The Impact of Shocks in Oil price and Exchange Rate on Inflation in Iran: The Application of the VAR Approach

Transcription

1 Environmenal Energy and Economic Research (2018) 2(1): DOI /eeer The Impac of Shocks in Oil price and Exchange Rae on Inflaion in Iran: The Applicaion of he VAR Approach Mohaddeseh Babajani Baboli a, Seyyed Abdul Majid Jalaei Esfandabadi b,*, Mohsen Zayandeh Roody a a Faculy of Managemen and Economics, Islamic Azad Universiy, Kerman Branch, Kerman, Iran b Faculy of Managemen and Economics, Shahid Bahonar Universiy of Kerman, Kerman, Iran Received: 1 Augus 2017 /Acceped: 25 December 2017 Absrac Iran is one of he oil exporing counries, so he oil price plays a remarkable role in he governmen budge and is a major source of foreign exchange. On he oher hand, he reliance of he governmen budge o oil income as well as is flucuaions is a fac referred o as he mos imporan cause of inflaion by many researchers. This paper explores he effec of shocks in he exchange rae, oil price, and producion as he hree main shocks in he economy on he mos imporan variable in Iran s macroeconomy, i.e. price level. So, he vecor auoregression (VAR) model is used wih seasonal daa for he period of Afer he model is esimaed, impulse response funcions are calculaed and analysis of variance is performed o figure ou he conribuion of each shock in he variance of he predicion error of hese variables. The resuls show ha he srong dependence of exchange rae on foreign exchange earnings of oil price allows he rapid growh of he prices in Iran and he effec of he shock is increasing over ime. Also, sancions in 2012 did no reduce oil price, bu hey influenced he exchange rae and inflaion significanly. Keywords: Shock, Oil, Inflaion. Inroducion Flucuaions in inflaion or prices may be a source of economic variaions wih some impacs on he overall economic performance. Thus, he inflaion rae is usually considered an imporan economic indicaor o assess he economic performance (Kun Sek, 2015).Since he 1970s, inflaion has emerged as one of he graves economic problems in various counries. Economiss and economic agencies in Iran have lised a wide range of facors in heir aemps o explain he causes of inflaion, including he srucure and foundaions of he economy, how exchange policies are implemened, and shocks. However, a look a inflaion as an acue problem shows ha all hese facors conribue o i and reinforce he impacs of one anoher and, in some cases, hey creae one anoher in a series (Yazdani and Zare, 2016). Public people are very familiar wih oil price as an economic indicaor as hey see a close relaionship beween oil price and heir daily life. People closely follow he variaions in he prices of oil producs, such as gasoline or gas. Also, heir expecaions have been firmly * Corresponding auhor Jalaee@uk.ac.ir

2 52 Babajani Baboli e al. shaped by he relaionship beween oil prices and economic cycles since he oil shocks of he 1970s (Yoshizaki, 2013).Iran possesses 11 percen of he oil reserves of he world so ha percen of oal expors and usually half of he governmen s annual budges are supplied from oil expors and oil sale accouns for over 20 percen of Iran s gross domesic produc (GDP). In his siuaion, any shock in he global oil marke can enail exensive impacs on Iran s economic srucure. Thus, o avoid economic crises and o use he poenial opporuniies, i is imperaive o explore he impacs of global oil price flucuaions on Iran s economy and o develop proper economic policies in order o uphold he sabiliy of he economy (Samei e al., 2017). Also, due o he over-reliance of Iran s economy o foreign exchange earnings of oil sale over he years, he flucuaions of exchange rae have emerged as a key facor in policymaking. So, when formulaing he developmen plans and annual budges, he policymakers need o undersand and predic how flucuaions of variables would influence economic inflaion in order o minimize he negaive impacs of shocks in oil price and exchange rae on inflaion by adoping appropriae policies. The presen paper evaluaes he effec of he shocks in oil price and exchange rae on inflaion in Iran using seasonal daa over he period. The paper is organized ino five secions. The firs secion is an inroducion o he subjec maer. The second secion briefly reviews he relevan lieraure. The model is inroduced in he hird secion followed by he presenaion of he resuls of model esimaion for Iran s economy in he fourh secion and finally, he concluding secion ha conains some recommendaions. Lieraure Review Exensive research has focused on describing he relaionship beween oil price shocks and he naional economy of he US. For example, Hamilon (1983) observed ha oil price shocks were he main facor responsible for he recession. This encouraged oher researchers in boh developed and developing counries o work in order o eiher suppor or refue his finding. Examples include Rasche and Taom (1981), Darby (1982), Hamilon (1983, 1996, 2003, 2008), Chang and Wong (2003), Cunado and Gracia (2005), Lorde e al. (2009), Dogrul and Soyas (2010), and Rasmussen and Roiman (2011) who have argued ha oil price shocks significanly influence economic aciviies, mainly economic growh.ragabooorur and Chicooree (2013) focused on he effec of exchange rae on consumer and producer price index and impor prices in Mauriius. They found ha he effec of exchange rae pass was srong on consumer price index bu weak on impor price. Analysis of variance poined o oil price shock as he main cause of impor price flucuaions. Madesha e al. (2013) empirically esed he relaionship beween exchange rae and inflaion in Zimbabwe over They employed he Granger causaliy es o esimae he long-erm relaionship beween exchange rae and inflaion rae. The resuls showed ha inflaion and exchange rae had Granger-causaliy over he sudied period. According o he resuls, if appropriae policies are adoped, hey can orien he aiudes owards deermining exchange rae wihou leading o inflaion. Fakieh (2013) addressed he relaionship beween inflaion and exchange rae policies in Saudi Arabia o specify an appropriae policy for his main oil exporing counry wih respec o inflaion rae using economeric echniques and ime series daa.si milar o Alowijiri (2011), Hassan and Alogell (2008) and Albarwani e al. (2010), he applied he economeric echniques of OLS, Dickey-Fuller es, uni roo es, inegraing, and error correcion models. He repored ha he fixed exchange rae sysem was no appropriae for Saudi Arabia. Accordingly, oher variables, e.g. growh, oil price, and expor and impor prices, should be considered when suiable exchange rae sysem was o be specified. Then, he argued ha he curren exchange rae sysem was he main cause of high inflaion in Saudi

