A Macro Assessment of China Effects on Malaysian Exports and Trade Balances

Transcription

1 MPRA Munich Personal RePEc Archive A Macro Assessmen of China Effecs on Malaysian Expors and Trade Balances Tze-Haw Chan and Hooi Hooi Lean and Chee Wooi Hooy Graduae School of Business, Universii Sains Malaysia, Economics Program, School of Social Sciences, Universii Sains Malaysia, Finance Secion, School of Managemen, Universii Sains Malaysia 1. April 2013 Online a hp://mpra.ub.uni-muenchen.de/48801/ MPRA Paper No , posed 2. Augus :22 UTC

2 A Macro Assessmen of China Effecs on Malaysian Expors and Trade Balances Tze-Haw Chan* Graduae School of Business Universii Sains Malaysia USM, Penang, Malaysia Tel: ; Fax: ; Hooi-Hooi Lean, Economics Program, School of Social Sciences, Universii Sains Malaysia USM, Penang, Malaysia Tel: ; Fax: ; Chee-Wooi Hooy Finance Secion, School of Managemen Universii Sains Malaysia USM, Penang, Malaysia Tel: ; Fax: ; ABSTRACT This paper focuses on he impac of China s expor expansion on Malaysian monhly rading wih o her 12 major rading parners over he liberalizaion era. Srucural break(s) found mosly coincides wih he Asia financial crisis and China s accession ino WTO and, regime shifs are eviden in he long run relaionship among he variables being sudied. While he income effecs are more apparen, real exchanges are raher insignifican and incorrecly signed for Malaysian bilaeral rading. An aemp o correc curren accoun imbalances by currency devaluaion would hereby inappropriae. In addiion, esimaion of he rade balance models is more superior ha complemenary China effecs are beer capured for Malaysia rading wih he advanced markes such as Ausralia, German, Japan, UK and he US. Such finding may parly due o he increase in global produc fragmenaion. Keywords: Malaysian expors, China effec, coinegraion, VAR, srucural break JEL Caegories: F31, C22, C51 * Corresponding auhor 1

3 1. INTRODUCTION The economic boom in Malaysia from he early 1990s unil he onse of he 1997 Asian financial crisis was underpinned by rapid expor growh o developed markes, mainly he US, Japan and he EU core members. The expor-lead policy has been quie successful and i has remained as growh impeus for Malaysia GDP in he recen decades. However, he rising of People Republic of China (PRC) as he world facory has raised a concern wheher he expor oriened growh policy will be susainable. PRC has consisenly achieved double-digi growh raes over he las wo decades. Likewise, PRC s expor has argeed on developed markes especially he US and he EU, which is also among he main expor desinaions of Malaysia. Moreover, he expor srucure of PRC is quie similar wih Malaysia as he base expor of boh counries is on labor-inensive manufacuring producs. The accession of PRC ino WTO in 2001 has furher enhanced he magniude of expor flow from PRC o he developed markes. Since hen PRC has become he larges final assembly base before consumer goods are expored o he developed naions. Mos of he pars and componens are shipped from Souheas Asia (ASEAN) and mos likely will shrink he direc expors of Souheas counries, including Malaysia, o he developed naions. No doub as a close neighbor of PRC, Malaysia expor performance is under pressure. Table 1 provides some general saisics o suppor such argumen. Wih he upward rending of expor value over GDP, one can ell ha boh PRC and Malaysia were geing more reliance on expors. Bu unlike PRC ha has experienced marvelous growh of expors afer he 1990s; Malaysian expors grow a slower pace since mid-1990s. The figures deceleraed in some of he years afer he 1997 Asian financial crisis and during he 2008 Subprime crisis. In recen world expor ranking, 2

4 PRC has been opped globally, while Malaysia is saying sagnan a he 20 h posiion. Looking a Table 2, boh PRC and Malaysia share similar desinaion of expor flows, wih slighly differen in he rade-parner ranking due o heir dependency on he neighboring world level enrepos, i.e. Hong Kong and Singapore. Boh counries also expor relaively more o heir neighboring counries, i.e. China o Souh Korea and Russia, while Malaysia o her ASEAN counerpars; bu generally boh counries expor o he same developed markes, i.e. he US, Japan and he Wesern EU naions. Inser [Table 1, Table 2] abou here Following he economy liberalizaion and recen rade expansion of China, he PRC compeiive hrea hypohesis has gained increasing aenion among scholars. A few sudies ha focus on non-neighboring counries, namely he Lain American and Caribbean markes, have repored some significan PRC crowding-ou effecs (see for example, Quinin, 2004; Lall e al., 2005; Jenkins a al., 2006). Ohers, on he oher hand, would sugges ha PRC expor expansion has a complemenary effec for is neighbors. Frenald e al. (1999), for insance, repored ha PRC expor expansion as a resul of Renmimbi devaluaion in 1994 did no poses any crowding-ou effec on NIEs (Korea, Singapore and Taiwan) and ASEAN-4 (Indonesia, Malaysia, Philippine, Thailand). Insead, PRC s real expor growh was posiively correlaed wih real expor growh of hose counries. Eichenggreen e al. (2007) employed panel regression of 13 Asian counries (including Malaysia) wih heir 180 imporing counries under an augmened graviy model also repored a posiive effec of PRC expor expansion. On op of ha, hey did find some exen of crowding-ou effec when disaggregaed daa were used, mainly for less developed Asian economies and in consumer goods marke which is considered as low-echnology expor; bu no in 3

