Macroeconomic Determinants of the Stock Market Index for a Major Latin American Country and Policy Implications

Transcription

1 Macroeconomic Deerminans of he Sock Marke Index for a Major Lain American Counry and Policy Implicaions Yu Hsing Deparmen of Managemen & Business Adminisraion, Souheasern Louisiana Universiy SLU 10813, Hammond, Louisiana 70402, U.S.A. Tel: yhsing@selu.edu Michael C. Budden Deparmen of Markeing & Supply Chain Managemen, Souheasern Louisiana Universiy SLU 10844, Hammond, Louisiana 70402, U.S.A. Tel: mbudden@selu.edu Anoinee S. Phillips Deparmen of Managemen & Business Adminisraion, Souheasern Louisiana Universiy SLU 10813, Hammond, Louisiana 70402, U.S.A. Tel: aphillips@selu.edu Received: December 8, 2011 Acceped: December 23, 2011 doi: /ber.v2i URL: hp://dx.doi.org/ /ber.v2i Absrac Applying he exponenial GARCH model and based on a quarerly sample during 1998.Q Q2, we find ha he Argenine sock marke index is posiively associaed wih real GDP, he raio of M2 money supply o GDP, he peso/usd exchange rae and he U.S. sock marke index. I is negaively influenced by he money marke rae, governmen spending as a percen of GDP and he inflaion rae. Hence, a srong domesic economy, a lower ineres rae, an increased money supply as a percen of GDP, lower governmen spending as a percen of GDP, depreciaion of he Argenine peso, a lower inflaion rae, or a robus U.S. sock marke would help he Argenine sock marke. Keywords: EGARCH, Moneary policy, Fiscal policy, Currency depreciaion, Conagion effec 1

2 1. Inroducion Due o he global financial crisis, a possible defaul of sovereign deb in Greece and several oher EU member saes, a huge naional deb, a poenial double-dip recession in he U.S. and oher concerns, sock indexes in many counries have experienced a high degree of volailiy. Argenina is no excepion. Before he global financial crisis, he Argenine Merval sock index reached a high of 2, on May 22, Due in par o he 2008 global financial crisis, he sock index plunged 63.13% o a low of on November 21, A one poin since hen, he sock index had risen %, reaching a high of 3, on January 20, Due o recen concerns abou poenial defaul of sovereign deb in Greece and oher financial relaed issues, he index dropped o 2, on December 7, Alhough he index rose slighly recenly, we expec ha he sock index would follow he rends in he U.S. and oher advanced counries and coninue o exhibi some degree of volailiy in he near fuure. The objecive of he paper is o deermine poenial impacs of several seleced macroeconomic variables on he Argenine Merval sock index. We selec Argenina as a case sudy mainly because of is being he second larges counry in Lain America, rising economic saus, and valuable experiences in economic developmen. I has several focuses. Firs, he sudy invesigaes he respecive impacs of a change in he ineres rae, governmen spending and he exchange rae on he Argenine sock marke index. Mainaining a lower ineres rae, fiscal prudence leading o less ax burdens or an appropriae exchange rae level promoing expors and price sabiliy would help he sock marke. Second, he sensiiviy of he Argenine sock index o a change in he U.S. sock marke index is invesigaed. I is widely believed ha mos sock marke indexes in Lain American counries are highly correlaed wih he U.S. sock marke index due o he conagion effec (Fernández-Serrano and Sosvilla-Rivero1, 2003). Third, he convenional GARCH model imposes he resricion of non-negaive parameers on he variance equaion. In his paper, he exponenial GARCH (EGARCH) model (Nelson, 1991) is employed in empirical work in order o have fewer resricions on he parameers and o allow he parameers o have negaive values in he variance equaion, and o respond o he good and bad news asymmerically. Therefore, he EGARCH model can capure he behavior of he sock marke beer han he convenional GARCH model. 2. Lieraure Survey There have been several recen sudies examining he impacs of macroeconomic facors on sock prices, inernaional linkages of sock markes, he random walk hypohesis, shor-run and long-run relaionships for Argenina and relaed counries. Spyrou (2004) analyzes he relaionship beween he sock price and inflaion for 10 seleced emerging markes. For Argenina, he finds ha he relaionship was posiive and significan during 1989M1 1995M12 and 1989M1 2000M8 bu posiive and insignifican during 1995M M6. Abugri (2008) examines he impac of macroeconomic variables on sock marke indexes for four seleced Lain American counries. For Argenina, he VAR esimaion reveals ha he sock marke index is posiively influenced by he MSCI global 2

