The Effects of usa Monetary Policy

Transcription

1 The Effecs of usa Moneary Policy on Cenral America and he Dominican Republic Absrac Ariadne M. Checo Salomé Pradel Francisco A. Ramírez This paper esimaes he impac of us moneary policy shocks on Cenral America and he Dominican Republic economies, using a facor augmened var model. A sign resricion approach is implemened for he idenificaion of such shocks. Our resuls indicae ha us moneary policy shocks affec hese economies mosly hrough is effecs on he real side of he economy due o is impac on exernal demand and he reduced role of he exchange rae as a shock absorber, where counries wih less flexible exchange rae regimes are more affeced. Likewise, he flow of remiances is also negaively influenced, revealing anoher channel hrough which foreign moneary shocks impac he Cenral American and he Dominican Republic economies. On he financial side, domesic ineres raes will rise and ne inernaional reserves will fall as cenral banks limi volailiy in exchange raes. Deparmen of Moneary Programming and Economic Sudies, Banco Cenral de la República Dominicana. Auhors hank he paricipans in he workshop organized by cemla in Mexico Ciy in Sepember 2015 for heir commens, paricularly hose of Claudia Ramírez, Jorge Fornero, Kolver Hernández, Albero Oriz, and Angel Esrada. They also hank paricipans in he IX Research Forum of he Consejo Moneario Cenroamericano in July 2015 and he XX Annual Meeing of he Cenral Bank Researchers Nework in November Opinions expressed in his paper do no reflec he poin of view of he Banco Cenral de la República Dominicana, he Banco de España or he Eurosysem. <a.checo@bancenral.gov.do>, <s.pradel@bancenral.gov.do>, <f.ramirez@bancenral.gov.do>. 189

2 Keywords: ransmission of moneary policy, normalizaion, ineres raes, Cenral America. jel classificaion: E50, E58, C55 1. INTRODUCTION A year afer he end of is unconvenional moneary policy sraegy, he Federal Reserve decided o increase he federal funds rae (ffr), even ha pus an end o seven years of policy ineres raes a he zero lower bound. This phenomenon, known as moneary policy normalizaion, has been a source of concern for policymakers of boh advanced and emerging economies, given ha a seep pah in ineres raes could increase financial marke volailiy. This decision reopens he quesion of how usa moneary policy shocks spillover o he res of he world, in paricular in he conex of hisorically low ineres rae levels. Of paricular ineres is he quesion of how his ype of shocks affecs economies wih a low degree of financial linkages wih inernaional capial marke flows, such as Cenral American and Caribbean economies. The main objecive of his paper is o quanify he effecs of foreign ineres rae shocks, measured hrough he usa ffr (a convenional moneary policy insrumen), on he economies of Cenral America and he Dominican Republic (hereafer cadr). This is a relevan subjec for policy makers in hese economies because of he imporan commercial linkage of cadr counries wih he usa economy, despie he low degree of financial developmen and linkages wih inernaional capial marke flows relaive o oher Emerging Marke Economies in Lain America. The empirical sraegy employed o sudy his phenomenon inends o measure he counry-specific effecs of usa moneary policy shocks. We esimae a facor-augmened vecor auoregressive model (favar) wih a foreign variables block, where he usa is he relevan foreign counry for hese economies. Common facors are exraced from a counry daa se of nearly 80 macroeconomic variables of cadr counries 1 for he period Counries include: Cosa Rica, El Salvador, Guaemala, Honduras, and he Dominican Republic. Nicaragua is excluded from he sample due o lack of daa prior o A. Checo, S. Pradel, F. Ramírez

3 Two empirical issues arise in he quanificaion of he effec of usa moneary policy shocks. One issue is he idenificaion of his ype of shock. The proper idenificaion is criical o undersanding he ransmission mechanism of his ype of shock o hese economies (see Canova and De Nicoló, 2003; Kim, 2001; Canova, 2005). We address his problem using sign resricions o idenify he effecs of a usa MP on he economies under sudy. Anoher issue is he decreasing variabiliy afer 2008 of he ffr as i adjuss o he zero lower bound. While he ffr has remained unchanged for he las seven years, he Federal Reserve has employed nonconvenional insrumens, known as quaniaive easing (qe) programs, which have led o a more expansive moneary policy han wha can be accouned for by he effecive ffr. Therefore, in order o address his issue, we use he shadow federal funds rae (Wu and Xia, 2016) as our measure of he moneary policy insrumen. To dae, his is one of he firs works ha addresses he effecs of usa moneary policy shocks for Cenral America and he Dominican Republic. Oher papers have used he favar mehodology o sudy he inernaional ransmission of moneary policy shocks. Mumaz and Surico (2008) exend he model of Bernanke e al.(2004) o he open economy case, analyzing he ransmission o seveneen indusrial counries. Meanwhile, Cruz-Zuniga (2011) sudies he effecs of a change in he usa moneary policy for he Mexican and Brazilian case. Summarizing he main findings, usa moneary shocks have conracive effecs on hese economies. The evidence suggess an unambiguous fall in real oupu for each of he considered economies, revealing ha foreign ineres shocks work as an imporan driver of he common business cycle in cadr counries. The relaive imporance of exchange rae sabiliy for moneary auhoriies in hese counries minimizes he response of his variable, hence rising ineres raes and falling ne inernaional reserves do mos of he adjusmen. On he real side, expors fall due o he dominance of he income absorpion effec over he expendiure swiching effec, backed by he limied flucuaion in real exchange raes. However, a recovery in rade balance is observed, as impors decrease more han expors, produc of a fall in domesic demand due o he conracionary effecs of moneary ighening. Finally, remiances, which are an imporan source of non-labor income in hese economies, respond negaively since he conracionary moneary shock is a signal of a fuure fall in usa aggregae demand. The Effecs of us Moneary Policy 191