5 markes for capial goods, which is considered as high-echnology expor. Using a graviy approach, Greenaway e al. (2008) hen documened empirical evidence ha PRC expor over displaced oal expor of high income Eas Asia counries o hird counry, bu for middle income Eas Asia counries, including Malaysia, PRC expor expansion was acually complemenary. Clearly, a presen sage, he lieraure has no been conclusive abou he PRC compeiive hrea and furher exploraion is o be aken. This sudy akes a differen approach o invesigae he issue peraining o Malaysian bilaeral rading wih her 12 major parners. Firs, we assess he impacs of PRC expor expansion via wo models: he Malaysian expor model and he Malaysian rade balance model, by having China expor as forcing variable. In oher words, we examine if he China effec homogeneous across counries or varies by he expor inensiy and rade balances (expor-impor raios). Such combined analysis is no ye o be occupied in he exising lieraure. 1 Second, we use higher frequency monhly daa (January 1990-June 2010) which allows us o capure he dynamic of bilaeral rade flows and hence a more 1 Mos sudies on Malaysia case employed he rade balance framework and none has assessed he China hrea hypohesis. Yusoff (2007) found ha he Malaysian rade balances are bounded by he real exchange rae, domesic and world incomes a aggregae level, and a delayed J-curve was suppored. Bu in Yusoff (2009), Malaysia's bilaeral rade balances are found o be responsive o he changes of he US and Singapore bilaeral exchange rae bu no he Japan case, and he J-curve only appeared for Malaysia-US. Bahmani-Oskooee and Harvey (2010) hen sudied he Malaysia s rade balances agains 14 rading parners during 1973Q1-2001Q3, bu wihou assessing he China effec and did no cover he recen period of boh do com crisis and Subprime crisis, which have significan impacs on he demand for Malaysian expors. They repored inconclusive suppor of income effecs and real exchange rae impacs on Malaysia rade balances. Moreover, ou of 14 bilaeral cases, he J-curve was only found for Malaysia versus Germany. Chan and Hooy (2012), on he oher hand, direcly examine he long run dynamics of exchange rae and bilaeral expor-impor flows beween China and Malaysia. Their finding reveals ha he Marshall-Lerner condiion holds in he long run bu he expor-impor demands do no adhere o he J-curve paern. Alhough he expansionary effec is of greaer evidence for Malaysia due o real exchange shocks bu inconclusive for China, hey found some evidence ha he China-Malaysia bilaeral rading is along he susainable pah. 4

6 efficien esimaion can be obained. As he lieraure is mosly relying on annual or quarerly daa, he esimaion of China s poenial crowding-ou effec could be relaively less efficien or no accuraely capured. Our analysis generally shows insufficien evidence of China s crowding-ou effec on Malaysian expor and bilaeral rading. Insead, here are some empirical evidences ha PRC s expor expansion gives a complemenary effec on Malaysia s rade wih her major parners, by and large, due o he increase in global producion neworks of ASEAN5 wih PRC over he las decade. In our sudy, he concern of srucural break(s) is also aken ino accoun. This is paricularly imporan as our sample period covers a number of economics shocks and policy changes ha may affec he srucure of global rading (e.g. Asian financial crisis, China s access o WTO and he Subprime crisis). Lieraure in recen years has also reached consensus ha he disregard of srucural changes leads o inefficien esimaion and herefore lower esing power of univariae uni roo ess (see Perron, 1989; Lee and Srazicich, 2004; among ohers). Likewise, lack of careful invesigaion of poenial srucural breaks may also lead o misspecificaion of he long-run properies of a dynamical mulivariae sysem and inadequae esimaion and esing procedures, e.g. coinegraion ess (Gregory e al., 1996; Gregory and Hansen, 1996; Esso, 2010). Thereby, he Lagrange Muliplier (LM) uni roo break es proposed by Lee and Srazicich (2004) and he Gregory and Hansen (1996) s coinegraion es wih break (GH) are uilized in our analyses. The LM uni roo es ends o esimae he endogenous (unknown) break poin correcly and is free of size disorions and spurious rejecions in he presence of a uni roo wih break (Acaravci and Ozurk, 2010). Meanwhile, he advanage of GH es lies on he abiliy o rea he issue of 5

7 endogenous break and coinegraion alogeher (Esso, 2010). The es procedure offers hree models corresponding o hree assumpions concerning he naure of shif in he coinegraing vecor: he level shif model (C), he level shif wih rend model (C/T), and he regime shif model (C/S). The res of paper is organized as follows: in secion 2, daa and mehodology are discussed; secion 3 repors our empirical resuls and finally in secion 4 we conclude. 2. DATA AND METHODOLOGY 2.1 Daa Our sudy primarily focuses on he Malaysian op-15 rading parners. Since China has been chosen as forcing variable, Malaysia-China expor is excluded from he dependen variable lis. Malaysia-Hong Kong expor is also dropped due o Hong Kong s unique saus as an enrepo for China. Malaysia-UAE is hen excluded because he series is incomplee. Thus, in oal, 12 rading parners are covered in his sudy, which are Ausralia, Germany, India, Indonesia, Japan, Souh Korea, Neherland, he Philippines, Singapore, Thailand, Unied Kingdom and Unied Saes. Monhly daa spanning from January 1990 o June 2010 are uilized in our analysis wih he excepion of Ausralia (quarerly daa). Daa used in his sudy include Malaysian expors o her op-12 rading markes, Chinese expors o he same desinaions, indusrial producion indexes of he welve rading parners, and bilaeral real exchange raes ha compiled from bilaeral nominal exchange rae and consumer price index. All daa are exraced from DaaSream and IMF Direcion of Trade Saisics (DOTS), and ransformed ino logarihmic scale prior o he analysis. Due o 6

8 daa unavailabiliy, he analysis period varies for some counries. The wo EU counries, i.e. Germany and Neherlands sar from January 1999, India sars from April 1994, Korea sars from January 1991 and he Philippines sar from January Empirical Model A simple demand funcion of Malaysian expor is given as follow: MALEX f Y, RER i * i i (1) * MALTB f Y, Y, RER (2) i i i i Equaion (1) saes ha Malaysian expor o a foreign counry i, ( dependen on he income level of he rading parner MALEX i ) is * Y i (e.g. China) and he relaive real prices of he wo currencies, i.e. he real exchange rae, RER. Whereas equaion (2) highlighs he convenional heory ha Malaysian rade balances ( MALTB i ) will be affeced by boh he domesic and foreign incomes as well as real exchange raes. The formula below is employed o compue he real exchange rae (RER) beween Ringgi Malaysia and a foreign currency: i RER i RM ER i CPI x CPI i M where RM is Ringgi Malaysia, ER i is he currency of foreign counry i, CPI i and CPI M are he consumer price index of counry i and Malaysia, respecively. To address he PRC effec on Malaysian expor demand, we consider he augmened version of model (1) and (2), which given by he following specificaions: 7