3 equiy index, negaively affeced by he U.S. Treasury bill rae, he domesic ineres rae and he money supply, and no impaced by indusrial producion and he exchange rae. Sudying real sock reurns for seven Lain American counries, Araújo (2009) shows ha for Argenina, he supply shock, he porfolio shock and he demand shock can explain 60.37%, 24.11% and 8.32% of he variaion in real sock reurns, respecively. He also indicaes he correlaion coefficien of beween real sock reurns in Argenina and he U.S. is significan a he 1% level. Invesigaing he linkage beween he U.S. and seleced Lain American sock markes, Fernández-Serrano and Sosvilla-Rivero1 (2003) find srong evidence of a long-run coinegraing relaionship beween he DJIA index and he Argenine, Venezuelan and Chilean indexes afer he 1998 financial crisis and a coinegraing relaionship beween he U.S. S&P 500 index and he Argenine, Mexican and Chilean indexes from Augus 1998, April 1999 and Ocober 1999, respecively. Diamandis (2009) indicaes ha 4 seleced Lain American sock markes including Argenina and he U.S. sock marke exhibi 4 common permanen componens, are parially inegraed, and have relaively small benefis in he long run wih inernaional diversificaion due o very sluggish adjusmens. Jawadi, Arouri and Nguyen (2010) find srong evidence ha he Argenine and Mexican sock markes are influenced by he U.S. sock marke in he shor run whereas here are no linkages found in he long run. Hence, hey conclude ha sock markes in Argenina and Mexico are deermined by he fundamenals in he long run. Hasan, Kadapakkam and Ma (2003) examine wheher he random walk hypohesis may hold for 8 seleced Lain American counries including Argenina. They reveal ha he random walk hypohesis applies o Argenina, does no apply o Peru, and may or may no apply o he remaining 6 counries. Some of he above findings may need o be furher sudied. For example, he posiive significan or insignifican coefficien for he inflaion rae (Spyrou, 2004), he relaively low correlaion beween sock marke indexes beween Argenina and he U.S. (Araújo, 2009), lack of a long-erm sable relaionship beween he sock marke indexes in Argenina and he U.S. (Jawadi, Arouri and Nguyen, 2010), he insignifican coefficiens of indusrial producion and he exchange rae and he negaive significan coefficien of he money supply for Argenina repored by Abugri (2008) need o be reexamined o deermine wheher a differen mehodology and a more recen sample wih 10 more years of daa may yield differen resuls. 3. The Model Exending previous sudies, we can express he Argenine sock marke index as a funcion of he following macroeconomic and global variables: MI f ( RO, IR, MS, GY, EXC, INF, SP ) (1) 3

4 where MI RO MI MI MI MI MI MI 0, 0, or 0, or 0, or 0, 0, 0 IR MS GY EXC INF SP MI RO IR MS GY EXC INF SP * = Argenine Merval sock marke index, = real GDP, = he ineres rae, = he money supply, = governmen consumpion spending as a percen of GDP, = he nominal exchange rae, = he inflaion rae, and = he sock marke index in he U.S. MI Noe ha indicaes he parial derivaive of he Argenine sock marke index wih X respec o any one of he explanaory variables. We expec ha he sign of real GDP (Chen, 1986; Fama, 1990; Abdullah and Hayworh, 1993; Mukherjee and Naka, 1995) and he U.S. sock price (Fernández-Serrano and Sosvilla-Rivero1, 2003; Araújo, 2009; Jawadi, Arouri and Nguyen, 2010) o be posiive, he sign of he ineres rae (Fama, 1981, 1990; Mukherjee and Naka, 1995; Raanapakorn and Sharma, 2007; Humpe and Macmillan, 2009)and he inflaion rae (Fama, 1981; Geske and Roll, 1983; Mukherjee and Naka, 1995) o be negaive, and he sign of he money supply (Raanapakorn and Sharma, 2007; Abugri, 2008; Humpe and Macmillan, 2009), he raio of governmen spending o GDP (Darra, 1990a, 1990b) and he exchange rae (Nieh and Lee, 2001; Kim, 2003; Raanapakorn and Sharma, 2007; Abugri, 2008) o be unclear. A moderae increase in he money supply would increase liquidiy and credi and be conducive o business operaions. On he oher hand, oo much money supply would cause inflaion o rise, reduce buying power, hur he economy, and harm sock prices When governmen spending is a a reasonable level, i would help raise aggregae demand and simulae he economy. On he oher hand, when governmen spending as a percen of GDP rises beyond a cerain hreshold, a large governmen defici or deb as a percen of GDP would cause a crowding-ou effec, raise he long-erm ineres rae, hur privae spending, and reduce sock prices. Currency depreciaion may help or hur sock prices. Currency depreciaion would help expors, creae more business opporuniies, generae more earnings and raise sock prices. On he oher hand, currency depreciaion would raise impor coss, increase domesic prices, cause higher inflaion, reduce inernaional capial inflows, and hur sock prices. 4