4 The paper is organized as follows: Secion 2 presens he lieraure review; Secion 3 describes he exchange rae arrangemens in hese economies. This is imporan because i is a characerisic feaure of cadr economies ha could influence he empirical responses o foreign moneary shocks. Secion 4 describes he empirical mehodology; Secion 5 compares he resuls for a posiive ineres rae shock o main Cenral American and Dominican indicaors; Secion 6 concludes. 2. LITERATURE REVIEW Lieraure relaed o convenional moneary shocks, measured hrough ineres rae changes, alhough exensive, focuses on normal imes, i.e., periods ha do no include hyperinflaion episodes, currency crises, or massive recessions (Canova, 2005). When sudying moneary shocks and heir inernaional ransmission, wo empirical sraegies can be disinguished: Those based on he esimaion of srucural (dsge) models, which by consrucion suggess expeced pahs for variables under his ype of shocks, and hose which are daa oriened, based on empirical relaions. In heoreical models, inspired by he Mundell-Fleming-Dornbusch (mfd) model and he Obsfeld-Rogoff exension (1996), he ransmission of moneary shocks o oher economies occurs hrough wo main channels: Curren accoun and exchange rae. A ighening shock in he counry of origin is associaed wih a fall in oupu and an appreciaion of he currency of ha counry. However, he impac of ha shock on oher counries is ambiguous, since wo offseing mechanisms work simulaneously, wih no clear evidence of which one would dominae: on one side, he exchange rae in he foreign counry depreciaes, having a posiive effec on economic aciviy (expendiure-swiching effec); meanwhile, he ineres rae hike shrinks domesic oupu in he counry of origin, leading o a fall in he demand for expors of foreign counries (income-absorpion effec; Kawai, 2015). Likewise, ineremporal models also show ambiguous resuls, even afer including fuure expecaions from economic agens as an addiional mechanism (Kim, 2001). Empirical models (see Lasrapes,1992; Eichenbaum and Evans, 1995; Grilli and Roubini, 1995; Kim and Roubini, 2000; Clarida and Galí, 1994) employ sraegies ha minimize resricions, using daa 192 A. Checo, S. Pradel, F. Ramírez

5 o idenify ransmission mechanisms for he exchange rae case. Kim (2001) compares he empirical resuls wih differen heoreical models, finding ha an expansive moneary shock in he usa, measured by a drop in he world ineres rae, has a posiive effec on growh for G6 economies, which maches he resuls suggesed by ineremporal models (see Svensson and van Wijnbergen, 1989; Obsfeld and Rogoff, 1995). Also, he rade link is no significan, which is no consisen wih he beggar-hy-neighbor heory of he mfd basic model. The paper concludes ha he exchange rae response does no depend on wheher he idenifying sraegies are recursive or no, as promped by Kim and Roubini (2000) and Cushman and Zha (1997). Oher findings of Kim (2001) include he exogeneiy of usa o non-usa moneary policy. The inernaional ransmission of moneary shocks o indusrial counries has been recenly addressed by Vespignani (2015). Mumaz and Surico (2008) explore he effecs of a decrease in he inernaional shor erm ineres raes on he Unied Kingdom, finding a posiive impac on gdp, invesmen and consumpion afer a year. On he oher hand, he sudy of Jannsen and Klein (1991) finds ha an increase in a foreign ineres rae (Eurozone, in his case) has a posiive impac on domesic ineres raes for a se of counries ha have no adoped he euro. 2 The increase in he ineres raes ranslaes ino a conracion in gdp hrough a reducion in domesic demand. Meanwhile, expors decline, exposing he imporance of he income-absorpion effec in hese economies. Since boh expors and impors decline, no significan changes are observed in he rade balance. The response of hese variables, as well as he negligible role observed in he exchange rae, is similar o he reacion of counries wih a fixed exchange rae regime, revealing he imporance of exchange rae sabilizaion for hese small open economies. For developing economies, he degree of ransmission of inernaional moneary shocks varies according o he currency regime, macroeconomic fundamenals and counry-specific srucural characerisics (see Borda e al., 2000; Arora and Cerisola, 2001; Mackowiak, 2007; Canova, 2005; Cruz-Zuniga, 2011). These auhors idenify, hrough differen var specificaions, wo key ransmission channels: Trade balance and ineres raes. 2 The se of counries include he Unied Kingdom, Denmark, Sweden, Norway, and Swizerland. The Effecs of us Moneary Policy 193

6 The research of Borda e al. (2000), relaed o he conribuion of usa moneary policy o Caribbean business cycles, concludes ha for counries wih a flexible exchange rae regime, a world ineres rae shock has a negaive effec on oupu due o an increase in he real exchange rae ha augmens he cos of inpus. However, i indicaes ha gdp for Caribbean counries is no mainly driven by he world ineres rae, bu raher by he exchange rae, highlighed as an imporan ransmission mechanism. This resul is consisen wih he conclusions of Mackowiak (2007), where he ypical response of an emerging marke economy o a ighening of he usa moneary policy is exchange rae depreciaion, inflaion and a fall in economic aciviy. 3 Meanwhile, he resuls provided by Canova (2005) sugges ha he ineres rae channel serves as an amplifier of usa moneary changes, conferring he rade channel an insignifican role in he ransmission of moneary shocks from he Unied Saes o Lain America. Since ineres raes remained a he zlb up o December 2015, he sudy of he inernaional ransmission of moneary policy focused on he impac of unconvenional insrumens adoped by indusrial counries afer he 2007 inernaional crisis. This approach has been used by differen auhors, who analyze is spillover effecs o emerging economies. Overall, heir resuls confer a more imporan role o financial linkages and rade channels. Hausman and Wongswan (2006) explore he channels of usa moneary policy ransmission hrough he Federal Open Marke Commiee announcemens, noing ha a counry wih a higher degree of real and financial inegraion wih he usa has a greaer ineres rae response, as well as hose wih less flexible exchange raes. In summary, unlike Ehrmann and Frazscher (2006), hey sugges ha real and financial linkages wih he usa are more imporan han hose wih he res of he world. Likewise, Bauer and Neely (2013) disinguishes he relaive imporance of he signaling and porfolio balance channels o explain he conribuion of unconvenional policy o he reducion of bond yields in mos counries afer he inernaional crisis of Through a dynamic erm srucure model, hey conclude ha boh channels are 3 Counries under analysis are Korea, Thailand, Malaysia, Philippines, Singapore, Hong Kong, Mexico, and Chile. 4 Ausralia, usa, Germany, Canada, and Japan 194 A. Checo, S. Pradel, F. Ramírez

