The imported price, inflation and exchange rate pass-through in China

Liu, Hai Yue; Chen, Xiao Lan A icle The impo ed p ice, in la ion and exchange a e pass- h ough in China Cogen Economics & Finance P o ided in Coope a ion wi h: Taylo & F ancis G oup Sugges ed Ci a ion: Liu, Hai Yue; Chen, Xiao Lan (2017) : The impo ed p ice, in la ion and exchange a e pass- h ough in China, Cogen Economics & Finance, ISSN 2332-2039, Taylo & F ancis, Abingdon, Vol. 5, Iss. 1, pp. 1-13, h ps://doi.o g/10.1080/23322039.2017.1279814 This Ve sion is a ailable a : h ps://hdl.handle.ne /10419/194641 S anda d-Nu zungsbedingungen: Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen Zwecken und zum P i a geb auch gespeiche und kopie we den. Sie dü en die Dokumen e nich ü ö en liche ode komme zielle Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich machen, e eiben ode ande wei ig nu zen. So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen (insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en, gel en abweichend on diesen Nu zungsbedingungen die in de do genann en Lizenz gewäh en Nu zungs ech e. Te ms o use: Documen s in EconS o may be sa ed and copied o you pe sonal and schola ly pu poses. You a e no o copy documen s o public o comme cial pu poses, o exhibi he documen s publicly, o make hem publicly a ailable on he in e ne , o o dis ibu e o o he wise use he documen s in public. I he documen s ha e been made a ailable unde an Open Con en Licence (especially C ea i e Commons Licences), you may exe cise u he usage igh s as speci ied in he indica ed licence. h ps://c ea i ecommons.o g/licenses/by/4.0/ Liu & Chen, Cogen Economics & Finance (2017), 5: 1279814 h p://dx.doi.o g/10.1080/23322039.2017.1279814 FINANCIAL ECONOMICS | RESEARCH ARTICLE The impo ed p ice, in la ion and exchange a e pass- h ough in China Hai Yue Liu 1 and Xiao Lan Chen 2 * Abs ac :This pape conside s he e ec o exchange a e (ER) le el on China’s do- mes ic p ices du ing he pe iod o 2003–2012. We examine China’s consume p ice index (CPI), impo p ice index (IPI) and p oduce p ice index (PPI) by using ime se ies ec o e o co ec ion analysis. The main inding o he pape is ha ER pass- h ough has had a limi ed bu g owing e ec on domes ic p ices and will con inue o do so. ER egime change ha was announced by Chinese go e nmen in 2005 led o an inc eased sensi i i y o ER pass- h ough o domes ic p ices. Subjec s: Economic Theo y & Philosophy; Mone a y Economics; In e na ional Finance Keywo ds: CPI; PPI; IPI; exchange a e JEL classi ica ion: F31; F41; E31 1. In oduc ion As China has been u he emb acing globaliza ion, exchange a e becomes an impo an economic index o e lec he s a us o he mac o economy. I is a key ac o linking in e na ional ade and capi al low. O e he pas wen y yea s, China has been mo ing om a managed exchange a e owa d a ully loa ing exchange a e. Since 2005, China has e o med he exchange a e egime o a managed loa ing sys em e e ing o a baske o cu encies. A p esen , China is u ging u he e o m o imp o e exchange a e lexibili y as pa o plans o libe a e China’s capi al accoun . In he lead up o es ablishing a ully loa ing exchange a e, he luc ua ion o he RMB will ha e mo e im- pac s on China’s economy, such as on he p ice le els in China. *Co esponding au ho : Xiao Lan Chen, School o Economics, Sichuan Uni e si y, Chengdu 