TESTING THE INTERNATIONAL TRANSMISSION OF INTEREST RATE SHOCKS USING VECTOR AUTO-REGRESSION: EVIDENCE FROM CAPE VERDE

Transcription

1 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 ISSN: Websie: TESTING THE INTERNATIONAL TRANSMISSION OF INTEREST RATE SHOCKS USING VECTOR AUTO-REGRESSION: EVIDENCE FROM CAPE VERDE Dr. Chibuike R. Oguanobi Deparmen of Economics, Chukwuemeka Odumegwu Ojukwu Universiy, Nigeria Absrac This paper examined if ineres rae shocks from he Unied Saes (U.S.) are ransmied o Cape Verde. I colleced daa on four Cape Verdean variables (real gross domesic produc, consumer price index (CPI), exchange rae and ineres rae) and wo foreign variables (U.S Federal fund rae and he world CPI) for he period The impulse response analysis of he VAR model shows ha he counry s variables respond insignificanly o shocks from foreign variables. I was herefore concluded ha macroeconomic shocks in Cape Verde are basically homemade. Key words: Ineres rae, shocks, Cape Verde, U.S.A., VAR.. Inroducion The global ransmission of moneary policy shocks and is effecs on macroeconomic environmen of counries has been a long-sanding debae in he field of inernaional finance. Over he years, one of he major problems facing moneary economiss across he globe, especially he developing counries has been he macroeconomic implicaion of ineres rae changes. Ineres rae is a major macroeconomic variable whose changes conribue immensely owards deermining he posiion of every economy. Hence, i is radiionally believed ha how a child behaves is always a funcion of how his/her paren(s) or elders behave. Consider a Cape Verdean residen enrepreneur who has enough money o inves in he home counry. On he verge of making his invesmen, he Unied Saes Federal Reserve announces an increase in ineres rae. Convenionally, he posiive slope of he LM curve would require a corresponding increase in oupu of he Unied Saes. Why? Because globalisaion and is accompanying openness of naional borders is expeced o allow ino he U.S., free inflow of invesmen capial, due o high ineres rae. Consequenly, he prospecive Cape Verdean invesor may auomaically be expeced o move his invesmen fund o he Unied Saes. The immediae effec of his financial cross over is an increase in he oupu of he Unied Saes and a decrease in Cape Verde s oupu. The decrease in Cape Verde s oupu would immediaely ake he counry s supply below is demand hereby causing demand-pull inflaion. However, afer several macroeconomic analysis and evenual confirmaion of he causes of he inflaion, he Cape Verde s moneary auhoriies in an aemp o curb inflaion via oupu increase, may be expeced o co-move wih he Unied Saes 70

2 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 by also increasing is ineres rae. When his is done, foreign invesors are araced ino Cape Verde. This would also increase he counry s oupu. In 2008, Cape Verde had her ineres rae increased from 3.3% o 3.8%. During his period, he ineres rae of he Unied Saes reduced from 5.02 o.92 percen. In 2009, during which he Unied Saes ineres rae decreased furher o 9%, Cape Verde s ineres rae decreased o 2.9%. Beween 2008 and 200, Cape Verde experienced grea changes in he growh rae of her GDP (from 6.2% in 2008 o 3.7% in 2009 and hen 5.4% in 200 before falling o 5.0% in 20). Cape Verde s inflaion rae decreased from 6.8% in 2008 o.0% in 2009, increased o 2.