Volume 31, Issue 2. Onto Exchange Rate's Short Run Impact on Oil Prices Dynamics: An OPEC Members' perspective

Transcription

1 Volume 31, Issue Ono Exchange Rae's Shor Run Impac on Ol Prces Dynamcs: An OPEC Members' perspecve Sadek Melhem LAMETA Mchel Terraza LAMETA Absrac In hs paper, we sudy he ol prce formaon for he purpose of undersandng prce reacons of OPEC member counres o changes n he exchange rae of he US dollar and prces of oher members n he shor run. The resuls suggesed ha here s a paral mpac of exchange raes volaly on ol prces dynamcs n shor runs. Moreover, he sudy demonsraed ha Saud Araba behaves as a leader n he OPEC srucure marke whle behaves dfferenly when lnked o oher reference markes. Generally, Saud Araba behaves more poenally and more moderaely han he oher OPEC members n respondng o change of references markes prces. 1- LAMETA, Deparmen of Economcs, Unversy of Monpeller I, Espace Rcher, Avenue Raymond DUGRAND, Monpeller, France. E-mal: Sadek.melhem@LAMETA.unv-monp1.fr. - LAMETA, Deparmen of Economcs, Unversy of Monpeller I, Espace Rcher, Avenue Raymond DUGRAND, Monpeller, France. E-mal: Mchel.erraza@LAMETA.unv-monp1.fr. Caon: Sadek Melhem and Mchel Terraza, (011) ''Ono Exchange Rae's Shor Run Impac on Ol Prces Dynamcs: An OPEC Members' perspecve'', Economcs Bullen, Vol. 31 no. pp Submed: Nov Publshed: Aprl 17, 011.

2 1- Inroducon For mos counres, he real prce of ol s dependen on wo man facors: he value of he US dollar and he nflaon rae. Snce ol prces are conraced n US dollars and ol exporers mpor her goods from counres ncludng he Uned Saes, a srong dollar wll decrease he real ol prce and a weak dollar wll ncrease (Melhem and Terraza, 008). A declne n he value of he dollar wll weaken an ol exporer s purchasng power snce effecvely makes mpors orgnang ode he Uned Saes more expensve and should lead he ol o be cheaper ode US (Amezegar, 1986). Several sudes have emprcally examned ol exporers reacons o changes n he exchange rae by adjng expor prces and holdng mpor prces seady n order o manan marke shares. The degrees of exchange rae pass hrough depend crucally on he marke srucure, marke share 1 ec. These leraures provde evdence ha mperfec compeon may be one major explanaon for he exsence of an ncomplee exchange rae pass hrough. In an aemp o fll he exsng gap n he leraure, Kneer (1993), Yoef and Wrjano (004, 005) ed he exchange rae as a deermnan elemen n he ol prce formaon and hey showed ha he expor prce mark-up mples a paral exchange rae pass hrough. Despe a volumno sudy on hs subjec, he queson on wheher and how mark-up adjmen n response o changes n exchange raes affecs he mpor prce, appears no o have been sudes much n he shor run. Therefore, he man objecve of hs paper s o analyze he reacons of OPEC counres o change n he exchange rae and changes n he prce of oher members n shor runs. The remanng pars of he paper are organzed as follows. A secon develops a model for he exchange rae pass hrough. Secon 3 presens he emprcal resuls and secon 4 concludes. - The model In order o nvesgae he mpac of exchange rae volaly on ol prces, we follow a model ed n Yoef-Wrjano (004), whch s compleed by a varance srucure as specfed by Froo and Sen (1991) o compue he exchange rae volaly. Ths modfcaon perms o hghlgh he mpac of exchange rae on ol prces n he shor run. Gven ha he deprecaon of he US dollar s expeced o arac he ol prce nflows a leas for he followng wo reasons: Frs, he volaly of he US dollar affecs he ol prce rend. Second, he deprecaon of he US dollar mpacs on he OPEC counres asses and weakens her purchasng power. Due o s Flexbly and uncerany n naure, real exchange rae volaly s assumed o conss of wo pars: one s a par 1 For an exensve sudes on hs subjec, see Kurgman (1987), Durnbh (1987), Govannn (1988), Froo and Klemperer (1989), Thnelda (1996), and Also for a comprehensve survey of curren sudes on he exchange rae pass hrough, Menon (1995). For he ol marke model; see Adelman (198, 1993), Cremer and Saleh-Isfahan (1991), Grffn and Teece (198), Johany (1980). Adams and Marquez (1984) develop a smple carel model for OPEC o explan opmal ol prce deermnaon. 1

