STUDYING RELATIONSHIP BETWEEN EXCHANGE RATE AND DIVIDEND (CASE STUDY: IRAN KHODRO COMPANY)

Transcription

1 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus STUDYING RELATIONSHIP BETWEEN EXCHANGE RATE AND DIVIDEND (CASE STUDY: IRAN KHODRO COMPANY) Ronak Mahdavi Dr.Aaollah Mohammadi 1 Deparmen of Accouning, Kurdisan Science and Research Branch, Islamic Azad Universiy Sanandaj, Iran 2 Assisan Professor, Deparmen of Accouning, Sanandaj Branch, Islamic Azad Universiy Sanandaj, Iran Absrac One of he characerisics of developed counries is he presence of efficien markes and financial insiuions which cause developmen of hese counries in addiion o heir imporan role in economy of hese counries. Since share marke value is affeced by macroeconomic variables, he presen research was conduced o deermine long-erm relaionship beween foreign currency variaions and dividend of Iran Khodro Company and explanaory variables of oil price and consumer price index have been used as explanaory variables. In his research, daa are sudied every year for ime period of 1991 o To sudy ime series saics, augmened Dicky Fuller es has been used. Resuls of his es showed ha variable of foreign currency was saic in he level and oher variables were saic in he firs order difference. In his research, relaions beween variables have been deermined based on exchange rae flucuaions have been deermined using Auo Regressive Wih Disribued Lags(ARDL) and resuls of his es indicae he presence of posiive and significan relaionship beween exchange rae and dividend and beween exchange rae and consumer price index and negaive and significan relaionship beween exchange rae and oil price. Keywords: exchange rae, oil price, consumer index, dividend, Iran Khodro Company 1. Inroducion One of he characerisics of coninual movemen oward susainable economic developmen is o obain necessary financial resources for economic aciviies by equipping saving resources in naional economy. In recen decades, expansion of capial marke in developing counries led o desirable economic growh. Developed counries believe ha major par of developmenal rend is due o financial markes and paricularly sock exchange. Financial markes are indicaive of economy of each counry. Boom and sagnaion of hese markes no only affec naional economy bu also regional and global economy. Sock exchange as one of he main financial markes is place of saving and liquidiy of privae secor o finance invesmen projecs on he one hand and formal and safe source in which sagnan savings can search for relaively suiable and safe place for invesmen and apply heir money for invesmen in companies. I is eviden ha boom and sagnaion of sock exchange can resul from differen facors in economy. In case his marke has no logical relaion wih oher secions, heir performance will have difficuly and defec. 2. Problem Saemen and Significance of he research Invesors should make invesmen wih hope of achieving more wealh. One of he imporan facors which invesors consider in heir decision making is dividend. Dividend in invesmen is he impeus which creaes moivaion and is regarded as a reward for invesors. In fac, any 347

2 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus invesors should ensure ha principal of invesmen will be reurned in he firs sage and hen he ineres is acquired o decide abou invesmen. Dividend is affeced by differen facors. One of hese facors can be exchange rae flucuaions. Theoreically, uncerainy of exchange rae flucuaions affecs local economic secion paricularly dividend in addiion o foreign rade secion. In an open economy, services and capial are applied considering exchange rae. Therefore, exchange rae can affec major variable of expor, impor, enry and exi of capials. In fac, i can be said ha exchange rae flucuaions creae a kind of risk in foreign ineracions secion which can disrup expor, impor and capial flows. Therefore, if exchange rae variaions are adjused in proper direcion, hey can provide suiable and agreeable environmen for producion, rade and invesmen, exchange rae flucuaions will change price of goods and services, producion and producion facors and will affec cash flows and expeced fuure flows and consequenly dividend of he economic agency. Hence, reducion of money value increases demand for local goods due o increase of relaive price of he foreign goods compared wih he local goods resuling in increased general level of prices. On he oher hand, i reduces impor of he inermediary and capial daa due o increase of heir price leading o increase of producion cos and reducion of invesmen, reducion of demand for share and finally reducion of dividend. Considering significance of he subjec, share marke acs as economic indices of he counry and increase of invesmen in his marke and aracion of capials o capial marke requires increase of dividend, reducion of risk and emergence of desirable condiions for invesmen. Since macroeconomic variables such as exchange rae, inflaion, oil price ec are effecive on dividend, clear long-run relaionship beween macroeconomic variable and dividend can help managers and fuure invesors make decision. 