% in 200 and furher o 4.5% in 20. Consequen o he analysis above, if we should follow he argumen of Edwards (200), ha ineres rae hikes by he Federal Reserve conribued significanly o some of he mos imporan currency crisis in recen imes, such as he Mexican Tequila crisis of and he Argenine Peso crisis of , i may no be unwise if we suspec ha he same ineres rae changes in he Unied Saes have conribued o major changes (flucuaions) in Cape Verde s macroeconomic indicaors, including ineres raes, prices and oupu. Research wise, he macroeconomic implicaion of such claims made by Edwards (200) have led o some empirical sudies aimed a undersanding he way in which possible reforms in he inernaional financial archiecure affec differen counries. Mos of hese sudies were carried ou using daa for he developed and developing counries of Europe, Asia and America. Currenly here is no such sudy compleed for he economy Cape Verde, despie huge macroeconomic volailiy in he counry. Hence, he problem a hand is wheher or no he ineres rae changes in he advanced counries are ransmied o he underdeveloped counries, in his paper proxied by Cape Verde. In responding o he above saed problem, his paper seeks o: ascerain if Cape Verde s ineres raes respond o ineres rae shocks in he Unied Saes? find ou if ineres rae shocks in he Unied Saes affec oher basic macroeconomic variables in Cape Verde. 2. Review of relaed lieraure Inernaional ransmission effecs of moneary policy shocks have been esed in many empirical works wih mixed resuls and suggesions. The small-open economy model of Mundell- Fleming-Dornbusch has coninued o be of influence in academic and poliical circles ill he mid 990s. Theoreically, Obsfeld and Rogoff (995) assume ha he purchasing power pariy (PPP) always holds. As a resul, he real ineres rae pariy holds beween wo counries. They predic ha he domesic moneary shocks raise he level of domesic oupu bu show an ambiguous effec on foreign oupu. A large empirical lieraure has invesigaed he inernaional ransmission of moneary and non-moneary shocks using small-scale srucural Vecor Auo-Regression (VAR) models. Clarida and Gali (994) idenified sources of real exchange rae flucuaions for pos- Breon Woods period for U.S., Japan, Germany, and Canada. The esimaion of srucural VAR in heir sudy produced consisen resuls wih he predicions of he Mundell-Fleming model showing ha demand shocks lead o appreciaion and moneary shocks leading o depreciaion of he home currency. Eichenbaum and Evans (995) also found he resuls similar o Clarida and Gali (994). Kim (200), on he oher hand, esimaed srucural VAR o idenify unidirecional effec of he US moneary policy shocks o he macroeconomic variables of G-7 counries and found ha he U.S. moneary expansion has a posiive spillover effecs on real GDP and indusrial producion of non-u.s. G-6 counries. Oher sudies ha found evidence of cross border 7

3 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 ransmission of moneary policies include Läufer K. and Sundararajan S. (994), Peek and Rosengren (996), Selover (998), Canover, Jensen, and Johnson (999), Forbes and Rigobon (999), Kollman (999), Frankel, Schmukler, and Serven (2000), Reinhar and Reinhar (200), Reside (2004), Campa and Goldberg (2005) as well as Boivin and Giannoni (2006) among ohers. Sudies ha are eiher no in suppor or no in full suppor of inernaional moneary ransmission include Schmi-Grohe (997), Geirega (2004) and Sahminan (2005). Evidence from previous sudies reviewed has shown mixed resuls. For a more accepable and consensus resul and conclusion, furher sudies need o be carried ou for differen economies wih differen economic condiions. However, since mos of he exising sudies in his area of inernaional Moneary Economics were done using daa for he advanced and developing counries of Europe, Asia and America, his sudy is an aemp a doing he same using daa for he economy Cape Verde. 3. Mehodological issues 3.: The model: I is globally argued ha in he era of globalizaion, moneary policies implemened in one counry (numeraire counry) can affec some oher counries, eiher posiively or negaively. Based on his argumen, Cenral Banks are always ineresed in deermining he exen o which heir counries domesic ineres raes diverge from world ineres raes. Hence, he vial variable in his sudy is he differenial beween he Cape Verde s ineres rae and U.S. ineres rae properly adjused by counry risk and currency risk. Specifically, ineres rae differenial is defined as: e r r...