Growth and Inflation Dispersions in EMU: Reasons, the Role of Adjustment Channels, and Policy Implications

Transcription

1 WP/7/67 Growh and Inflaion Dispersions in EMU: Reasons, he Role of Adjusmen Channels, and Policy Implicaions Emil Savrev

2

3 27 Inernaional Moneary Fund WP/7/67 IMF Working Paper European Deparmen Growh and Inflaion Dispersions in EMU: Reasons, he Role of Adjusmen Channels, and Policy Implicaions Prepared by Emil Savrev Auhorized for disribuion by Jörg Decressin July 27 Absrac This Working Paper should no be repored as represening he views of he IMF. The views expressed in his Working Paper are hose of he auhor(s) and do no necessarily represen hose of he IMF or IMF policy. Working Papers describe research in progress by he auhor(s) and are published o elici commens and o furher debae. This paper s analysis of growh and inflaion dispersions in he euro area reveals several findings. Firs, hese dispersions have declined appreciably since EMU; remaining dispersions are small bu persisen, relaing mainly o counry-specific shocks, no differences in he ransmission of common shocks. Second, he differen behavior of ineres raes jus before and afer he inroducion of he euro has conribued significanly o growh dispersions. However, his has been a one-off shock whose effecs, paricularly on consrucion, should be declining over ime. Third, financial secor inegraion could do much more o insure counries agains shocks and increase consumpion smoohing. JEL Classificaion Numbers: C3, C33, E3, F5, F2, F36 Keywords: Common and counry-specific shocks, oupu and inflaion dispersions, convergence, risks sharing Auhor s Address: esavrev@imf.org I am graeful o Jörg Decressin for exended commens and suggesions, Wim Foneyne for useful suggesions, Naalia Tamirisa for inspiring discussions, and seminar paricipans a wo seminars held a he EC and he ECB for heir commens.

4 2 Conens Page I. Inroducion...3 II. Lieraure Review...4 III. Analyical Framework and Esimaion Resuls...5 A. Common versus counry-specific shocks...5 B. Counry-specific developmens and income and price level convergence...6 C. Persisence of counry-specific developmens...7 D. Shock ransmission and relaive imporance of channels...8 E. Cross-counry consumpion and income smoohing...9 IV. Conclusions...2 References...4 Figures. Growh and Inflaion Dispersions Uni Labor Coss and Curren Accoun Conribuion of Common Shocks o Inflaion and Growh, Pre-EMU and EMU...7 4a. Conribuion of Counry-specific Shocks o Growh Dispersions, Pre-EMU and EMU...8 4b. Conribuion of Counry-specific Shocks o Growh Dispersions, Pre-EMU and EMU (Concluded)...9 5a. Conribuion of Counry-specific Shocks o Inflaion Dispersions, Pre-EMU and EMU..2 5b. Conribuion of Counry-specific Shocks o Inflaion Dispersions, Pre-EMU and EMU (Concluded) Ineres Raes and Growh Dispersions House Prices and Growh Dispersions Impulse Response of Oupu o House Price Shock...24 Appendixes Appendix Dynamic General Equilibrium Model...25

5 3 I. INTRODUCTION Growh and inflaion dispersions in he euro area have declined since 99 and are now comparable o hose among US saes, bu hey are longer-lasing. As business cycles have become more synchronized afer euro adopion, he conribuion of he cyclical componen o growh dispersions has declined. However, he conribuion of he rend componen has increased, parly as a resul of differen degrees of srucural reform implemenaion among euro-area members (Figure ). The persisence of remaining inflaion dispersions has come wih cos dispersions and diverging exernal posiions (Figure 2). While emporary differences in inflaion dynamics in a moneary union can be benign, e.g., reflecing income convergence or adjusmen o counry-specific shocks, hey can also be associaed wih risks for fuure growh and incomes. Accordingly, growh and inflaion dispersions among euro-area counries have araced aenion. Several quesions arise from a policy sandpoin. In paricular, wha facors are behind growh and inflaion dispersions? Wha has been he role of he real exchange rae, he real ineres rae, and he financial secor in adjusmen? Wha role could furher financial inegraion play in insuring agains counry-specific shocks ha rigger dispersions? The paper proposes o sudy hese quesions by using a variey of disinc approaches. Firs, he paper examines he conribuion of counry-specific versus asymmerically ransmied common shocks in explaining divergences. Second, i sudies he role of convergence facors in growh and inflaion heerogeneiy in he European Moneary Union (EMU). Third, i assesses he role of he ineres rae and he housing secor in he adjusmen process before and afer euro inroducion, using Bayesian echniques o fi a small heoreically founded model o he daa. Finally, he paper analyzes he conribuion of he financial sysem in sharing counry-specific risks, by smoohing consumpion and income across euro-area counries. The analysis reveals several findings. Firs, growh and inflaion dispersions have declined appreciably since he onse of EMU; remaining dispersions are small bu persisen, relaing mainly o counry-specific shocks, no differences in he ransmission of common shocks. Second, he differen behavior of ineres raes jus before and afer he inroducion of he euro has conribued significanly o growh dispersions. However, his has been a one-off shock whose effecs are declining over ime. Third, and relaed o he ineres raes shock, he consrucion secor has conribued o growh dispersions. Fourh, financial secor inegraion could do much more o insure counries agains shocks and increase consumpion smoohing. The res of he paper is organized as follows. Secion II provides an overview of he findings in he lieraure. Secion III describes he mehodological approaches used in he analysis and discusses he resuls, while Secion IV concludes.