3 Environmenal Energy and Economic Research (2018) 2(1): Arabia. Aloui and Jammazi (2010) used a wo-regime model (MS-EGARCH) o relae he flucuaions of he oil marke and sock price in France, he UK, and Japan for he period. The resuls showed ha he increase in oil price significanly influences boh sock reurn flucuaions and ransfer possibiliy across he regimes. Wih respec o he impacs of oil shocks on sock marke dynamics, he moneary economy is based on he premise ha here is no a long-erm relaionship beween inflaion and real economic aciviies. Samei e al. (2017) addressed he impacs of oil income growh shocks on Iran s economy using he TVP-VAR models for he period. They included he variables of added-value growh of indusry and mining secor, unofficial exchange rae, inflaion rae, he governmen s real consumpion expendiure growh, real impors growh, real oil income growh, and oil marke shocks in he model. They found variaions in he relaionships of hese variables over ime. Also, hey repored ha how he curren condiions affec he economy influences how model variables affec one anoher. Asna Ashari e al. (2015) employed Qu and Perron (2007) s suggesed model o explore he effec of oil price shocks on inflaion, growh, and money in Iran s economy in They idenified five srucural shocks happened in Sepember 1973, July 1979, May 1990, July 1994, and May The highes coefficiens of oil price impac on producion, inflaion, and money growh were relaed o he firs, firs, and fifh regimes, respecively. Furhermore, he longes period of oil price impac on producion, inflaion, and money growh was found in he fourh, second, and fifh regimes, respecively. Yazdani and Zare (2016) invesigaed he effec of exchange rae shocks in Iran s economy using he seasonal ime series saisics for he period of They employed srucural vecor auo-regression model (SVAR). They found ha in he sudied period, exchange rae variaions and exchange policies have been he main causes of inflaion so ha hey have creaed a srucural inflaion in Iran s economy. Thus, he exchange policies adoped in his period, including he policy of exchange rae sabilizaion, have fomened he inflaion. Ghafari e al. (2013) focused on he effec of exchange rae increase on main macroeconomic variables of Iran in a srucural macroeconomeric model by assessing he impac of he gradual or sudden increase in exchange rae wihin some simulaion scenarios. They repored ha he derimenal effecs of exchange rae increase on main macroeconomic variables are less severe in case of gradual growh of exchange rae han in case of is sudden increase. Maerials and Mehods We seleced vecor auo-regression (VAR) model for daa analysis as i is he bes model o analyze he empirical model of he presen sudy because i uses a very easy procedure and does no require he researcher o deermine wheher he variable is indigenous or exogenous (Sarzaeem, 2007). VAR models yield beer predicions han simulaneous equaions, and heir poenial o represen he dynamic srucure of he model and he raional expecaions in shor run enables ackling he consrains and limiaions usually associaed wih economic heories(kani, 1989). For hese reasons, i is said ha he VAR mehod does no need an explici economic model o esimae he model. The model of he presen sudy ha is derived from Samadi e al. (2010) s empirical model employs he VAR mehod o examine and calculae how much inflaion is accouned for by he shocks. The variables included gross domesic produc (GDP), oil price (POIL), exchange rae calculaed as USD o IRR (Ex), and inflaion a general level of prices (CPI). Time series daa of Iran s economy for he ime period of were employed and hey were analyzed wih Eviews sofware package. The research esimaed he following equaions o relae he variables using he VAR model:

4 54 Babajani Baboli e al. y c p i 1 AY i 1 (1) CPI GDP CPI GDP POIL EX (2) CPI GDP POIL EX (3) POIL CPI GDP POIL EX (4) EX CPI GDP POIL EX (5) Model Esimaion To esimae he model, we should firs deermine wheher he variables are saionary. Uni roo es is a popular mehod o idenify he saionariy of a ime series process. This es can be conduced in wo forms: Dickey-Fuller es and augmened Dickey-Fuller (ADF) es. The presen paper uses he laer es. Table 1. Resuls of augmened Dickey-Fuller uni roo es for he variables included in he model Variable Level Firs-order differeniaion Wih y-inercep Wih y-inercep and rend Wih y-inercep Wih y-inercep and rend Calculaed Criical Calculaed Criical Calculaed Criical Calculaed Criical Saisic saisic Saisic saisic Saisic saisic saisic saisic Inflaion Exchange rae Oil price Producion Noe: Criical values a p < 0.05 level. Source: Compuer appendix (he oupu of Eviews sofware package) The resuls show ha he log of he ime series variables ha were included in he model was non-saionary a daa level and he -values calculaed by ADF was smaller han he MacKinnon criical values a he 1%, 5%, and 10% levels. So, he null hypohesis of uni roo was no refued and i was confirmed ha he variables were no saionary. Thus, i was necessary o perform firs-order differeniaing on hem. All variables became saionary afer performing one-ime differeniaing. The firs sep in Johansson long-run coinegraion mehod is o deermine he opimal lag lengh in he VAR model. This was esimaed by he Schwarz Bayesian crierion (SBC) saisic. The opimal lag was seleced o be 3. In he sudy, before examining he presence of a long-erm relaionship, he correlaion of he residuals was esimaed by he LM es. The resuls showed he lack of correlaion beween residual erms up o eigh lags. Also, his shows ha he order by which he variables were included in he model had no effec on he esimaions. In he nex sep, o deermine wheher here was a long-run relaionship beween he variables of he model, Johansson es was firs performed o esimae he number of long-run

5 Environmenal Energy and Economic Research (2018) 2(1): coinegraion vecors. This was deermined wih wo likelihood raio ess (Max-Eigen saisic max and race saisic race) whose resuls are presened in Tables 3 and 4. Accordingly, here was a single long-run coinegraion vecor. Table 2. The resuls of he LM es Lag LM saisic Probabiliy Table 3. Resuls of he Johanssen coinegraion es according o race saisic Number of coinegraion vecors in null hypohesis Number of coinegraion vecors in opposie hypohesis Tes saisic Criical value a level Probabiliy level *r 0 r = r 1 r = r 2 r = r 3 r = * Rejecion of null hypohesis abou he lack of coinegraion vecors a significance level Table 4. Resuls of he Johanssen coinegraion es according o Max-Eigen saisic Number of coinegraion vecors in null hypohesis Number of coinegraion vecors in opposie hypohesis Tes saisic Criical value a level Probabiliy level *r 0 r = r 1 r = r 2 r = r 3 r = * Rejecion of null hypohesis abou he lack of coinegraion vecors a significance level Afer i was ensured ha here was a coinegraion vecor beween he variables in long run, he response of he variables o he shocks induced by oher variables was examined in order o sudy he dynamic behavior of he variables in he conex of he sudied model. This was performed by response funcions. Figures 1-3 illusrae he resuls for a shock a he magniude of one sandard deviaion (SD) and he response of oher variables o his shock. Figure 1 displays ha when a posiive shock wih he magniude of 1 SD is applied o he price index, is impac gradually increases over ime so ha GDP is increased and oil price and exchange rae are decreased. This shock had significan posiive effecs over he sudied periods (p < 0.05). According o Figure 2, a posiive shock o GDP wih he magniude of 1 SD enails a gradually decreasing impac resuling in he increase in price index, oil price, and exchange rae. This shock had posiive effecs over he sudied period and is effecs were significan (p < 0.05). Figure 3 depics ha a 1SD posiive shock o oil price resuls in a gradually decreasing impac. The shock o oil price has a negaive effec on GDP, bu is effec is posiive on price index and exchange rae. The shock had posiive effecs on he sudied period, bu hey were no significan a p < 0.05 level. Figure 4 shows he effec of a posiive 1SD shock o he exchange rae. This shock enails a gradually descending effec. The effec of shock o exchange rae is negaive on GDP and oil