() where r is he domesic currency nominal ineres rae for securiies of a cerain mauriy. r is he inernaional, nominal ineres rae on foreign currency denominaed securiies of he same mauriy. is he expeced rae of depreciaion of he domesic currency, is a measure of counry risk. However, he seady sae equilibrium wih perfec capial mobiliy requires ha he ineres rae differenial should be approximaely zero. In his case, he speed a which convergence o long erm equilibrium akes place is purely a funcion of specific counries condiions, bu under free capial mobiliy, should raher be very fas. Hence, in seady sae equilibrium, e e...(2) Where e is he domesic ineres rae. The possibiliy of finding an equaion (2) ha is differen from zero can only be deermined empirically. I is only a funcion of several variables such as, degree of capial mobiliy and oher forms of marke fricions and ransacion coss. Wih full capial mobiliy and in he absence of ransacion coss of any form, we would expec ha e would be approximaely zero. However, he mos imporan issue I seek o address in his sudy is wheher e is acually a funcion of inernaional ineres rae, proxied by he Unied Saes of America (numeraire counry) Federal Fund Rae. Hence, 72

4 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 e i FFR 0 i i j m K where FFR is he level of U.S. Federal Fund Ineres Raes. i 73...(3) e is Cape Verde s policy ineres raes a ime. (i represens Cape Verde) 0, and i are coefficiens, K i is a vecor of oher possible deerminans of i is sill an error erm wih is usual characerisics. However, following Mojon and Peersman (200), Frankel, e al, (2004), ec, I have decided e ( o represen he se of oher deerminans of i K i ) by world consumer price index. The use of price is imporan because i is relaed o he concep of moneary independence. Moreover, i is believed ha much of he variaions in nominal ineres raes could jus reflec variaions in inflaion differenials (Frankel, e al: 2004). I also helps o solve he so-called price puzzle: empirical finding in he Vecor Auoregression (VAR) mehodology lieraure ha prices rise following an ineres rae ighening (Mojon and Peersman, 200). As a resul, equaion (3) could be ransformed ino: e FFR WCPI...(4) Where WCPI i e i i i represens he world consumer price index i is sill an error erm wih zero mean and independenly disribued across counries a differen imes,., and are coefficiens. However, o examine if ineres rae shocks in he Unied Saes affecs oher macroeconomic variables in Cape Verde, I inroduce he Srucural Vecor Auoregressive (SVAR) mehodology used by Weber, eal (20) o idenify he moneary policy shocks in he Euro area. The Baseline VAR Model Vecor Auoregressive (VAR) models are widely used in he empirical analysis of moneary policy ransmission. This mehodology has undoubedly he meri of avoiding he need for a complee specificaion of a srucural model of he economy (Bagliano and Favero, 997). Evidence from lieraure shows he mehodology o be appropriae in measuring he impac of moneary policy shocks on macroeconomic performance of counries. In his sudy, my ineres is in asceraining if ineres rae shocks in he Unied Saes (U.S.A) are ransmied o Cape Verde. As such, I proceed on a bilaeral basis wih he USA on one side and Cape Verde on he oher side. In his seup, any correlaion beween USA and Cape Verde is likely o be unidirecional. Consequen o his, he specificaion and he esimaion of he saisical model could be significanly simplified (Canova, 2005) In order o reflec he above seup in our model, I assume a block-exogenous VAR model in which he foreign variables (FFR and WCPI) are reaed as exogenous o Cape Verdean variables. In oher words, hese exogenous foreign variables influence Cape Verdean variables bu here is no feedback from Cape Verdean variables o hese foreign variables. I herefore make allowance for a conemporaneous impac of hese exogenous variables on he endogenous Cape Verdean variables. Moreover, he consrucion of a block-exogenous VAR model saves he degree of freedom. Therefore, using he conen of equaion (4) in specifying a SVAR model and inroducing o he lef-hand side oher Cape Verdean macroeconomic variables, yields equaion

