EXCHANGE RATE PASS-THROUGH INTO IMPORT PRICES: EVIDENCE FROM CENTRAL ASIA

Size: px

Start display at page:

Download "EXCHANGE RATE PASS-THROUGH INTO IMPORT PRICES: EVIDENCE FROM CENTRAL ASIA"

Adela Hart
5 years ago
Views:

1 EXCHANGE RATE PASSTHROUGH INTO IMPORT PRICES: EVIDENCE FROM CENTRAL ASIA Gulnara Moldasheva KIMEP Universiy/Bang College of Business Absrac This paper esimaes he degree of exchange rae passhrough (ERPT) ino impor prices for Cenral Asia counries over period 1995q112012q11 years. The addiional sudy was done for Kazakhsan o deermine he impac of money regulaing policy ino shorrun and longrun ERPT. The sudy of ERPT was done by employing coinegraion analysis across five cenral Asia counries: Kazakhsan, Kyrgyzsan, Tajikisan, Turkmenisan and Uzbekisan. The paper analyzes hree opporuniies of he ransmission of he expor cos ino ERPT using producion price indices (PPI) of hree major rading parners of Cenral Asia counries: PPI of Russia, PPI of Turkey and PPI of China. In case wih PPI of Russia he resuls showed, ha ERPT is exended beyond of 1 for Kazakhsan and Kyrgyzsan. Incomplee ERPT is indicaed for Tajikisan and Turkmenisan, and zero ERPT is indicaed for Uzbekisan. In case wih PPI of Turkey, he elasiciies of exen ERPT ino impor prices are more higher hen in case wih PPI of Russia, and elasiciies for all Cenral Asia counries are negaive and significan. In case wih PPI of China, incomplee ERPT is indicaed for all Cenral Asia counries, excep he Kyrgyzsan, which has exended ERPT. In all cases his sudy showed he significance of ERPT, excep only wo cases for Kazakhsan and Tajikisan wih PPI of China. Also ineresing, hose elasiciies of ERPT wih PPI of China are lower comparing wih oher cases. In he special sudy of money regulaion policy only for Kazakhsan he shorrun and longrun elasiciies of ERPT are incomplee wih PPI of Russia and PPI of Turkey, and hey are exended wih PPI of China. Key Words: Exchange Rae Pass Trough, impor prices, inflaion JEL Classificaion: E31, F31, G15 157

2 1. Inroducion According o he heory, exchange rae pass hrough (ERPT) measures he percenage change in domesic prices of goods resuling from one percenage change in he exchange rae. If 1 percenage change in exchange rae resuls in 1 percenage change in domesic goods prices, hen pass hrough is 100% or a complee pass hrough. Less han oneoone change in domesic price compared o exchange rae is an incomplee or parial pass hrough. A complee exchange rae pass hrough indicaes ha PPP holds, i.e. ha he prices of radable goods when expressed in he same currency are same across counries. Many empirical sudies have focused on undersaing he degree of hese pass hrough and came o he conclusion ha complee pass hrough never occurs, hough incomplee pass hough occurs wih varying degree hus refuing he purchasing pariy hypohesis. More recenly, he empirical analysis on exchange rae ransmission broadly relaes he passhrough esimaes o he efficiency of moneary policy, sabiliy of economic environmen, opimizaion of exchange rae regime and inflaion performance. Based on previous researches sudies, his paper examines boh ERPT ino impor prices and esimaes wheher passhrough raes are endogenous o inflaion performance for he economy of Cenral Asia counries. The primary objecive of his paper is o esimae he degree of ERPT across Cenral Asia counries. From he counries perspecives we provided he coinegraion analysis of ERPT in he shor run and in he long run using he quarerly daa from 1995 o A widely known hypohesis posed by John Taylor (2000) is ha he degree of passhrough is deermined by inflaion performance, a counry wih relaively lower and more sable inflaion rae experiences relaively lower passhrough elasiciy. We analyzed hree opporuniies of he ransmission of expor cos ino ERPT using producion price indices (PPI) of hree major rading parners of Cenral Asia counries: PPI of Russia, PPI of Turkey and PPI of China. In case wih PPI of Russia he resuls showed, ha ERPT is exended beyond of 1 for Kazakhsan and Kyrgyzsan. Incomplee ERPT is indicaed for Tajikisan and Turkmenisan, and zero ERPT is indicaed for Uzbekisan, or no ERPT ino impor prices is indicaed for Uzbekisan. This resul is new, since ERPT was no sudied in he lieraure for Tajikisan, Turkmenisan and Uzbekisan due o lack of informaion even in inernaional and local daa bases. In case wih PPI of Turkey he elasiciies of exen ERPT ino impor prices are higher hen hose wih PPI of Russia, and elasiciies for all Cenral Asia counries are negaive and significan. The 158