610064, China E-mail: [email p o ec ed] Re iewing edi o : Ca oline Ellio , Hudde s ield Uni e si y, UK Addi ional in o ma ion is a ailable a he end o he a icle ABOUT THE AUTHORS Hai Yue Liu is a associa e p o esso in he Business School o Sichuan Uni e si y, China. He esea ch in e es s a e in e na ional economics, money and banking, C oss-Bo de Me ge and Acquisi ion, e c. Xiao Lan Chen is a lec u e in he School o Economics o Sichuan Uni e si y, China. She specialized in econome ics. He esea ch in e es s a e en i onmen al economics and in e na ional economics. PUBLIC INTEREST STATEMENT How does exchange a e pass- h ough a ec domes ic p ice le el? I has long been a opic o in e es in academia. In China, Schola s uses mul iple heo ies and empi ical me hods only o p o e ha he impe ec ions o exchange a e pass- h ough channels. While in ou esea ch, we ind ha a shock on nominal e ec i e exchange a e (NEER) has nega i e e ec on CPI which depic s ha RMB app ecia ion will cause CPI dec ease. Exchange a e pass- h ough also p o ed o ha e limi ed bu g owing impac on domes ic in la ion since he o eign exchange a e egime e o m in 2005. In ligh o hese indings, China’s in la ion can’ only be explained as “domes ic economic imbalance”. Mo e lexible exchange a e egime b ings s onge ansmission e ec om exchange a e and impo ed p ices o domes ic in la ion in China. Recei ed: 23 Oc obe 2016 Accep ed: 03 Janua y 2017 Published: 02 Feb ua y 2017 Page 1 o 13 © 2017 The Au ho (s). This open access a icle is dis ibu ed unde a C ea i e Commons A ibu ion (CC-BY) 4.0 license. Page 2 o 13 Liu & Chen, Cogen Economics & Finance (2017), 5: 1279814 h p://dx.doi.o g/10.1080/23322039.2017.1279814 The luc ua ion o exchange a e will cause changes on he impo ed p ice and hen pass on o he p oduce p ice index (PPI) and consume p ice index (CPI) o a coun y. T adi ional heo y s a es ha he app ecia ion o a coun y’s cu ency will elie e he p essu e on he in la ion in ha coun y. Since 2003, he RMB has app ecia ed o e 24% and his app ecia ion didn’ help b ing down he p ice le el while China was acing he high p essu e o in la ion(Yue, Qiang, & Kai, 2015). The eal e ec i e exchange a e (REER) inc eased 17% om 2003 o 2008, while CPI inc eased o e 8% du ing ha same ime. I aised an in e es ing poin ega ding how e ec i e he exchange a e pass- h ough is o he impo ed p ice and p ice le el in China. This pape ies o s udy he e ec s o he exchange a e luc ua ion on he p ice le els in China. Using mon hly da a om 2003 o 2012, he sensi i i y o he impo ed p ice and consume p ice in China esponding o he exchange a e luc ua ion was examined h ough he ec o e o co ec ion (VEC) model. The es o he pape is o ganized as ollows: Sec ion 2 e iews he ele an heo e ical and empi i- cal li e a u e. Sec ion 3 desc ibes he heo e ical backg ound. Sec ion 4 se s up he empi ical model and explains he esul s. Sec ion 5 concludes he pape . 