5 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 (5). In (5), Z FR is a vecor of foreign variables (FFR and WCPI) while Z CV is a vecor of endogenous Cape Verdean variables (we shall define hese variables laer). where Z Z i FR Z Z and, FR CV FR is an m m marix, A A A( L) A 2 L 2 is an m ( L) ( L) 2 m A 22 are m vecors, Z 0, ( L) CV and marix, and 74 FR i......(5) CV CV are m2 vecors, A L A L is an m m marix. L is a lag operaor. However, due o parameer idenificaion problems associaed wih he SVAR models, Ordinary Leas Squares (OLS) esimaion of he models yields inconsisen parameer esimaes. This problem can be avered by re-wriing he SVAR model (5) in a reduced form hus: RGDPi CPRi WCPI Zi 2 ( L). Zi j i...(6) HEXR i FFR HINTi where HINT replaces e as Cape Verde s ineres rae. Z is a vecor of endogenous variables and i represen Cape Verde. Here, he baseline model consiss of four endogenous and wo exogenous variables for he home counry, Cape Verde. The endogenous variables are he counry s key macroeconomic indicaors of real gross domesic produc (RGDP); consumer prices (CPI); exchange rae (HEXR) and ineres rae (HINT). These endogenous variables depend on heir own lags and a consan erm. The exogenous variables included in he model are he already defined WCPI and FFR which as indicaed earlier, are expeced o have conemporaneous impac on he endogenous variables. The objecives of his sudy would be achieved by esimaing he VAR equaion (6) and analysing he Impulse Response Funcions (IRF). Tradiionally, he IRF have been widely used as a means of analysing an esimaed VAR model (Hamilon, 994). Here, he IRF is expeced o expose he degree o which domesic ineres raes responds o shocks in foreign ineres rae as well as how oher domesic macroeconomic variables respond o foreign ineres rae. 3.2: Daa The all iems annual averages of Consumer Price Indices (CPI) are used as proxies for measuring he price levels of Cape Verde's oupu. The Unied Saes is considered as he foreign counry. All CPI daa were obained from he Inernaional Moneary Fund (IMF) World Economic Oulook daabase. The ineres rae daa of he USA were obained direc from he U.S. Bureau of Labour Saisics (BLS), while all oher daase were obained from he World Bank. All series are annual and span he period, The choice of he period is based on daa availabiliy. 4. EMPIRICAL RESULT. Our resuls reflec he Cape Verde s response o an ineres rae shock in he Unied Saes which is idenified as a one-sandard-error increase in he Federal Fund Rae (FFR). The resul furher shows he response of his counry s variables o shocks in he global consumer prices. Table : Response of Cape Verdean variables o foreign shocks RGDP CPI HEXR HINT FFR

6 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 WCPI Noe: Two sided arrows denoe insignifican responses of he endogenous variables o shocks in he exogenous variables. There were however responses (someimes posiive and oher imes negaive bu hey are insignifican). Source: Generaed by he auhor from he IRF esimaes. Generally, an ineres rae policy shock in he Unied Saes as well as a shock in he world consumer prices leads o insignifican flucuaions (increases and decreases) in Cape Verde s macroeconomic variables. In more deail, an ineres rae shock in he Unied Saes led immediaely o sligh increases in he Cape Verde s real gross domesic produc and consumer prices wihin he firs hree years of our sample space. Afer his period, hese wo variables decreased in values unil hey responded negaively (see figures A and C of he appendix). The exchange rae of he counry responded negaively unil he 7 h year when i responded slighly posiively before going down negaively once again (see figure