CHAPTER 3. REAL EXCHANGE RATE AND COMMODITY EXPORTS PRICES

Transcription

1 CHAPTER 3. REAL EXCHANGE RATE AND COMMODITY EXPORTS PRICES 3.1 Inroducion Many sudies ha have been conduced on he behaviour of he real exchange rae in developing counries emphasised he imporance of he movemens in he erms of rade in explaining variaion in he real exchange rae. Despie his, i is naurally assumed ha changes in he real commodiy prices have he poenial o explain a greaer par of he changes in he real exchange raes for developing counries. This is due o he fac ha mos developing counries rely on he expor of commodiies, and in some cases, on a single commodiy for heir foreign exchange revenues. Cashin, Liang and McDermo (2000) classify developing counries based on he composiion of heir expors. The IMF similarly classifies counries based on he composiion of expors. Counries whose commodiies accoun for a significan share of oal expor are classified as commodiy exporing counries. Namibia is a commodiy exporing counry since is expor is dominaed by wo main commodiies. In 2003 and 2004, diamonds accouned for 41 percen of he oal expor and fish accouned for 20 percen. Oher minerals producs such as copper, uranium, zinc, gold accouned for 15 percen of oal expors. Alhough Namibia is a commodiy exporing counry here is no sudy invesigaing he relaionship beween real exchange rae and real commodiy prices. This chaper will examine wheher prices of commodiy expors are key deerminans of he real exchange rae of Namibia. The chaper sar wih a review of he heory and empirical sudies on he relaionship beween he real exchange rae and prices of expor commodiies. I hen discusses he empirical model o be applied for Namibia. I is imporan o noe ha he purpose of his chaper is no o develop a model, bu review he exising empirical and heoreical models and sugges he appropriae one for Namibia. The chaper is organised as follows. Secion 3.2 discusses 42

2 he lieraure and empirical sudies. Secion 3.3 presens he empirical model for Namibia and Secion 3.4 concludes. 3.2 Lieraure and Empirical Sudies One of he opics ha have caused a lo of disagreemens in inernaional economics is he relaionship beween he real exchange rae and i fundamenal deerminans. This demonsraed a number of empirical puzzles such as ha of purchasing power pariy (PPP). The lieraure idenifies numerous problems in linking he behaviour of he real exchange rae and shocks o i fundamenal deerminans. For example, Cashin, Cespedes and Sahay (2004: 240) saes ha: A variey of srucural exchange rae models failed o forecas more accuraely han a naive random walk model for boh real and nominal exchange raes and heir key findings have no been overurned in he succeeding hree decades. This demonsraes ha if he real exchange rae follows a random walk, he shocks o i remain and ime series can move wihou limis. This is no consisen wih PPP. PPP which is also referred o as he law of one price saes ha any good should sell for he same price in differen counries when prices are convered in he same currency (Rogoff, 1996). The conroversial resuls of various models of he real exchange rae have shown ha he PPP is no a good model for he long-run real exchange rae. Numerous problems in linking he behaviour of he real exchange rae o shocks in fundamenals were idenified. In an effor o deal wih hese problems, Taylor and Peel (2000) and Taylor (2001) incorporaed non-lineariies o model he dynamics of he real exchange rae. I has also been noed by Chen and Rogoff (2003) ha if real shocks which are volaile can be found, hese empirical problems could be poenially solved. Rogoff (1996) poins ou some facors ha can help o solve he PPP conundrum. Among hese facors are real fundamenals such as he Balassa-Samuelson effec, differenials in ineres raes and porfolio models. A large number of sudies (such as Edwards, 1988a, 43

3 1988b; Elbadawi, 1994; Baffes e al. 1997; Aron e al., 2000) emphasised imporance of he fundamenals such as erms of rade, capial flows, openness, governmen expendiure in explaining movemens in he real exchange rae. These sudies emphasised ha erms of rade is a key facor in explaining variaion of he real exchange rae. However, i is imporan o noe ha mos developing counries (including some developed counries such as Ausralia, Canada and New Zealand) expor mainly commodiies. The expors of hese counries are dominaed mainly by primary commodiies. This implies ha movemens in he prices of commodiies have a poenial o explain more variaion in he real exchange rae. Alhough erms of rade has been regarded as key facor in explaining he behaviour of he real exchange rae in developing counries, i is very broad and may no be an appropriae explaining facor. I is imporan o invesigae how changes in commodiy prices influence he real exchange rae in commodiy exporing counries such as Namibia. Empirical research on he mechanism hrough which changes in he real exchange rae is affeced by he commodiy prices is limied, bu here are some noable sudies ha examine he link beween he real exchange rae and commodiy prices. Edwards (1985) was probably he firs sudy ha invesigaed wheher he prices of commodiies have an imporan effec on he real exchange rae. The invesigaion was done o es wheher he price of coffee has an effec on he real exchange rae of Colombia. A model was developed o analyse he relaionship beween he price of coffee, creaion of money, inflaion and real exchange rae. The emphasis was on he link beween he real exchange rae and he price of coffee. The model has he following assumpions. The firs one is ha here are hree goods in he economy and hese goods are coffee, nonradable and non-coffee radable. The second is ha he economy (Colombia) has a crawling exchange rae regime in which he nominal exchange rae is adjused wihin a band according o some indicaors. The hird assumpion is ha capial conrols are exogenous. Edwards developed and esed his model for Colombia. The invesigaion confirmed ha he prices of coffee have an imporan effec on he real exchange rae. A rise in he price of coffee leads o higher disposable income and he demand for radable and noradable 44