3 explaned by he falures of esablshng he law of one prce and he oher s an unexplaned par. In order o foc on he unexplaned par of he real exchange rae of USD volaly, VOL, we defne as he devaon of he acual value from he value explaned by he falure of he law of one prce: VOL e p e p var vâr (1) p p e p e p Where, var s he acual volaly and vâr p s he volaly explaned by he p e p falure of law of one prce. The exchange rae represens he rao of he domesc p prce n me, P, o he prce n counry n ha me. The large magnude of VOL means ha oher facors han he falure of he law of one prce play an mporan role n he real exchange rae volaly. The emprcal model proposed o explan prces dynamcs s jfed by equaon (1). We assume ha exchange raes varaons affec he changes n he margnal cos and all expor desnaons prces. Therefore, changes n he exchange rae and n compeor s prces are he prmary explanaory varables of he desnaon-specfc bes response funcon: P, VOL P D, () Where P, s he logarhm of crude prce charged by he source naon o he desnaon marke, VOL s he volaly of effecve exchange rae of dollar, P s he logarhm of crude prces charged by he compeors. s he random error erms assumed o be..d. (0, ) and D s a dummy varable se o 1 for 008:07-008:1 and 0 oherwse., The desnaon specfc nercep erm capures all non prce deermnan facors. The prcng o marke can be judged from he sgn and magnude of he parameer esmae of β only. The slope coeffcens as specfed n he model are allowed o vary across expor desnaon markes. Therefore, we confne our analyss n order o show he ol expor prce reacon o an exogeno change n exchange raes n he shor run. A hghlgh of he mporance of he wo sources of prces changes s obaned hrough he nerpreaon of he sgn and magnude of he parameers and. s he expor prce markup parameer. When 1 exchange raes pass hrough does no ake place becae he ndvdual producer adjs s mark-up o absorb he full n he exchange rae o keep s expor prce consan. When 0 mples ha changes n he exchange rae have no mpac on he prce charged by he exporng naon. When 0 1 here

4 would be a paral exchange rae pass hrough. Tha s, changes o he exchange rae affec he way n whch crude prces are formed. In he case where s negave, he ndvdual producer adjs s mark-up o absorb he full change n he exchange rae o keep s expor prce consan n erms of foregn currency. In he case s posve he ndvdual producer does no need o adj s mark-up becae hey keep s purchasng power. Tha s, he prof margn remans unchanged and change n he exchange rae would be fully passed on o he mporng naon, ne of overall prce changes. The value of provdes a clear ndcaon of he relave sze of each ndvdual exporng naon. Therefore, he large value of ndcaes, close o one, for a large marke, whle he value can ndeed devae from one for small marke. The expor prce reacon o rval s prce s capured by he parameer esmae of ha measures he elascy of he bes response funcon of an ndvdual exporng counry. A posve sgn of mples ha reacons are sraegc complemens, whle a negave sgn ndcaes a sraegc subsuon. The value of provdes a clear ndcaon of he marke power of each ndvdual exporng naon. In parcular, whle one may expec a large value, possbly close o one, for a small exporng naon, such value can ndeed devae from one for a large exporng naon. 3- Emprcal resuls Before underakng a sascal analyss ha examnes he reacon of spo prces for ol o he behavour of he exchange rae, we shall descrbe he daa ng n hs paper. The daa we e are daly observaons, over he frs of January 1999 o end of Dec 008, of he real ndex of he effecve exchange rae of he dollar (EERD). The real ndex of he effecve exchange rae s he prce adjed major currences ndex of Dollar. The ol prce seres are he US dollar daly spo prces of Wes Texas Inermedae Crude Ol reference (WTI), Norh Sea Bren reference (Bren), OPEC baske prce reference (OPEC), Iranan ol prce reference (Iran lgh), he ol prce of Venezuela reference (Ta Juana Lgh), he ol prce of Saud Araba reference (Arab Lgh) and he ol prce of Ngera (Bonny Lgh) deflaed by he US consumer prce ndex. The varables are ed n logarhmc form of frs dfference. The daa employed are aken from he Federal Reserve s, European Cenral Bank, OPEC organzaon and Energy Informaon Admnsraon (EIA). We sar our emprcal examnaon wh he un roo es by employng he augmened Dckey Fuller (1981) ess. The resuls are presened n able 1. We e Akake s nformaon creron o selec he approprae lag lenghs. For all seres, we are unable o rejec he un roo null hypohess and he seres are saonary n frs order dfferencng of raw seres. 3