7 imporan. 5 Noneheless, Chen e al. (2014) indicae ha he spillovers o asse prices and capial markes are larger if hey come from signal surprises. They highligh ha even if unconvenional moneary policies have a greaer impac han convenional ones, characerisics such as beer fundamenals and a more liquid marke srucure help o miigae he effecs. Bowman e al. (2014) also demonsraes ha alhough flucuaions of asse prices in emerging markes afer a usa moneary shock are bigger han flucuaions in he counry of origin (usa), weaker fundamenals explain, in par, his overreacion. For he effecs of unconvenional moneary policy o oher counries, see also Craine and Marin (2008). More recenly, he expecaions of an ineres rae hike in he usa promped he sudy of he inernaional impac of such an even. In his conex, research analyzing he spillover effecs on foreign counries of his convenional moneary policy insrumen has resurged. For he Cenral American region, Valle and Morales (2016) employ a recursive idenificaion sraegy (Cholesky) for a foreign ineres rae shock (usa, in his scenario). A var is consruced for each economy, where he usa block of variables is exogenous. Their main resuls include a muliple shock approach (including as well separae growh and remiances shocks), summing an overall posiive effec for he normalizaion of usa moneary policy. Noneheless, as Fornero e al. (2016) indicae, he idenificaion of foreign moneary shocks is no sraighforward in recursive var models. For his reason, hose auhors compare he resuls from a svar model wih sign and zero resricions (szr) and a dsge model for he Chilean economy o sudy he effecs of foreign moneary policy on Chilean oupu and he overall economy. For he szr model, a one percen posiive shock of he foreign ineres provokes a saisically significan decrease in local aciviy and exchange rae depreciaion, while inflaion (alhough wih no significan change) firs increases by he depreciaion and laer on decreases by he weak demand. The impulse responses derived from his scheme provide resuls in line wih macroeconomic heory. The main differences wih he dsge model come from he lengh of he propagaion of he shock and he impac on inflaion, where in his scheme he impac on inflaion is saisically significan. 5 The signal channel is more imporan for counries wih a srong response o convenional moneary policy surprises in he usa; and he porfolio balance is consisen wih he degree of subsiuion of inernaional bonds beween counries. The Effecs of us Moneary Policy 195

8 3. EXCHANGE RATE ARRANGEMENTS IN cadr ECONOMIES One of he peculiariies of hese economies is he imporance of exchange rae sabiliy as a policy objecive. For he region, de faco exchange regimes for mos counries are classified beween differen degrees of managed floaing o dollarizaion. According o he Annual Repor on Exchange Rae Arrangemens and Exchange Resricions 2014 by he Inernaional Moneary Fund, Guaemala has shown greaer flexibiliy, being classified as floaing for differen years in he period under consideraion, even hough i shares he volailiy of is inernaional reserves wih he oher exchange rae argeers (Jácome and Parrado, 2007). 6 Honduras and he Dominican Republic follow a crawl-like arrangemen, while Cosa Rica has he leas flexible regime afer El Salvador, which is a dollarized economy. Table 1 CLASSIFICATION OF EXCHANGE RATE ARRANGEMENT FOR CADR COUNTRIES Counry Exchange rae arrangemen 1 Cosa Rica Oher managed arrangemen 2 El Salvador Honduras Guaemala Dominican Republic No separae legal ender Crawl-like arrangemen Crawl-like arrangemen Crawl-like arrangemen 1 Classificaion according o he Annual Repor on Exchange Rae Arrangemens and Exchange Resricions 2014 by he imf. 2 As he repor saes, his exchange rae arrangemen is characerisic of periods when volaile foreign exchange marke condiions hinder he use of more clearly defined exchange rae arrangemens. I was previously classified as sabilized arrangemen in The Annual Repor on Exchange Rae Arrangemens and Exchange Resricions 2014 reclassified Guaemala as crawl-like arrangemen, previously considered a floaing regime. 196 A. Checo, S. Pradel, F. Ramírez

9 The exchange rae regime of a counry deermines he conduc of is moneary policy. Even hough price sabiliy is he aim of all regimes, heir primary shock absorber is no he same; herefore, i shapes he degree of ransmission mechanisms of foreign moneary policy shocks. Likewise, many counries claim o be floaers, while acually adhering o an exchange rae regime. As Canova (2005) explains, he lack of a differeniaed ransmission mechanism of usa moneary shocks beween groups of floaers and non-floaers, for a se of Lain America counries, 7 may arise because floaers may suffer from fear of floaing, see Calvo and Reinhar (2000), hus using inernaional reserves o offse exchange rae volailiy. 4. EMPIRICAL METHODOLOGY In his secion we describe he empirical sraegy used o characerize he ransmission mechanism of usa moneary policy shocks o cadr economies. 8 The approach consiss of wo seps. In he firs sep, we use a mulicounry daase comprising 76 macroeconomic variables for all cadr counries o esimae common facors hrough Principal Componens. These facors sum up he macroeconomic informaion for he whole sample of abovemenioned counries and are used as indicaors of he sae of he economy (business cycle) for he cadr region. In he second sep, we specify a dynamic model beween he esimaed common facors and a block of foreign variables, where he laer includes he ffr. Once he model is esimaed, we address he issue of proper idenificaion of he impac of usa moneary policy shocks on foreign economies and esimae he effecs on cadr macroeconomic variables. 4.1 Firs Sep: Daa Descripion and Common Facors Esimaion This secion explains how we collec and rea daa of he economies under analysis. Firs we describe he daase used and is characerisics. Then we discuss he procedure for daa reducion hrough facor esimaion. 7 Counries under analysis include Argenina, Brazil, Chile, Ecuador, Mexico, Panama, Peru, and Uruguay. 8 Counries include: Cosa Rica, El Salvador, Guaemala, Honduras, and he Dominican Republic. Nicaragua is excluded from he sample due o lack of daa prior o The Effecs of us Moneary Policy 197