6 4 II. LITERATURE REVIEW The lieraure has analyzed several aspecs of he heerogeneiy of economic performance in EMU: (i) he degree of cyclical synchronizaion and he role of idiosyncraic versus asymmeric propagaion of common shocks; (ii) he relaive imporance of various ransmission channels in he adjusmen process, noably he role of price and wage rigidiies for adjusmens in compeiiveness; (iii) he role of fiscal policy in divergences; and, (iv) he role of he financial secor in promoing economic convergence and improving cross-counry risks sharing and consumpion and income smoohing has also feaured exensively. The lieraure finds ha cycles have become more synchronized afer he launch of EMU. In paricular, Giannone and Reichlin (26) find ha business cycles are similar across EMU counries and movemens in oupus are mainly explained by common shocks wih similar propagaion mechanisms, while idiosyncraic shocks are relaively small bu persisen and accoun for he bulk euro-area dispersions. They also noe ha shock propagaion is more persisen in he euro area han in he Unied Saes, bu ha cycles are less volaile. Eickmeir (26) also finds ha, in general, oupu and inflaion responses o common shocks (demand, supply, moneary policy, and exernal) across euro-area counries are similar, bu long-lasing idiosyncraic shocks are responsible for oupu and inflaion variaions across counries. EC (26) finds ha counry-specific shocks, including a fall in risk premia following he inroducion of he euro, relaxaion of credi consrains, and produciviy in raded and nonraded goods are imporan explanaory facors for divergences. However, despie similariies in he ransmission of shocks, differences remain. For example, van den Noord (24) finds ha he decline in he ineres raes afer he launch of he euro had a differen impac on he housing markes in he small and large counries and, via his channel, he shock has had a differen effec on economic aciviy in he wo groups. Also, Hoeller and ohers (24) argue ha, as he cyclical posiion of housing prices in he small counries may be ou of line wih he common moneary policy, he consrucion secor raises dispersions via is impac on aciviy. The resuls in he lieraure sugges ha he compeiiveness channel dominaes he adjusmen process in he medium run, bu operaes slowly. EC (26) finds ha he procyclical effec of he real ineres channel has been somewha less imporan han previously hough and dominaes in he iniial phase of he expansion, while in he medium erm adjusmens in compeiiveness are more imporan. The sudy also finds ha wage and price rigidiies influence he efficiency of he adjusmen process and could lead o slow correcion in compeiiveness and resul in proraced economic divergences. The lieraure concludes ha fiscal policy has conribued o he reducion of oupu volailiy over ime, bu elemens of procyclicaliy remain. Darvas and ohers (25) find evidence ha fiscal convergence (persisenly similar GDP raios of governmen balances) is associaed wih synchronizaion of business cycles. They also observe ha he Maasrich fiscal crierion may have moved he EMU closer o an opimal currency area by reducing counries scope o cause idiosyncraic shocks. Darvas and ohers recognize ha by imposing convergence of budge deficis, he crierion could make fiscal policy less effecive in couneracing