5 Table 1 : Un roo es of ADF. Sa-Tes WTI Bren OPEC Iran Venezuela Saud Ngera EERD ADF [0.00] * [0.00] * [0.00] * [0.00] * [0.00] * [0.00] * [0.00] * [0.00] * ADF(-1) [0.35] [0.4] [0.1] [0.37] [0.55] [0.43] [0.39] [0.57] * The model whou nercep nor rend. [] s he P-value. Null hypohess are acceped a 5% sgnfcan level. Table repors he summary sascs of boh ol seres and exchange raes shows sgnfcan dfference beween he sandard devaons of he prces seres of he WTI reference, Bren reference and hose of OPEC baske (members of OPEC). We noe a low varaon of he OPEC member s prces seres (excep Ngera) compared o hose of he WTI and Bren reference. Moreover, dfferences appeared amongs OPEC members, wh low varaons of Souh Amercan members (Venezuela), hgh varaons of he Afrcan members (Ngera) and fnally he Mddle Eas members (Arab Saud and Iran) sayed n he mddle. Table : Summary Sascs Mean Mn Max SD Skew Kur Level Δ Level Δ Level Δ Level Δ Level Δ Level Δ WTI Bren Opec Iran Vnzla Saud Ngra eerd Δ= Frs dfference of he logarhm In esmang Eq. (), we ake no accoun he possbly ha changes n crude prces charged by compeve exporers are endogeno o changes n crude prce charged by he source naon. Ths suggess ha prce varables may be correlaed wh he equaon s error erm, sgn of co-lneary problem beween he varables. Table (3) show ha he hgh correlaon beween he varables s close o one. 4

6 Table 3: prce correlaon marx (level) WTI Bren OPEC Iran Venezuela Saud Ngera EERD WTI 1 Bren OPEC Iran Venezuela Saud Arab Ngera EERD Table 4: comparson of he seres n level and n FD wh he Varance Inflaon Facor VIF WTI Bren OPEC Iran Venezuela Saud Ngera R (Level) R ( ) VIF (Level) VIF ( ) VIF= s he Varance nflaon facor = 1/1- R In order o examne hs problem, we e he varance nflaon facor o deec he presence of mul co-lneary amongs seres. Table 4 esablshes comparson beween he seres n level and n frs dfference; we noe ha VIF sascs are relavely low wh he frs dfference (less han ). Ths suggess ha mul co-lneary, f s presen, s very weak and does no seem o nduce sero rsks. Therefore, we consder frs order dfferencng of seres, denoed, ( P ) and defned by P ln P ln P 1 for each counry. For hs reason, we e he OLS esmaor o esmae he Eq. (). The esmaed exchange rae mark-up elasces are almos sascally sgnfcan and carry negave sgns whle he expor prce reacon elasces o rvals prces are sascally sgnfcan and carry posve sgns (reacons are sraegc complemens). As can be seen from able 5, he expor prce mark-up elascy esmaes vary beween and mplyng a paral exchange rae pass hrough. For a 10% deprecaon of he effecve exchange rae of he US dollar durng he sample perod, for nsance, expor prce n US dollar have been mark-up by % o parally recoup he declne n he nernaonal purchasng power of ol revenues. The esmaed exchange rae expor prce mark-up elasces llrae a paern as hey are low for he small counres and hgh for he bgger naons. The resuls show a hgh value for Saud Araba (-0.07) and a small value for Ngera (-0.011) whch ndcaes he sze of each exporng naons marke. Th, he hgh expor prce mark-up elascy of Saud Araba ndcaes ha has a bgger sze of ol marke whle Ngera has a smaller sze of ol marke compared o OPEC members. The answer o wheher or no he 5