10 4.1.1 Daa Descripion We ake a broad sample of daa, consising of he main macroeconomic indicaors for a se of small open economies on a monhly basis: Cosa Rica (cr), El Salvador (es), Honduras (hn), Guaemala (g), and he Dominican Republic (dr), for he period. The complee se of variables and he ransformaions performed are shown in Annex A. All variables are expressed in welve-monh variaion, and sandardized by subracing he sample mean and dividing by he sample sandard deviaion. The daase comprises hree main groups: a) Real Indicaors This group conains variables from he real secor of he economy, i.e. real aciviy indicaors, 9 expors, impors, rade balance and remiances, all in real erms. From he fiscal secor, we incorporae oal fiscal revenue and expendiure, boh in real erms. By including his group, we aim o capure he varying responses across secors and periods o business cycles, and how hey migh respond differenly o a foreign ineres shock. b) Prices and Relaive Prices This group consiss of real exchange raes and consumer price indexes (cpi). Finally, nominal and real exchange raes (local currency price of usa dollar) are included. c) Financial and Moneary Secor Indicaors This se is composed of several measures of ineres raes, including lending and deposi raes (in nominal erms). We also include credi growh o he privae secor in real erms as an indicaor of he business cycle. Finally, o capure he overall evoluion of money supply, we include m Common Facor Esimaion Insead of esimaing a srucural var model for each counry, we address he research quesion using a daa reducion approach o deal wih he dimension of he by-counry daase described in he las secion. 9 We uilize a monhly indicaor of economic aciviy called Indicador Mensual de Acividad Económica (imae, for is acronym in Spanish). 198 A. Checo, S. Pradel, F. Ramírez

11 Our mehodology employs he esimaion of common facors hrough principal componens analysis summarizing he se of variables described above. This mehodology inroduced o forecasers by Sock and Wason (2002) and o macroeconomics by Bernanke e al. (2004) exracs from a large se of daa a smaller group of facors ha drive he dynamics of he whole sample. This mechanism allows he researcher o summarize big daa nealy, avoiding he curse of dimensionaliy, while a he same ime accouning for he crucial informaion. We use he principal componens analysis o esimae hese common facors. This analysis exracs a series of facors from N number of variables, which are linear combinaions of his daa se, and aemps o: a) minimize noise, since he exraced facors conain he mos imporan informaion, leaving aside noisy deviaions and b) minimize redundancy, since wo facors should no conain he same informaion from he daase, bu should express differen dimensions along which he daa varies. Suppose we have M series spanning T periods, colleced in M 1 vecors X, from which we exrac N facors spanning he same T periods in a N 1 vecor F, where N < M. These facors resume he informaion shared by he variables in X. X and F are relaed by he measuremen equaion: 1 X =ΛF, where he marix Λ is M N. Is elemens are called facor loadings; hese associae he value of he facors o he measured variables of he model. For he empirical exercise, we choose he firs four esimaed facors, which accoun for 53% of he common variance of he whole se (76 series). Since he complee daase is used, we inerpre hese facors as he sae of he economy or common cycles beween cadr economies. Afer a visual inspecion (Figure 1) we observe a srong correlaion beween he firs facor and gdp growh raes in hese economies. Likewise, he second facor could be relaed o he common behavior of cpi inflaion in he counries under sudy. The Effecs of us Moneary Policy 199

12 Figure 1 ESTIMATED PRINCIPAL COMPONENTS FACTORS PC PC IMAE_CR IMAE_G IMAE_H IMAE_CR IMAE_SV PC1 IPC_CR IPC_G IPC_H IPC_RD IPC_SV PC PC3 2.5 PC A. Checo, S. Pradel, F. Ramírez

13 4.2 Second Sep: favar Specificaion and Esimaion In his sep we specify a favar model beween he se of esimaed facors, F, as discussed in Secion 4.1.2, and a block of foreign variables. The block of foreign variables includes he usa cpi, usa Indusrial Producion Index (ipi), and Real Balances (m1), which are he ypical se of variables used o analyze he impac of mp shocks in he usa (Sims, 1992). As for he measure of he usa moneary policy insrumen, he effecive ffr remained unchanged for he las seven years. Noneheless, he Federal Reserve has employed nonconvenional insrumens, known as quaniaive easing (qe) programs, which have led o a more expansive moneary policy han wha can be accouned for by he effecive ffr. Therefore, in order o address his issue, we consider he Wu-Xia Shadow Federal Funds Rae as our measure of he moneary policy insrumen (Wu and Xia, 2016). We also consider he Volailiy Index (vix) as a measure of he inernaional risk premium. Following Canova (2005), we assume ha domesic variables (summarized in he common facors from he firs sep) do no have an impac on foreign variable dynamics (he small open economy assumpion). In addiion, we assume ha vix has no impac on usa macroeconomic variables, bu he laer have influence on he level of risk percepion. This assumpion is jusified under he argumen ha he macroeconomic impac of financial risk shocks is difficul o race, because 1) i is difficul o rule ou he conemporaneous response of uncerainy shocks from financial shocks, and 2) ha he effecs of uncerainy shocks seem significan only in cases of ighening financial condiions (Caldara e al., 2016). Expression 2 summarizes he specificaion of he favar model: 2 W p = C + AiW V i = (), Y Y C VIX where W = VIX, C = c, F F C A10 0 O Y V Ai () = a20 a21 O VIX, V = v. A30 A31 A F 32 V Here, Y includes usa macroeconomic variables menioned above. Exogeneiy resricions are represened by he marix O. V is he reduced form error erm wih mean zero and covariance marix Σ V. This error is a linear combinaion of srucural shocks. The Effecs of us Moneary Policy 201

14 To assess he dynamic responses of he measuremen variables o foreign ineres shocks we rewrie Equaion 2 in erms of a vecor moving average, vma ( ) : W B iv i = 1. = () From he relaion beween reduced form residuals and srucural shocks: W B i DE i = 1 or W = G () i DE i = 1, = () where D is he marix of srucural coefficiens and E is he vecor of srucural shocks. In paricular, E includes he usa moneary policy FFR shock of ineres,. Therefore, he impulse response of common facors vecor o he shock of ineres is: 3 F + s FFR = G( s) for s = 0,1, K and G(s) a vecor wih he response of each facor in F o he srucural innovaion on he federal funds rae. Our concern is on he dynamic response of observables X o he moneary shock, so using 1 and 3, X, F =Λ =ΛGs ( ). FFR + s + s FFR For example, he response of variable i o he foreign ineres rae shock is: xi, s f1 s f 2 s f = λ FFR 1i λ FFR 2i λ FFR Ki K + s FFR = ( )+ ( )+ + ( ) λ g s λ g s λ g s. 1i 1 2i 2 Ki K Idenifying usa Moneary Policy Shocks To complee he explanaion of our empirical mehodology, we now discuss he idenificaion sraegy of usa moneary policy shocks. To draw a coheren characerizaion of he ransmission mechanism of 202 A. Checo, S. Pradel, F. Ramírez