4 goods will increase. This is he income effec of he increase in he prices of commodiies (coffee) and i causes higher relaive price of nonradable and appreciaion of he real exchange rae. All his depends on he exen o which he price of oher non-coffee radable goods is given by he price in he world marke and he exchange rae. Improvemen in he price of coffee causes a surplus in he balance of paymens and an increase in he level of foreign exchange reserves. According o Edwards, if he reserves are no serilised compleely i will resuls in he rise in he moneary base. An increase in he moneary base leads o high inflaion and appreciaion of he real exchange rae. Chen and Rogoff (2003) sudied he relaionship beween he real exchange rae and commodiy prices for Ausralia, Canada and New Zealand. These hree counries expors are dominaed by commodiies and i is logical o hink ha he prices of commodiies have a poenial o explain variaion in heir real exchange raes. According o Chen and Rogoff, his could help o solve empirical PPP conundrum. Prices of commodiies could be a missing link ha has he poenial o solve he PPP conundrum. Chen and Rogoff (2001, 2003) esimaed he relaionship beween real exchange rae and he real prices of commodiies as: RER = γ lnrealcom + v (7) ln 0 + γ1 where RER is he real exchange rae, REALCOM is real commodiy prices and v is he error erm. Chen and Rogoff wen furher o include real commodiy prices in sandard moneary models o es if he addiion of commodiy prices variable affec he significance of some variables in he real exchange rae deerminaion. Real commodiy prices were included in he moneary sandard models as follows: ln 0 + γ1lnrealcom + γ2lnrealdiff + γ3 RER = γ lnoutdiff + v (8) where differenials REALDIFF is he differenials in real ineres rae and OUTDIFF is real oupu 45

5 The sudy found evidence of close relaionship beween he real exchange rae and prices of commodiies. A rise in commodiy prices causes real exchange rae appreciaion. Alhough real commodiy prices do no solve he enire PPP empirical conundrum, hey play an imporan role in he deerminaion of he real exchange raes of commodiy exporing counries. A recen model of real exchange rae and commodiy has been developed by Cashin e al. (2002, 2004) o es wheher he prices of commodiies have an imporan effec in he deerminaion of he real exchange rae for commodiy exporing counries. To discuss he relaionship beween he real prices of commodiies and he real exchange rae, a model of a small open economy ha produces wo ypes of goods is developed. These are nonradable and exporable goods. Producion of he exporable good is associaed wih he producion of he primary commodiy. This primary commodiy can eiher be agriculural or mineral produc. The local economy has one secor responsible for producing a primary produc and he oher one producing a nonradable good. In his model, labour is he only inpu in hese wo secors. There are wo assumpions in his model. The firs one is ha producion akes place in a compeiive marke where firms have access o consan reurns o scale. The second assumpion is ha labour is mobile beween secors and hence wages are equal in all secors and here is no arbirage. The local economy is characerised by consumers ha are suppliers of labour inelasically. These consumers consume nonradable and radable goods. I is imporan o noe ha he radable good is sourced from he res of he world and i is no produced locally. The primary commodiy and he inermediae commodiy (produced abroad only) are used joinly by foreign firms o produce a final radable good. This final radable good is consumed by foreign consumers. The algebraic deails of his model are presened in Cashin e al. (2004: ). This model can be referred o as he commodiy price and relaive produciviy model. According o his model, he real exchange rae for a commodiy exporing counry is deermined as follows: 46

