Exchange rate misalignment and economic growth

Exchange a e misalignmen and economic g ow h Vic o Ribes Segu a [email p o ec ed] Supe ised by: Juan Ca los Cues as Oli a es EC-1049 Bachelo ’s Thesis Academic yea 2018-2019, deg ee in Economics July 2019 Abs ac This pape s udy he ela ion o RER misalignmen s o h ee sou he n coun ies and o h ee no he n coun ies, ob aining he misalignmen s om a VEC model. Concluding ha misalignmen s a e bigge in he no he n coun ies wi h one excep ion. The nex s ep done is h ough a basic equa ion o economic g ow h do a simila VEC model, la e wi h impulse esponse analyses he e ec o misalignmen s in economy ac i i y o each coun y. The esul a e posi i e e ec s in sou he n coun ies wi h he excep ion o Spain and nega i e o no he n wi h he excep ion o Denma k. Finally mixing he wo esul s he esul s a e no clea because a e con adic o y esul s and is no possible say i unde alued exchange a e is good o bad. Anyways, concluding, he misalignmen s a ec he economy ac i i y and i is g ow h o each coun y so exchange a e becomes an impo an ins umen o conside o he coun ies. 2 Con en s 1. In oduc ion ............................................................................................................... 3 2. Li e a u e e iew ....................................................................................................... 6 3. RER Misalignmen s .................................................................................................. 7 4. Coun y G ow h ....................................................................................................... 15 5. Conclusions ............................................................................................................ 22 6. Appendix ................................................................................................................. 25 7. Da a Appendix ........................................................................................................ 29 8. Bibliog aphy ............................................................................................................ 30 Figu es and Tables Figu e 1. GDP pe capi a .............................................................................................. 4 Figu e 2. HDI ................................................................................................................ 5 Figu e 3. RER ............................................................................................................... 5 Figu e 4. Spain eq RER .............................................................................................. 11 Figu e 5. I aly eq RER ................................................................................................. 12 Figu e 6. G eece eq RER............................................................................................ 12 Figu e 7. Denma k eq RER ......................................................................................... 13 Figu e 8. Sweden eq RER .......................................................................................... 13 Figu e 9. Finland eq RER............................................................................................ 14 Figu e 10. Misalignme s .............................................................................................. 15 Figu e 11. Spain impulse esponse ............................................................................. 18 Figu e 12. I aly impulse esponse ............................................................................... 19 Figu e 13. G eece impulse esponse .......................................................................... 19 Figu e 14. Denma k impulse esponse........................................................................ 20 Figu e 15. Sweden impulse esponse ......................................................................... 21 Figu e 16. Finland impulse esponse .......................................................................... 21 Table 1. VEC o RER.. ................................................................................................ 