2. Li e a u e e iew Al hough exchange a e pass- h ough has long been o in e es , he ocus o his in e es has e ol ed conside ably o e ime. Beginning in he la e 1980s, a e a long deba e o e he law o one p ice and con e gence ac oss coun ies, exchange a e pass- h ough s udies emphasized indus ial o ganiza- ion heo ies. The ole o segmen a ion and p ice disc imina ion ac oss geog aphically dis inc p od- uc ma ke s was also hea ily s udied. Mo e ecen ly, pass- h ough issues ha e played a cen al ole in hea ed deba es o e app op ia e mone a y policies and an op imal exchange a e egime in gen- e al equilib ium models. These deba es ha e b oad implica ions o he conduc o mone a y policy, o mac oeconomic s abili y, o in e na ional ansmission o shocks, and o e o s o con ain la ge imbalances in ade and in e na ional capi al lows. 2.1. Explana ions om heo ies o indus ial o ganiza ion Subs an ial wo k, bo h heo e ical and empi ical, has d awn on models o indus ial o ganiza ion o explain he links be ween exchange a es and p ices in e ms o ma ke concen a ion, impo pen- e a ion, he subs i u abili y o o eign and domes ic p oduc s, and he na u e o oligopolis ic com- pe i ion. Dohne (1984) de eloped a dynamic model o p icing by o wa d-looking compe i i e p o i -maximizing expo e s in which consume s adjus slowly o p ice changes. The le el o pass- h ough is de e mined by he speed o consume adjus men and expec a ions o he du a ion o exchange- a e changes. Mann (1986) and K einin, Ma in, and Sheehey (1987) all p esen some e i- dence sugges ing ha indus y cha ac e is ics in luence he ex en o which exchange- a e mo e- men s a e passed on o he dolla p ice o o eign goods. K ugman (1986) poin ed ou ha he eason o incomple e exchange a e ansmission o he impo ed p ice is he p icing- o-ma ke (PTM) be- ha io o i ms. Feinbe g (1986, 1989) ound ha ma ke concen a ion and impo pene a ion, di - e ing by indus y, in luenced he pass- h ough o changes in exchange a es o p ices o Wes Ge man domes ic goods om 1977 o 1983. He also explained he pass- h ough in o domes ic p ices o he Uni ed S a es o e he 1974–1987. F oo and Klempe e (1989) p o ide bo h heo y and e i- dence sugges ing ha expec a ions o exchange a es play a dominan ole in p ice de e mina ion. Do nbusch (1987) de eloped se e al s a ic models o explain pass- h ough wi h e e ence o he na u e o compe i ion be ween i ms and he ela i e numbe o home and o eign i ms. The esul s showed ha incomple e pass- h ough can occu e en in he case o pe ec subs i u es. Baldwin (1988) p esen ed a model, wi h sunk ma ke -en y cos s, in which empo a y exchange- a e luc ua- ions al e domes ic ma ke s uc u e su icien ly o in luence he pass- h ough om eal exchange a es o eal impo p ices. Fishe (1989) laid ou a pa ial equilib ium model in which domes ic and o eign i ms se p ices based on expec ed exchange a es; he showed ha exchange- a e changes could imply changes in impo p ices, bu he deg ee o pass- h ough depended on domes ic and o eign ma ke s uc u es. As su eyed by Goldbe g and Kne e (1997), mos o he a ailable e i- dence is om e y na owly de ined expo indus ies, wi h an emphasis o en placed on he p icing- o-ma ke beha io o expo e s. Ma s on (1990), Kne e (1992, 1993) and Goldbe g and Kne e Page 3 o 13 Liu & Chen, Cogen Economics & Finance (2017), 5: 1279814 h p://dx.doi.o g/10.1080/23322039.2017.1279814 (1997) used expo p ices o expo uni alues om speci ic coun ies o mul iple des ina ions wi h he in en o iden i ying p ice disc imina ion o p icing- o-ma ke ac i i y. Whe eas impo p ices a e by de ini ion jus he local-cu ency alue o ano he p oduce ’s expo p ices, he impo p ice se ies agg ega e ac oss p oduce s om all sou ce coun ies and ac oss a b oade a ay o p ices. 2.2. Explana ions om mone a y heo ies