3 negaive elasiciies can be explained by decreasing demand for Turkish goods, reflecing o decrease of prices for Turkish goods, and herefore, o negaive ransmission ono ERPT. In case wih PPI of China, incomplee ERPT was indicaed for all Cenral Asia counries, excep he Kyrgyzsan, which has exended ERPT. Concerning he saisical significance of ERPT, in all cases we had significan ERPT, excep only wo cases for Kazakhsan and Tajikisan wih PPI of China. Also ineresing, hose elasiciies of ERPT wih PPI of China are lower comparing wih oher cases. Special coinegraion analysis and dynamic OLS was done for ERPT of Kazakhsan in shor run and in longrun reflecing he impac of money regulaing policy ino ERPT. The shorrun and longrun elasiciies of ERPT showed ha ERPT is incomplee wih PPI of Russia and PPI if Turkey, and is exended in case wih PPI of China. The longrun elasiciy of ERPT wih PPI of China is almos complee and close o 1. The paper is organized by nex sricure: in secion 2 we presen he lieraure review, in secion 3 we presen our economeric model, in secion 4 we presen he daa descripion and mehodology, in secion 5 we show our empirical resuls of he coinegaraion analysis of ERPT elasiciies, and in secion 6 we presen our conclusions. 2. The Lieraure Review The phrase passhrough was firs used in economics lieraure by Seve Magee (1973) in explaining he impac of currency depreciaion. Since hen he concep has been widely used in he lieraure. Dourbusch (1987) has considered he DixiSigliz (1977) and he Salop (1979) model of compeiion o confine he effecs of imperfec subsiuabiliy and produc differeniaion on price response o he changes in he exchange raes. He concluded ha he degree of passhrough is direcly relaed o he degree of subsiuion beween he impored goods and domesic produced goods. Nex concern relaes o he presence of foreign firms in he domesic marke in an open economy, which also affecs he degree of passhrough.the exisence of foreign firms in he domesic marke leads o he role of mulinaional corporaions (MNC) and inrafirm rade in influencing he degree of passhrough. The inracorporae exchange rae may deviae significanly from he passrough for a long period, as his is used as a clearing mechanism for inrafirm rade. Shapiro (2003) showed ha he MNCs use his sor of pricing mechanism o opimize profi in heir global operaions. According o he 159

4 heory of absolue purchasing power pariy, he price levels should be equal worldwide when expressed in a common currency. Relaive purchasing power pariy is he exension of absolue purchasing power pariy, which uses nominal exchange rae o adjus he difference of inflaionary level among he counries and mainain law of one price worldwide. Krugman (1987) and Dornbusch (1987) suggesed ha one moivaion for deviaions from law of one price is pricingomarke. Feensra and Kendall (1997) also found he empirical evidences ha pricingomarke conribues he big porion of incomplee passhrough and deviaion from law of one price. Devereux and Engel (1998) and Devereux and Yeman (2008) invesigaed he opimal choice of he exchange rae regime and i s relaion o currency price seing. They found ha when prices are se in local currency (LCP), adoping floaing exchange rae is always beer han fixed exchange rae because i makes he variance of domesic consumpion is no influenced by foreign moneary shocks. When prices are se in producer currency (PCP), here is a radeoff beween floaing exchange rae and fixed exchange rae. Under his circumsance, floaing exchange rae allows for lower variance of consumpion. Campa and Goldberg (2002) sudied he exchange rae passhrough ino impor prices of 25 OECD counries. They found ha here is empirical evidence of incomplee exchange rae passhrough in he shor run rejecing boh PCP and LCP. Over he long run, he degrees of passhrough are closer o one, so PCP is more prevalen for many ypes of impored goods. These resuls are consisen also wih he resul of Campa, Goldberg, and GonzalzMinguez (2005). Devereux, Engel and Sorgaard (2004) argued ha he degree of passhrough is relaed o sabiliy of moneary policy. Counries wih low volailiy of money growh will have relaively low raes of exchange rae passhrough, while counries wih relaively high volailiy of money growh will have relaively high passhrough rae. Frankel, Parsley and Wei (2005), Mumaz, Oomen and Wang (2005), Ihrig, Marazzi and Rohenberg (2006) found empirical suppor ha developing counries experienced a rapid downward rend in recen years in he degree of shor run passhrough, and in he adjusmen speed. Sulz (2006), Korhonen and Wachel (2005) sudied passhrough of exchange rae and impor price using VAR framework insead of radiional OLS or ECM model o esimae elasiciy of exchange rae passhrough. The degree of passhrough under heir framework is esimaed by means of impulse response funcion from VAR model ha is highly depending on he specificaion of he empirical model. Eike (2011) and Boug, Cappelen 160