15 ineres, i is imporan o insrumen he proper idenificaion of his shock. Recursive (Cholesky) ordering for he foreign variables block leads o wrong measuremen of he shock of ineres revealed in he radiional puzzles, as discussed in Fornero e al. (2016). Therefore, we adop a sign resricion approach, as is common in he lieraure on he ransmission mechanism of foreign moneary shocks. According o he heory, a conracionary foreign ineres rae shock leads o a fall in oupu, diminishing inflaion pressures, whereas exchange rae appreciaes, as expeced from heoreical models. 10 We rely on his sraegy popularized by Canova and De Nicoló (2003), Uhlig (2005) and Gerler and Karadi (2014) for our idenificaion sraegy. 11 Our goal is o esimae srucural shocks associaed wih models ha produce he expeced response of usa variables o exogenous moneary policy movemens hrough he ffr. In paricular, we impose he following sign resricions in he spiri of Canova and De Nicoló (2003), where prices are sluggish and oupu has a lagged response o moneary innovaions. As in Uhlig (2005), we limi sign resricions on he impulse responses o provide a minimalisic idenificaion, herefore no imposing furher views beyond he sign resricions hemselves. We impose resricions on he foreign variables block only on impac, where he horizon for he sign resricion o hold is one period, hus: ffr > 0, =1 usa ip growh < 0, =2 usa cpi inflaion < 0, =2 usa real balance growh < 0, =2, where denoes he period in monhs where he sign resricion is imposed. The raionale for his idenificaion sraegy for he usa moneary policy shocks is ha he ransmission of moneary policy innovaions o he economy occurs wih lags. 10 Uhlig (2005) employs an agnosic idenificaion procedure o sudy he effecs of moneary policy on oupu. He finds no clear effec of ineres rae hikes on real gdp. 11 However, as emphasized by Fry and Pagan (2011), we recognize he muliple model issue arising from he ransformaions of he new se of srucural shocks. The Effecs of us Moneary Policy 203

16 5. RESULTS In his secion we discuss he response o a foreign ineres rae innovaion of domesic variables (hrough he associaed facor loadings o each of he esimaed facors included in he favar model). The shock is calibraed by a one-ime 25 basis poin unexpeced increase o he shadow ffr, our proxy of moneary policy rae in he usa. Table 2 summarizes he qualiaive response of macroeconomic variables for each economy. Complee resuls in erms of impulse response funcion are shown in Annex B. 12 Table 2 RESULTS OVERVIEW Variables Cosa Rica El Salvador Guaemala Honduras Dominican Republic Oupu Expors Impors Trade balance - Remiances - cpi inflaion - - Real exchange rae Nominal exchange rae Ne inernaional reserves m1 - Privae credi - - Ineres rae - embi Source: Auhor s esimaion. ( ) represens a saisically significan increase (decrease). 12 In Annex B we also include impulse responses assuming a recursive idenificaion sraegy using Cholesky decomposiion. The problems o idenify moneary policy shocks arise when such approach is used. 204 A. Checo, S. Pradel, F. Ramírez

17 According o he esimaed impulse response funcions, a posiive shock o he ffr has a negaive impac on main real domesic variables. For all counries under analysis, oupu, expor and impor growh raes fall. In addiion, financial secor variables such as ineres raes and risk premium increase, while money and credi demand decrease. There is no evidence of significan nominal and real exchange rae adjusmens o he shock, while we find a decrease in inernaional reserves for hree of hese economies. The empirical lieraure on ransmission mechanisms of usa moneary policy shocks (see Canova, 2005) emphasizes he role of he exchange rae regime and he degree of financial inegraion in he magniude of he pass-hrough o domesic macroeconomic variables (real and nominal) of hese ype of innovaions. Therefore, counries wih flexible (less-flexible) exchange rae regimes and relaive high (low) inegraed financial markes show less (more) volailiy in domesic variables such as oupu and ineres raes. Despie ha, impulse response resuls sugges depreciaion pressures afer a foreign ineres shock in cr, g, and hn are no saisically significan. Insead, our resuls illusrae ha cenral banks reac o he exernal shock by increasing ineres raes across all counries and reducing ne foreign reserves in cr, hn and he dr. Risk premium rises in es and he dr, evidence of a ighening in foreign financial condiions. 13 Likewise, posiive inflaion pressures are no observed due o ineres rae reacion and hus a limied exchange rae pass-hrough effec. On he real side, our resuls show a negaive effec on oupu growh. Similarly, expor and impor growh fall in all counries. These resuls are in line wih Jannsen and Klein (2011) which emphasizes he imporance of he income-absorpion effec over he expendiureswiching effec in counries wih acive exchange rae policies oriened o sabilize his variable. Neverheless, he fall in impor growh exceeds he fall in expors; herefore, rade balance improves for mos counries considered, excluding es whose resuls are no significan. This finding is opposie o he predicion from heoreical open economy dsge lieraure, such as Galí and Monacelli (2005), where he real depreciaion induced by a foreign ineres rae shock riggers an expor increase. Behind his heoreical ransmission mechanism is he assumpion of relaive flexibiliy in exchange rae markes. 13 Daa for he sample period are only available for hese wo counries The Effecs of us Moneary Policy 205

18 Finally, remiances are an imporan inflow of foreign resources o cadr economies, up o 16% of gdp for es and hn in This inflow depends on economic and labor marke condiions where domesic labor force emigraes. Our resuls highligh he negaive response of remiances flow in all counries (excluding g where he response is no significanly differen from zero). This consiues an addiional channel hrough which foreign ineres shocks impac domesic aciviy. 6. CONCLUSION In his documen we analyzed he impac of usa moneary policy shocks on he developing economies of Cenral America and he Dominican Republic. As we menioned, hese economies are differen from oher emerging economies given heir lower financial deepening, heir lesser exposure o capial flows and higher weigh of exchange rae sabiliy in cenral bank loss funcions. Using a mulicounry daase of macroeconomic variables which includes real secor and moneary indicaors, we idenify he ransmission mechanism of foreign (usa) ineres rae shocks o he domesic economy. Impulse response analysis suggess ha his ype of shock pushes down real oupu, expors and impors. In addiion, a usa moneary policy shock will have low impac on nominal exchange raes, a he cos of increasing ineres raes, falling ne inernaional reserves and rising risk premium. ANNEXES Annex A. Daa Descripion All series were direcly aken from he Consejo Moneario Cenroamericano/Secrearía Ejecuiva Daabase, excep for he Miscellaneous series (sources a he end of he Annex). Forma is presened as follows: Series name; daa span and series descripion as appears in he daabase. Nominal variables, excep ner and ineres raes, were cpi deflaed. As for he ransformaion, he ineres raes are presened as year-on- year firs-difference values. The res were one year logged differeniaed. All ransformed variables are mean derended and expressed in erms of heir sandard deviaion. 206 A. Checo, S. Pradel, F. Ramírez