Mos no ably, heo e ical wo ks a gue ha ola ili y in mone a y agg ega es and exchange a es o coun ies should in luence he choice o in oice cu encies in ade (De e eux & Engel, 2001). In equilib ium, coun ies wi h low ela i e exchange a e a iabili y o s able mone a y policies would ha e hei cu encies chosen o ansac ion in oicing. The low-exchange- a e- a iabili y coun ies would also be hose wi h lowe exchange a e pass- h ough. Taylo (2000) also has no ed he po en- ial complemen a i y be ween mone a y s abili y and mone a y e ec i eness as a policy ins u- men . The idea is ha i pass- h ough a es a e endogenous o a coun y’s ela i e mone a y s abili y, pe iods o mo e s able in la ion and mone a y pe o mance will also be pe iods when mone a y policy may be mo e e ec i e as a s abiliza ion ins umen . Fo he pu pose o he ele an mac oe- conomic deba e, impo p ice se ies wi h agg ega ion a e he app op ia e uni s o analysis. Taylo (2000) and Gold ajn and We lang (2000), among o he s, ha e a gued ha pass- h ough a es may ha e been declining o e ime. 2.3. Some e idence ound in Chinese li e a u e Recen empi ical esea ch by some Chinese schola s has d awn con o e sial conclusions on he exchange pass- h ough o China’s in la ion. Jixiang, Yancheng, Liping, and Yuan eng (2011) ound ha RMB app ecia ion would inc ease he domes ic p ice le el h ough a G ange causali y es by using mon hly da a o 2005–2010. Zhu and Liu (2012) s udied he exchange a e pass- h ough and i s ela ionship wi h agg ega e domes ic demand and wo ld commodi y p ice om he yea 1980 o 2010. The esul shows a posi i e ela ionship wi h RMB app ecia ion and domes ic in la ion by using an ARDL model. They d ew a conclusion ha China’s domes ic in la ion is a “domes ic issue” a he han an “impo ed” one. Ha (2008)a gued ha RMB app ecia ion would es ain he domes ic in la- ion. Shi, Fu, and Xu (2008) ound he nega i e ela ionship be ween RMB app ecia ion and in la ion by using qua e ly da a om 1994 o 2007. They also ound ha he exchange a e pass- h ough is impe ec in China. Jiang and Kim (2013) u he p o ed hei esul s. Bei and Zhu (2008) esea ch showed ha he e is no di ec ela ionship be ween exchange a e and domes ic p ice. This conclu- sion is simila o ha o Li and He (2007). This pape looks a China’ ER pass- h ough o domes ic p ice le els by ocusing on he exchange a e egime e o m. Al hough many exis ed li e a u e concluded ha China does no ha e a s ong exchange a e pass- h ough channel. We a gue ha a sho - e m pass- h ough channel om ex- change a e o IPI and PPI and hen o CPI is p o ed o be e ec i e and smoo h. Fu he he ex- change a e e o m in 2005 s eng hened he sensibili y o domes ic p ice le els esponding o exchange a e luc ua ions. 3. Exchange a e pass- h ough and domes ic p ice le el Acco ding o S ensson (2000)’s in la ion a ge ing model, exchange a e can a ec domes ic p ice le el h ough di ec and indi ec channels hence ha ing u he impac on mone a y policy. Fi s ly, exchange a e ola ili y a ec s he impo ed p ice le el which is a pa o he CPI calcula ion baske . Secondly, exchange a e ola ili y a ec s he ela i e p ice le el o he impo ing and expo ing coun ies, which in u n changes he demand le el o he impo ed and expo ed p oduc s. Thi dly, aking o eign exchange a e as an asse , he luc ua ion o he exchange a e expec a ions a ec s he domes ic p