9 ln (3) * MALEX i, 1 ln IPi, 2 ln RERi, 3 ln CHNEX i, ln (4) * MALTB i, 1 ln IPi, 2 ln IPi, 3 ln RERi, 4 ln CHNEX i, where lnmalex i and lnmaltb i are he naural logarihmic of Malaysian expors o foreign counry i and rade balance wih counry i respecively. The rade balance is usually defined as he bilaeral expor-impor raio. lnrer i is he naural logarihmic of real exchange rae beween Ringgi Malaysia and he currency of counry i, lnip i and lnip * i are he respecive naural logarihmic of domesic and foreign indusrial producion as proxy for he domesic (Malaysia) and foreign income of rading parner, and ε as error erm. For model (3), we expec > 0 and 1 2 > 0 since an increase in foreign income would lead o an increase demand for Malaysian expors, whereas real Malaysian ringgi depreciaion (RER posiive) is also expeced o increase Malaysian expors. For model (4), similar jusificaion is applied. When Malaysian income rises, domesic impors will increase and hence rade balance deerioraes. If Malaysian ringgi depreciaes, rade balance will improve due o cheaper bu beer-demanded expors and declined impors which are more expensive. Such siuaion will occur only if Marshall-Lerner condiion is me. So, we expec < 0, > 0, > Unlike he exising lieraure, our sudy incorporaes an addiional facor, lnchnex i (China s expor o he same desinaion marke) o capure he poenial long run crowding-ou effec of China expor expansion. If he Chinese expors exhibi subsiuing effec on Malaysian expors o he same desinaion, 3 or 4 will repor a negaive sign. This implies ha he more Chinese goods are expored, he lesser he demand for Malaysian expors in he similar markes. Oherwise, 3 or 4 8

10 will display a posiive sign o imply complemenary effec. The above seing allows us o esablish and examine he China crowding-ou hypohesis. 2.3 Uni roo ess Following he sandard pracice for ime series analysis; we begin our empirical evaluaion by examining he saionary properies of all variables concerned. Many of he exising sudies used he Augmened Dickey-Fuller (ADF) or Phillips-Perron (PP) uni roo es o ascerain he order of inegraion of he series. A problem wih hese ess is neiher allows for he possibiliy of a srucural break. Perron (1989) showed ha he power o rejec a uni roo decreases when he saionary alernaive is rue and a srucural break is ignored. In oher words, he failure o allow for poenial break leads o a bias ha reduces he abiliy o rejec a false uni roo null hypohesis. We employ he Lagrange Muliplier (LM) uni roo es wih one srucural break proposed by Lee and Srazicich (2004). LM uni roo es wih one srucural break has he major advanage ha i is unaffeced by he exisence of a srucural break under he null (see Lee and Srazicich, 2001). The LM uni roo es can be explained using he following daa generaing process: y Z e, e e 1 (5) where Z consiss of exogenous variables and is an error erm wih classical properies. Model A which is also known as he crash model, allows for a one-ime change in he inercep under he alernaive hypohesis. Model A can be described by ' Z 1,, D, where D 1 for T 1, and zero oherwise; T B is he dae of he B srucural break, and δ' = (δ 1, δ 2, δ 3 ). The LM uni roo es saisic is obained from he following regression: 9

11 y Z S 1 (6) where S y ˆ ˆ x Z, 2,..., T ; ˆ are coefficiens in he regression of y on Z ; ˆ x is given by y Z ; and y 1 and Z 1 represen he firs observaions of y and Z respecively. The LM es saisic, is given by -saisic for esing he uni roo null hypohesis ha 0. The locaion of he srucural break T B is deermined by selecing all possible break poins for he minimum -saisic as follows: inf i inf, where T B T. (7) The search is carried ou over he rimming region (0.1T, 0.9T), where T is sample size. To selec he lag lengh, we used he general o specific procedure proposed by Hall (1994). We se he maximum number of lags equal o 12 and used he 10% asympoic normal value of o ascerain he saisical significance of he las firs-differenced lagged erm. Afer deciding he opimal lag lengh for each breakpoin, we deermine he break where he endogenous LM -es saisic is a a minimum. Criical values for he LM uni roo es wih one srucural break are abulaed in Lee and Srazicich (2004). 2.4 Coinegraion es Once he order of inegraion of each variable is ascerained, we es for coinegraion for he long-run relaionship beween Malaysian expors o counry i wih he respecive independen variables as saed in equaion (3) and (4). Gregory and 10

12 Hansen (1996) proposed hree models for esing coinegraion where he exisence of a srucural break in he coinegraing vecor is allowed. Model C:, = 1,, n (8) Model C/T:, 1,..., n (9) Model C/S:, 1,..., n (10) Model C conains a level shif, Model C/T conains level shif and rend, and Model C/S allows for regime shif. Then, D 0 for and D 1 for. X can be referring o he marix of all relaed dependen variables and β is he vecor of esimaed coefficiens. Here, β denoes he coinegraing slope coefficiens before he regime shif and γ denoes he change in he slope coefficiens. In order o es for coinegraion beween dependen and independen variables wih srucural change, i.e. he saionariy of u in Equaions (8) (10), Gregory and Hansen (1996) proposed a suie of ess. These saisics are he commonly used ADF saisics and exensions of he Z and Z es saisics proposed by Phillips (1987), which defined as: * ADF ADF (11) inf T * Z Z (12) inf T Z inf Z (13) * T As he break poin,, is unknown a priori, he model is esimaed recursively allowing he break poin o vary beween (0.1T, 0.9T), where T is he sample size. The null hypohesis of no coinegraion is examined using he hree saisics wih ineres in he smalles values for he hree saisics across all break poins required o 11