6 RER = f( PROD/ PROD, PROD / PROD, COM) (9) N N where RER is he real exchange rae, PROD/ PROD is he differences in produciviy beween local and foreign radable secors, PROD N/ PROD N represens he differences in produciviy beween foreign and local nonradable secors and COM is he world prices of commodiies. The variables PROD/ PROD and PROD N/ PRODN represen he Balassa Samuelson effec. In his famous aricle, Balassa (1964) claimed ha deviaion from he PPP is relaed o he relaive produciviy in one counry o anoher. Taking GDP per capia as a proxy for produciviy or echnology he real exchange rae will appreciae in response o increase in GDP per capia. In oher words, if a counry is more producive i will experience overvaluaion in is real exchange rae. The variable COM represens he effec of he real world prices of commodiies on he real exchange rae. I is expeced ha a rise in real commodiy prices resul in higher wages and his will cause a rise in he prices of nonradable goods and he real exchange rae will appreciae. Cashin e al. developed and applied heir model o a number of developing counries and he resuls revealed evidence of close relaionship beween he real exchange rae and real prices of commodiies as well as produciviy for nearly half of he counries. Increase in boh real commodiy prices and produciviy causes he real exchange rae o appreciae. Spaafora and Savrey (2003) esimaed he real exchange rae for a commodiy exporing counry, Russia using produciviy and commodiy prices (oil prices) as explanaory variables. The sudy found evidence ha boh commodiy prices and produciviy causes real exchange rae appreciaion. These findings were confirmed by Koranchelian (2005) when he Cashin e al. model was applied o Algeria. Spaafora and Savrey (2003) and Koranchelian (2005) found evidence in suppor of he Cashin e al. model. Relaed sudies such as Hazinikolaou and Polasek (2005) invesigaed he link beween he real exchange rae and commodiy prices for Ausralia for he period 1984:1 o 2003:1. The invesigaion revealed ha commodiy prices have an imporan effec on he 47

7 exchange rae of Ausralia. A rise in he prices of commodiies causes he real exchange rae o appreciae. Bjørnland and Hungnes (2005) analysed he behaviour of he real exchange rae in Norway, which is a commodiy exporing counry. Oil accouns for more han 20 percen of Norway s expors and i can be expeced o have a poenial o explain a greaer par of he variaion in he counry s real exchange rae. The equilibrium real exchange rae for Norway was esimaed as a funcion oil prices. No coinegraion was found beween he real exchange rae and commodiy prices. This promped Bjørnland and Hungnes o examine furher oher variables ha cause a susained movemen of he real exchange rae from PPP. Some more variables such as domesic ineres rae, foreign ineres rae, domesic and foreign prices were added as addiional explanaory variables. When hese variables were added, coinegraion was obained. The resuls indicae ha he exchange rae appreciaes when he price of oil is increasing, and depreciaes when he price of oil price is decreasing. In summary, all he reviewed empirical lieraure found evidence of a close relaionship beween he real exchange rae and commodiy prices as well as produciviy or echnology. Therefore his sudy expecs o find evidence ha he real commodiy prices and produciviy cause he real exchange rae for Namibia. 3.3 Empirical Model for Namibia The lieraure review in he previous secion revealed ha here are hree main models for invesigaing he relaionship beween he real exchange rae and prices of commodiies. These are Edwards (1985) model developed and applied o Colombia, Chen and Rogoff (2001, 2003) and he Cashin e al. model (2004) ha was developed and esed for a number of developing counries. Edwards (1985) model included he creaion of money and inflaion in he deerminaion of he real exchange rae. The model assumes ha he counry has a crawling exchange rae regime. This makes i inappropriae for Namibia, 48

8 because he counry does no have a crawling exchange rae regime. The counry s nominal exchange rae is no adjused wihin a band of indicaors as in crawling exchange rae regime. The nominal exchange rae of he Namibia dollar wih respec o oher currencies is deermined by he Souh African Reserve Bank. The Chen and Roggoff model is a moneary sandard ha include real commodiy prices ineres rae and produciviy differenials. Since Namibia has pegged is dollar o he Souh African rand, a model ha includes some shor-run macroeconomic policy fundamenals such as moneary is no appropriae for Namibia. Pegging requires ha moneary and fiscal policies are consisen wih he chosen nominal exchange rae o mainain an equilibrium which is susainable. The Cashin e al. model has ineresing feaures for a developing economy such as Namibia. I assumes a small open economy producing nonradable and exporable goods which are primary commodiies. The final radable good is consumed by consumers abroad. This model fis Namibia because he counry is a small open economy producing primary producs (such as uranium, copper, diamonds, fish) which are consumed abroad. Looking a he feaures of he hree models, his sudy considers he Cashin e al. model as he suiable model o invesigae he relaionship beween commodiy prices and he real exchange rae for Namibia. The relaionship beween he real exchange rae and he prices of commodiies for Namibia is esimaed as: LREER β β LRCOMP β LPERCAPI + ε = (10) where LREER is he log of he real effecive exchange rae, commodiy prices, produciviy or echnology and LRCOMP is he log real LPERCAPI is log of real GDP per capia which is a proxy for ε is he error erm. All variables are expeced o be posiively relaed o he real exchange rae. Hence, β 1>0 and β 2 >0. 49