E of he same appendix). The ineres rae in Cape Verde iniially did no respond o his shock. I responded afer hree years (hough negaively). I coninued on his negaive bu insignifican rend unil he end of he period reviewed. The Cape Verdean oupu had a sable insignifican negaive response o he shock in global consumer prices. This lased for abou six years of he sudy period. Afer his period, he response became posiive for he nex nine years. This posiive response may be aribued o he measures aken o figh inflaion which evenually resuled in oupu growh. Wihin he firs hree years of he sudy period, figure D of he appendix shows ha consumer prices in his counry did no respond o global price shocks. Beyond his period, he response flucuaed beween posiive and negaive. Figure F of he same appendix shows he response of exchange rae o global price shock. As observed, he response was negaive hough, insignifican almos hroughou he sample period. I was afer he weny-fourh year ha he variable responded posiively. This however jusifies he basic economic heory on he relaionship beween inflaion and he exchange rae. Finally, ineres rae in Cape Verde iniially had an insignifican posiive response o shocks in global consumer prices. This lased for abou hree and half years. This is shown in figure H of he same appendix. Afer abou fory-wo monhs, ineres rae sared responding negaively hough insignificanly unil nearly he end of he sample period. In a nushell, he macroeconomic variables in Cape Verde responded insignificanly o foreign shocks. This is eviden from he impulse response analysis conduced. By implicaion, he flucuaions in macroeconomic variables in he counry may be a funcion of inernal shocks arising from he implemenaion of wrong policies (boh fiscal and moneary). This is however subjec o invesigaion via furher research. 5. Conclusion and policy recommendaions Our sudy invesigaed he ransmission of Unied Saes ineres rae shocks o Cape Verde using a VAR approach. We found ha he macroeconomic variables of he Cape Verdean economy responded only insignificanly o he foreign variables shocks. Overall, our findings sugges ha he Cape Verdean economy is no highly responsive o foreign shocks. However, his may imply ha he huge macroeconomic volailiy experienced by he counry is home-made. Only an insignifican proporion of he domesic shocks can be linked o exernal shocks Based on he findings of his sudy and he conclusion ha followed, we hereby recommend ha moneary auhoriies in Cape Verde should coninue wih independen moneary policy 75

7 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 making, especially as i concern ineres rae conrol. By implicaion, hey should ignore o a large exen wha happens exernally. They are expeced o base heir policy making on domesic shocks, as considering wha happens in oher counries migh confuse policy makers in heir decision making process. This is sequel o he fac ha domesic variables responds insignificanly o foreign shocks. This recommendaion is indeed in line wih he view ha he cenral issue for policy making in developing counries is how o sabilize he economy in response o foreign shocks provided ha hose foreign shocks have impacs on he developing counries. In Cape Verde, hey do have impacs bu insignificanly. References Bagliano F. and Favero C. (997) Measuring Moneary Policy wih VAR Models: An Evaluaion, IGIER Working Paper 32. Biovini J. and Giannoni M. (2006) Has Moneary Policy Become More Effecive? Review of Economic and Saisics, Vol. 88, Campa, J. and Goldberg, L. (2005), Exchange Rae Pass-Through ino Impor Prices, The Review of Economics and Saisics, Vol. 87, No. 4, December, Pages Canova, F (2005), The Transmission of US Shocks o Lain America, Journal of Applied Economerics, Vol. 20, No. 2, Pages Clarida R. and Gali J. (994), Sources of Real Exchange-Rae Flucuaions: How Imporan are Nominal Shocks? Carnegie-Rocheser Conference Series on Public Policy 4, 56. Conover