10 Table 2. VEC o g ow h.. ............................................................................................. 16 Table 3. Sou h RER coin eg a ion ............................................................................... 26 Table 4. No h RER coin eg a ion ............................................................................... 27 Table 5. Sou h g ow h coin eg a ion ............................................................................ 28 Table 6. No h g ow h coin eg a ion ............................................................................ 29 3 1. In oduc ion Fo cen u ies, he cu encies o he wo ld we e backed by gold. This means, ha a cu ency bill issued by a wo ld go e nmen ep esen ed a eal amoun o gold ha go e nmen kep in a aul . In he 30s, he US es ablished he alue o he dolla a a single and unal e able le el: an ounce o gold was wo h $ 35, he gold s anda d. A e Wo ld Wa II, he coun ies signed he B e on Woods pac , whe eby he IMF was c ea ed. Th ough ixed exchange a es e e yone knew how much gold was wo h a US dolla , he alue o any o he cu ency agains he dolla could be based on i s alue in gold. A cu ency whose alue was wice he alue in gold o a dolla , was wo h, he e o e, wo dolla s. Un o una ely, he eal wo ld o economics o e came his sys em. The US dolla su e ed in la ion (i s alue ela i e o he goods i could buy dec eased), while o he cu encies e alued and became mo e s able. In he end, he USA hey could no longe p e end ha he dolla was wo h as much as i had been wo h, so i s alue was o icially educed so ha an ounce o gold would hen ha e a alue o $ 70. Finally, in 1971, he gold s anda d was o e and coun ies applied a lexible exchange a e. This mean ha he dolla no longe ep esen ed a eal quan i y o p ecious ma e ial, and change he model o one whe e supply and demand adjus he p ice, and in some cases wi h cen al banks keeping he p ice be ween some alues. A e 1971 he lexible exchange a e became he mos used ype o exchange a e. Today, he US dolla con inues o domina e many inancial ma ke s. In ac , in e es a es a e o en exp essed in US dolla s. Cu en ly, he US dolla and he eu o accoun o app oxima ely 50 pe cen o all he wo ld's o eign exchange ope a ions. Including B i ish pounds, Canadian dolla s, Aus alian dolla s and Japanese yen, we ha e mo e han 80 pe cen o all cu ency changes. On he o he hand, he las yea s he impo ance o he exchange a e policies was undamen al o many coun ies, and has become a huge deba e. Coun ies like China a e he example o his policies, which a e ela ed wi h cu en accoun su pluses unde alua ion his cu encies o gain compe i i eness. As o he eme ging coun ies ha adap he exchange a e policies o each he de eloped coun ies. The Balassa- Samuelson e ec ela e he exchange a e and economic ac i i y, whe e a ia ion in exchange a es may ha e an impo an e ec o economy. 4 In his pape , I s udy he ela ionship be ween eal exchange a e misalignmen s and economic ac i i y o h ee sou he n Eu opean coun ies (Spain, I aly and G eece) and h ee no he n coun ies (Denma k, Sweden and Finland). The eason o choose hese coun ies is compa e wo eali ies, du ing he inancial c isis he di e ence be ween he no h and he sou h has inc eased. In he no h he s anda d o li e is on he op compa ing all he EU coun ies, while he sou h is on he bo om (wi h he excep ion o Spain ha is in he middle), capi al ligh om he sou h o he no h and he e ec in unemploymen was so much bigge in he sou h. Looking a he GDP pe capi a he di e ence is eno mous, we see how Denma k leads he able o he no he n coun ies wi h he highes alues ollowed by Sweden and Finland. While in he sou h, I aly leads ollowed by Spain and G eece wi h he lowes alues. Obse ing well he da a we see how he dis ance is abysmal be ween hese coun ies, whe e Finland has wice he GDP as I aly, whe e also Sweden and Denma k ge high di e ences wi h I aly. Figu e 1. GDP pe capi a The Human De elopmen Index is o he clea example be ween he di e ences in he no h and he sou h. This index in eg a es li e expec ancy, educa ion and pe capi a income o c ea e a anking o he coun ies. In he Figu e 1, clea ly, we can