9 3.4 Conclusion This chaper discussed he relaionship beween real exchange rae and he prices of commodiy expors. Despie he fac ha he mos developing counries rely on he expor commodiies, mos sudies used he fundamenal approach o real exchange rae deerminaion. They idenified erms of rade as a key facor in explaining movemens in he real exchange rae. However, erms of rade is very broad, and commodiy prices could be a beer explanaory variable for he changes in real exchange raes of commodiy exporing counries. Three main models were idenified o invesigae he relaionship beween he real exchange rae and commodiy prices. These models of real exchange rae and commodiy prices suggess ha since developing counries rely more on he expor of primary commodiies, commodiy prices have a poenial o explain a significan componen of he variaion in heir real exchange rae. All reviewed empirical sudies revealed ha here is a close link beween he real exchange rae and prices of commodiies. The empirical model o be esimaed laer in his sudy is also expeced o provide evidence ha commodiy prices have a significan effec on he real exchange rae for Namibia. 50

10 CHAPTER 4. REAL EXCHANGE RATE MISALIGNMENT 4.1 Inroducion This chaper highlighs he lieraure on he real exchange rae misalignmen. I surveys he empirical lieraure on he impac of real exchange rae on economic performance, and he models or mehodologies for esing he effec of real exchange rae misalignmen on economic performance. The res of he chaper is organised as follows. Secion 4.2 provides heory and lieraure. Secion 4.3 reviews he lieraure, models and mehodologies for assessing he impac of real exchange rae misalignmen on economic performance. Secion 4.4 discusses he measures o correc real exchange misalignmens, and Secion 4.5 presens he conclusion. 4.2 Theory and Lieraure The real exchange rae responds o real and moneary variables, alhough he equilibrium real exchange rae is a funcion of real variables only. Since he equilibrium real exchange rae exiss, i does no necessarily mean ha he acual real exchange rae is always equal o his equilibrium value. Zhang (2001: 83) defines real exchange rae misalignmen as he difference beween he acual and equilibrium real exchange rae. The observed real exchange rae depends on he values of he fundamenals as well as aggregae macroeconomic pressures generaed by an excess supply of money or a fiscal defici, a any given momen. Real exchange rae misalignmens occur in boh fixed and floaing exchange rae regimes. Asfaha & Huda (2002:1) poined ou ha in fixed and flexible exchange rae regimes, real exchange rae misalignmen reflec poor policy fundamenals which prevens he exchange rae from responding o changes in he fundamenals. In floaing exchange rae 51

11 regimes variables such as speculaive aacks ha move he exchange rae a lo relaive o economic fundamenals are he primary cause of real exchange rae misalignmens. According o Asfaha & Huda (2002: 1) a large real exchange rae misalignmen can be aribued o poor policy fundamenals. Real exchange rae misalignmen has become a cenral issue in he analysis of macroeconomic policies in developing counries. As Kamnisky e al. (1997: 10) sae, susained or coninuous overvaluaion of he currency is seen as an early warning of a currency crisis. The real exchange rae misalignmen has a derimenal effec on he performance of he economy. A real exchange rae misalignmen can resul in welfare and efficiency coss. According o Pfeffermann (1985: 17-18) and Edwards (1989:12) real exchange rae misalignmen especially overvaluaion hurs expors and can wipe ou he agriculural secor. I can also cause capial fligh which may be opimal from a privae perspecive bu a subsanial cos in erms of social welfare. A real exchange rae misalignmen, especially overvaluaion undermines expors. I is well recognised ha a dynamic expor secor is imporan in he course of developmen. According Pfeffermann (1985: 18), real exchange rae misalignmen such as overvaluaion reduces oher counries incenive o impor from he counry wih an overvalued real exchange rae, and his srikes a he core of he process of developmen. In addiion o is conribuion o oal producion, expors are imporan in mos counries because he availabiliy of foreign exchange is one of he main deerminans of he overall level of economic aciviy. Even in counries where expor accouns for a small percenage of he GDP, a shorfall in foreign exchange reserves can reduce economic growh. Misalignmen undermines incenives o produce for expors and for impor subsiues. This resuls from he fac ha expors lose compeiiveness, and impors become relaively cheaper because of misalignmen (mainly overvaluaion). This can happen if impor resricions have no been imposed. According o Pfeffermann (1985) if impor resricions are imposed, impors may no become relaively cheaper. Expors are discriminaed agains because of inefficiencies and he high coss associaed wih impor 52