5 and Erika (2012) found ha magniude of passhrough a all sages has decreased during 1990s by using subsample analysis. Ghosh and Rajan (2008) examined he exen and evoluion of exchange rae passhrough (ERPT) ino Korea s and Thailand s impor prices a he aggregae level for he period over he las wo decades. They found ha ERPT appears o be consisenly higher for Thailand compared o Korea; while for boh naions ERPT of heir respecive bilaeral raes wih respec o he US dollar is higher han wih respec o he Japanese yen. Bhaacharya, Karayalcin and Thomakos (2004) explored he exen of exchange rae passhrough for he USA, UK and Japan using a posbreon Woods indusrylevel daase. Their resuls were suggesive of wo channels of ransmission and hey found considerable variaion in he exen of passhrough across indusries and counries. From presened lieraure review, i can be concluded ha he incomplee ERPT ino impor prices is prevalen in he shor run; while he passhrough elasiciy is higher and close o one over he long run. Moreover, here is srong saisical evidence ha he downward rend of passhrough exiss in many counries and indusries hrough pas decades alhough he explanaion of his phenomenon is ill conroversial. However, mos of hese sudies only focus on cross counry analysis of large and developed economies. This paper invesigaes he degree of ERPT ino impor prices in Cenral Asia counries considering hem as small open economies wih he emerging markes. 3. The Model. We examined ERPT ino he aggregae impor prices of Cenral Asia counries wih regard o heir bilaeral nominal exchange rae wih he US dollar (USD), using he esimaing equaion in our empirical specificaion for he exchange raes: imp (1) ln( P ) = β 0 + β1 ln( S) + β2 ln( GDP) + β3 ln( MS) + β4 ln( PPIcounry) + ε Here S denoes he nominal exchange rae for each naion. A rise in S denoes a depreciaion of he currency. PPI counry denoes he producer price index (PPI) eiher of he Russia, Turkey or China, respecively. We ake hese counries as main rader parners of five CA counries in almos balk values of rade. Money supply (MS) variable was inroduced in he equaion as a proxy for money regulaion policy rough exchange rae. If β 1 = 0, here is no ERPT ino impor prices, while if β 1 = 1 here is complee ERPT. If he coefficien lies anywhere in beween 0 and 1 here is parial or incomplee ERPT. In 161

