welsum D. (003). Unyng he Gordan kno: he mulple lnks beween exchange raes and foregn drec nvesmen. Journal of Common Marke Sudes 4(5) Pedron P. (996). Fully Modfed OLS for heerogenous conegraed panels and he case of purchasng power pary. Workng paper orh Amercan Economerc Socey Summer Meeng Pedron P. (999). Crcal values for conegraon ess n heerogeneous panels wh mulple regressors. Oxford Bullen Econonomc Sascs Specal Issue o Pedron P. (000). Fully Modfed OLS for heerogeneous conegraon panel n nonsaonary panels panel conegraon and dynamc panels. Advances n Economercs Pedron P. (00). Purchasng power pary ess n conegraed panels. he Revew Economc Sascs 83(4) Pedron P. (004). Panel conegraon: Asympoc and fne samples properes of pooled me seres ess wh an applcaon o he PPP hypohess. Economc heory Phllps S. and Fredoun Z.A. (008). Exchange raes and foregn drec nvesmen: heorecal models and emprcal evdence. he Ausralan Journal of Agrculural and Resource Economcs Phllps P.C.B. and Hansen B.E. (990). Sascal nference n nsrumenal varables regresson wh I() process. Revew of Economc Sudes 57() Sarans. and Sewar C. (999). Is he consumpon-ncome rao saonary? Evdence from panel un roo ess. Economc Leers Shu Y. and Zeng K. (006). FDI flows beween Chna and ASEA: Emergng facors and prospec. Chna & World Economy. 4(6)

27 Journal of Economc Cooperaon and Developmen 7 Schneder F. Bruno S. and Frey B.S. (985). Economc and polcal deermnans of foregn drec nvesmen World Developmen 3() Solow R.M. (956). A conrbuon o he heory of economc growh. Quarerly Journal of Economcs Sumner A. (005). Is foregn drec nvesmen good for he poor? A revew and sock ake. Developmen n Pracce 5(4) oda H.Y. and Yamamoo. (995). Sascal nference n vecor auoregressons wh possble negraed processes. Journal of Economercs sa P. (994). Deermnans of foregn drec nvesmen and s mpac on economc growh. Journal of Economc Developmen 9() Ullah S. Hadeer S. Z. and Azm P. (0). Impac of exchange rae volaly on foregn drec nvesmen: A case sudy of Paksan Paksan Economc and Socal Revew 50() 38. UCAD. Uned aons Conference on rade and Developmen) (996) World Invesmen Repor. ew York: Uned aons 97. Walsh J.P. and Yu J. (00). Deermnans of foregn drec nvesmen: A secoral and nsuonal approach. Inernaonal Moneary Fund Workng Paper WP/0/87 7. Wang Z.Q. and Swan.J. (995). he deermnans of foregn drec nvesmen n ransformng economes: Evdence from Hungary and Chna. Welwrschaflches Archv We Y. and Lu X. (00). Foregn drec nvesmen n Chna: Deermnans and mpac Edward Elgar UK. Xu G. and Wang R. (007). he effec of foregn drec nvesmen on domesc capal formaon rade and economc growh n a ranson economy: Evdence from Chna. Global Economy Journal 7().

28 8 Analyss of FDI Inflows no Chna from ASEA-5 Counres: Zhang K.H. (999). How does FDI nerac wh economc growh n a large developng counry? he case of Chna. Economc Sysems 3(4) Zhang K.H. (000). Why s U.S. drec nvesmen n Chna so small? Conemporary Economc Polcy 8() Zhang K.H. (00). Why does Chna receve so much foregn drec nvesmen? Chna & World Economy