13 rejec he null. As robusness es, we also perform he Johansen-Juselius (1990) coinegraion es based on he following vecor auoregressive (VAR) specificaions: and (14) (15) 3. EMPIRICAL RESULTS AND DISCUSSION 3.1 Uni Roo and Breaks The LM uni roo es wih break generally suppors ha all variables in each of he counries are inegraed of order one a he 5 per cen level or beer excep he Malaysian expor o Ausralia. 2 Basically no all he series has srucural break and he significan break for differen variables of a counry is also differ. We do no inend o examine he deail of he break poins for all he series bu some discussion on he break poins of expor series is relevan o our research focus. We noe ha he locaion of he significan breaks for Malaysian expor series ook place mainly around he period of 1997 Asian financial crisis, while he Chinese expor series have many breaks surrounding 1992/93. For Malaysia, he break for expor o her wo larges ASEAN members, i.e. Singapore and Thailand was happened afer he implemenaion of capial conrol and Ringgi pegging in 2 We find ha For Malaysian expor, he break in he inercep is saisically significan a he 5 per cen level or beer in eigh ou of he welve counries. For Chinese expor, he break in he inercep is saisically significan a he 5 per cen level or beer in seven ou of he welve counries. Ausralia, Indonesia and Neherland are he hree ha do no have any significan break in he inercep for boh expors from Malaysia and China. Deails of he uni roo ess are available upon reques. 12

14 Sepember The break for Souh Korea, he Philippines and India was happened before he oubreak of he crisis in July 1997 while he break for Germany occurred in November 2005, exacly he monh when Malaysia ended her Ringgi pegging regime. For Japan, he break was in 2008, he ransiion period which China came close o replace Japan as Malaysia s hird larges expor desinaion since he las wo decades. 3 Malaysian expor o US has a break as early as in This break was happened when Malaysia s elecronic and elecrical (E&E) expor, which accouned for nearly 40% of her oal manufacured expor, had been shifed o Singapore in The expor share of elecronic and componens o US for example, which accouned for nearly 50% of Malaysian E&E expor since he mid-1980, has dropped o 30% in On he oher hand, many of he breaks in 1992 and 1993 ha have been deeced for China s expor were mainly due o is major economic ransformaion which sared in The break for Chinese expor o he Philippines was happened in 1997 due by large o he devaluaion of Peso in July. 3.2 Long run Esimaes Given ha all variables are I(1) for each of he counries we proceed o es for coinegraion wih a srucural break in he coinegraion vecor using he Gregory and Hansen (1996) es. The resuls are presened in Table 3a and Table 3b respecively. In general, we find coinegraion relaionship for expor model and rade balance model beween variables sudied in each of he 12 counries. Even hough he null 3 In fac, based on he monhly expor daa from IMF DOTS, in 2008, China expor value was higher han Japan from April o Sepember, excep June. Afer March 2009, Malaysia expor o China has surpassed hose o Japan and he US and since Augus 2009 he value has come close o Singapore which has been he larges desinaion of Malaysian expor since he drop in Malaysian expor o he US in Ocober

15 hypohesis of no coinegraion is no able o be rejeced wih he ADF* saisic for all hree expor models in India, Singapore and UK (see Table 3a), boh Z and Z α saisics show srong evidence of significan a 1 per cen level for all hree models in all counries. Hence, we conclude ha here exiss long-run relaionship beween Malaysia s expor and real exchange rae, economic growh and China s expor in all welve counries. If we refer o Table 3b for rade balance models, similar jusificaion is applied o he Malaysia-UK case. For oher 11 rade balance cases, hypohesis of no coinegraion is highly rejeced and long run relaionships among he variables are confirmed. Inser [Table 3a, Table 3b] abou here There are a range of break poins across he es saisics for expor models. In general he break in Ausralia, Japan, and he EU counries occurs in 2001/02 coincides wih China s accession ino WTO. Only for Singapore he break was happened during he Asian financial crisis. Indonesia and Thailand have a break in 1993/94. The break for Korea was in 2006 while for he Philippines i was in The break dae of he US was in he mid of 1996 and India showed a diver break across models. As for he rade balance models, break daes happen mosly during he 1997/98 Asia crisis and he 2001do com bubble. The break dae of 2008 Subprime crisis was deeced only for he Malaysia-Philippines case. The unrepored resuls of Johansen coinegraion es also show a leas one coinegraing equaion is found in all counries infers ha here exiss long-run relaionship beween he variables. Hence, we are confiden o examine he long-run relaionship beween Malaysian expor and Chinese expor. The normalizaion resuls 14

16 from coinegraion es wih lnmalex as he dependen variable or in oher words he long-run elasiciies of Malaysian expors wih respec o oher variables are repored in Table 4a. We find ha China s expor posiively affecs he expor of Malaysia o five of her major rading parners including India, Indonesia, Japan, he Philippines, and he UK. China s expor posis a significan negaive impac only on Malaysian expor o Singapore. Ye, Ausralia, Germany, Korea, Neherland, Thailand and he US do no show any significan impac due o Chinese expors. Excep for Singapore and India, he significan coefficiens of China effec (CHNEX) are generally less elasic, ranging from o If we refer back o Table 2, he sum of oal expors o hese naions (excluding Japan) are less han 9% of Malaysian oal expors over , which mainly consis of refined peroleum, palm oil, iron & seel producs. These long run posiive effecs may be offse by he negaive effecs on expors o Singapore (15% of Malaysian expors, ), which are mainly elecrical and elecronic producs. Bu since Singapore is an enrepor ha provide bes springboard for Malaysian exporers o venure abroad, i is sill early o conclude if Malaysian expor could suffer from China s rea in he high-ech and low-ech manufacuring expors as highlighed by Halmaier e al. (2007). A he same ime, i is worh noing ha he long run resuls have no provided sufficien suppor o Walmsley and Herel (2000) who claimed ha China's compeiors in he labor-inensive apparel indusry would experience significan losses in real income, parly due o declining erms of rade. Inser [Table 4a] abou here 15