see he di e ence while he no he coun ies s ay in he op wi h Sweden in he head ollowed by Denma k and Finland, he sou he n coun ies s ay in he middle o a anking wi h 58 coun ies. As Spain as in he head ollowed by I aly and G eece. 5 Figu e 2. HDI Fi s , using he eal exchange a es, in he Figu e 1 a e ep esen ed o he coun ies selec ed since 1995 un il 2018 in qua e s using he 2010 as base yea (2010=100), I will p edic he heo e ical equilib ium o exchange a e and ob ain he misalignmen s espec he o iginal alue. Poin ou ha hey a e all coun ies o he Eu opean Union, bu Sweden and Denma k ha e hei own cu encies. Figu e 3. RER Each coun y will be analysed indi idually using ime se ies, he p ocess o ob aining he misalignmen s will be explained in he co esponding sec ion 3, when we see he a iables used and he econome ic p ocess. Secondly, using he esul s p e iously ob ained I will make an equa ion o g ow h including he misalignmen s calcula ed in he p e ious s ep o ob ain he e ec in he economic g ow h in each coun y. A e h ough impulse eac ion we will analyse he compo men o economic g ow h espec RER misalignmen s; concluding wi h a compa a i e be ween he no h and he sou h. Summa izing, he objec i e is o explain how misalignmen s in he exchange a e a ec in he coun ies o he No h and he Sou h. On he o he hand, see how he exchange 6 a es a e impo an nowadays o he economic ac i i y using coun ies o he Eu opean Union as he bes example o desc ibe his impo ance. Whe e he eu o a ea ( o med by nine een coun ies) is he mos impo an a ea in he Eu opean Union o main ain he s abili y, nowadays Spain, G eece, I aly and Finland a e membe s o he eu ozone. Denma k and Sweden, ha e hey own cu encies, bu Denma k ha e linked his cu ency o he eu o wi h he ERM II (Eu opean Exchange Ra e Mechanisms II) whe e he exchange a e o a non-eu o a ea Membe S a e is ixed agains he eu o and is only allowed o luc ua e wi hin se limi s. On he o he hand, Sweden s ill ou o his mechanism because he popula ion did no app o e i by e e endum. The es is o ganized i s wi h a e iew o he li e a u e used o he pape in sec ion 2. In he sec ion 3 we analyse he RER equilib ium equa ion analysing he componen s and he esul s o misalignmen s. The sec ion 4 use he misalignmen s ob ained in sec ion 3 o c ea e an equa ion o economic g ow h o la e analyse he coe icien and do an impulse esponse, o know how g ow h espond o misalignmen s in each coun y. Following in he sec ion 5 a e he gene al conclusions and in sec ion 6 and 7 appendix and da a appendix. 2. Li e a u e e iew The e a e many s udies ha ela e exchange a e misalignmen s (de ined as de ia ions o he exchange a e om he equilib ium le el) and economic g ow h, whe e he use o coin eg a ion models a e common. The mos di e ences be ween he s udies a e he a iables. One o he s udies bases use he pu chasing powe pa i y om Rod ik (2008) whe e unde alua ion ha e a good e ec o e he g ow h, bu in he long- un his suppose does no hold. O he g oup use he long- un ela ionship o ob ain he misalignmen using coin eg a ion ime se ies o panel da a, based on a model o de e mining he exchange a e. Agui e and Calde ón (2006) is an example, he s udy o he e ec s o he misalignmen s in he eal exchange a e (RER) o 60 coun ies du ing 1965-2003 using coin eg a ion me hods o ime se ies and panel. Based on he model o Obs eld and Rogo ´s (1995) whe e exchange a e equilib ium is p oduc i i y, ne o eign asse s, he e ms o ade and go e nmen spending. Concluding ha dep ecia ion ha e a posi i e e ec o he g ow h, like Rod ik (2008). Simila , Razin and Collins (1999) pose a di e en model concluding ha o e alued cu encies ha e a nega i e e ec o he g ow h. 7 O he s udies a e based in he sea ch o which a iables include as undamen als. Be g and Miao (2010) compa e he esul s be ween he “Washing on Consensus” who a gues ha RER misalignmen s imply imbalances and in consequence bad o he g ow h. Al hough, Rod ick (2008) ela e he unde alua ion ela i e o pu chasing powe pa i y is good o g ow h. This s udy concludes he heo y o Rod ik bu he iewpoin o WC is mo e di icul o con i m. Comunale (2017) ob ain he same esul s wi h he analysis o