19 Real Secor 1. IMAE_CR 2003M01:2014M12 Monhly indicaor of economic aciviy (imae): rend - cycle, index - Cosa Rica 2. IMAE_SV 2003M01:2014M12 Monhly indicaor of economic aciviy (imae): rend - cycle, index - El Salvador 3. IMAE_G 2003M01:2014M12 Monhly indicaor of economic aciviy (imae): rend - cycle, index - Guaemala 4. IMAE_H 2003M01:2014M12 Monhly indicaor of economic aciviy (imae): rend - cycle, index - Honduras 5. IMAE_RD 2003M01:2014M12 Monhly indicaor of economic aciviy (imae): rend - cycle, index - Dominican Rep. 6. EXPORTS_CR 2003M01:2014M12 Expors of goods: millions of usd, oal fob - Cosa Rica 7. EXPORTS_SV 2003M01:2014M12 Expors of goods: millions of usd, oal fob - El Salvador 8. EXPORTS_G 2003M01:2014M12 Expors of goods: millions of usd, oal fob - Guaemala 9. EXPORTS_H 2003M01:2014M12 Expors of goods: millions of usd, oal fob - Honduras 10. EXPORTS_RD 2003M01:2014M12 Expors of goods: millions of usd, oal fob - Dominica Republic 11. IMPORTS_CR 2003M01:2014M12 Impors of goods: millions of usd, oal fob - Cosa Rica 12. IMPORTS_SV 2003M01:2014M12 Impors of goods: millions of usd, oal fob - El Salvador 13. IMPORTS_G 2003M01:2014M12 Impors of goods: millions of usd, oal fob - Guaemala 14. IMPORTS_H 2003M01:2014M12 Impors of goods: millions of usd, oal fob - Honduras 15. IMPORTS_RD 2003M01:2014M12 Impors of goods: millions of usd, oal fob - Dominica Republic 16. REMESAS_CR 2003M01:2014M12 Remiances income: millions of usd - Cosa Rica 17. REMESAS_SV 2003M01:2014M12 Remiances income: millions of usd - El Salvador 18. REMESAS_G 2003M01:2014M12 Remiances income: millions of usd - Guaemala 19. REMESAS_H 2003M01:2014M12 Remiances income: millions of usd - Honduras 20. REMESAS_RD 2003M01:2014M12 Remiances income: millions of usd - Dominica Republic The Effecs of us Moneary Policy 207

20 Exchange Rae 21. TCR_CR 2003M01:2014M12 Real exchange rae - Cosa Rica 22. TCR_SV 2003M01:2014M12 Real exchange rae - El Salvador 23. TCR_G 2003M01:2014M12 Real exchange rae - Guaemala 24. TCR_H 2003M01:2014M12 Real exchange rae - Honduras 25. TCR_RD 2003M01:2014M12 Real exchange rae - Dominica Republic 26. TCN_CR 2003M01:2014M12 Nominal exchange rae: local currency per usd - Cosa Rica 27. TCN_SV 2003M01:2014M12 Nominal exchange rae: local currency per usd - El Salvador 28. TCN_G 2003M01:2014M12 Nominal exchange rae: local currency per usd - Guaemala 29. TCN_H 2003M01:2014M12 Nominal exchange rae: local currency per usd - Honduras 30. TCN_RD 2003M01:2014M12 Nominal exchange rae: local currency per usd - Dominica Republic Money and credi quaniy aggregaes 31. BMR_CR 2003M01:2014M12 Narrow moneary base: millions of local currency - Cosa Rica 32. BMR_SV 2003M01:2014M12 Narrow moneary base: millions of local currency - El Salvador 33. BMR_G 2003M01:2014M12 Narrow moneary base: millions of local currency - Guaemala 34. BMR_H 2003M01:2014M12 Narrow moneary base: millions of local currency - Honduras 35. BMR_RD 2003M01:2014M12 Narrow moneary base: millions of local currency - Dominican Republic 36. M1_CR 2003M01:2014M12 Moneary aggregae m1: millions of local currency - Cosa Rica 37. M1_SV 2003M01:2014M12 Moneary aggregae m1: millions of local currency - El Salvador 38.M1_G 2003M01:2014M12 Moneary aggregae m1: millions of local currency - Guaemala 39. M1_H 2003M01:2014M12 Moneary aggregae m1: millions of local currency - Honduras 40. M1_RD 2003M01:2014M12 Moneary aggregae m1: millions of local currency - Dominican Republic 208 A. Checo, S. Pradel, F. Ramírez

21 41. RIN_CR 2003M01:2014M12 Ne inernaional reserves: millions of usd - Cosa Rica 42. RIN_SV 2003M01:2014M12 Ne inernaional reserves: millions of usd - El Salvador 43. RIN_G 2003M01:2014M12 Ne inernaional reserves: millions of usd - Guaemala 44. RIN_H 2003M01:2014M12 Ne inernaional reserves: millions of usd - Honduras 45. RIN_RD 2003M01:2014M12 Ne inernaional reserves: millions of usd - Dominican Republic 46. CREDITOPRIV_CR 2003M01:2014M12 Credi: privae secor, millions of local currency - Cosa Rica 47. CREDITOPRIV_SV 2003M01:2014M12 Credi: privae secor, millions of local currency - El Salvador 48. CREDITOPRIV_G 2003M01:2014M12 Credi: privae secor, millions of local currency - Guaemala 49. CREDITOPRIV_H 2003M01:2014M12 Credi: privae secor, millions of local currency - Honduras 50. CREDITOPRIV_RD 2003M01:2014M12 Credi: privae secor, millions of local currency - Dominican Republic Ineres Raes 51. TASA_ACTIVA_CR 2003M01:2014M12 Ineres rae: nominal, lending (per annum) - Cosa Rica 52. TASA_ACTIVA_SV 2003M01:2014M12 Ineres rae: nominal, lending (per annum) - El Salvador 53. TASA_ACTIVA_G 2003M01:2014M12 Ineres rae: nominal, lending (per annum) - Guaemala 54. TASA_ACTIVA_H 2003M01:2014M12 Ineres rae: nominal, lending (per annum) - Honduras 55. TASA_ACTIVA_RD 2003M01:2014M12 Ineres rae: nominal, lending (per annum) - Dominican Republic 56. TASA_PASIVA_CR 2003M01:2014M12 Ineres rae: nominal, deposi (per annum) - Cosa Rica 57. TASA_PASIVA_SV 2003M01:2014M12 Ineres rae: nominal, deposi (per annum) - El Salvador 58. TASA_PASIVA_G 2003M01:2014M12 Ineres rae: nominal, deposi (per annum) - Guaemala 59. TASA_PASIVA_H 2003M01:2014M12 Ineres rae: nominal, deposi (per annum) - Honduras 60. TASA_PASIVA_RD 2003M01:2014M12 Ineres rae: nominal, deposi (per annum) - Dominican Republic The Effecs of us Moneary Policy 209