17 On he oher hand, he impac of real exchange raes and demand side effec owing o income changes (proxy by indusrial producion) has been raher mixed. Expors are responding negaively o he depreciaion of he US, Singapore and he Philippines real exchange raes (RER negaive = ringgi appreciaes) bu responding posiively agains he depreciaion of real Indian Rupee, Indonesian Rupiah and Korean Won (RER posiive = ringgi depreciaes). Such mixed resul has been repored by Baharumshah (2001), Bahmani-Oskooee and Wang (2006) and Ahmad and Yang (2007), who found limied suppor of posiive long-run effec of foreign exchange on rade. As for he income effecs (demand side) in long run, Malaysia expors are mosly affeced by producion growh (as proxy of income) in he rading parner excep for he Philippines, Korea and UK. Demand for Malaysian expor rises when producion expands a Ausralia, Germany, Neherland, Singapore, Thailand and he US. Oherwise, expors respond negaively o producion growh a India, Indonesia and Japan. The effec varies among rading parners may be due o he producion and expor diversificaion. Togeher, naions ha show posiive income effecs have accouned for abou half of he oal Malaysian expors during and, mos of he Malaysian expors o hese developed naions (excep Thailand) comprise of medium- and high-ech manufacuring expors. Such finding is in line wih Eichengreen e al. (2004, 2007) ha Chinese expor growh is likely o have negaive consequences for relaively less echnologically advanced consumer goods-exporing Asian counries, bu posiive consequences for exporers of more sophisicaed capial goods. As for he Japan case, wo saes of affairs are of concern. Firs, he Japanese economic downurn in 1990s and sluggish recovery in he 2000s; second, duy-free 16

18 goods consiued 86% of Japan s impors from Malaysia (including duy-free reamen under he Generalized Sysem of Preferences (GSP)) and 67% of Malaysia s impors from Japan during 1990s-2000s (see Japan-Malaysia Economic Parnership, 2003). Togeher, boh scenarios fail o provide posiive momenum on Japanese income-demand for Malaysian expors. However, o verify he deails of such inconsisen effecs among he rading parners, we have o esimae horoughly a micro-level, which a presen sage is beyond he scope of presen paper. Unlike he expor models, he long run esimaes of rade balance models are more consisen and sraighforward (see Table 4b). Nine ou of welve cases, he domesic income coefficiens have been significan wih expeced negaive sign. Malaysian incomes are repored insignifican only for he India, Indonesia and Neherland cases. As for he foreign income variables, all coefficiens are posiive of which only hree cases are insignifican (Germany, India and Neherland). Such finding is consisen wih hose prediced by economic heory where income and purchasing power have been he main deermining facor of expors and impors. So, as far as domesic and foreign incomes are concerned, he Malaysian bilaeral rading wih major parners are demand-driven. Bu when we assess he real exchange rae effec on rade balance, he inconsisency remains. Half of he coefficiens are repored significan wih only five carry he expeced negaive sign. Such resul does no allow he Malaysian auhoriy o coun on devaluaion as ool of curren accoun correcion. On he oher hand, Table 4b also shows ha he China s expors have exhibied complemenary effecs o Malaysian rading in he long-run, mainly for advanced expor desinaions such as Ausralia, Germany, Japan, UK and he US. 17

19 Subsiuing effecs are only repored for he Malaysia-Singapore and Malaysia-Korea cases. Inser [Table 4b] abou here 3.3 Shor run Esimaes and Adjusmens One needs o be careful when inerpreing he consisency of long- and shor-run resuls of China effec. For expor models, only India and UK repored posiive and significan Chinese effec in boh long-run and shor-run, he coefficiens of he res are differen (see Table 5a). Occurrence of such inconsisency is possibly owing o he various marke and policy responses in shor-run due o he series of global and regional economic shocks during s, which include he 1997/98 Asia financial crisis, 2001 IT bubble, 2008 subprime crisis, as well as global food and crude oil shocks, ec. Bu overall, he shor run dynamics are fairly consisen wih long run esimaes ha he crowding ou effec is ligh. Only wo rading parners have repored compeing role (negaive effec) while six ohers have repored significan complemenary role (posiive effec) of Chinese expors (see Table 5a). In addiion, he coefficiens of DCHEX range from o , suggesing a much possible scenario ha China effec could be over-projeced in he lieraure. As for he rade balance model repored in Table 5b, Ausralia and Japan have shown complemenary China effecs in he shor run, Indonesia has shown subsiuing effec while he res are insignifican. Unil his poin, i seems like complemenary China effec has been favoring he Malaysian rading in he advanced markes raher han her ASEAN neighboring naions. Inser [Table 5a, Table 5b] abou here 18

20 Quie surprising, however, he exchange rae effec almos disappears in he shor run excep for Malaysia-US (expor model-table 5a) and Malaysia-Ausralia (rade balance model-table 5b). Such finding is suppored by Rose and Yellen (1989) who rejeced he exchange rae-rade nexus among US-G7. Then, he income effecs are also rivial in shor run and only found significan in he US, Singapore, Germany and India wih smaller coefficiens. In he rade balance model, only Thailand and he US cases have shown significan domesic income, bu he coefficiens are posiive (unexpeced) in he shor run and negaive (expeced) in he long run. A possible explanaion provided by Yusoff (2007) is ha he rise in domesic income is due o he increase in he producion of imporables and exporables hrough he uilisaion of he impored capial and inermediae goods. The imporing of hese goods worsens he rade balance in he shor run bu i helps improve he rade balance in he long run as invesmen and inermediae goods suppor he manufacure of domesic exporables and imporables. In addiion, en, ou of welve cases in he expor models (Table 5a) have shown negaive and significan error correcion erms (ECT) excep for Germany, and Neherland. Of all, Indonesia and Ausralia repored higher coefficiens of ECT a and respecively. This would imply greaer and faser adjusmens in he Malaysia-Indonesia and Malaysia-Ausralia expor markes owards long run equilibrium, once being shocked. As for he rade balance models, all major rading parners have repored negaive and significan ECT, ranging from o All in all, he resuls show ha he rade balance model is superior o he expor model in erms of consisency, shor run adjusmens and he China effecs evidence. 19