ice le el. Finally, he mac oeconomic condi ions o o eign coun ies would also a - ec he domes ic p ice le el. Buil on S ensson’s amewo k, his pape se up a VAR model o he dynamic exchange a e pass- h ough o domes ic p ice le el aking Chinese economic eali y in o conside a ion. Page 4 o 13 Liu & Chen, Cogen Economics & Finance (2017), 5: 1279814 h p://dx.doi.o g/10.1080/23322039.2017.1279814 The ansmission s a ed om he o eign supply shocks which we use IM (China’s o al impo ) as a p oxy a iable. TIM ep esen s he o eign supply. We use IM as a p oxy a iable. M is he b oad money. We ake M2 as a p oxy a iable. NEER is he nominal e ec i e exchange a e. ES , EM , ED and EE ep esen he impac om shocks o he han o he supply le el, he money supply, domes ic demand o he ex- change a e. We assume ha he IPI, PPI and CPI ha e posi i e ansmission ela ionships wi h each o he . Exp essed as lows: PIPI , PPPI and PCPI ep esen he impo ed p ice index, p oduce ’s p ice index and consume ’s p ice index, espec i ely. These h ee a iables e lec he p ice le el o a coun y. EIPI , EPPI and ECPI ep e- sen he impac om shocks o IPI, PPI, and CPI, espec ully. We ake hese se en impac s as a i- ables o he VAR model o e alua e hei e ec on in la ion. 4. Empi ical analysis Basically, he analysis s a s om he educed o m o an un es ic ed VAR model wi h k lags ha can be w i en down as: whe e D is a (n×1) ec o o all de e minis ic a iables such as in e cep s, ends, dummies e c., A1, ⋯ ,Ak a e (n×n) ma ix o coe icien s, Z = (Z1 , Z2 , ⋯, Zn )’ is a ec o o n a iables, u is a (n×1) column ec o o inno a ions, ha is, se ially unco ela ed dis u bances ha ha e ze o mean and a a iance-co a iance ma ix a iance-co a iance ma ix E� u1,u� � = ∑u , i.e. u ∼N � 0, ∑� . The es ima ion o a,A1, ⋯ ,Ak , a e ob ained by applying o dina y leas squa es (OLS) o each pa o Equa ion (8) sepa a ely, and he es ima e o ∑u is gi en by he sample co a iance ma ix o he OLS esiduals. I co-in eg a ion be ween he a iables exis s, he VAR model can be ew i en in a VEC ( ec o e o co ec ion) model as: whe e Δ is a di e ence ope a o , is a (n×n) ma ix o long- un mul iplie s and [ Γi ]k−1 i−1 is a (n×n) ma ix o coe icien s ha con ain he sho - un esponses among a iables. (1) TIM =E +1( T IM ) +E S (2) M =E +1( M ) +𝛼 1 E S +E M (3) 𝛾 GDP =E +1 ( 𝛾GDP ) +𝛽1ES +𝛽2EM +E M (4) NEER =E +1 ( 𝛾GDP ) +𝛾1ES +𝛾2EM +𝛾3ED +E E (5) PIPI =E −1( P IPI −1) +𝛿 1 E S +𝛿 2 E M +𝛿 3 E D +𝛿 4 E E +E IPI (6) PPPI =E −1( P PPI −1) +𝜃 1 E S +𝜃 2 E M +𝜃 3 E D +𝜃 4 E E +𝜃 5 E IPI +E PPI (7) P CPI =E −1 ( PCPI −1 ) +𝜂1ES +𝜂2EM +𝜂3ED +𝜂4EE +𝜂5EIPI +𝜂6EPPI +E CPI (8) Z =a0D + k ∑ i=1 Ai+Z −1+u (9) Δ Z =a0D + k ∑ i=1 ΓiΔZ −i+∏Z −k+u Page 5 o 13 Liu & Chen, Cogen Economics & Finance (2017), 5: 1279814 h p://dx.doi.o g/10.1080/23322039.2017.1279814 In his s udy, Z = ( LCPI,LIPI,LPPI,LM2, LIM,LNEER,LGDP )� , a (7×1) ec o o I(1) a iables a e conside ed as endogenous. The da a employed o he s udy a e mon hly se ies o CPI, IPI, PPI, NEER, M2, GDP, and IM, ep e- sen ing in la ion, impo p ice index, p oduce ’s p ice index, nominal e ec i e exchange a e, money supply, domes ic demand and o eign supply, espec i ely. The ime span chosen o analysis is om 2003 o 2012. Th ee majo in la ions happened du ing his pe iod. The da a sou ces o a iables a e lis ed in Table 1. Unless o he wise s a ed, all a iables ha