22 Fiscal Balance 61. ING_FISCALES_CR 2003M01:2014M12 Governmen income: oal, millions of local currency - Cosa Rica 62. ING_FISCALES_SV 2003M01:2014M12 Governmen income: oal, millions of local currency - El Salvador 63. ING_FISCALES_G 2003M01:2014M12 Governmen income: oal, millions of local currency - Guaemala 64. ING_FISCALES_H 2003M01:2014M12 Governmen income: oal, millions of local currency - Honduras 65. ING_FISCALES_RD 2003M01:2014M12 Governmen income: oal, millions of local currency - Dominican Republic 66. GASTOS_FISCALES_CR 2003M01:2014M12 Governmen expendiure: oal, millions of local currency - Cosa Rica 67. GASTOS_FISCALES_SV 2003M01:2014M12 Governmen expendiure: oal, millions of local currency - El Salvador 68. GASTOS_FISCALES_G 2003M01:2014M12 Governmen expendiure: oal, millions of local currency - Guaemala 69. GASTOS_FISCALES_H 2003M01:2014M12 Governmen expendiure: oal, millions of local currency - Honduras 70. GASTOS_FISCALES_RD 2003M01:2014M12 Governmen expendiure: oal, millions of local currency - Dominican Republic Miscellaneous 71. EMBI_SV M01:2014M12 Emerging marke bond index (jp Morgan Chase) - El Salvador 72. EMBI_RD M01:2014M12 Emerging marke bond index (jp Morgan Chase) - Dominican Republic 73. USA_CPI_SA M01:2014M12 Consumer price index for all urban consumers: all iems - usa 74. FFR M01:2014M12 Effecive federal funds rae (no seasonally adjused) - usa 75. USA_IP_SA M01:2014M12 Indusrial producion index(2007=100) - usa 76. USA_M M01:2014M12 m1 money sock, billions of dollars, seasonally adjused - usa 77. SHADOW_FFR M01:2014M12 Shadow federal funds rae (Wu-Xia) - usa 78. VIX M01:2014M12 Volailiy index, vix - usa Sources: 1 jp Morgan Chase; 2 Bureau of Labor Saisics; 3 fred 210 A. Checo, S. Pradel, F. Ramírez

23 Annex B. Impulse Response Funcions Figures Oupu Expors Impors XN Remiances 0.03 CPI RER Figure B.1 FAVAR WITH SIGN RESTRICTIONS CR ES GT Noe: All resuls are expressed in erms of a 25-basis poins shock o he Wu-Xia Shadow FFR. The Effecs of us Moneary Policy 211

24 Figure B.1 (con.) FAVAR WITH SIGN RESTRICTIONS HN DR Oupu Expors Impors XN Remiances 0.03 CPI RER 0.06 Noe: All resuls are expressed in erms of a 25-basis poins shock o he Wu-Xia Shadow FFR. 212 A. Checo, S. Pradel, F. Ramírez

25 Figure B.1 (con.) FAVAR WITH SIGN RESTRICTIONS CR ES GT NER NIR M Credi Ineres rae EMBI Noe: All resuls are expressed in erms of a 25-basis poins shock o he Wu-Xia Shadow FFR. The Effecs of us Moneary Policy 213

26 Figure B.1 (con.) FAVAR WITH SIGN RESTRICTIONS HN DR NER NIR M Credi Ineres rae EMBI Noe: All resuls are expressed in erms of a 25-basis poins shock o he Wu-Xia Shadow FFR. 214 A. Checo, S. Pradel, F. Ramírez

27 Figure B.2 CHOLESKY DECOMPOSITION CR ES GT Oupu Expors Impors XN Remiances CPI RER Noe: All resuls are expressed in erms of a 25-basis poins shock o he Wu-Xia Shadow FFR. The Effecs of us Moneary Policy 215

28 Oupu Expors Impors XN Remiances CPI RER Figure B.2 (con.) CHOLESKY DECOMPOSITION HN Noe: All resuls are expressed in erms of a 25-basis poins shock o he Wu-Xia Shadow FFR. DR 216 A. Checo, S. Pradel, F. Ramírez

29 Figure B.2 (con.) CHOLESKY DECOMPOSITION CR ES GT NER NIR M Credi Ineres rae EMBI Noe: All resuls are expressed in erms of a 25-basis poins shock o he Wu-Xia Shadow FFR. The Effecs of us Moneary Policy 217

30 Figure B.2 (con.) CHOLESKY DECOMPOSITION HN DR NER NIR M1 Credi Ineres rae EMBI Noe: All resuls are expressed in erms of a 25-basis poins shock o he Wu-Xia Shadow FFR. 218 A. Checo, S. Pradel, F. Ramírez