7 dfferen responses are he resul of he dfference n he sock reserve, producon capacy ulzaon, or he ncreased cos producon n he ol ndry caed by deprecaon of he dollar canno be nferred from our resuls. The resuls, however, provde an ndcaon of he dfferen reacons of OPEC members o an exogeno shock such as change n he exchange rae of USD. Table 5: he OLS esmaon for OPEC Members Consan eerd Saud ran Venzla Ngera Dummy Saud (P-value) (0.1) (0.04) * (0.00) * (0.00) * (0.00) * (0.00) * Iran (P-value) (0.0) (0.05) * (0.00) * (0.0) * (0.00) * (0.00) * Venzla (P-value) (0.31) (0.06) ** (0.00) * (0.0) * (0.00) * (0.00) * Ngera (P-value) (0.30) (0.06) ** (0.00) * (0.00) * (0.00) * (0.00) * * Resuls acceped a 5% sgnfcan level. ** Resuls acceped a 10% sgnfcan level. The dummy varable se equal o 1 for July 008 o December 008 and 0 oherwse. The esmaed rval s prces elasces carry he expeced posve sgns and are all sgnfcan a he 5% level. The resuls showed ha he measure of he rval prce elascy esmaes s as low as 0.13 for Saud Araba and as hgh 0.35 for Venezuela whch ndcaes a subsanal dfference n he reacons of prces o he prce changes of oher members. We dsngh ha he reacon of ol prces for Saud Araba o changes n he prces of ohers members are smaller ohers reacons of oher members prces. However, we observe agan ha he Saud Araba has he bgges power/prcng Sraegy amongs member naons of OPEC. In suppor of hs sylzed fac, we consder ha he subsanal dfference among he esmaed rval prce elasces s an ndcaon of he dfferen marke Power/Prcng sraegy of member naons of OPEC n respec of crude prces ndexes of OPEC reference baske, WTI reference and Bren reference ( 1 for a small naon and he devaes from 1 for a bg naon). Durng he perod of sudy, able 6 suggess ha esmaes of new measures of rvals prce elasces, of OPEC members relaed n OPEC marke, range from o ( opec ). Ths resul ndcaes ha a subsanal dfference n he reacons of prces o he prces changes of oher members. The low value for Saud Araba expor prce elasces provdes bg sze of marke power, snce s reacons o he change of oher member s he smalles (0.049). 6