M., Jensen G. and Johnson R. (999), Moneary Environmens and Inernaional Sock Reurns, The Journal of Banking and Finance, Vol. 23, Number 9, Pp , Sepember. Edwards S. (200), The Inernaional Transmission of Ineres Rae Shocks: The Federal Reserve and Emerging Markes in Lain America and Asia, Journal of Inernaional Money and Finance, Vol. 29, Issue 4, Pp Eichenbaum M. and Evans C. (995), Some Empirical Evidence on he Effecs of Shocks o Moneary Policy on Exchange Raes, The Quarerly Journal of Economics, Vol. 0, No. 4, November, Pages 975,009. Ezeibe O. (2008), An Empirical Tes of Uncovered Ineres Pariy Theory: Evidence from Nigeria, Unpublished M.Sc. Thesis. Forbes K. and Rigobon R. (999), No Conagion, Only Inerdependence: Measuring Sock Marke Co-Movemens, NBER Working Paper Series, No. 7267, July. Frankel J., Schmukler, S. and Serven N. (2000), Global Transmission of Ineres Raes: Moneary Independence and Currency Regime, World Bank Working Paper No and NBER. Working Paper No Geirega C. (2004), Transmission of Exernal and Inernal Shocks in Argenina during he Converibiliy Period: Some Empirical Findings from VARs, Williams College, Deparmen of Economics Working Paper Number Hamilon J. (994), Time Series Analysis, Princeon, Princeon Universiy Press. Kim S. (200), Inernaional Transmission of U.S. Moneary Policy Shocks: Evidence from VARs, Journal of Moneary Economics 48,

8 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 Kollman R. (999), Explaining Inernaional Co-Movemens of Oupu, and Asse Reurns: The Role of Money and Nominal Rigidiies, The IMF Working Paper Series, WP/99/84, June. Läufer N. and Sundararajan S. (994), The Inernaional Transmission of Economic Shocks in a Three-Counry World Under Mixed Exchange Raes, Journal of Inernaional Money and Finance, Vol. 3, Issue 4, Pp Mojon B. and Peersman G. (200), A VAR Descripion of he Effecs of Moneary Policy in he Individual counries of he Euro Area, ECB Working Paper Series Obsfeld M. and Rogoff K. (995), Exchange Rae Dynamics Redux, Journal of Poliical Economy, Vol. 03, Pp Peek J. and Rosengren E. (996), The Inernaional Transmission of Financial Shocks: The Case of Japan, Federal Reserve Bank of Boson Working Paper Number 96-, January. Reinhar C. and Reinhar V. (200), Wha Hurs Mos? G-3 Exchange Rae or Ineres Rae Volailiy, MPRA Paper No Reside R. (2004), Inernaional Transmission of US Moneary Shocks o Asia, UPSE Discussion Paper Number 0406, June. Sahminan U. (2005), Ineres Raes and he Role of Exchange Rae Regimes in Major Souheas Asian Counries, Bank of Indonesia Inernaional Economics/Economic Developmen Seminar, March Schmi Grohe S. (997), The Inernaional Transmission of Economic Flucuaions: Effecs of U.S. Business Cycles on he Canadian Economy, Journal of Inernaional Economics, Vol. 44 Pp Selover D. (998), Business Cycle Transmission Beween The Unied Saes and Japan: A Vecor Error Correcion Approach, Japan and he World Economy Vol. 9, Issue 3, Pp ENDNOTE Oguanobi, Akamobi and Agu (204) published a paper iled cross-border ransmission of ineres rae shocks: a VAR analysis of he Nigerian economy. Oguanobi, Akamobi, Ifebi and Maduka (205) also published a work iled does ineres rae shocks ransmi from Unied Saes o Ghana? Evidence from vecor auoregression. The only difference beween hese papers and his curren paper is he counry of sudy. As a resul, his curren paper only changed he counry of sudy, he empirical daa used and he resuling findings. All oher secions of his paper remain as used in hese previous papers. I herefore hank he auhors (Oguanobi Chibuike, Akamobi Anhony, Agu Cleus, Ifebi Ogonna and Maduka Anne) for wriing hose papers and allowing me o copy pars of hem. 77

9 Journal of Policy and Developmen Sudies Vol. 0, No., February 206 APPENDIX: IMPULSE RESPONSE FUNCTIONS RESPONSE TO FFR RESPONSE TO WCPI 03 FIG. A: RGDP 03 FIG. B: RGDP FIG. C: CPI FIG. D: CPI FIG. E: HEXR FIG. F: HEXR FIG. G: HINT FIG. H: HINT