3. Hisory of Research Relaionship beween exchange rae and shares has been one of he conroversial issues among he researchers since a long ime ago. For his reason, i has long hisory in his research. In his Secion, some of hese sudies which relae o he sudied subjec are reviewed. Mohammad Barzandeh (1997) showed ha role of variables of share price of Tehran Sock Exchange, exchange rae, vehicle price index and house price index in variaions of share price index. Hassan Ghalibaf Asl(2002) showed ha variaions of exchange rae had negaive effec on sock reurn bu variaions of exchange rae wih a ime lag had posiive effec on sock reurn of Tehran sock exchange companies. Eslamion and Zare (2006) showed ha exchange rae had posiive effec and money volume had negaive effec on share price index in Tehran sock exchange. Heidari (2001) showed Granger s Bidirecional Causaliy relaionship beween index of share price, exchange rae, auomobile share price index and house price index in Iran. Mosafa Karim Zadeh (2005) showed ha here was a co-inegraion vecor beween sock exchange index and moneary variables. Long-run relaionship confirms significan posiive effec of liquidiy and significan negaive effec on exchange rae and bank ineres rae on sock exchange index. Barzani e al. (2009) found ha marke share value had direc relaionship wih expendiure of governmen and money volume and reverse relaionship wih ax and exchange rae in long run. Sudy of shor-run relaionship using vecor error correcion model also showed ha shor-run flucuaions of variables were relaed o long-run equilibrium value. Najar Zadeh e al. (2009) found ha effec of exchange rae flucuaions increased share price in shor and medium erm and reduced share price in long erm in Tehran Sock Exchange. Bogari (2003) showed ha behavior of capial markes also could be effecive in addiion o macroeconomic variables. Wali (2005) found ha rade, financial merger, economic srucure of counries, informaion dissymmery and Exchange Rae Policy are he facors which can be effecive on behavior of share marke. Pen e al. (2007) found ha here was significan relaionship beween exchange rae and share price for Hong Kong, Japan, Malaysia and Thailand before financial crisis of There was also relaionship beween share marke and exchange rae marke for Hong Kong, Korea and Singapore. During financial crisis, no counry shows significan relaionship beween share price 348

3 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus and exchange rae bu here is relaionship beween exchange rae and share price for all of he sudied counries excep Malaysia. Morley (2009) in England, Japan and Swizerland showed ha here was long-run relaionship beween exchange rae and share price for he said counries. Resuls of esimaing error correcion models sugges posiive relaionship beween exchange rae and share price. Bir and Habin (2008) sudied relaionship beween share price and exchange rae for wo groups of developed counries and developing counries. Resuls showed ha here were no sable flucuaions in share marke in he developed counries while he menioned flucuaions are sable in developing counries. Yan and Nieh (2009) sudied effecs of exchange rae of Taiwan compared wih Japan, share price in Japan and Taiwan during using hreshold error correcion model. Resuls of Granger Causaliy Tes show ha here is no shor run relaionship beween wo financial asses for boh counries bu hey have asymmeric long-run relaionship. Zhao(2010) in China showed ha here was no sable long-run equilibrium relaionship beween real exchange rae and share price and he pas changes in share marke had larger effec on he fuure exchange rae flucuaions. There are also bidirecional spillover effecs of flucuaions beween wo markes. The model used in his research has been used as main model for he research. Alajideh e al. (2010) found ha here was causal relaionship beween exchange rae and share price for Canada, Swizerland and England and here was causal relaionship beween share price and exchange rae for Swizerland. Sobari and Salehiv(2010) in Nigeria showed ha exchange rae flucuaions had negaive and significan effec on share price while ineres rae and inflaion rae had no long-run relaionship wih share marke. Chiza (2011) in Souh Africa showed bidirecional relaionship beween macroeconomic variables and share price and also concluded ha uncerainy of macroeconomic variables had significan effec on share marke flucuaions. 