he EU coun ies, in he same way Habib e al. (2017) o a la ge panel o almos 150 coun ies. On he o he hand, Sch öde (2013) and Agui e and Calde on (2005) conclude wi h an in e se esul . The unde alua ion is no posi i e o he economic g ow h o he coun ies in compa ison wi h he o he s udies. Looking a ound he li e a u e, we can see how misalignmen s a ec s g ow h, unde alua ion s imula e g ow h in con as han o e alua ion ha ha m he economy. Fo his, he impo ance o sea ch asymme ic e ec s gained impo ance due o policy. Rod ick (2008), Be g and Miao (2010), Comunale (2017) ba ely ind asymme ies wi h unde alua ion and o e alua ion, bu Agui e and Calde on (2006) and Sch öde (2013) concludes ha bo h a ec nega i ely o he economy g ow h whe e o e alua ion wi h he s onges e ec . O he o he s udies, Cues as, Mou elle and Reges (2019) ob ain he same esul s as Be g and Miao (2010), Sch öde (2013), Comunale (2017) whe e o e alued exchange a e is bad o he economic ac i i y o a g oup o CEE coun ies. Besides, o e alua ion is much s onge han unde alua ion. 3. RER Misalignmen s Fi s , as Cues as, Mou elle and Reges (2019) do in hei pape , o ob ain he RER misalignmen s we selec ed a g oup o a iables o c ea e an equilib ium equa ion o he RER. The a iables a e GDP pe capi a (pibpc), balance o paymen s (bp), go e nmen expendi u e (go ), in es men (in ), bond yields (in ) and consume p ice index (pc). This a iable is selec ed based o li e a u e p e iously explained and da a a ailable, see he da a appendix o a de ini ion and he da a sou ces o he a iables. Ob ained he a iables, he equilib ium equa ion is: 8 e = α0 + α1 pibpc + α2 bp + α3 go + α4 in + α5 in + α6 pc (1) This equa ion is analysed o each coun y sepa a ely, as we know ha Pu chasing Powe Pa i y does no hold. The RER is p oposed as a coin eg a ion ela ionship o he a iables p e iously exposed o ob ain he misalignmen . Applying he coin eg a ion es by Johansen (1988, 1991) o each equa ion he esul s show a clea ela ionship in he long- un, he e is coin eg a ion in he a iables. To see he esul s o he es look a he appendix. Nex , applying he VEC model o coin eg a ed and non-s a iona y I ob ain he coe icien s o he coin eg a ed model o each coun y. The e a e alues omi ed due o di icul o ind all he da a o he speci ic coun y and yea . The es ima ion in he Table 1 ollows he s a egy o es ima e a long- un exchange a e o ime se ies since 1995 un il 2018 in qua e s. The coe icien o exchange a e is no malized o 1 and he le el o signi icance is indica ed wi h *. Analysing he Table 1 o each coe icien , i s we can see ha GDP pe capi a is signi ican in all coun ies bu he sign is no he same, wi h ou coun ies in posi i e and wo in nega i e. In his case he esul s a e ambiguous, he sign expec ed is he nega i e like G eece and Denma k because acco ding o he Balassa-Samuelson e ec when a coun y is mo e de eloped should ha e a mo e alued cu ency. This e ec ela es he di e ences in o a coun y be ween adable and non- adable ma ke , whe e he “Penn E ec ” says ha RER ollow he same di ec ion: I he incomes a e high, he p ices le els a e high compa ing o in e na ional a e age, and when a e low he con a y. Acco ding o Cues as, Mou elle and Reges (2019): “The Balassa-Samuelson e ec is he eal app ecia ion gene a ed by he inc ease in he ela i e p ice o non- adable goods ha ollows an inc ease in p oduc i i y in he mo e compe i i e adable ma ke . This is d i en by he upwa d p essu e on wages in he non- adable sec o ha a ises because wages in he adable sec o a e highe since p oduc i i y g ow h is as e in ha sec o han in he non- adable sec o ”, so in he case o Spain, I aly, Sweden and Finland may be o he no inc emen in he ela i e p ice o and a lowe incomes and p ice le els compa ed o in e na ional a e age. The nex coe icien o analyse i is only a ailable o Spain, Sweden and Finland due o he lack o da a in he o he yea in espec i e qua e s. The balance o paymen s alone does no se e o explain luc ua ions in he eal exchange a e, i is composed o a 9 cu en accoun and capi al accoun . The balance o paymen s egis e s all mone a y ansac ions be ween a coun y and he es o he wo ld, he cu en accoun