e been adjus ed seasonally using X12 me hod and adop ed he na u al loga i hm in o de o elimina e he in luence o seasonal in e up ions and he e oscedas ici y (Findley, Monsell, Bell, O o, & Chen, 1998). Table 2 p esen s summa y s a is ics on he key a iables used in his analysis. The es ima ion p ocess is o ganized as ollows: i s ly, i p esen s he uni oo es esul o he s abili y o all a iables. Secondly, i shows he co-in eg a ion ela ionship examined by he Johansen co-in eg a ion es . A las , a VEC model will be se up o analyze he impulse esponse and a iance decomposi ion. 4.1. Uni oo es To de e mine he app op ia e speci ica ion o he VAR es ima e, he a iables mus be es ed o he s a iona y p ocess. The p esence o non-s a iona y beha io in he au o eg essi e ep esen a ion o he a iable, i.e. he o de o in eg a ion o each a iable is de e mined using he Augmen ed Dickey-Fulle (ADF) uni oo es . The esul s o he ADF es s o six key a iables e eal ha GDP, IPI, IM and PPI a e s a iona y a le el (I(1)p ocess), while CPI, M2 and NEER a e i s di e ence s a- iona y (I(2) p ocess) a 10% c i ical alue (Table 3). Table 2. Summa y s a is ics o key a iables Va iables Obse a ion Mean S d. de . Minimum Maximum LCPI 120 4.6347 0.0227 4.5870 4.6886 LIPI 120 4.6642 0.0955 4.3770 4.8097 LPPI 120 4.6346 0.0407 4.5196 4.7014 LNEER 120 4.5521 0.0708 4.4308 4.6818 LM2 120 12.961 0.4929 12.155 13.789 LGDP 120 2.6473 0.2928 1.6864 3.1442 LIM 120 6.6628 0.4805 5.4714 7.4242 Table 1. Va iable desc ip ion and da a sou ce Va iables Desc ip ion Rep esen a ion Sou ce CPI Consume p ice index (2005 = 100) Domes ic in la ion Na ional Bu eau o S a is ics IPI Impo p ice index (2005 = 100) P ice le els o impo ed p oduc s Na ional Bu eau o S a is ics PPI P oduce p ice index (2005 = 100) P oduce ’s manu ac u ing cos Na ional Bu eau o S a is ics NEER Nominal e ec i e exchange a e (2005 = 100) Exchange a e le el IMF In e na ional Finance S a is ics M2 B oad money (in million US dolla s) Domes ic money supply Na ional Bu eau o S a is ics GDP G oss domes ic p oduc (using indus ial alue-added as p oxy a iable) Agg ega e domes ic demand Na ional Bu eau o S a is ics IM China’s o al impo om he wo ld (in million US dolla s) Fo eign supply Na ional Bu eau o S a is ics Page 6 o 13 Liu & Chen, Cogen Economics & Finance (2017), 5: 1279814 h p://dx.doi.o g/10.1080/23322039.2017.1279814 4.2. Co-in eg a ion es and VEC model se up Conside ing his s udy is o analyze he pass- h ough o he a iables ela ed o exchange a e ola- ili y o he domes ic p ice le el, all hese mac o-economic a iables migh sha e he same long- un end and a ec each o he in a ce ain way, o while hey exhibi sho un luc ua ions a ound a long un equilib ium. The ADF es indica es ha all he ime se ies a e no s a iona y a he same le el. A Johansen co-in eg a ion es was employed in o de o p o e he e is co-in eg a ion ela ion- ship be ween hese a iables. Table 4 shows ha he e a e a leas wo co-in eg a ing equa ions a he 5% le el o bo h ace es and maximum eigen alue es . The e o e, CPI, GDP, IPI, PPI, NEER, M2, and IM we e included o es ablish a VEC model. The esul shows ha all he in e se oo s a e less han one, i.e. all he model oo s a e loca ed in a uni ci cle (see Figu e 1), and he op imal lag pe iod is 2 acco ding o AIC esul (Table 5). The model is s able so ha we can p oceed o impulse esponse and a iance decomposi ion analysis. 