4- Research Variables The variables used in he research model are divided ino hree classes: dependen, independen and conrol Dependen variable EPS : is dividend of Iran Khodro Company a ime. in capial marke and in financial managemen erm, EPS means he amoun of ineres which is obained by dividing ne profi by he number of published shares of company (public join sock) Independen Variables GXE : Variaions of exchange rae a ime. OP : Oil price a ime. (oil price has been calculaed in USD and is mean price of each barrel). CPI : Consumer price index (CPI) a ime. Consumer price index is he average price of goods and services which a family purchases. In Iran, Laspeyers mehod is used for calculaing consumer price index. In his mehod, base year consumer price index in curren price should be divided by base year consumer price index and hen muliplied by 100. In his case, variaion of price in he base year is sudied Conrol Variable CAP is capial of company a ime. capial of join sock company means sum of nominal price of is share. Therefore, he capial is divided ino equal shares and each one of he parners will have one or more shares. 4. Research Model: in he model which relaionship beween dividend of Iran Khodro Company (as dependen variable ) and independen variable of exchange rae variaions along wih independen variables of oil price, consumer price index (CPI ) and capial of he company can be sudied, Auo Regressive Wih Disribued Lags(ARDL) can analyze his model in he bes manner. In addiion, ARDL model as a dynamic equaion wih equilibrium soluion can give suiable response for he raised quesions abou he discussed issue. Therefore, he research model is as follows: 349

4 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus EPS 0 1GXE 2OP 3CPI 4 CAP I is worh noing ha is error erm in regression models which indicaes effec of oher facors on dividend bu hey have no been given in he model. 5. Research Hypoheses To achieve goals of he research, he following main hypoheses can be exraced: 1- Variaions of exchange rae have significan effec on dividend of Iran Khodro Company. 2- Oil price has significan effec on dividend of Iran Khodro Company. 3- Consumer price index CPI has significan effec on dividend of Iran Khodro Company. 6. Research Mehod This research is of applied ype and is included in correlaing sudies. This research mehod is applied for conducing he sudies which inend o invesigae cause or causes of definie relaionships which have occurred in he pas and have been compleed. This ype of research mehod has relaively high validiy because i seeks o achieve causal relaion or cause and effec relaion beween facors of he research. Therefore, i is of Ex-pos faco ype. In hese researches, variables canno be manipulaed by researcher or arificial or experimenal condiions canno be creaed by him for differen reasons (Seyed Abbas Zadeh, 2001). 7. Informaion Collecion Mehods Informaion used in his research can be divided ino wo classes. The firs class includes informaion relaing o heoreic fundamenals and lieraure which were provided by reviewing Iranian and Foreign papers and heses available in inerne and local and foreign publicaions hrough library sudies. The second class includes informaion and saisics necessary for esimaing he model in which saisically reliable references (Cenral Bank of Islamic Republic of Iran, financial repors of he company from sie of Tehran Sock Exchange Organizaion and Rahavard Novin Daabase Sofware) have been used. 8. Saisical Populaion In his research, Iran Khodro Company was considered as saisical populaion and was sudied in 22-year period from 1991 o Tesing Hypoheses 9.1. Esimaing Model Using ARDL Mehod We firs selec he maximum number of lags 2 o esimae he model based on a mos 2 lags for each variable. In he nex sep, we should selec one ou of four crieria which include adjused coefficien of deerminaion, Akaike, Schwarz Bayesian and Hannan-Quinn crieria. Our selecive crierion for esimaing he model is Schwarz Bayesian ( (SBC)). We used his crierion