21 3.4 Furher Discussion Our analysis ha based on he expor models and rade balance models have so far shows general suppors of he complemenary China effecs, and, in some cases, some subsiuing effecs. Ye, our empirical evidence reveals ha he fear for China effec migh be over-projeced. The complemenary effec of PRC expors expansion could be explained by he increase in global producion neworks or so called produc fragmenaion in he global expor secor. Wih he accession ino WTO in 2001, PRC enerprises are now more specializing and in coordinaion wih her regional counerpars and his has resuls in raising inra-indusry rade in differeniaed producs in he region. On his ground, Lall and Albaladejo (2004) have conduced a lenghy examinaion on expor flows of PRC and Eas Asia manufacuring secor in he 1990s and highlighs ha he new Tigers (mainly Souheas Asia markes, including Malaysia) facing PRC rea in low echnology manufacuring expors in hird markes, bu no in medium and high-echnology manufacuring. In a more recen sudy, Halmaier e al. (2007) found ha PRC expor hrea on Eas Asia counries was more pronounced in he medium echnology manufacuring raher han he low echnology secor, bu heir individual resul on Malaysia revealed ha PRC hrea is fel a boh exreme of in high-ech and low ech manufacuring expors. This finding seems o sugges ha PRC may have moved up he ladder chain from low-echnology expor o medium echnology expor as prediced by Lall and Albaladejo (2004), and soon PRC migh as well dominae he high-echnology expor in he near fuure. Neverheless, Ahukorala (2009), again, poined ou ha such worries migh be over-projeced. Though PRC has experienced rapid increase in high- 20

22 echnology expors over , he focus was on he bulk of labor-inensive high-echnology produc a he expense of he more high-wage Eas Asia NIEs, bu no on he ASEAN (including Malaysia). The fac is being suppored when produc fragmenaion is accouned (he componen rade is need ou). 4. CONCLUSION This paper addresses he crowding-ou effec of People Republic of China (PRC) on Malaysian expors and rade balances over he period of wih monhly rade series. We employ uni roo and coinegraion ess wih break o encouner he problem of srucural changes in ime series. The corresponding long run elasiciies and shor-run dynamics of he relaed variables are being examined. While he income effecs are more apparen in mos cases, he real exchange raes are raher insignifican and incorrecly signed for Malaysian bilaeral rading. Such finding does no favor he pracice of ringgi devaluaion as ool of curren accoun correcion. In addiion, esimaion of he rade balance model is generally more consisen ha he China s expors have exhibied complemenary effecs in he long-run, mainly for advanced expor desinaions such as Ausralia, Germany, Japan, UK and he US. Our finding ha China s expor expansion has been complemenary for emerging marke like Malaysia is consisen wih oher empirical lieraure using panel regression via graviy modeling by Frenald e al. (1999), Eichenggreen e al. (2007) and Greenaway e al. (2008). Such posiive effec could be due o he increase in global producion neworks in he expor secor over he las decade. As observed by Ahukorala (2009), he producion fragmenaion in Eas Asia region which pushes PRC s fas expor expansion happens a he expenses of he more high-wage Eas 21

23 Asia NIEs, bu no on he medium-wage ASEAN counries, including Malaysia. Our resul is consisen wih he lieraure looking a he componen rade daa. Fuure research migh wan o look ino he various differen echnology level of he componen rade in he Eas Asia region o gauge a more complee picure of he China effecs on Malaysia in specific, and on Eas Asia and oher emerging markes in general. 22

24 REFERENCES Ahmad, J. and J. Yang, Esimaion of he J-Curve in China. Eas Wes Cener Working Papers No. 67. Ahukorala, P., The rise of China and Eas Asian expor performance: is he crowding-ou fear warraned? The World Economy, 32 (2), Acaravci, A. and Ozurk, I., Tesing purchasing power pariy in ransiion counries: evidence from srucural breaks. Amfiearu Economic, 27, Baharumshah, A. Z., The Effec of Exchange Rae on Bilaeral Trade Balance: New Evidence from Malaysia and Thailand. Asian Economic Journal, 15 (3), Bahmani-Oskooee, M. and H. Harvey The J-curve: Malaysia versus her major Trading Parners. Applied Economics, 42: Bahmani-Oskooee, M. and S. Hegery, The J- and S-Curves: A Survey of he Recen Lieraure. Journal of Economic Sudies, 37 (6), Bahmani-Oskooee, M and Y. Wang, The J-Curve: China versus Her Trading Parners. Bullein of Economic Research, 58 (4), Chan, T. H. and Hooy, C. W. (2012) Role of exchange rae on China-Malaysia long run rading: complemenary or conflicing? Malaysian Journal of Economic Sudies, 49(2), Eichengreen, B., Rhee, Y., and Tong, H., The impac of China on he expors of oher Asian counries. Working Paper 10768, Naional Bureau of Economic Research, Cambridge, MA. Eichengreen, B., Rhee, Y. and Tong, H., China and he expors of oher Asian counries. Review of World Economics, 143 (2), Esso, L. J., Re-examining he finance-growh nexus: srucural break, hreshold coinegraion and causaliy evidence from he Ecowas. Journal of Economic Developmen, 35(3), Fernald, J., Edison, H. and Loungani, P., Was China he firs domino? Assessing links beween China and oher Asian economies. Journal of Inernaional Money and Finance, 18, Greenaway, D., Mahabir, A. and Milner, C., Has China displaced oher Asian counries expors? China Economic Review. 19 (2), Gregory, A.W. and Hansen, B.E., Tess for coinegraion in models wih regime and rens shifs. Oxford Bullein of Economics and Saisics, 58 (3), Gregory, A.W., J.M. Nason, and D. Wa, Tesing for srucural breaks in coinegraed relaionships. Journal of Economerics, 71, Hall, A., Tesing for a uni roo in ime series wih prees daa-based model selecion. Journal of Business and Economic Saisics, 12,