12 resricions, and aemp o offse ani-expor bias hrough subsidies may be unsuccessful because he budge defici may be widened. Since he real exchange rae affecs he price of radable goods, i is of immediae imporance o farmers who will be influenced in deciding wha par of heir effors is o be devoed o growing crops and wha par will be desined for he marke and wha is o be reained for heir own consumpion. The effec of real exchange rae misalignmen especially overvaluaion on agriculure was given a special menion by Pfeffermann (1985: 18) because in he early sages of developmen and in many developing counries he agriculural secor is he key employer. The poores people live in he rural areas and hey are dependen on agriculure and a real exchange rae which is no realisic harms hem. A realisic real exchange rae is conducive o rural prosperiy and can have a posiive effec on growh and he disribuion of income. According o Pfeffermann, where he inernal erms of rade are biased agains agriculure, causing migraion o he urban areas, he need for impored foodsuffs rises and more pressure will be pu on he balance of paymens. If here are no adequae incenives on agriculure he impac on developmen can be negaive, because here is a close relaionship beween agriculure and he overall economic developmen. This can happen even if agriculure accouns for a smaller share of he economy. Pfeffermann exends his argumen o oher resource-based aciviies, ha misalignmen undermines incenives in foresry, mining and agro-indusries. If impors are made relaively cheaper, misalignmen no only discriminaes agains he developmen of domesic echnologies, i also encourages relaively capial inensive mehods of producion hrough cheaper impors of capial goods. This discourages employmen creaion. Real exchange rae misalignmen can also cause capial fligh, which may be opimal from privae perspecive bu a subsanial cos in erms of social welfare. Alhough mos analyses are more concerned abou he impac of overvaluaion, undervaluaion of he currency can also affec he economy negaively hrough higher inflaion and hrough discouraging consumpion and invesmen. Kahn (1992: 13) argues ha alhough 53

13 undervaluaion resuls in a build up of reserves ha can be used o repay previous deb or as a buffer agains fuure adverse shocks, curren accoun surpluses come a he expense of domesic absorpion of resources. Consumpion and invesmen are lower han hey would have been. I is no a good developmen policy o run curren accoun surpluses in order o finance privae capial expor. Developmen policy should focus on simulaing invesmen in he domesic economy insead of invesing abroad. Kahn (1992) argues furher ha an undervalued real exchange rae has an impac on income disribuion, in he sense ha i redisribues income from labour o capial, bu he exen of his would depend on how powerful rade unions are. Through is impac on he compeiiveness of he radable secor versus he res of he world and subsequen effec on invesmen, real exchange rae misalignmen affecs growh. A counry is compeiive if i is able o produce producs a lower prices in comparison wih oher counries ha are in he inernaional marke. Producing producs a lower cos han compeing counries plays a role in he deerminaion of he exernal posiion of he counry. The impac of real exchange rae misalignmen on he compeiiveness of he counry can be a susained problem and herefore i is crucial for hose in policy making posiions o regularly assess and adjus subsanial real misalignmens. This would help o avoid poenial economic problems. Misalignmen of he real exchange rae can be a serious problem for economiss in addiion o he damage i causes o he economy (Devarajan, 1999: 359). If he real exchange rae is overvalued everyone urns o economiss for a quaniaive esimae of he degree of misalignmen. This may sugges a muli-year research projec o answer quesions such as he magniude of currency devaluaion. Daa and oher informaion wihou he aid of a model or a consisency check o develop quick esimaes of he real exchange rae misalignmen have o be used. Devarajan (1999) illusraes his siuaion by saing ha in he CFA Franc zone, prior o he 1994 devaluaion, mos observers agreed ha he real exchange rae was overvalued bu hey disagreed on he exen of overvaluaion. Daa were scarce and robus esimaes and formal models were hard o come by. This problem was made worse by he paricular naure of he CFA franc zone in 54