because i esimaes coefficiens wih he leas lag. In Auo Regressive Wih Disribued Lags(ARDL), we can use shor-erm model dynamics. In his Secion, CUSUM and CUSUMSQ graphic ess are used for shor-erm model of residuals. These ess are graphically presened. Now, if plo of cumulaive sum of recursive residuals is locaed inside area beween wo criical lines in level of 5% for CUSUM es, he long-run relaionship will be sable and sabiliy of he esimaed coefficiens is confirmed. Bu if cumulaive sum of recursive residuals is locaed ouside area beween wo criical lines in level of 5%, he long-run relaions will be unsable. In oher words, sabiliy of he long-run relaionship in differen ime periods will be jeopardized. I also holds rue for CUSUMSQ and he only difference is ha square of cumulaive sum of recursive residuals is used in he es based on CUSUMSQ mehod Sudying Long-run relaionship Angel-Granger Two-Seps Tes (1978) has limied esimaes bias in small samples, herefore, esimaion of he model and es of hypoheses are invalid wih helps of common saisics. Angel- Granger Two-Seps Tes is based on premise of coinegraion vecor and when here is more han one coinegraion vecor, use of his mehod will no be efficien. To remove hese limiaions, 350

5 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus Auo Regressive wih Disribued Lags (ARDL) has been presened by Pesaran and Shin (1998). This mehod is no sensiive o cumulaive sum of explanaory variables and compaible esimae of he long-run coefficiens of he model can be obained by selecing he number of suiable lag in he model. The second mehod which has been presened by Pesaran, M.H. and e al. (1996) ess he presence of long-run relaionship beween he sudied variables wih F saisic of bounds es for esing significance of levels wih lag of he variables in error correcion form. The imporan poin is ha he above F disribuion is no sandard. Pesaran, M.H. and e al. have calculaed suiable criical values corresponding o he number of repressor wheher he model includes inercep and rend or no. They have presened wo groups of criical values: one is based on he fac ha all variables are saionary and anoher one is based on he fac ha all are non-saionary (hey have become saionary wih one differencing. If compuaional F of he bounds es are ou of his bound, definie decision will be made wihou need o know ha variables are I(0) or I(1). If compuaional F of bounds es exceeds he upper bound, null hypohesis ha here is no long-run relaionship is no acceped and if i is below he lower bound, null hypohesis will no be rejeced. If compuaional F of he bounds es is beween wo bounds, resuls of inference will be uncerain and dependen on wheher variables are I(0) or I(1). Under hese condiions, we have o perform uni roo ess on he variables Esimaing Long-run Equaion: Model esimaion mehod is Auo Regressive Wih Disribued Lags(ARDL). To esimae he model wih his mehod, i is necessary o deermine he number of opimal lags of he variables. To deermine opimal lags in he equaions, Schwarz Bayesian Crierion(SBC) has been used. Resuls of esimaing long-run equaions are given in Table Esimaing Error Correcion Model To esimae his model, a long-run relaion should be esimaed and in case ha i is no false, lagged residual coefficien of he long-run relaion is regarded as error correcion coefficien and he following relaion is esimaed: Y a b X cu 1 e Error correcion coefficien means esimaion of c wih negaive mark which will indicae speed of error correcion speed and endency o long-run equilibrium. This coefficien shows ha nonequilibrium of he dependen variable is adjused and approaches he long-run relaion (Tashkini, 2005). 10. Research Resuls Table 1- Resuls of Shor-run Esimaion Lags of he model wih Schwarz Bayesian Crierion(SBC) have been seleced as ARDL(1,0,1,0,0). Firs model Variable Saisic Coefficiens EPS ( 1) GXE OP OP ( 1) CPI CAP Inercep R 2 1 Tashkini, 2005, P