6 56 Babajani Baboli e al. price, bu posiive on price index. This shock had significan posiive effecs over he sudied period (p < 0.05). Figure 1. Impulse response funcions when shock is applied o inflaion Figure 2. Impulse response funcions when shock is applied o GDP

7 Environmenal Energy and Economic Research (2018) 2(1): Figure 3. Impulse response funcions when shock is applied o oil price Whils he response funcions draw he effec of a shock o an indigenous variable on oher variables of he VAR model, analysis of variance (ANOVA) disinguishes he variaions in an indigenous variable wih he changes in oher indigenous variables. Therefore, ANOVA informs on he relaive imporance of each random shock in influencing he model variables. Table 5. Analysis of variance for price index Lag Analysis of variance of CPI CPI EX GDP POIL

8 58 Babajani Baboli e al. Figure 4. Impulse response funcions when here is a shock wih he magniude of one sandard deviaion As is eviden in Table 5, he greaes par of he predicion error of he variable price index is accouned for by he variable iself in he shor run. Bu, over ime he oher variables gain increasing imporance. In his respec, he variable exchange rae is more influenial han he oher variables. Afer 10 periods, 34.63% of he predicion variance error of price index is capured by exchange rae and 15.69% is accouned for by oil price. The variable GDP had he lowes share in explaining he predicion error of price index so ha i amouns o 0.36% afer 10 lags. Table 6. Analysis of variance for GDP Lag Analysis of variance of GDP GDP CPI EX POIL Table 6 shows ha in he shor run, he predicion error of he variable GDP is mosly explained by he variable iself. Bu, as ime elapses, he conribuion of he oher variables grows. In his respec, he variable oil price has a higher share han he oher variables so ha

9 Environmenal Energy and Economic Research (2018) 2(1): afer 10 periods, i accouns for 42.21% of he predicion variance error of GDP while 3.70% is explained by GDP. The variable price index has he lowes share in explaining he predicion error of GDP so ha i accouns for only 0.063% afer 10 lags. Table 7. Analysis of variance for oil price Lag Analysis of variance of GDP POIL GDP EX CPI According o Table 7, he variable oil price has he highes share in accouning for he predicion error of his variable in he shor run, bu over ime he oher variables become more imporan. Among hese variables, GDP has a higher share han he oher variables so ha i capures 5.18% of he predicion variance error of oil price afer 10 periods, while 0.89% is accouned for by GDP. The leas imporan variable in explaining he predicion error of GDP is price index ha accouns for 0.083% of his error afer 10 lags. Table 8. Analysis of variance for exchange rae Lag Analysis of variance of GDP EX CPI GDP POIL Table 8 reveals ha in he shor run, he mos imporan variable in explaining he predicion error of he exchange rae is he variable iself. However, over ime oher variables gain imporance among which price index is more imporan han he oher variables and accouns for 37.07% of he predicion variance error of price index afer 10 periods. The share of GDP in explaining he predicion variance error of exchange rae is 1.73%. The variable oil price is he leas imporan in explaining he predicion error of exchange rae and accouns for 1.29% of his error afer 10 lags.