5 Mahemaically, hese respecive effecs can be expressed as: MI EXC MI EP EP EXC MI IC IC EXC MI P P EXC MI CF CF EXC or 0, (2) where EP, IC, P, and CF denoe expors, impor coss, he price level, and he inernaional capial inflow. Excep for he inflaion rae wih negaive values, oher variables are measured in he logarihmic scale so ha esimaed coefficiens are elasiciies. The regression o be esimaed can be wrien as: Log MI log RO log IR log MS loggy log EXC INF log SP (3) where ' s are he parameers o be esimaed, is he error erm of he regression, and is ime. 3. Daa and Empirical Resuls Daa was colleced from he Ocober 2011 issue of he Inernaional Financial Saisics, which is published by he Inernaional Moneary Fund. MI is represened by he share price index wih 2005 as he base year. RO or real GDP is measured in millions of pesos using 2005 as he base year. IR is represened by he money marke rae, which is a policy rae ha can be influenced by he cenral bank. MS is measured as he raio of he M2 money supply o nominal GDP. GY is defined as governmen spending as a percen of GDP. EXC is measured as unis of he peso one U.S. dollar can exchange. An increase in EXC means a deprecaion of he Argenine peso, and a decrease in EXC indicaes an appreciaion of he Argenine peso. INF is represened by he inflaion rae derived from he percen change in he consumer price index. The S&P 500 index in he U.S. is chosen o represen SP *. The quarerly sample ranges from 1998.Q Q2 wih a oal of 54 observaions. The daa for M2 before 1998.Q1 are no available. Monhly daa are no used mainly because he daa for real GDP are available on a quarerly basis. According o he uni roo es, all he variables have uni roos in he level form and are saionary in firs difference. According o he Augmened Dickey-Fuller (ADF) es on he regression residuals wih a lag lengh of 4, he es saisic is esimaed o be , which is greaer han he criical value of in absolue values a he 1% level. This resul suggess ha variables in equaion (1) are coinegraed and have a long-erm sable relaionship. Table 1 presens esimaed parameers and oher relaed saisics. The exponenial GARCH model is applied in empirical esimaion in order o properly esimae he error variance and capure he behavior of he sock marke. 5

6 Table 1. Esimaed regression of he sock marke index for Argenina Business and Economic Research Variable Coefficien z-saisic Consan Real GDP 1.958* Money marke rae * M2/GDP raio 0.640* Governmen spending/gdp raio * Peso/USD exchange rae 0.521* Inflaion rae * U.S. sock marke index 1.168* Adjus R-squared Akaike informaion crierion Schwarz crierion Esimaion mehod EGARCH Sample period 1998.Q Q2 Sample size 54 Noe: * indicaes ha he coefficien is significan a he 1% level. EGARCH is he exponenial GARCH model. Because he inflaion rae has negaive values, he logarihmic scale is no used. As shown, he adjused R-squared is esimaed o be 0.952, suggesing ha 95.2% of he variaion in he Argenine sock marke index can be explained by he seven righ-hand side variables. All he coefficiens are significan a he 1% level. The Argenine sock marke index has a posiive relaionship wih real GDP, he M2/GDP raio, he peso/usd exchange rae and he U.S. sock marke index and a negaive relaionship wih he money marke rae, he raio of governmen spending o GDP and he inflaion rae. Specifically, a 1% increase in real GDP, he M2/GDP raio, he peso/usd exchange rae and he U.S. sock marke index leads o an increase in he Argenine sock marke index by 1.958%, 0.640%, 0.521% and 1.168%, respecively. A 1% increase in he money marke rae and he raio of governmen spending o GDP reduces he sock marke index by 0.092% and 0.746%, respecively. The negaive significan impac of a higher inflaion rae on he Argenine sock marke index is relaively small. According o he esimaed coefficiens or elasiciies in absolue values, he sensiiviy of he Argenine sock marke index o hese explanaory variables in a descending order is real GDP, he U.S. sock marke index, he raio of governmen spending o GDP, he raio of M2 o GDP, he peso/usd exchange rae and he money marke rae. A srong economy as measured by real GDP growh has he highes percenage influence on he Argenine sock marke index. As he U.S. sock marke changes, he Argenine sock marke would respond and change more proporionally. 6