6 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus Prob[.009] F DW-saisic According o Table 1, all variables excep capial of company have expeced signs and are saisically significan in high confidence level. Coefficien of he main variable i.e. variaions of exchange rae has been evaluaed posiive. In oher words, increase of exchange rae has led o increaser of dividend of Iran Khodro Company. Oil price has negaive and significan effec on dividend of Iran Khodro Company. In oher words, wih increasing oil price, dividend of Iran Khodro Company has decreased. General level of prices has posiive and significan effec on dividend of Iran Khodro Company because cash purchasing power has decreased wih increasing general level of prices and inflaion rae and people inves heir money in durable goods (such as house, shares and bonds). Hence, dividend also increases wih increasing demand of sock exchange. Variable coefficien of capial has been evaluaed negaive and significan indicaing ha increase of he company s capial has reduced dividend of he company and sign of his coefficien was no expeced. In Table 2, resuls of bounds es are given. Therefore, i can be said ha bounds es confirms coinegraion relaionship among he model variables in significance level of 5%. For his reason, he presence of long-run relaionship among he variables canno be rejeced. Table 2- Bounds es for sudying long-run relaionship 95% 90% Upper Bound Lower Bound F-saisic Lower Bound Upper Bound Table 3- Resuls of Long-run esimaion Lags of he model wih Schwarz Bayesian Crierion(SBC) have been seleced as ARDL(1,0,1,0,0). Firs model Saisic Coefficiens Variable GXE CAP Inercep According o Table 3, all variables excep capial of company have expeced signs in long-run model and are saisically significan. Coefficien of he main variable i.e. variaions of exchange rae has been evaluaed posiive in long-run model. In oher words, increase of exchange rae has led o increaser of dividend of Iran Khodro Company. Oil price has negaive OP CPI 352

7 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus and significan effec on dividend of Iran Khodro Company. In oher words, wih increasing oil price, dividend of Iran Khodro Company has decreased. General level of prices has posiive and significan effec on dividend of Iran Khodro Company because cash purchasing power has decreased wih increasing general level of prices and inflaion rae and people inves heir money in durable goods (such as house, shares and bonds) for compensaion. Hence, dividend also increases wih increasing demand of sock exchange. Variable coefficien of capial has been evaluaed negaive and significan indicaing ha increase of he company s capial has reduced dividend of he company and sign of his coefficien was no expeced. Considering confirmaion of long-run relaionship beween economic variables in he model, shor-run relaionships beween financial shocks and real consumpion of he privae secor have been esimaed using error correcion mehod (ECM) and resuls of esimaion are given in Table 8-4: Table 4- Resuls of esimaing error correcion mehod (ECM) Lags of he model wih Schwarz Bayesian Crierio n(sbc) have been seleced as ARDL(1,0,1,0,0). Dependen variable is Households Consumpion expendiure and he number of observaion is 34. Saisic Coefficien Variable dgxe dop dcpi dcap ecm(-1) error correcion model According o Table 4, shor-run relaionships among he variables are also confirmed and all coefficiens are saisically significan. ecm(-1) coefficien indicaing adjusmen speed of nonequilibrium process is equal o -0.60, hence, 60% of he deviaions of Households Consumpion expendiure from long-run roue in each period are adjused in he nex period and move oward heir long-run equilibrium. In Table 5, diagnosic ess of ARDL model are given. Based on resuls of his able, hypohesis of variance consisency among he error erms canno be rejeced considering high probabiliy of saisic, herefore, here is no heeroskedasiciy among he error erms. Considering high probabiliy of saisic, hypohesis of lack of serial auocorrelaion among error erms canno be rejeced, herefore, here is no serial auocorrelaion among error erms. Table 5- (0,1,1,2,0,0) ARDL model diagnosic ess serial auocorrelaion es Heeroskedasiciy es Saisic Probabiliy Saisic Probabiliy CHSQ (0.187) CHSQ (0.118) F (0.170) F (0.127) 11. Conclusion Using he obained resuls, all hree hypoheses were confirmed. To analyze direc relaionship beween variaions of dividend exchange rae of Iran Khodro Company, i can be said ha Iran Khodro Company has beer compeiive power in expor wih increasing exchange rae and is in beer condiion wih increasing revenues from expor of goods and services. In his case, demand for share of company increases wih increasing exchange rae and improving revenue condiion and finally dividend of he socks of his company increases. In he esimaed models wih dependen variable of dividend which has been applied in he presen research, increase of 353