8 Table 6: he OLS esmaon for WTI, Bren and OPEC references Consan Dummy eerd w Bren opec Saud (P-value) (0.03) * (0.05) * (0.00) * (0.00) * (0.0) * (0.00) * Iran (P-value) (0.10) ** (0.09) ** (0.00) * (0.00) * (0.0) * (0.00) * Venzla (P-value) (0.11) (0.07) ** (0.00) * (0.00) * (0.03) * (0.00) * Ngera (P-value) (0.07) ** (0.0) * (0.03) * (0.0) * (0.03) * (0.00) * * Resuls are acceped a 5% sgnfcave levels. ** Resuls are acceped a 10 % sgnfcan level. The dummy varable se equal o 1 for July 008 o Sepember 008 and 0 oherwse. We conclude ha Saud Araba behaves as a prce leader n he OPEC marke srucure. Moreover, Saud Araba has a bgger marke power and perfecly compeve marke srucure n respec of he OPEC prce reference. Ths resul appears as a logcal fac consderng he geopolcal realy of all naons, relave o producon and ol expor scale. To nspec he robness of he prce behavor of OPEC members, we employ alernave measures of crude ol reference prce ndexes from dfferen regons of he world ol marke. Table 6 llrae he fac ha when Saud Araba s expor prce s lnked o anoher prce references (as WTI and Bren), as a resul, a sgnfcan bu very weak relaonshp was obaned for rval prce elasces, whch also dffers from ha of OPEC reference. Ths means ha Saud Araba s rval-prce elasces vary from one marke o anoher (0.049, and respecvely for OPEC, WTI and Bren prce references). Gven s rval prce elasces esmaes convergence for each marke reference, Saud Araba s consdered o be more dscplned and more moderae n s reacons o changes n exchange raes, when compared o oher OPEC members. Furhermore, we noe ha Iran s sascal elascy esmaes n respec of WTI and Bren references are hgh (0.064 and 0.17, respecvely). 4- Conclon In hs paper we analysed he role of US exchange rae on ol prce formaon n he shor run, and reached a man conclon: The paper showed ha he volaly of exchange raes nfluences drecly daly ol expors prces. We analysed he prce reacons of he seleced OPEC members o responses on exchange raes changes and he prce changes of oher members. Specfcally, our emprcal resuls sugges ha, n response o change n exchange raes, exporers adj her prces o acheve hree man nerrelaed objecves: a) To secure a sable nernaonal purchasng power of ol revenues. b) To avod suppressng marke 7

9 demand and losng marke share. c) To geopolcal reasons. Therefore, each member of OPEC s ryng o exercse some degree of marke power n seng s expor prces. Accordng o our resuls, Saud Araba seems o behave que dfferenly n seng s prces compared o ohers members of OPEC, who se her own prces whou beng consderably nfluenced by oher members prces. Th Saud Araba behaves as a prce leader n he OPEC srucure marke, whle behavng dfferenly when lnked o oher reference marke. Furhermore, Saud Araba proved o be more dscplned and more moderae n s reacons o changes n exchange raes han oher OPEC members markes. References 1. Alman, D. (007) The ol prce currency queson, Inernaonal Herald Trbune.. Amezegar, J. (1986) OPEC and he dollar dlemma, Foregn Affars, 56, Engel, C. and Rogers, J. H. (1996) How wde s he border? Amercan Economcs Revew, 86, Froo, K. and Sen, J. (1991) Exchange raes and foregn drec nvesmens, board of governors of he Federal Reserve Sysem, Inernaonal fnance, Workng paper N Hooper, P. and Mann, C.L. (1989) Exchange rae pass-hrough n he 1980s: he case of US mpors of manufacures, Brookng Pap. Econ. Ac., 1, Hosn, D. (003) Les mécansmes de formaon des prx du pérole, L energe en France, 18 annales des mnes. 7. Kneer M. (1993) Inernaonal Comparsons of prcng o marke behavor, The Amercan Economc Revew, 83, Melhem, S. and Terraza, M. (008) The ol prce and he dollar Reconsdered, Commody Modelng and Prcng: Mehods for analyzng Resource Marke Behavor. Chp. 4, John Wlley & Sons, Inc. June Regner, E. (006) Ol and energy prce volaly, Energy economcs, 9, Verleger, P. (003) Implcaon of he weaker dollar, The peroleum Economcs, 0, Yang, C.W, Hwang, M.J and Huang, B.N. (00) An analyss of he facors affecng prce volaly of he US ol marke, Energy Economcs, 4, Yoef, A. and Wrjano, T. (004) The emprcal role of he exchange rae on he crude ol prce formaon, Energy Economcs, 6, Yoef A. and Wrjano T. (005) A sylzed exchange rae pass-hrough model of crude ol prce, OPEC Revew. Sepember