includes ne ansac ions o goods and se ices, while capi al accoun he in lows and ou lows o capi al. The sum o hese wo has o add ce o, i cu en accoun has su plus he capi al accoun has de ici and ice e sa. Con inuing wi h he main pa , i he cu en accoun has de ici du ing a long ime he cu ency ends o dep ecia e. The sing o he coe icien s is he expec ed in Spain and Finland whe e an imp o emen in he balance o paymen s end o app ecia e he cu ency bu in Sweden he esul s a e con adic o y may be due o ha his o ically Sweden always has su plus. The go e nmen expendi u e and in es men a e all signi ica i e in all he cases excep he go e nmen expendi u e in I aly. The in e p e a ion o he signs depends he policy o each coun y, when a coun y spends mo e in non- adable goods he sign is nega i e. As we can see Sweden and Finland ha e a nega i e sign because he speci ic policies whe e spend a lo o money in non- adable goods. On he o he hand, Spain, G eece and Denma k ha e spen mo e in adable goods so o his he sign is posi i e. Wi h he excep ion o Denma k ha is a in e es ing di e ence be ween no he n and sou he n coun ies. The explana ion o in es men is simila , a posi i e sign indica es ha he in es depends mo e on non- adable goods and posi i e i depends mo e on adable goods. In his case ou coun ies ha e posi i e sign and wo nega i es, wi h he excep ion o Spain and I aly (due o go e nmen expendi u e is no signi ica i e) is su p ising ha in he o he coun ies go e nmen expendi u e and in es men goes o he same di ec ion. To adable goods in G eece and Denma k and non- adable goods in Sweden and Finland. Finally, he in e p e a ion o he las wo coe icien s a e con using and depend o he cha ac e is ics o each coun y. The coe icien s o bond yields a e signi icance wi h he excep ion o Spain and oscilla e be ween nega i e o posi i e depend he coun y. The consume p ice index has a li le signi icance and i is only signi ica i e in ou coun ies, whe e he signs a e posi i e o wo coun ies and nega i e o wo oo. 16 Table 2. VEC o g ow h. No es: S anda d e o s in pa en heses. *** p<0.01, ** p<0.05, * p<0.1. 17 A e applying VEC model I ob ain he coe icien s o analyse i . The es ima ion in he Table 2 ollows he s a egy o es ima e a long- un g ow h o ime se ies since 1998 un il 2018 in qua e s. The coe icien o g ow h is no malized o 1 and he le el o signi icance is indica ed wi h *. The i s coe icien go e nmen expendi u e, is only signi ican o Spain wi h a nega i e sign ha signi ica e when mo e spend mo e g ow h. I is s ange ha only o one coun y is signi ica i e bu ha is no he main componen o analyse. Following wi h in es men , his is signi ica i e o ou coun ies. Spain I aly Sweden and Finland, he sign is posi i e wi h he excep ion o Spain. In his case he expec ed sing is posi i e because when mo e in es mo e economic g ow h, bu in his case I aly, Sweden and Finland ha e a con a y e ec wi h a educ ion o g ow h when in es men inc ease. Lea ing he misalignmen s o la e , employmen s is signi ica i e o 5 coun ies. The sign is posi i e in h ee o hem and nega i e in wo, i is cu ious in Spain and I aly because a ec nega i ely in he g ow h may be o he big impac o he c isis he esul is no he expec ed. Al hough, in G eece, Swede and Finland he e ec is he expec ed whe e an augmen o employmen is ela ed posi i ely o g ow h. The nex coe icien , consump ion o households is only signi ican o he sou he n coun ies. In Spain and I aly an augmen o consump ion inc emen s he g ow h while in G eece educes i , so i is a s ange compo men o he a iable ha will be nega i e. As we can see he sign o coe icien s in some cases a e no he expec ed, i does no ma e o much because he analysed coe icien s o a iables a e only a complemen o c ea e a be e equa ion o economic g ow h wi h a misalignmen as main componen . Be o e explain he compo men o g ow h when misalignmen a y using impulse esponse unc ion o VEC models, le s wa ch he coe icien s and he sign o coin eg a ion equa ion. Fo all he coun ies he coe icien is signi ican , he sign is nega i e in I aly, G eece, Denma k and Finland meaning a posi i e ela ionship be ween misalignmen s and g ow h. On