10 60 Babajani Baboli e al. Conclusion and Recommendaions The facs abou Iran s economy show ha any flucuaion in oil price and he exchange rae is influenial on inflaion whereas foreign exchange earnings from oil expors have dominaed he oal foreign exchange earnings of his counry despie he emergence of diverse socioeconomic shocks, nuclear issues, and sancions. Our resuls show ha Iran has faced a drasic decline of oil incomes in 2012 and 2013 due o he severe sancions on he banking sysem and oil exper by he European Union, he US and he UN given he fac ha oil accouns for abou 80% of foreign exchange incomes. In his period, oil price did no change considerably, bu he producion and exporaion raes were decreased and since sancions were applied on foreign exchange ransfer via he banking sysem, Iran sared o suffer from he shorage of exchange supply. Subsequenly, he exchange rae was muliplied and mismanagemen aggravaed he urmoil of he exchange marke. The increase in he exchange rae and he loss of naional exchange value resuled in he reducion of impors, and since capial and inermediae commodiies mosly consiue he imporing iems, he loss of impors impaired manufacuring secion subsanially so ha some indusrial unis were closed and some kep operaing wih heir minimum capaciy. Also, he consrucion aciviies of he governmen were sopped in hese years and mos projecs were abandoned unfinished. Consequenly, he counry sank ino a deep recession so ha he inflaion rae exceeded 10 percen and economic growh rae became negaive. The main cause of his sagflaion was he severe loss of oil incomes due o he sancions. According o our resuls, he governmen should manage oil incomes soundly and preven he swif change of exchange incomes which aggravaes inflaion. I should also direc oil incomes owards manufacuring projecs o hinder he inflaion pressure induced by he increase in general level of he prices. I is, also, imperaive o reform economic policies of axes in he naional economy o shif he reliance of he governmen expendiures o oil incomes. Furhermore, acions should be aken o manage and subsiue oher commodiies o avoid he impacs of he shocks in he oil marke. References Aloui, C. and Jammazi, R. (2010). The Effecs of Crude Oil Shocks on Sock Marke Shifs Behavior: A Regimes Wiching Approach. Energy Economics, 31(5), Ashari, A., Nadri, K., Abolhasani, A., Mehregan, N., and Babaei, M. (2015). The impac of oil price shocks on inflaion, growh and money: A case sudy of Iran. Quarerly Journal of Economic Growh and Developmen Research, 6 (22), Cashin, P., Mohaddes, K., Raissi, M., and Raissi, M. (2014). The Differenial Effecs of Oil Demand and Supply Shocks on he Global Economy. Energy Economics, 44(3), Ghafari, H., Ashiyani, A., and Jaloli, M. (2013). Invesigaion and forecasing of he effecs of exchange rae increase on major macroeconomic variables of Iran.Quarerly Journal of Applied Economics Sudies in Iran, 2 (8), Gojarai, D. N. (2003). Basic Economerics. Singapore: Me Grew, Fifh Ediion. Hamilon, J. D. (2008). Undersanding crude oil prices. In: NBER Working Papers Naional Bureau of Economic Research, Inc. Hamilion, J. D. (1983). Oil and he macroeconomy since world war II. Journal of Poliical Economy, 91, Hamilion, J. D. (2003).Wha is an oil shock? Journal of Economerics, 113(2), Huang, S., An, H., Gao, X., Wen, S., and Hao, X. (2016). The muliscale impac of exchange raes on he oil-sock nexus (Evidence from China and Russia). Applied Energy,194(1), Johansen, S. (1988). Saisical Analysis of Coinegraion Vecors. Journal of Economic Dynamics and Conrol,12, Kilian, L. (2008). The Economic Effecs of Energy Price Shock. Journal of Economic Lieraure, 46,

11 Environmenal Energy and Economic Research (2018) 2(1): Kun Sek, S., Qi., Teo, X. and Nee Wong, Y.(2015).A Comparaive Sudy on he Effecs of Oil Price Changes on Inflaion, Proedia Economic and Finance Madesha, W., Chidoko, C., and Zivanomoyo, J. (2013). Empirical Tes of he Relaionship beween Exchange Rae and Inflaion in Zimbabwe. Journal of Economics and Susainable Developmen, 4(1), Rasmussen, N. R. and Roiman, A. (2011), Oil shocks in a global perspecive: are hey really ha bad? IMF working paper. 11(3), Ragooobur, V., and Chicooree, A. (2012). Exchange Rae Pass-Through o Domesic Prices: Evidence From Mauriius. Journal of Economics Research, Samadi, S., Yahyabadi, A., and Moallemi, N. (2010). An analysis of he effec of price shocks on macroeconomic variables in Iran. Quarerly Journal of Economic Research and Policies, 17 (52), Samei, M., Kasraei, K., Ranjbar, H., and Ghobadi, S. (2017). Analyzing non-linear effecs of oil price volailiies on Iran's macroeconomic variables wih emphasis on indusry and mine secor (An applicaion of sae-space model). 11 (22), Sarzaeem, A. (2007). The effec of oil shocks on economic variables in a VAR model. Quarerly Energy Economics Review, 4(12), Siok Kun, S., Qi, X., and Nee Wong,y(2015),A comparaive sudy on he effecs of oil prices change on inflaion. Procedia Economics and Finance, 26, Yazdani, M., and Zare, S. (2016). Invesigaing effec of exchange rae shocks on inflaion in Iranian economy during seasonal period Quarerly Journal of Applied Economics Sudies in Iran, 5(17), Yoshizaki, Y., Hamori, Sh. (2013). On he Influence of Oil Shocks on Economic Aciviy,Inflaion and Exchange Raes, Inernaional Journal of Financial Research, (4)2, Environmenal Energy and Economic Research is licensed under a "Creaive Commons Aribuion 4.0 Inernaional (CC-BY 4.0)"