8 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus general level of prices has posiive effec on dividend of Iran Khodro Company. In oher words, sock reurn of sock exchange companies averagely increases wih increasing price indices (increase of inflaion ). This phenomenon can be jusified such ha cash purchasing power is reduced wih increasing general level of prices and increasing inflaion rae and people inves heir money in durable goods (such as house, shares and bonds). In case of inflaion, sockholder will adjus his expeced reurn based on inflaion rae. Therefore, he expeced reurn of invesor increases wih increasing inflaion rae. In he research model, oil price has significan effec on sock reurn. Considering major role of oil revenue in governmenal coss, expor and naional income, oil secor plays dominaing role in economy of Iran and has disruped economic sysem of Iran. Considering ha oil price is a poliical economic variable and is beyond conrol of Iran, i creaes moivaional and posiive effec on capial marke of Iran following economic boom and consequenly increase of oil price. Based on experiences of capial marke flucuaions and he direc relaionship wih oil price, i seems reasonable o jusify posiive effec of oil price on sock reurn. As documened evidence, global crisis and sagnaion 2009 led o considerable reducion of oil price and sock reurn of mos companies. Bu as observed above, income flucuaions lack such cerain and significan effec. I can be said ha alhough oil price has significan effec on sock reurn, oil revenues lack such effec due o facor of oil producion rae. In oher words and based on he exising evidence paricularly performance of he member saes of OPEC, when global oil price increases, general decision is abou reducion of oil producion rae. In oher words, effec of increased price is neuralized by reducing oil producion rae and his led o he absence of effec of oil incomes on sock reurn of Iran Khodro Company. 12. Conclusion Since rade volume is no effecive on sock reurn, policymakers should expor and impor he goods which lead o rise and moivaion of capial marke and inernaional economy of he economic policymakers should be planned based on improvemen of rade relaions and ineracions beween Iran and hese counries. As observed above, one of he resuls of his research is posiive effec of oil price on sock reurn and generally boom of sock exchange. Therefore, i is recommended o preven excessive dependency of naional economy on oil because capial marke will decline considerably wih decreasing oil price. In he used research model, annual daa has been used due o inaccessibiliy of monhly daa. To achieve he research goals, monhly daa of oher companies can be used for esimaing he model, if available. In he presen research, here was no enough ime for dealing wih relaionship beween inflaion and dividend. Hence, i is of special imporance o sudy ineracion beween hese moneary variables. The used research model was Auo Regressive Wih Disribued Lags(ARDL). Trend of his work can be sudied using oher dynamic models such as vecor auo Regression mehod (VAR), if necessary. 13. References - Cenral Bank of he Islamic Republic of Iran, Economic Sudies Deparmen, Iran Naional Accouns, differen years -Noferesi, Mohammad (1999), uni roo and coinegraion in economerics, Rasa Culural Services Insiue, Tehran - Gojarai, Damodar(1999), economeric fundamenals, second volume, ranslaed by Hamid Abrishami, Universiy of Tehran Publicaion, Tehran - Tashkini, Ahmad (2005), applied economerics wih help of Microfi, Tehran: Dibagaran Arisic Culural Insiue, firs ediion - Dickey, D.A. and W.A. Fuller (1979), Disribuion of Esimaors for Auoregressive Time series wih a uni Roo. Journal of American Saisical Associaion, Vol. 74, pp Granger, C.W.J. (1988), Developmen in he sudy of co-inegraed Economic Variables, Oxford Bullein of Economics and Saisics, Vol. 48, p Gujarai, D.N. (2004), Basic Economerics, 4 h Ediion, McGraw-Hill, New York. 354

9 Kuwai Chaper of Arabian Journal of Business and Managemen Review Vol. 3, No.12; Augus Pesaran, M.H., Shin, Y., Smih, R.J., (2001), Bounds esing approaches o he analysis of level relaionships, Journal of Applied Economerics. No.16, pp Pesaran, M. H., & Shin, Y. (1995). An auoregressive disribued lag modeling approach o Coinegraion analysis. In: S. Srom, A. Holly, & P. Diamond (Eds.), Cenennial volume of Rangar Frisch. Cambridge Universiy Press, Cambridge. - Pesaran, M.H and Pesaran, B. (1997), Working wih Microfi 4.0: Ineracive Economeric Analysis. Camfi Daa Limied, Cambridge. - Inder, B. (1993), Esimaing long-run relaionships in economics: A Comparison of differen approaches, Journal of Economerics, No.57, pp