4.3. The dynamic beha io o he VEC model The dynamic beha io o he VEC model is analyzed using he impulse esponse unc ion and a i- ance decomposi ion. 4.3.1. Impulse esponse unc ions The impulse esponse unc ion (IRF) aces he e ec o a one-s anda d-de ia ion shock in a a iable on cu en and u u e alues o o he s in he sys em. I indica es he size and cha ac e is ics o he shock’s e ec s. I is well known ha he esul s o he IRF and a iance decomposi ion based on Cholesky’s decomposi ion a e sensi i e o he o de o he a iables and he lag leng h. A VEC model Table 4. Johansen co-in eg a ion es esul No es: T ace es indica es 2 co-in eg a ing equa ion(s) a he 0.05 le el. Maximum eigen alue es indica es 2 co-in eg a ing equa ion(s) a he 0.05 le el. *Rejec ion o he hypo hesis a he 0.05 le el. Un es ic ed co-in eg a ion ank es ( ace) Un es ic ed co-in eg a ion ank es (maximum eigen alue) Hypo hesized No. o CE(s) Eigen alue T ace s a is ic 0.05 C i ical alue P ob. Hypo hesized No. o CE(s) Eigen alue T ace s a is ic 0.05 C i ical alue P ob. None* 0.4239 166.6119 125.6154 0.0000 None* 0.4239 62.3116 46.2314 0.0005 A mos 1* 0.3087 104.3002 95.7537 0.0113 A mos 1* 0.3086 41.7067 40.0776 0.0325 Table 3. ADF uni oo es esul s No e: Lag leng h selec ion is au oma ically based on Schwa z in o C i e ion (SIC), maxlag=13. *Rejec ion o he null hypo hesis (a a iable has a uni oo ) a 1% signi icance le el. **Rejec ion o he null hypo hesis (a a iable has a uni oo ) a 5% signi icance le el. Va iables A le el 1s di e ence - alue (cons an and end) p- alue (cons an and end) - alue (cons an and end) p- alue (cons an and end) LCPI −2.3969 0.3793 LGDP −3.6875** 0.027 LIPI −4.2898* 0.0046 LPPI −2.5369** 0.0101 LM2 −2.5369 0.3101 LNEER −3.1075 0.1094 LIM −5.6911* 0.0000 LCPI −12.3134* 0.0000 LM2 −5.5843* 0.0000 LNEER −7.4046* 0.0000 Page 7 o 13 Liu & Chen, Cogen Economics & Finance (2017), 5: 1279814 h p://dx.doi.o g/10.1080/23322039.2017.1279814 can be used o es he G ange causali y among a iables o he model, i.e. o asce ain he di ec- ion o causali y. G ange ’s de ini ion o causali y is a widely used concep o causali y (G ange , 1969). In o de o analyze he causal ela ionships be ween CPI and he o he a iables, a G ange causali y es wi hin a VEC amewo k is used. The es es ima es he χ2 alue o he coe icien on he lagged endogenous a iables. The null hypo hesis is ha he lagged explana o y a iables o he model and also hei join signi icance do no G ange -cause he dependen a iable. Table 6 epo s p- alues o F-s a is ics a lag h ee o he VEC G ange causali y es esul s. The esul s p esen ed in Table 6 indica e ha among six a iables, excep dLM2 o he s do “cause” CPI indi idually and he six a iables in luence CPI join ly. Combined wi h he G ange Causali y Tes esul s and he economic eali y, we can decide he Cholesky o de as ollows: GDP, M2, NEER, IM, PPI, IPI, CPI. Figu e 2 shows a shock on NEER has nega i e e ec on CPI. The impac om exchange a e o he CPI added up g adually and eached he highes poin o 0.36% a ound he se en h mon h. A shock on PPI has posi i e e ec on CPI. The e ec is quick ( eaches he highes poin in he i s mon hs a abou 0.36%) and hen aded ou o 0.1% a e 12 mon hs. A shock om impo ed p ice also has posi i e e ec on CPI. I eaches he highes poin a e 3 mon h o 0.18% and slowly ades