25 Halmaier, Jane. T., Ahmed, S., Coulibaly, B., Knippenberg, R., Leduc, S., Marazzi, M., and Wilson, B.A., The role of China in Asia: engine, condui, or seamroller? Inernaional Finance Discussion Papers, No. 904, Board of Governors of Federal Reserve Sysem, Washingon, D.C. Japan-Malaysia Economic Parnership, Join Sudy Group Repor, accessed 1/8/2012, hp:// Jenkins, R., Dussel Peers, E. and Mesquia Moreira, M., 2006, The impac of China on Lain America and he Caribbean. World Developmen, 36 (2), Johansen, S. and Juselius, K., Maximum likelihood esimaion and inference on coinegraion wih applicaions o he demand for money. Oxford Bullein of Economics and Saisics, 52, Lall, S. and Albaladejo, China compeiive performance: a hrea o Eas Asia manufacured expors? World Developmen, 32 (9), Lall, S., Weiss, J. and Oikawa, H., China s compeiive hrea o Lain America: an analysis for Oxford Developmen Sudies, 33 (2), Lee, J. and Srazicich, Mark C., Minimum LM Uni Roo Tes wih One Srucural Break. Working Paper, Deparmen of Economics, Appalachain Sae Universiy. Lee, J. and Srazicich, Mark C., Break poin esimaion and spurious rejecions wih endogenous uni roo ess. Oxford Bullein of Economics and Saisics, 63, Perron, P., The grea crash, he oil shock and he uni roo hypohesis. Economerica, 57, Phillips, P.C.B., Time series regression wih a uni roo. Economerica, 55, Quinin, E., Mexico s expor woes no all China-induced. Souhwes Economy, Issue 6, November/December, Federal Reserve Bank of Dallas. Rose, A. K. and J.L. Yellen,, Is here a J-curve? Journal of Moneary Economics, 24 (July), Walmsley, T. and Herel, T., China's accession o he WTO: Timing is everyhing. Working paper, Cener for Global Trade Analysis, Purdue Universiy. Yusoff, M.B., The Malaysian real rade balance and he real exchange rae. Inernaional Review of Applied Economics, 21 (5), Yusoff, M.B., Bilaeral rade balance, exchange raes, and income: evidence from Malaysia. Global Economy Journal, 9 (4), Aricle 7. 24

26 Table 1: GDP, Expor Performance and Expor Srucure of PRC and Malaysia, Aggregae GDP (in billion USD) Average GDP Growh (% Growh) Aggregae Expor (in billion USD) Average Expor Growh (%) Expor/GDP (%) Average World Expor Rank Year China Mal China Mal China Mal China Mal China Mal China Mal * Source: GDP and GDP growh are obained from World Bank, available only for only. Toal expor value is obained from IMF for he full sample Table 2: Top 15 Expor Desinaion of Malaysia and China, Malaysia Percenage China Percenage Aggregae for Top % Aggregae for Top % 1. US 15.98% 1. US 19.39% 2. Singapore 15.25% 2. Hong Kong 15.12% 3. Japan 10.12% 3. Japan 9.74% 4. China 9.33% 4. Souh Korea 4.64% 5. Hong Kong 5.00% 5. Germany 4.10% 6. Thailand 4.69% 6. Neherland 3.16% 7. Neherland 3.56% 7. UK 2.52% 8. Souh Korea 3.50% 8. Singapore 2.26% 9. Ausralia 3.16% 9. India 1.83% 10. India 2.77% 10. Russia 1.78% 11. Indonesia 2.64% 11. Ialy 1.72% 12. Germany 2.31% 12. France 1.64% 13. UK 1.93% 13. Ausralia 1.54% 14. UAE 1.40% 14. Canada 1.47% 15. Philippine 1.38% 15. Malaysia 1.46% Source: Auhors calculaion from expor series obained from IMF Direcion of Trade Saisics (DOTS). Noe: Ranking based on cumulaive expor value o he desinaion counry over Malaysia oal world expor value over

27 Table 3a: Gregory-Hansen Coinegraion Tes for Malaysian Expors wih Srucural Break * ln MALEX i, 1 ln IPi, 2 ln RERi, 3 ln CHNEX i, Expor Desinaion Model ADF * k TB Z * TB Z * α TB Ausralia C *** 1 95Q *** 01Q *** 01Q4 C/T *** 1 01Q *** 01Q *** 00Q4 C/S *** 1 01Q *** 01Q ** 01Q3 Germany C M *** 01M *** 01M04 C/T *** 0 01M *** 01M *** 01M01 C/S M *** 01M ** 01M04 India C M *** 96M *** 98M01 C/T M *** 00M *** 00M12 C/S M *** 96M *** 96M08 Indonesia C *** 1 93M *** 93M *** 93M06 C/T *** 1 94M *** 94M *** 94M02 C/S *** 1 94M *** 94M *** 94M01 Japan C *** 1 01M *** 01M *** 01M07 C/T M *** 01M *** 01M09 C/S ** 1 02M *** 02M *** 02M02 Korea C *** 1 98M *** 98M *** 98M01 C/T *** 0 05M *** 06M *** 06M11 C/S *** 1 06M *** 06M *** 06M09 Neherland C *** 1 02M *** 02M *** 02M02 C/T *** 1 02M *** 02M *** 02M04 C/S *** 0 02M *** 02M *** 02M02 Philippines C *** 1 99M *** 99M *** 99M10 C/T *** 1 99M *** 99M *** 99M10 C/S ** 1 99M *** 99M *** 99M10 Singapore C M *** 97M *** 97M09 C/T M *** 97M *** 97M09 C/S M *** 98M *** 98M08 Thailand C *** 1 94M *** 93M *** 94M04 C/T *** 1 94M *** 94M *** 94M05 C/S *** 0 99M *** 97M *** 97M03 UK C M *** 01M *** 01M01 C/T M *** 01M *** 01M03 C/S M *** 01M *** 01M01 US C ** 1 06M *** 07M *** 07M02 C/T M *** 96M *** 96M06 C/S ** 1 96M *** 96M *** 96M07 Noe: TB is he endogenous srucural break dae deeced. ( ** ) ( *** ) denoes saisical significance a he (5)(1)% level. 26