14 he sense ha i was fully converible and here was no parallel marke in foreign exchange which was ofen used as a guide for esimaing real exchange rae misalignmen. Members (Benin, Burkina Faso, Cameroon, Cenral African Republic, Chad, Congo- Brazzaville, Ivory Coas, Gabon, Mali, Niger, Senegal, and Togo) of he CFA Franc zone share he same currency across wo moneary unions. The degree of misalignmen could be differen across he counries of he CFA Franc zone. According o Devarajan (1999: 370) he CFA has differen represenaive counries. Some counries are high income and have a high componen of primary producs in heir expors, while some are low income or he componen of primary producs in expors is low. Counries wih a larger componen or percenage of primary expors in oal expors can experience high real exchange rae misalignmens because of shocks ha affec primary producs. These counries are expeced o have high overvalued real exchange raes compared o hose wih low share of primary producs in oal expors. 4.3 Impac of Real Exchange Misalignmen on Economic Performance A number of empirical sudies invesigae he effec of he real exchange rae misalignmen on economic performance. Coani, Cavallo & Khan (1990) was probably he firs sudy o invesigae he impac or effec of real exchange rae misalignmen on economic performance. Coani e al. (1990) examine he view ha real exchange rae behaviour and economic performance are correlaed, using empirical evidence from a cross-secion of less developed counries. The impac of real exchange rae misalignmen on economic performance was examined by regressing per capia growh on real exchange rae misalignmen, and invesmen on real exchange rae misalignmen as well as agriculure on real exchange rae misalignmen. Expor growh was also regressed on real exchange rae misalignmen. The resuls showed a srong negaive relaionship beween real exchange rae misalignmen and hese measures of economic performance. 55

15 However, Coani e al. (1990) acknowledges ha he resuls do no imply ha real exchange rae behaviour is he main deerminan of economic performance. Dollar (1992) invesigaes he relaionship beween disorion in real exchange rae and GDP growh for he period 1976 o The invesigaion was done by running regression and correlaion beween real exchange rae disorion and economic growh for a cross-secion of 95 developing counries (43 in Africa, 16 in Asia, 24 in Lain America, 12 in Europe/Middle Eas) and 22 developed counries. The invesigaion revealed ha ouward-oriened policies and exchange rae levels ha encouraged expor growh in Eas Asian counries generaed a boos in heir growh raes. Real exchange rae disorion has a negaive impac on economic growh, and mainenance of sable real exchange rae can improve growh in many poor counries. Ghura and Grennes (1993) examine he impac of real exchange rae insabiliy and real exchange rae misalignmen on economic performance for 33 counries in Sub-Saharan Africa for he period 1972 o The measures of economic performance were GDP per capia, expor o GDP, impor o GDP, invesmen o GDP and savings o GDP. The pooled ime series and cross-secion daa confirmed ha here is a negaive relaionship beween real exchange rae misalignmen and economic performance. Higher levels of real exchange rae misalignmen are associaed wih higher levels of macroeconomic insabiliy. Lowering real exchange rae misalignmen and real exchange rae insabiliy will improve economic performance. Easerly, Loayza and Moniel (1997) esimaed he growh equaion for 81 Lain American counries for he period 1960 o The dependen variable was growh in real GDP per capia, and amongs ohers, real exchange rae misalignmen proxied by black marke premium was an explanaory variable. The esimaion revealed ha real exchange rae misalignmen has a negaive impac on real GDP per capia. 56

16 In sudying he effecs of real exchange rae misalignmen on growh for Egyp, Jordan, Morocco, and Tunisia, Domac and Shabsigh (1999) consruced hree alernaive measures of real exchange rae misalignmen. These are (PPP) measure, black marke rae measure and a model based measure which capures policy-induced misalignmen. The firs measure of real exchange rae misalignmen is based on he PPP heory. The real exchange rae misalignmen is compued as he deviaions of he acual real exchange rae from some base year in which he real exchange rae is believed o have been in equilibrium (Domac and Shabsigh, 1999: 12). The average of he hree highes values of he real exchange rae is used as a proxy for he equilibrium real exchange rae. According o Domac and Shabsigh, selecing he hree highes values of he real exchange rae as a reference, one chooses he years of devaluaion which may no necessarily be equilibrium years. The highes values are chosen because generally devaluaion occurs during balance of paymen crises and when he exernal secor is ou of equilibrium. I is assumed ha he real exchange rae is closer o is equilibrium value when devaluaion akes place. Misalignmen calculaed using he PPP is defined as (Domac and Shabsigh, 1999: 13): ( j max RERij ) / 3 RERMIS = 1 i (11) RERi where RERMIS is real exchange rae misalignmen, [( j max RER ij ) / 3] ( j = 1,2,3) is he average of he hree highes values of he real exchange rae for he i h counry. The PPP measure has a major drawback. I fails o capure changes in he susainable equilibrium real exchange rae produced by changes in economic fundamenals such as erms of rade and commercial policies. The black marke exchange rae premium measure uilises he premium of he nominal black marke exchange rae (B) over he official rae (E) as a proxy for real exchange rae misalignmen. I is calculaed as: 57