ou . The esul s show a sho - e m pass- h ough channel om exchange a e o he CPI h ough impo ed p ice and PPI. This is he i s empi ical e idence ha ou hypo hesis con o ms o eal-wo ld ends. Figu e 1. In e se oo s o AR cha ac e is ic polynomial. Table 5. VAR lag o de selec ion c i e ia No es: FPE: Final p edic ion e o ; AIC: Akaike in o ma ion c i e ion; SC: Schwa z in o ma ion c i e ion; HQ: Hannan- Quinn in o ma ion c i e ion. *Lag o de selec ed by he c i e ion: LR: sequen ial modi ied LR es s a is ic (each es a 5% le el). Lag LogL LR FPE AIC SC HQ 0 1,006.884 NA 4.15E-17 −17.855 −17.685 −17.786 1 1,870.569 1,603.985 2.00E-23 −32.403 −31.044* −31.852 2 1,962.539 159.305 9.37E-24* −33.170* −30.622 −32.136* 3 1,996.708 54.914 1.25E-23 −32.905 −29.168 −31.389 4 2,044.992 71.565* 1.33E-23 −32.893 −27.965 −30.894 5 2,081.281 49.250 1.81E-23 −32.666 −26.549 −30.184 6 2,124.255 52.949 2.29E-23 −32.558 −25.252 −29.594 Page 8 o 13 Liu & Chen, Cogen Economics & Finance (2017), 5: 1279814 h p://dx.doi.o g/10.1080/23322039.2017.1279814 Figu e 3 shows ha a shock on NEER has nega i e e ec on CPI, IPI, and PPI. The impo ed p ice index has been mos a ec ed by he changes in he exchange a e compa ed wi h he o he wo p ice indices. The e ec inc eases g adually and eaches he highes poin a e 7 mon hs (2.2%). I indica es he pass- h ough e ec om exchange a e o impo ed p ice le els is ela i ely smoo h. The NEER has he smalles impac on he CPI (0.27% a he highes ). I explains he exchange a e pass- h ough o domes ic in la ion is impe ec and lagged-behind. 4.3.2. Va iance decomposi ion Va iance decomposi ion measu es he pe cen age o he o ecas a iance in in la ion. This a iance can be a ibu ed o shocks o changes o each explana o y a iable o e a se ies o ime ho izons. In addi ion, i shows how his p opo ion changes o e ime. The a iance decomposi ion in CPI o he VEC model by using he Cholesky decomposi ion me hod is epo ed in Table 6 ( he Cholesky o de has been discussed ea lie ). The a iance decomposi ion esul s suppo he empi ical indings om he impulse esponse unc ions (Table 7). CPI i sel caused 60% o i s a iabili y. Then ou side supply shocks (IM) accoun s o almos 16% o he CPI’s a iance a e en lags, which is he highes among all hese explainable a iables. Exchange a e (NEER) claims 9.6% o he CPI’s a iance a he end o 10 mon hs which is suppo ed by he IRF es as well. PPI has a signi ican con ibu ion o CPI’s a iance in he beginning (15.6%) bu he pe cen age dec eases g adually. IPI makes up 2% o CPI’s a iance in he i s mon h and keeps a s eady pe cen age o 1.75% a e 10 lags. M2 has immedia e in luence in he second Figu e 2. Response o LCPI o Cholesky one S.D inno a ions. Table 6. VEC g ange causali y/block exogenei y wald es esul s-p obabili ies *Rejec ion o he null hypo hesis (a a iable has a uni oo ) a 1% signi icance le el. **Rejec ion o he null hypo hesis (a a iable has a uni oo ) a 5% signi icance le el. Dependen a iables Excluded a iables (d =3) dLCPI dLNEER dLIPI dLPPI dLM2dLGDP dLIM Block exogenei y (d =10) dLCPI 0.01* 0.01* 0.01* 0.34 0.00* 0.07* 0.00* dLNEER 0.48 0.63 0.01* 0.75 0.65 0.26 0.21 dLIPI 0.86 0.44 0.00* 0.64 0.84 0.35 0.00* dLPPI 0.03** 0.00** 0.14 0.77 0.05** 0.48 0.00 dLM2 0.02** 0.36 0.26 0.20 0.62 0.40 0.18 dLGDP 0.60 0.55 0.63 0.92 0.18 0.56 0.29 dLIM 0.25 0.77 0.13 0.23 0.00* 0.86 0.00*