he o he hand, Spain and Sweden ha e a nega i e ela ion be ween misalignmen s and g ow h, an inc emen o misalignmen s educes he g ow h. The sign o coe icien may be explained how is he g ow h compo men bu , in some cases, can no be ue his ela ionship. Fo ob ain a mo e accu a e esponse o he G ow h when misalignmen s changes, we use he Impulse Response o Cholesky One S.D. Inno a ions. Th ough his i is possible ob ain in a be e way he e ec on he 18 g ow h. A e all, using he VEC model i is possible ob ain he long- un ela ionship o misalignmen s wi h he g ow h, in his case o a pe iod o 40 qua e s. The i s coun y o analyse is Spain, once ob ained he Figu e 9 he e is a clea end. The e ec in he g ow h is posi i e, du ing he i s pe iods he in luence o misalignmen s on o g ow h g ow up om 0% in he pe iod 1 o 0.06% in he pe iod 3 when he in luence s abilizes un il pe iod 5. S a ing o his pe iod he in luence s a s o dec ease pe iod a e pe iod un il pe iod 12 ha we can see a change, a e ha he in luence dec ease mo e un il -0.04% when s abilizes. Anyways, o a long pe iod o en yea s he pa e n, e en hough is posi i e a he beginning, is nega i e misalignmen s a ec nega i e o g ow h as he sign o coe icien . Figu e 11. Spain impulse esponse The nex coun y is I aly, in he Figu e 10 we can see he esul s o he esponse o g ow o misalignmen s. S a ing om 0% in he i s pe iod he ela ion goes down un il -0.09% app oxima ely in pe iod 2, since his pe iod he end changed. A e pe iod 2 he e ec o misalignmen s o g ow h eco e he ini ial alue 0%, bu a e his he ela ionship s a s o be posi i e. As he highes poin in he pe iod 7 wi h almos 0.08%, a e some posi i es pe iods bu wi h a big a iance he alue s abilizes a ound 0.03%. In his case, as he sign o he coe icien , he ela ionship be ween he g ow h and misalignmen s is posi i e, e en hough has a ew pe iods o nega i e ela ion. 19 Figu e 12. I aly impulse esponse Following wi h he las sou he n coun y, G eece. The esul s a e in he Figu e 11. In his case he ela ion be ween g ow h and misalignmen s is always posi i e, highligh he pe iod 2- 3 when he alue is almos 0.12%. A e his maximum he alue all o 0.02% and s a o oscilla e a ound 0.06% and 0.04%, o s abilizing in he las pe iods a ound 0.05%. In he same o m as Spain and I aly he esponse o g ow h o misalignmen s is like he sign o he coe icien ob ained p e iously in VEC model. Figu e 13. G eece impulse esponse 20 S a ing wi h he i s no he n coun y, Denma k. The esul s a e in he Figu e 12. Tha esul s a e simila o G eece bu wi h less oscilla ion, a he beginning he alue g ow bi a bi he i s 2-3 pe iods o jump o 0.12% a pe iod 4. A e , he alue alls o 0.06% in pe iod 6 o in he ollowing pe iods oscilla e be ween 0.11 and 0.09, and s abilize a ound 0.10% in he las 20 pe iods. As que can se he ela ion in his case is posi i e all he ime wi h a big impac on he g ow h, ollowing he same sign ob ained in he coe icien . Figu e 14. Denma k impulse esponse Sweden has a cu ios compo men , o m he i s pe iod all o -0.06% in pe iod 2 o g ow up o 0.03% in pe iod 3 and o inally all o -0.08% in pe iod 4. A e his he alue s abilizes in -0.08% some pe iods and all wi h a li le oscilla ion o -0.10%. In his case he ela ion is almos all he ime nega i e, whe e misalignmen s a ec nega i ely o economic g ow h. As he o he coun ies he coe icien o VEC model coincide wi h he p edic ion o en yea s e en ha ing a pe iod o posi i e ela ion. 21 Figu e 15. Sweden impulse esponse The las is Finland. In he Figu e 14 a e he esul s. As Sweden, Finland has a s ange compo men he i s pe iods. In he pe iods be ween 2-3 he alue is almos 0.12% o a e all o -0.01% in pe iod 4, he nex pe iods he oscilla ion is big a ound -0.01% and -0.04%. In his case he coe icien o VEC model i does ma ch wi h he esul s, may be due o he big posi i e alue ha we can see in he pe iod 3, anyways he conclusion is ha misalignmen s ha e a nega i e e ec o he economic g ow h. Figu e 16. Finland impulse esponse 22 Once he coun ies a e analysed, we can ge some conclusions. The equa ion o ela e economic g ow h and he misalignmen s has wo ked ine o ob ain a consis ence esul . Al hough, in he VEC