17 Bi RERMIS = 1 i (12) Ei This black marke exchange rae premium measures misalignmen in he real exchange rae, disorion in he foreign exchange marke exchange conrol and raioning of impor in he economy. The hird measure of real exchange rae misalignmen is based on he formal model of equilibrium real exchange rae deerminaion which was developed by Edwards (1988b). The model-based measure has an advanage over he firs wo measures. I capures he effecs of changes in he fundamenals and domesic macroeconomic, rade and exchange rae policies on he equilibrium real exchange rae. Empirically i can be obained by using he link beween acual real exchange rae and equilibrium real exchange rae (Domac and Shabsigh, 1999: 13). I is compued as: log e RER i [ log ( ERER ) log ( RER )] log ( ERER ) (13) e i e i where he erm in square brackes on he righ hand side of he ideniy indicaes he gap beween he acual real exchange rae (RER) and equilibrium real exchange rae (ERER), which is real exchange rae misalignmen. Regression analysis are employed o deermine he empirical relaionship log e (RER i ) and log e (ERER i )-[log e (ERER i )-log e (RER i )]. e i Afer consrucing he hree measures of real exchange rae misalignmen, Domac and Shabsigh (1999) invesigae he impac of real exchange rae misalignmen on economic performance in conjuncion wih oher variables. The following equaion was esimaed o invesigae he impac of real exchange rae misalignmen on economic performance: i = β 0 + β1rervi + β 2RERMISi + β3siyi + β 4TOTGi + β5 PCGR POPG + v i i, (14) where PCGR is growh in real GDP per capia, RERV is real exchange rae variabiliy, RERMIS is real exchange rae misalignmen, SIY is invesmen o GDP raio, TOTG is 58

18 erms of rade growh and POPG is populaion growh. The resuls showed ha he real exchange rae misalignmen has a negaive impac on economic growh. Bleaney and Greenaway (2001) used daa from a sample of 14 Sub-Saharan African counries ha are highly relying on he expor of primary commodiies for he period 1980 o The daa were used o analyse he effecs of real exchange rae volailiy and real exchange rae misalignmen on invesmen and GDP growh. The invesigaion revealed ha real exchange rae volailiy and misalignmen have a negaive effec on boh invesmen and GDP growh. Asfaha and Huda (2002) esimaed he degree of real exchange rae misalignmen and is effec on inernaional rade compeiiveness for he economy of Souh Africa using quarerly daa for he period 1985 o The esimaion was based on Edwards ineremporal general equilibrium model of a small open economy. The error correcion approach was used o esimae equilibrium real exchange rae and he resuling misalignmen, and he impulse response and variance decomposiion of he vecor auo regression (VAR) have been esablished o es he impac of misalignmen on rade compeiiveness. Inernaional rade compeiiveness was proxied by uni labour cos and expor. The esimaion showed ha real exchange rae misalignmen affecs negaively Souh Africa s compeiiveness accouning for 20 percen of he variaion in compeiiveness. The growh effecs of real exchange rae misalignmen and heir volailiy were evaluaed by Aguire and Calderόn (2005). The evaluaion was done for 60 counries for he period 1965 o 2003 using boh ime series and panel coinegraion mehods. Afer esimaing real exchange rae misalignmen indices, he impac on growh using dynamic panel daa echniques, conrolling for oher radiional growh deerminans, was evaluaed. The evaluaion found ha developing counries have a higher degree of real exchange rae misalignmen han he developed counries. There is a negaive relaionship beween real exchange rae misalignmen and economic growh. Real exchange rae misalignmen was separaed ino overvaluaion and undervaluaion. The impac of overvaluaion on 59

19 economic growh is more negaive compared o ha of undervaluaion. Reducing overvaluaion and undervaluaion would improve economic growh. In summary, all he above empirical sudies regardless of he mehod used, poined ou clearly ha real exchange rae misalignmen has a negaive impac on growh of boh developing and developed counries. 4.4 Correcion of Real Exchange Rae Misalignmens The las secion discussed he negaive impac of real exchange rae misalignmen on economic welfare and efficiency. This secion discusses wha should be done o correc real exchange rae misalignmens. Correcion of real exchange rae misalignmen is discussed in deail by Edwards (1988a: 23-39). In he case of macroeconomic induced real exchange rae misalignmen he imporan sep is o eliminae he source of macroeconomic disequilibrium, which is he inconsisency beween macroeconomic policy and he nominal exchange rae. This policy can be supplemened wih oher measures or he auhoriies can simply wai for he economy o adjus on is own so ha he real exchange rae reurns o is own equilibrium value. As Edwards (1988) noes, his policy, which is referred o as disinflaion wih auomaic adjusmen, can be severe and has limiaions under predeermined nominal exchange rae. The real exchange rae will differ from is equilibrium value even afer policymakers have conrolled he inconsisen macroeconomic forces which generae he macroeconomic induced misalignmen. If he real exchange rae is overvalued he counry will loose is compeiiveness in he inernaional marke. Under fixed exchange rae in his case a fas reurn o equilibrium requires a decline in he nominal domesic prices of nonradables. A fas reducion in hese nominal prices is no likely under many siuaions. Hence an auomaic adjusmen could ake a long ime and prolong real exchange rae misalignmens. According o Edwards (1988a), if nominal domesic prices and wages are no flexible an auomaic 60