7 Aemps were made o use differen measuremens or variables o deermine wheher he resuls may vary. Due o lack of he daa for he lending rae in 2002.Q2, i canno be included in empirical work. When he raio of M1 o GDP replaces he raio of M2 o GDP, is coefficien is esimaed o be and significan a he 1% level. However, he coefficien of he inflaion rae is negaive bu insignifican. Because M2 is a broader measure of he money supply han M1, he choice of he raio of M2 o GDP is appropriae. Because he daa for he governmen defici are unavailable afer 2005.Q2, he raio of he governmen defici o GDP canno be used o subsiue for he raio of governmen spending o GDP. The U.S. T-bill rae was included and esed. Is negaive coefficien is found insignifican a he 10% level parly because he Argenine money marke rae may conain he informaion. 4. Summary and Conclusions This paper has examined he macroeconomic facors ha are expeced o influence he Argenine sock marke index. These variables include real GDP, money policy, fiscal policy, he exchange rae, he world sock marke as represened by he U.S. sock marke index, and he inflaion rae. The exponenial GARCH model was employed in order o properly esimae he parameers in he variance equaion and capure sock marke movemens. A higher real GDP, a lower money marke rae, a higher raio of M2 o GDP, a lower raio of governmen spending o GDP, a higher exchange rae or depreciaion of he peso agains he U.S. dollar, a lower inflaion rae or a srong U.S. sock marke would increase he Argenine sock marke index. There are several policy implicaions. Auhoriies need o pursue fiscal discipline by keeping he governmen defici or deb as a percen of GDP relaively low in order no o crowd ou privae spending and harm sock prices. The posiive significan coefficien of he peso/usd exchange rae suggess ha depreciaion of he Argenine peso has a posiive impac on sock prices. The exchange rae policy of gradual depreciaion of he peso agains he U.S. dollar seems o be consisen wih empirical findings. According o The Economis, he growh rae of real GDP, he inflaion rae, and he prime lending rae for Argenina in 2011 are projeced o be 6.3%, 10.6% and 12.5%, respecively. These saisics sugges ha a relaively srong economy will be conducive o he sock marke whereas auhoriies need o reduce he relaively high inflaion rae and prime lending rae in order o increase real wealh, he buying power and he value of he peso, encourage household consumpion and business invesmen spending, and increase business opporuniies and sock prices. References Abdullah, D. A. & Hayworh, S. C. (1993). Macroeconomerics of sock price flucuaions. Quarerly Journal of Business and Economics, 32, Abugri, B. A. (2008). Empirical relaionship beween macroeconomic volailiy and sock reurns: Evidence from Lain American markes. Inernaional Review of Financial Analysis, 17, hp://dx.doi.org/ /j.irfa Adrangi, B., Charah, A., Pamplin, R. B., & Sanvicene, A. Z. (2002). Inflaion, oupu, and sock prices: Evidence from Brazil. Journal of Applied Business Research, 18,

8 Araújo, E. (2009). Macroeconomic shocks and he co-movemen of sock reurns in Lain America. Emerging Markes Review, 10, hp://dx.doi.org/ /j.ememar Beer, F., Rafiq, R. B., & Robbani, M. G. (2008). Sock marke and exchange rae dynamics: An empirical invesigaion on Brazil and US. Journal of Inernaional Business & Economics, 8, Bulmash, S. B. & Trivoli, G. W. (1991). Time-lagged ineracions beween sock prices and seleced economic variables. The Journal of Porfolio Managemen, 17, hp://dx.doi.org/ /jpm Chaudhuri, K. & Smiles, S. (2004). Sock marke and aggregae economic aciviy: Evidence from Ausralia. Applied Financial Economics, 14, hp://dx.doi.org/ / Chen, N.-F., Roll, R., & Ross, S. A. (1986). Economic forces and he sock marke. Journal of Business, 59, hp://dx.doi.org/ / Chen, G.-M., Firh, M., & Rui, O. M. (2002). Sock marke linkages: Evidence from Lain America, Journal of Banking & Finance, 26, hp://dx.doi.org/ /s (01) Darra, A. F. (1990a). Sock reurns, money and fiscal deficis. Journal of Financial and Quaniaive Analysis, 25, hp://dx.doi.org/ / Darra, A. F. (1990b). The impac of federal deb upon sock prices in he Unied Saes. Journal of Pos Keynesian Economics, 12, Dhakal, D., Kandil, M., & Sharma, S. C. (1993). Causaliy beween he money supply and share prices: A VAR invesigaion. Quarerly Journal of Business & Economics, 32, Diamandis, P. F. (2009). Inernaional sock marke linkages: Evidence from Lain America. Global Finance Journal, 20, hp://dx.doi.org/ /j.gfj Fama, E. F. (1981). Sock reurns, real aciviy, inflaion and money. American Economic Review, 71, Fama, E. F. (1990). Sock reurns, expeced reurns, and real aciviy. Journal of Finance, 45, hp://dx.doi.org/ / Fama, E. F. & French, K. R. (1989). Business condiions and expeced reurns on socks and bonds. Journal of Financial Economics, 25, hp://dx.doi.org/ / x(89) Fernández-Serrano, J. L. & Sosvilla-Rivero1, S. (2003). Modelling he linkages beween US and Lain American sock markes. Applied Economics, 35, hp://dx.doi.org/ / Geske, R. & Roll, R. (1983). The fiscal and moneary linkages beween sock reurns and inflaion. Journal of Finance, 38, hp://dx.doi.org/ / Hasan, T., Kadapakkam, P.-R., & Ma, Y. (2003). Tess of random walk for Lain American sock markes: Addiional evidence. Lain American Business Review, 4, hp://dx.doi.org/ /j140v04n02_03 8