6 esimaing equaion (1) we conrolled for possible shifs in domesic demand and coss changes in he exporing naions. A rise in income implies an increase in demand for impored goods, hereby raising he impor price. Thus β 1 are generally expeced o be posiive. However, a rise in oupu could also imply less demand for impored goods and a decline in he impor prices. So i is plausible ha β 1 migh be negaive as well. 4. Daa and Mehodology 4.1 Descripion of he Daa The daa for impored prices, nominal exchange raes, gross domesic produc, money supply and producion price index were used in his sudy from ADB, NB of RK, IFS, IMF and WB daa bases on Quarerly base from 1995Q1 o 2012Q3. Excep for Kazakhsan, if daa on money supply were no available for oher CA counries, hen his variable was excluded from he analysis. We used impor prices (P imp ) o measure he price level for impor goods in domesic marke and use he aggregaed producion price index (PPI) o measure exporer s coss. As an exchange rae variable, we used nominal exchange rae vise a vise o USD. The nominal GDP in USD for all cenral Table 1. Tesing for Inegraion Order counries was used as a conrol variable o influence passhrough elasiciy. 4.2 Mehodology Since all variables have ime series properies, Augmened DickeyFuller Tes (ADF) and PhillipsPerron (PP) ess were applied o es he exisence of uni roo. For example, he resuls only for impor prices are repored in Table 1. The resuls of esing of uni roo for GDP, nominal ER, and hree PPIs are available upon reques from he auhor. Boh ess failed o rejec he null hypohesis of uni roo in he variables in heir level form, suggesing ha hey are saionary in heir firs differenced form. Given ha, he variables are I (1). Variable ADF PP Impor prices ( lpimp=log(prices of impor) Level Firs Difference Level Firs Difference Kazakhsan * Kyrgysan * * 162

7 Tadjikisan * * Turkmenisan * * Uzbekisan * * * Significance a he 5% level Nex, we performed coinegraion analysis among he variables in equaion (1) using he mehodology developed by EngleGranger (EG) and PhillipOuliaris (PQ) (1990). Coinegraion analysis was done four each of five counries separaely, and he resuls of only wih PPI of Russia are presened in Table 2. To obain he longrun exchange rae passhrough elasiciies, we used a recen mehodology developed by Sock and Wason (1993). The dynamic OLS (DOLS) procedure involves he regressing any variable wih he regressors iself bu also he leads and lags of he firs differences of he regressors. Table 2. Tesing for Coinegraion wih PPI of Russia Kazakhsan Kyrgyzsan Tajikisan Turkmenisan Uzbekisan β (0.482) (0.366) (0.561) (0.078) (0.163) β (0.209) (0.400) (0.244) (0.046) (0.00) β (0.285) (0.195) (0.219) (0.047) (0.123) β (0.156) (0.166) (0.312) (0.086) (0.095) ( β0=0) * 2.374* 0.301* ( β1=0) 6.103* 2.917* * 5.244* ( β2=0) 2.574* 2.749* * (β3=0) 6.302* 7.821* 3.146* 8.911* 4.310* EG PQ EG PQ EG PQ EG PQ EG PQ Ζˆ 4.00 Ζˆ α 29.35* * ** ** 5.40** 5.42** 4.21** 4.309** The empirical esimaion of (1) for he longrun ERPT is given by equaion (2) ** 38.59* * * Significan a he 5% level, **MacKinnon (1996) onesided pvalues 39.08* * 29.93* * ** 163

8 The resuls, referred only for Kazakhsan wih one period leads and lags of he regressors, are presened in secion 5. (2) ln( P where B = k=+ 1 k=+ 1 k=+ 1 imp ) = B X + η kδln( S) k + λkδln( GDP) k + γ kδln( PPI) k k= 1 k= 1 k= 1 [ β, β, β, β ], X = [ 1,ln( S ),ln( GDP),ln( PPI )] ζ 5. Empirical Resuls: Exchange rae pass rough ino impor prices We esed he coinagraion regression (3) for all five counries wih hree PPIs: PPI of Russia, PPI of Turkey and PPI of China. The resuls are shown in secions 5.1. imp (3) ln( P ) = β 0 + β1 ln( S) + β 2 ln( GDP) + β3 ln( PPI counry ) + ε In all cases, including he PPI of Turkey and China, he H o of no coinegraion was rejeced, and herefore, impor prices have long run relaionship wih he underlined explained he ERPT variables excep only one case wih PPI of China for Kyrgyzsan. The summarizing elasicizes of exchange rae pass rough ino impor prices are shown in Table 3. Wih PPI of Russia he ERPT is exended beyond of 1 for Kazakhsan and Kyrgyzsan. The exen of ERPT in hese counries can be explained by very ough inegraion in economy wih Russia. Negaive sign for ERPT in Kyrgyzsan migh be explained by lower demand for impored goods from Russia because of increasing naional oupu, and herefore he decline in impor prices. Incomplee ERPT is indicaed for Tajikisan and Turkmenisan, and zero ERPT is for Uzbekisan. This resul is also consisen for Uzbekisan, since his counry, from all of CA, is more close 164