20 adjusmen can generae addiional unemploymen and reduce domesic oupu. A cu in aggregae expendiure which resuls from he correcive macroeconomic measures will generae an excess supply for all goods and asses. This low demand for radables will be refleced in a low curren accoun defici and reducion in ne foreign indebedness. Disinflaion generaes an excess supply of nonradables, and his will require a drop in relaive prices o resore equilibrium. If nominal prices are sicky or inflexible, his relaive price realignmen will no happen and his will resul in high unemploymen. Devaluaion can also be used o correc real exchange rae misalignmen. This concenraes on affecing he nominal price of nonradables o resore equilibrium real exchange rae by adjusing he domesic price of radables. Since he real exchange rae is defined as he exchange rae muliplied by he raio of prices of radables o prices of nonradables (P T /P N ), devaluaion aims a generaing a higher real exchange rae hrough an increase in he price of radables (EP T ). Devaluaion can help o sidesep he adjusmen coss associaed wih auomaic adjusmen policies by affecing he real exchange rae direcly and avoid he necessary reducion in he prices of radables. I is imporan o noe ha moneary arrangemens in he CMA make i impossible for Namibia o use devaluaion for realigning he real exchange rae. There are oher policies ha can have he effec similar o ha of devaluaion, bu i is no easy o replicae all resuls of devaluaion. These are impor ariffs and expor subsidies, and income policies (see also Edwards, 1989: 12-18). The combinaion of impor ariffs and expor subsidies replicaes some effecs of devaluaion. Impor ariffs increase he price of imporables, while expor subsidies increase he domesic price exporables. If impor ariffs and subsidies are of he same rae he relaive price beween imporables and exporables will remain unchanged, bu heir relaive price wih respec o nonradables will increase. As saed by Edwards (1989: 15), his is he same resuls as is achieved under devaluaion. The difference beween devaluaion and impor ariffs and expor subsidies is ha devaluaion affecs he visible and invisible rade as well as he domesic currency price of radable goods, services and radable asses. Impor ariffs and expor subsidies affec only he domesic price of radable goods and services. The oher 61

21 difference is ha devaluaion affecs domesic ineres raes if i generaes expecaions of furher devaluaions. Even if he capial accoun is closed parially some fracion of he expeced devaluaion will be passed ono he domesic ineres raes in his siuaion. Tariffs and subsidies do no have any effec on ineres raes. Devaluaion does no have direc effecs on he fiscal budge, while ariffs and subsidies have a general effec on fiscal imbalances. The problem wih imposiion of ariffs and subsidies is ha hey will encourage various ineres groups o claim exempions for heir paricular indusry. According o Edwards (1989: 15) his is poliical and can be avoided by using devaluaion. Income policies can also be used o realign he real exchange rae. These policies aemp o conrol wage and price increases hrough some form of direc inervenion. This policy can succeed if he domesic goods prices fall relaive o hose of foreign goods. According o Edwards (1989) income policies ha are no accompanied by resrain on demand will fail o bring down inflaion. Correcing real exchange rae misalignmen hrough income policies is inefficien and very risky. 4.5 Conclusion The purpose of his chaper was o review he lieraure and models for assessing he impac of real exchange rae misalignmen on economic performance. The chaper also discussed measures o correc real exchange rae misalignmen. Real exchange rae misalignmen, which is a susained deparure of he real exchange rae from is equilibrium value, has now become a cenral issue in he economies of developing counries. I has a derimenal effec on he economy and can resul in welfare and efficiency coss. I can cause reducion in expor and can wipe ou he agriculural secor. I can also cause capial fligh. Counries ha have expors dominaed by primary commodiies experience he highes real exchange rae misalignmen compared o hose wih few primary commodiies in heir expors. The lieraure has come up wih differen mehodologies and models ha can be used o invesigae he impac of real exchange 62

22 rae misalignmen on economic performance. A number of sudies es he impac of real exchange rae misalignmen on economic performance and compeiiveness. Regardless of which measure is used, real exchange rae misalignmen has a negaive impac on economic performance. Real exchange rae misalignmen can be correced hrough price and wage flexibiliy, devaluaion and income policies. 63