31 References Arora, Vivek, and Marin Cerisola (2001), How Does usa Moneary Policy Influence Sovereign Spreads in Emerging Markes?, imf Economic Review, Vol. 48, No. 3, pp , <doi: / >. Bauer, Michael D., and Chrisopher J. Neely (2013), Inernaional Channels of he Fed s Unconvenional Moneary Policy, Working Paper Series, No , Federal Reserve Bank of San Francisco. Bernanke, Ben S., Jean Boivin, and Pior Eliasz(2004), Measuring he Effecs of Moneary Policy: a Facor-Augmened Vecor Auoregressive (favar) Approach, nber Working Paper, No , January, <doi: /w10220>. Bernanke, Ben S., and Alan S. Blinder (1992), The Federal Funds Rae and he Channels of Moneary Transmission, American Economic Review, Vol. 82, No. 4, Sepember, pp Boivin, Jean, and Marc Giannoni (2008), Global Forces and Moneary Policy Effeciveness, nber Working Paper, No , January. Borda, Parice, Olivier Manioc, and Jean Gabriel Monauban, (2000), The Conribuion of usa Shocks o Caribbean Business Cycles, Social and Economic Sudies, Vol. 49, No. 2, pp Bowman, David, Juan M. Londono, and Horacio Sapriza (2014), usa Unconvenional Moneary Policy and Transmission o Emerging Marke Economies, Inernaional Finance Discussion Papers, No. 1109, Board of Governors of he Federal Reserve Sysem. Caldara, Dario, Crisina Fuenes-Albero, Simon Gilchris, and Egon Zakrajšek (2016), The Macroeconomic Impac of Financial and Uncerainy Shocks, European Economic Review, Vol. 88, Sepember, pp Calvo, Guillermo, and Carmen Reinhar (2000), Fear of Floaing, nber Working Paper, No. 7993, November. Canova, Fabio (2005), The Transmission of usa Shocks o Lain America, Journal of Applied Economerics, Vol. 20, No. 2, pp Canova, Fabio, and Gianni De Nicoló (2003), On he Source of Cyclical Flucuaions in he G7, Journal of Inernaional Economics, Vol. 59, pp The Effecs of us Moneary Policy 219

32 Chen, Jiaqian, Tommaso Mancini Griffoli, and Rana Sahay (2014), Spillovers from Unied Saes Moneary Policy on Emerging Markes: Differen This Time?, imf Working Paper, No. 14/ 240. Clarida, Richard, and Jordi Galí G. (1994), Sources of Real Exchange Rae Flucuaions: How Imporan are Nominal Shocks, Carnegie-Rocheser Series on Public Policy, Vol. 41, pp Craine, Roger, and Vance L. Marin (2008), Inernaional Moneary Policy Surprise Spillovers, Universiy of California, Berkeley. Cruz-Zuniga, Marha (2011), Inernaional Moneary Transmission, A Facor-augmened Vecor Auoregressive (favar) Approach: The Cases of Mexico and Brazil, Business and Economics Journal, Vol. 26. Cushman, David O., and Tao Zha (1997), Idenifying Moneary Policy in a Small Open Economy under Flexible Exchange Raes, Journal of Moneary Economics, Vol. 39, No. 3, pp Dahlhaus, Tajana, and Garima Vasishha (2014), The Impac of us Moneary Policy Normalizaion on Capial Flows o Emerging-Marke Economies, Working Paper, No , Bank of Canada. Ehrmann, Michael, and Marcel Frazscher (2006), Global Financial Transmission of Moneary Policy Shocks, Working Paper Series, No. 616, European Cenral Bank. Eichenbaum, Marin, and Charles L. Evans (1995), Some Empirical Evidence on he Effecs of Shocks o Moneary Policy on Exchange Raes, The Quarerly Journal of Economics, Vol. 110, No. 4, pp Fornero, Jorge Albero, Markus Kirchner, and Andrés Yani (2016), Reassessing he Effecs of Foreign Moneary Policy on Oupu: New Evidence from Srucural and Agnosic Idenificaion Procedures, Working Paper, No. 773, Cenral Bank of Chile. Fry, Renée, and Adrian Pagan (2011), Sign Resricions in Srucural Vecor Auoregressions: A Criical Review, Journal of Economic Lieraure, Vol. 49, No. 4, pp Galí, Jordi, and Tommaso Monacelli (2005), Moneary Policy and Exchange Rae Volailiy in a Small Open Economy, Review of Economic Sudies, No. 72, pp Gerler, Mark, and Peer Karadi (2014), Moneary Policy Surprises, Credi Coss and Economic Aciviy, nber Working Papers, No A. Checo, S. Pradel, F. Ramírez

33 Grilli, Viorio, and Nouriel Roubini (1995), Liquidiy and Exchanges Raes: Puzzling Evidence from G7 Counries, Working Paper, Yale Universiy. Hausman, Joshua, and Jon Wongswan (2006), Global Asse Prices and fomc Announcemens, Inernaional Finance Discussion Papers, No. 886, Board of Governors of he Federal Reserve Sysem. Inernaional Moneary Fund (2014), Annual Repor on Exchange Arrangemens and Exchange Resricions, Washingon. Jácome, Luis H., and Eric Parrado (2007), The Ques for Price Sabiliy in Cenral America and he Dominican Republic, imf Working Paper, No. 07/54. Jannsen, Nils, and Melanie Klein (2011), The Inernaional Transmission of Euro Area Moneary Policy Shocks, Working Paper, No.1718, The Kiel Insiue for he World Economy. Kawai, Masahiro (2015), Inernaional Spillovers of Moneary Policy: usa Federal Reserve s Quaniaive Easing and Bank of Japan s Quaniaive and Qualiaive Easing, Working Paper Series, No. 512, January, Asian Developmen Bank Insiue. Kim, Soyoung (1999), Do Moneary Policy Shocks Maer in he G7 Counries? Using Common Idenifying Assumpions abou Moneary Policy across Counries, Journal of Inernaional Economics, Vol. 48, No. 2, Augus, pp Kim, Soyoung, (2001), Inernaional Transmission of us Moneary Policy Shocks: Evidence from var s, Journal of Moneary Economics, Vol. 48, No. 2, Ocober, pp ,<doi: / S (01) >. Kim, Soyoung and Nouriel Roubini (2000), Exchange Rae Anomalies in he Indusrial Counries: A Soluion wih a Srucural var Approach, Journal of Moneary Economics, Vol. 45, No. 3, June, pp , <doi: /S (00) >. Lasrapes, William D. (1992), Sources of Flucuaions in Real and Nominal Exchange Raes, Review of Economics and Saisics, Vol. 74, No, 3, Augus, pp , <doi: / >. Loayza, Norman, J. Humbero Lopez, and Angel J. Ubide (1999), Secorial Macroeconomic Inerdependencies: Evidence for Lain America, Eas Asia and Europe, imf Working Paper, No. 99/11. The Effecs of us Moneary Policy 221