9 Table 3. Elasiciy of ERPT ino impor prices Kazakhsan Kyrgyzsan Tajikisan Turkmenisan Uzbekisan wih PPI of Russia Elasiciy of ERPT (β1) 1.279* 1.168* 0.589* 0.283* 0.000* Elasiciy of PPI (β3) 0.984* 1.299* 0.983* 0.768* 0.410* wih PPI of urkey Elasiciy of ERPT (β1) 4.550* 3.611* 0.569* 0.174* 0.000* Elasiciy of PPI(β3) 0.684* 1.191* 0.894* 0.438* 0.317* wih PPI of China Elasiciy of ERPT (β1) * * Elasiciy of PPI(β3) * * *) significance a 5% o inernaional rade due o specific governmen resricions in rade policy. Wih PPI of Turkey we have negaive signs, excep for Uzbekisan, where ERPT is almos zero. Comparing wih case of PPI of Russia, he increasing values of ERPT rae for Kazakhsan and Kyrgyzsan migh be explained by greaer influence of demand for Turkish goods, han for Russian goods. Bu elasiciies of PPI, comparing wih PPI Russia, were declined. Wih PPI of China, only for Kyrgyzsan and Turkmenisan he elasiciies of ERPT are significan, and hey were no increased, even he PPI elasiciies became much higher han in previous cases wih PPI of Russia and PPI of Turkey. Special coingarion analysis of equaions (1) was done using money supply reflecion for Kazakhsan wih hree PPIs. In all cases he H o of no coinegraion was rejeced, and, herefore, impor prices have longrun relaionship wih underlined explained he ERPT variables. The summarized elasiciies of exchange rae pass rough ino impor prices are shown in Table 4. As we can see from Table 4, he incomplee ERPT Kazakhsan has wih Russia and Turkey s PPI, bu coefficiens are insignifican. Complee and significan ERPT Kazakhsan has wih China PPI. Money supply is posiively relaed o impored prices, and wih PPI of China almos 100% ransferred o ERPT. As shown in Table 4, elasiciy of PPI of China, expor coss of China 165

10 four imes ransferred ino ERPT of Kazakhsan rough impor prices. Since we have complee ERPT wih PPI China, finally, we run he Dynamic Ordinary Leas Squares (DOLS) for he coinegraed regression (2), including one lag and one lead ino equaion (2). These resuls of summarized elasiciies are shown in las column of Table 4. Table 4. Elasiciy of ERPT ino impor prices of Kazakhsan in shorrun and in longrun PPI of Russia PPI of Turkey PPI of China Shorrun Longrun Elasiciy of ERPT (β 1 ) * Elasiciy of MS(β 3 ) * * * * Elasiciy of PPI (β 4 ) * * * * *) significance a 5% 6. Conclusion This paper esimaed he degree of ERPT across CA. Were analyzed hree opporuniies of he ransmission of expor cos ino ERPT, using PPI of hree major rading parners of CA: Russia, Turkey, and China. The resuls showed ha ERPT wih PPI of Russia is exended beyond of 1 for Kazakhsan and Kyrgyzsan. Incomplee ERPT was indicaed for Tajikisan and Turkmenisan, and zero ERPT was indicaed for Uzbekisan. The elasiciies of exen ERPT ino impor prices wih PPI of Turkey are higher hen hose wih PPI of Russia and hese elasiciies are negaive and significan. This can be explained by decreasing demand for Turkish goods, reflecing o decrease of prices for Turkish goods, and herefore, o negaive ransmission ono ERPT. Incomplee ERPT wih PPI of China, were indicaed for all CA, excep he Kyrgyzsan, which has exended ERPT. All cases showed he significan ERPT, excep only wo cases for Kazakhsan and Tajikisan wih PPI of China. Also ineresing, hose elasiciies of ERPT wih PPI of China are lower comparing wih oher cases. Shorrun and longrun elasiciies of ERPT wih he reflecion of money supply for Kazakhsan showed ha ERPT is incomplee wih PPI of Russia and PPI of Turkey, and is exen in case wih PPI of China. The longrun elasiciy of ERPT in case wih PPI of China is almos complee. REFERENCES 166