9 Humpe, A. & Macmillan, P. (2009). Can macroeconomic variables explain long-erm sock marke movemens? A comparison of he US and Japan. Applied Financial Economics, 19, hp://dx.doi.org/ / Jawadi, F., Arouri, M. E. H., & Nguyen, D. K. (2010). Sock marke inegraion in Mexico and Argenina: are shor- and long-erm consideraions differen? Applied Economics Leers, 17, hp://dx.doi.org/ / Kerr, R. B., L. Perera, C. J., Kimura, H., & Lima, F. G. (2009). Evidence of sock marke realignmen in Brazil, France and he Unied Saes. Journal of Inernaional Finance & Economics, 9, Kim, K.-h. (2003). Dollar exchange rae and sock price: Evidence from mulivariae coinegraion and error correcion model. Review of Financial Economics, 12, hp://dx.doi.org/ /s (03) Kerr, R. B., Perera, L. C. J., Kimura, H., & Lima, F. G. (2009). Evidence of sock marke realignmen In Brazil, France and he Unied Saes. Journal of Inernaional Finance & Economics, 9, Mukherjee, T. K. & Naka, A. (1995). Dynamic relaions beween macroeconomic variables and he japanese sock marke: an applicaion of a vecor error correcion model. The Journal of Financial Research, 18, Nelson, D. B. (1991). Condiional heeroskedasiciy in asse reurns: A new approach. Economerica, 59, hp://dx.doi.org/ / Nieh, C.-C. & Lee, C.-f. (2001). Dynamic relaionship beween sock prices and exchange raes for G-7 counries. Quarerly Review of Economics and Finance, 41, hp://dx.doi.org/ /s (01) Ozun, A. (2007). Are he reacions of emerging equiy markes o he volailiy in advanced markes similar? comparaive Evidence from Brazil and Turkey. Inernaional Research Journal of Finance & Economics, 9, Raanapakorn, O. & Sharma, S. C. (2007). Dynamic analysis beween he US sock reurns and he macroeconomic variables. Applied Financial Economics, 17, hp://dx.doi.org/ / Rivas, A., Verma, R., Rodriguez, A., & Verma, P. (2008). Inernaional ransmission mechanism of sock marke volailiies. Lain American Business Review, 9, hp://dx.doi.org/ / Robiaille, P. & Roush, J. (2006). How do FOMC acions and U.S. macroeconomic daa announcemens move Brazilian sovereign yield spreads and sock prices? Working Papers, US Federal Reserve Board's Inernaional Finance Discussion Papers, 1-52, Slaer, J., Lyons, J., Bara, P., & Davis, A. (2007). Conagion effec: Markes fear U.S. woes will hi global growh; overseas socks dive, commodiies down; meals venures a risk. Wall Sree Journal - Easern Ediion, Augus 17, A1. Spyrou, S. I. (2004). Are socks a good hedge agains inflaion? Evidence from emerging markes. Applied Economics, 36, hp://dx.doi.org/ / Tabak, B. M. (2006). The dynamic relaionship beween sock prices and exchange raes: 9

10 Evidence for Brazil. Inernaional Journal of Theoreical & Applied Finance, 9, hp://dx.doi.org/ /s Wang, G. & Lim, C. (2010). Effecs of macroeconomic facors on share prices. Journal of Inernaional Finance and Economics, 10, Copyrigh Disclaimer Copyrigh reserved by he auhor(s). This aricle is an open-access aricle disribued under he erms and condiions of he Creaive Commons Aribuion license (hp://creaivecommons.org/licenses/by/3.0/). 10