coe icien s some o he a iables ha e no signi ica ion he misalignmen s a e signi ican in all he coun ies, whe e he sign wi h he excep ion o Finland a e in he same way as he impulse esponse. Respec he no h and he sou h, in he wo cases a e one excep ion, I aly and G eece ha e a posi i e e ec om misalignmen s o he economic w o h while Spain has a nega i e. Sweden and Finland ha e bo h a nega i e e ec om misalignmen s o economic g ow h while Denma k has a posi i e. Using he in o ma ion ob ained in he sec ion 3 whe e we calcula e, h ough coin eg a ion ela ionship, he hea ical equilib ium o each coun y, ha we ha e ob ained o he coe icien s o he equa ion o RER. I is possible con i m one o he esul s p e iously saw in he li e a u e. In he g aph compa ing he RER and que heo e ical equilib ium om Sweden we see ha he cu ency is unde alued, linking his o esponse impulse analysis can conclude ha unde alua ion has a nega i e e ec o he economy. As Agui e and Calde on (2006) and Sch öde (2013) concludes, ha o e - alua ed and unde alued a ec nega i ely o he economy g ow h. On he o he hand, looking he g aph o I aly can see an asymme y because in he g aph he endency o I alian cu ency is o e - alua ed while in he g ow h analysis in he case o I aly he misalignmen s a ec posi i ely. So, he heo y o Agui e and Calde on (2006) and Sch öde (2013) does no hold in his case. 5. Conclusions The objec i e o his p ojec is unde s anding how exchange a e a e impo an o he economy and he possible e ec s. In a wo ld whe e he exchange a es a e inc easingly impo an , wi h he clea example o China and i is policy o unde alua ion o gain compe i i eness. This added o he Balassa-Samuelson e ec we e a ia ions o exchange a e may ha e impo an consequences o he economy. Selec ing h ee no he n coun ies (Denma k, Sweden and Finland) and h ee sou he n (Spain, I aly and G eece), wi h he simple eason o compa e wo eali ies inside he EU as eu o a ea as a e e ence due o eu o is he second mos impo an cu ency a he wo ld. The misalignmen s a e he main componen o he p ojec , h ough an equilib ium equa ion o RER. Applying coin eg a ion by VEC model we ob ain some esul s ha 23 using he coe icien s o coin eg a ion can compu e he heo e ical RER equilib ium o each coun y. Spain be o e he c isis had an unde alued exchange a e bu a e he c isis he misalignmen s had educed and he las yea s o eco e y o m he c isis he exchange a e s a ed o s ay o e - alua ed. I aly had an o e - alua ed exchange a e du ing all he analysed pe iod and G eece has a simila esul o Spain bu wi h a high a iance and he las yea s he exchange a e s a ed o s ay unde alued. Denma k is di ided in o h ee pe iods, he i s he cu ency is unde alued, in he second he end change and is o e - alua ed and inally in he las he cu ency is s a ing o s ay unde alued. In he Sweden case he cu ency is clea ly unde alued du ing all he pe iod analysed. Finally, Finland is ha d o analyse because a high a iance du ing all he pe iods ha makes impossible analyse he s a e o exchange a e. The esul s o his is much mo e misalignmen s in he no he n coun ies han sou he n wi h he excep ion o Denma k, wi h Finland and Sweden on he head ollowed by G eece, I aly, Spain and Denma k espec i ely. So, we can see he i s di e ences be ween no h and sou h. These misalignmen s a e c ucial because we include i in o he nex eg ession o g ow h, he p ocess is he same as p e iously, highligh he big signi icance o he coe icien misalignmen s in all he coun ies. Th ough impulse esponse in he VEC model we ob ain he las esul s. In Spain he ela ion o misalignmen s and g ow h is posi i e he i s pe iods bu la e his begins o be nega i e li le by li le. I aly wi h a nega i e peak a he i s s pe iods becomes posi i e wi h he ollowing pe iods wi h a big posi i e peak o la e s abilize o a lowe alue. G eece cha ac e ized o a a big posi i e peak a he beginning o s abilizes o a lowe posi i e alue in he nex pe iods a e some a iance. Denma k simila o G eece peak a he beginning and s abilizes in he nex pe iods bu a high alue. Sweden has a s ange beha iou , i s he ela ion is nega i e wi h a peak in he nex o pe iods he ela ion is posi i e wi