11 Bhaacharya, P., Karayalcin, C.,and. Thomakos,D.(2004): Exchange Rae PassThrough and Relaive Prices: An IndusryLevel Empirical Invesigaion, Rerieved from Deparmen of Economics a Florida Inernaional Universiy, Miami, FL Boug,P., Cappelen,A. and Erika, T. (2012):Exchange Rae Passhrough in a Small Open Economy: he Imporance of he Disribuion Secor, Rerieved from Saisics Norway, Research Deparmen, P.O.B Dep Oslo, Norway. Campa, J. and Goldberg, L. (2002): Exchange Rae Passhrough ino Impor Prices: A Macro or Micro phenomenon? NBER working paper No Campa, J., Goldberg, L. and GonzalezMinguez, J. (2005): Exchange Rae Passhrough o Impor Prices in Euro Area, NBER Working paper No Dornbusch, R. (1987): Exchange Raes and Prices, American Economic Review 77, Devereux, M. and Engel, C. (2001): Endogenous Currency of Price Seing in a Dynamic Open Economy Model, NBER working paper No Devereux, M., Engel, C. and Sorgaard, P. (2004): Endogenous Exchange Rae Passhrough When Nominal Prices are se in Advance, Journal of Inernaional Economics, Vol.63, No. 2, July, PP Dixi, A. K., and J. E. Sigliz (1977), Monopolisic compeiion and opimum produc diversiy Devereux, M.,Yeman,J. (2008): Price Adjusmen and Exchange Rae Passhrough.,July 2008 Eike,B.(2011):Exchange rae passhrough: New evidence from German micro daa Economics working paper / ChrisianAlbrechsUniversiä Kiel, Deparmen of Economics, No. 2011,01 Feensra, R. and Jon D. Kendall. (1997): Passhrough of exchange raes and Purchasing Power Pariy, Journal of Inernaional Economics 43, pp Frankel, J., Parsley, D. and Wei, S. (2005): Slow Passhrough around he World: A New Impor for Developing Counries? NBER working paper No Ghosh, A., Rajan, R.S., Exchange rae passhrough in Korea and Thailand: Trends and deerminans, Japan World Economy (2008), doi: /j.japwor Ihrig, J., Marazzi, M. and Rohenberg, A. (2006): Exchange Rae Passhrough in he G7 Counries, Inernaional Finance Discussion Papers No Korhonen,I. and Wachel,P.(2005): A noe on exchange rae passhrough in CIS counries, 2/2005 Krugman, P. (1987): Pricing o Marke When he Exchange Rae Changes, in Sven Arnd and J. David Richardson, eds., RealFinancial Linkages Among Open Economics, (MIT Press: Cambridge), Magee, S. (1973), "Currency Conracs, Passhrough and Devaluaion," Brookings Papers On Economic Aciviy, 1, 1973, Mumaz, H., Oomen, O zlem and Wang, J. (2005): Exchange Rae Passhrough ino U.K. Impor Prices: Evidence from Disaggregaed Daa, mimeo, July. Salop, Seven, (1979), Monopolisic compeiion wih ouside goods, Bell Journal of Economics IO, no. I, Shapiro, A. (2003): Mulinaional Financial Managemen, 7h ED, John Wiley & Sons, Inc. PP.121. Sulz, J. (2006): Exchange Rae PassThrough in Swizerland:Evidence from Vecor Auoregressions, Taylor, J. (2000): Low Inflaion, Passhrough, and he Pricing Power of Firms, European Economic Review, June volume 44 issue 7, PP

Estimating a Time-Varying Phillips Curve for South Africa

Estimating a Time-Varying Phillips Curve for South Africa Esimaing a Time-Varying Phillips Curve for Souh Africa Alain Kabundi* 1 Eric Schaling** Modese Some*** *Souh African Reserve Bank ** Wis Business School and VU Universiy Amserdam *** World Bank 27 Ocober