28 Table 3b: Gregory-Hansen Coinegraion Tes for Malaysian Trade Balance wih Srucural Break * ln MALTB i, 1 ln IPi, 2 ln IPi, 3 ln RERi, 4 ln CHNEX i, Expor Desinaion Model ADF * k TB Z * TB Z * α TB Ausralia C *** 1 98Q *** 98Q * 98Q1 C/T *** 1 98Q *** 98Q Q1 C/S ** 1 98Q *** 98Q Q1 Germany C *** 1 01M *** 01M *** 01M02 C/T *** 1 00M *** 00M *** 00M11 C/S *** 1 03M *** 03M ** 03M09 India C * 2 07M *** 98M *** 98M01 C/T * 2 05M *** 00M *** 00M03 C/S M *** 98M *** 98M01 Indonesia C *** 1 97M *** 97M *** 97M11 C/T *** 1 06M *** 06M *** 06M07 C/S *** 1 06M *** 06M *** 06M09 Japan C ** 2 98M *** 98M *** 98M01 C/T *** 1 96M *** 98M *** 98M01 C/S *** 0 98M *** 97M *** 97M11 Korea C *** 1 99M *** 99M *** 99M09 C/T *** 1 94M *** 99M *** 99M09 C/S *** 0 96M *** 03M *** 95M02 Neherland C *** 1 00M *** 00M *** 00M12 C/T *** 1 06M *** 06M *** 06M03 C/S *** 1 02M *** 02M *** 02M02 Philippines C M *** 08M *** 08M04 C/T *** 0 08M *** 08M *** 08M04 C/S *** 0 03M *** 03M *** 03M09 Singapore C *** 2 01M *** 01M *** 01M08 C/T *** 2 01M *** 96M *** 96M03 C/S *** 2 01M *** 93M *** 93M05 Thailand C *** 0 96M *** 96M *** 96M05 C/T *** 0 96M *** 96M *** 96M05 C/S *** 1 96M *** 96M *** 96M03 UK C M *** 01M *** 01M01 C/T M *** 97M *** 97M11 C/S M *** 01M *** 01M01 US C *** 1 96M *** 96M *** 96M10 C/T *** 1 98M *** 98M *** 98M06 C/S *** 1 98M *** 98M *** 98M06 Noe: TB is he endogenous srucural break dae deeced. ( ** ) ( *** ) denoes saisical significance a he (5)(1)% level. 27

29 Expor Desinaion Table 4a: Long-run Esimaes of Malaysian Expors Independen Variables Consan IP RER CHNEX Ausralia *** ( ) *** (0.4597) (0.1976) (0.0530) Germany *** (2.3405) *** (0.5923) (0.4522) (0.1115) India *** (4.6861) *** (1.6016) *** (0.6718) *** (0.2946) Indonesia ** (5.0515) * (1.2686) ** (0.7553) *** (0.1716) Japan *** (3.3701) *** (0.7245) (0.3672) *** (0.0489) Korea ** (5.8443) (0.9696) ** (0.8836) (0.3711) Neherland (6.1649) ** (1.4619) (0.5652) (0.1626) Philippines (7.2524) (1.5418) *** (0.6896) * (0.1854) Singapore *** (2.0064) *** (1.2380) *** (1.6688) *** (0.3879) Thailand ** (3.8572) *** (1.1581) (1.5046) (0.4556) UK ** (2.9245) (0.6688) (0.1668) *** (0.0303) US *** (2.1000) *** (0.6420) ** (0.3293) (0.0679) Noes: ( * )( ** ) ( *** ) denoes saisical significance a he (10)(5)(1)% level respecively. Repored in parenheses are sandard errors. IP Indusrial Producion, RER bilaeral real exchange rae, CHNEX Chinese expors o Malaysian Trading Parners. 28

30 Expor Desinaion Table 4b: Long-run Esimaes of Malaysian Trade Balance Independen Variables Consan IP IP* RER CHNEX Ausralia *** (1.5651) *** (0.3078) *** (0.3696) (0.1927) *** (0.0907) Germany (1.0925) *** (0.1947) (0.2667) *** (0.0985) ** (0.0778) India *** (3.7152) (0.7318) (0.7709) (0.9608) (0.1848) Indonesia * (1.5012) (0.2235) * (0.1833) *** (0.0802) (0.0654) Japan *** (2.7934) *** (0.5120) *** (0.5040) (0.3470) *** (0.2165) Korea ** (1.4314) *** (0.3908) *** (0.3608) * (0.2198) *** (0.0392) Neherland * (1.8541) (0.3677) (0.4928) *** (0.1183) (0.1163) Philippines *** (7.4304) * (0.8124) *** (0.7663) *** (0.3768) *** (0.2382) Singapore (0.7823) *** (0.1035) *** (0.1222) (0.0809) ** (0.0289) Thailand ** (0.4721) *** (0.1707) ** (0.1731) (0.1424) (0.0595) UK *** (1.8715) ** (0.3579) *** (0.4543) (0.3080) ** (0.1012) US *** (1.1308) *** (0.2277) *** (0.3868) * (0.1765) ** (0.0554) Noes: ( * )( ** ) ( *** ) denoes saisical significance a he (10)(5)(1)% level respecively. Repored in parenheses are sandard errors. IP Domesic Indusrial Producion, IP* - Foreign Indusrial Producion, RER bilaeral real exchange rae, CHNEX Chinese expors o Malaysian Trading Parners. 29

31 Expor Desinaion Table 5a: Error Correcions and Shor run Dynamics of Malaysian Expors Independen Variables DMALEX -1 DMALEX -2 DRER -1 DRER -2 DIP -1 DIP -2 DCHEX -1 DCHEX -2 ECT -1 Ausralia [ ] [ ] [ ] ** [ ] *** [ ] Germany *** [ ] [ ] * [ ] [ ] [ ] India *** [ ] [ ] ** [ ] ** [ ] *** [ ] Indonesia [ ] [ ] [ ] ** [ ] *** [ ] Japan *** [ ] [ ] [ ] [ ] * [ ] Korea *** [ ] [ ] [ ] [ ] *** [ ] Neherlands *** [ ] [ ] [ ] ** [ ] [ ] Philippines *** [ ] [ ] [ ] [ ] * [ ] Singapore *** [ ] * [ ] [ ] [ ] *** [ ] [ ] ** [ ] [ ] ** [ ] Thailand *** [ ] [ ] [ ] ** [ ] * [ ] UK *** [ ] [ ] [ ] * [ ] * [ ] US *** [ ] ** [ ] ** [ ] *** [ ] *** [ ] Noes: ( * )( ** ) ( *** ) denoes saisical significance a he (10)(5)(1)% level respecively. Repored in parenheses are - saisics. DMAYEX Changes of Malaysian expors o rading parners, DIP Changes of Indusrial Producion, DRER Changes of Bilaeral Real Exchange Rae, DCHNEX Changes of Chinese expors o Malaysian Trading Parners. 30