h a big alue and inally all o a nega i e alue bigge han he i s peak. Finally, Finland s a wi h a big posi i e alue and all o a nega i e in he nex pe iods whe e a y be ween 0.01% and 0.04%. In he no h and he sou h, we see a endency wi h an excep ion in bo h. Sweden and Finland ha e a nega i e e ec o misalignmen s in he economic g ow h while Denma k has a posi i e ela ion being he bigges espec he posi i e coun ies. I aly and G eece ha e bo h posi i e e ec o misalignmen s o e he economic g ow h bu Spain has a nega i e ela ion. I is ha d ge ing a conclusion o his esul s because we would need mo e coun ies o analyse he beha iou o disca d ha he case o Spain and Denma k 24 a e isola ed cases o conclude. In he case o Denma k, he ac ha his cu ency is linked o he eu o wi h ERM II, ha does no allow luc ua e mo e han 15% om he eu o can ha e he explana ion o he di e en esul espec he no he n coun ies. In he case o excluding Spain and Denma k, he di e ences be ween he no h he sou h a e clea . The no he n coun ies ha e a nega i e ela ion be ween misalignmen s and g ow h and he sou he n ha e a posi i e ela ion be ween g ow h and sou h, o he pa e n is ha Sweden and Finland ha e mo e high misalignmen s compa ed o I aly and G eece. So, in his heo e ical case we can conclude ha i he e a e o much misalignmen s p obably he e ec o his in he economy will be bad, as he case as Sweden and Finland. Mixing he esul s o he RER and g ow h sec ions, we can see a ela ion saw in he li e a u e. I we compa e he heo e ical RER equilib ium o Sweden wi h he conclusion o all he ime he cu ency is unde alued linked o he conclusion o nega i e e ec o misalignmen s in o he g ow h. We can p o e on pa o he heo y o Agui e and Calde on (2006) and Sch öde (2013), ha o e - alua ed and unde alued a ec nega i ely o he economy g ow h. Al hough, looking he case o I aly we ha e an in e s esul espec Agui e and Calde on (2006) and Sch öde (2013), in his case he exchange a e o I aly is o e alued and he e ec o misalignmen s is posi i e so he heo y does no hold. Highligh ha his mix o he wo sec ions was done wi h hese coun ies because in he heo e ical exchange a e ha e a clea endency. Anyways, ha heo y does no hold, one hing is clea in his p ojec . Exchange a es play a c ucial ole in he g ow h o coun ies and mo e in an inc easingly globalized wo ld, whe he nega i e o posi i e hese misalignmen s cause as we ha e seen eac ions in coun ies and hei g ow h o his he impo ance o he policies will inc ease mo e. In his wo k we ha e no been able o ob ain a clea answe abou he possible di e ences be ween he coun ies o he no h and he sou h, because o he lack o analysing mo e coun ies, o h ough a mo e selec selec ion o a iables ha would allow us o ge close o eali y. Al hough he e is an indi idual esponse o each coun y and he e ec s ha he misalignmen s ha e on hem. 25 6. Appendix In he i s momen in he p ojec we wan o use a VAR model and apply an impulse esponse o he a iables because is a pe ec model o a ime se ies. Bu o use his model we need s a iona y a iables, ha d hing o ge . Con inuing, checking he s a iona i y, all he a iables a e in eg a ed e go no s a iona y. A e he i s analysis we can no use he VAR model, in his case que can do wo hings. T ans o m he a iables in s a iona y, o example using g ow h a es o he a iables bu i we do his he possible ela ions o coin eg a ion a e elimina ed and he we lose impo an in o ma ion. O using he coin eg a ion es by Johansen (1988, 1991) check i he a iables ha e a long- un ela ion o use a Vec o E o Co ec ion (VEC) model, ha is an ex ension o VAR model o coin eg a ed and in eg a ed se ies. We use he coin eg a ion es by Johansen (1988, 1991) in he RER equa ion om sec ion 3. In he Table 3 we ha e he es o Spain, I aly and G eece espec i ely, as we can see he p- alue o MacKinnon-Huag-Michelis a 0.05 le el indica e a leas ou coin eg a ion equa ions o Spain and I aly and h ee o G eece. As he Table 3, Table 4 shows he esul s o Denma k, Sweden and Finland. Whe e he p- alue a 0.05 we can see Denma k wi h h ee coin eg a ion equa ions, Sweden wi h wo coin eg a ed equa ions and Finland wi h se en coin eg a ed equa ions.