7 5 asymmeric shocks, bu he resuls sugges ha he synchronizaion effec of fiscal policy has dominaed. The lieraure finds ha risk sharing has increased over he pas decade, bu he share of idiosyncraic shocks smoohed by he financial sysem is significanly lower in he EMU han in he Unied Saes. In paricular, Kalemli-Ozcan, Sørensen, and Yosha (24) find ha abou percen of idiosyncraic (counry-specific) shocks o he per capia gross domesic produc of he euro area counries (over 993-2) are smoohed hrough capial markes, while 55 percen are in he Unied Saes (over ). Marinheiro (23) esimaes a somewha higher share of smoohed counry-specific shocks for he euro area (25 percen) bu sill significanly lower han in he Unied Saes. He also finds ha if financial sysem inegraion in he euro area reaches he level of he Unied Saes, is conribuion o smoohing idiosyncraic shocks could increase by abou 2 percenage poins. III. ANALYTICAL FRAMEWORK AND ESTIMATION RESULTS A. Common versus counry-specific shocks The relaive imporance of common and counry-specific shocks for inflaion and growh was esimaed using a bi-variae VAR for each counry. Specifically, wo separae bi-variae VARs were esimaed for each counry one for inflaion and one for growh: x i x = x A x + A x x + A ε i ε 2 i 2 i 3, 2 i where x is he euro area growh/inflaion, x is counry i growh/inflaion, ε is a common i shock, and ε is a counry-specific shock. The VARs were esimaed wih quarerly daa for wo periods pre-emu (98Q-998Q4) and EMU (999Q-26Q4). To idenify he srucural shocks, he euro area growh/inflaion were assumed o be affeced by counryspecific shocks wih a lag. The impac of common and counry-specific shocks on growh and inflaion was calculaed using he esimaed impulse response funcions from he above VARs. The above sysems can be rewrien in erms of he impulse response funcions and he srucural shocks, as follows: x i x j φ( j) φ 2( j) φ j ε 2 ( ) φ ( j) ε 22 j i j, where φ ( j) is he impulse response funcion of euro area growh/inflaion o common shocks, φ 2( j) is he impulse response funcion of growh/inflaion in each counry o common shocks, φ ( ) is he impulse response of euro area growh/inflaion wih respec o 2 j

8 6 counry-specific shocks, and φ 22 ( j) are he impulse response funcions of growh and inflaion in each counry wih respec o counry-specific shocks. The main findings from he esimaion resuls poin o significanly more euro-area member counry inegraion since he sar of he currency union. The conribuion of common shocks o inflaion and growh has increased since he inroducion of he euro (Figure 3). While, on average common shocks accouned for 2 percen growh and 3 percen of inflaion before EMU creaion, heir conribuion increased o around 6 percen for boh growh and inflaion during EMU. Also, common shocks rigger increasingly similar responses across member counries. Nex, using he above decomposiion of he shocks, he conribuion of counry-specific shocks o growh/inflaion dispersions is relaed o he esimaed impulse response funcions as follows: [ φ ( j) φ ( j)] [ φ ( j) φ ( j)] [ φ ( j) φ ( j)] 2. This relaionship was used o calculae how much of he dispersions is due o counry specific shocks. Remaining growh and inflaion dispersions have largely been driven by counry-specific shocks since he euro inroducion, no differen counry responses o common shocks. Counry-specific shocks accoun for more han 7 percen of growh dispersions (wih he excepion of Ausria, 45 percen, and Greece, 4 percen, Figure 4) and more han 75 percen of inflaion dispersions (excep Ialy, 4 percen, Figure 5). B. Counry-specific developmens and income and price level convergence The impac of income and price level convergence on growh and inflaion dispersions is assessed using wo panel regressions: For inflaion dispersions, π i, π = α + γ log( P i, / P ) + ε i,, where π i, π is he deviaion of inflaion in each counry from he euro-area average and log( Pi, / P ) is he deviaion of he price level in a member counry from he euro area average a he beginning of he sample. For growh dispersions: g i, g = α + γ log( YPi, / YP ) + ε i,, where gi, g is he deviaion of oupu growh in each counry from he euro-area average and log( YP i, / YP ) measures he percen difference of member counries per capia GDP from he euro-area average a he beginning of he sample. The above equaions were esimaed for wo periods: pre-emu (98-998) and EMU (999-26).

9 7 The esimaion resuls sugges an increasing imporance of income convergence and a declining role of price level convergence in accouning for dispersions over ime. Price level convergence was associaed Income and Price Level Convergence and Dispersions / wih 6 percen of inflaion dispersions during bu his halved under Inflaion Growh Inflaion Growh EMU (o slighly above 3 percen of inflaion dispersions). Findings elsewhere in he lieraure sugges similar resuls. For Speed of convergence, Adjused R 2 γ example, Rogers (27) concludes ha he price / Impac of percen price/income level difference from euro area average on inflaion/growh dispersions, in percenage poins. levels of raded goods in EMU have converged mosly prior o he euro adopion, wih heir dispersion hereafer similar o ha in he Unied Saes. A he same ime, income convergence acceleraed under EMU and accouned for a larger share of growh dispersions compared o he pre-emu period. However, from a policy sandpoin i is imporan o bear in mind ha accouning for does no mean causing. Accordingly, emporary raher han fundamenal convergence facors could also explain he remaining growh and inflaion dispersions. C. Persisence of counry-specific developmens Persisence of remaining growh and inflaion dispersions is esimaed using he following panel regression: x i, x = α + ρ ( xi, x ) + γ log( X i, / X ) + ε i,, where xi, x is he deviaion of inflaion/growh in each counry from he euro area average and log( X i, / X ) is he deviaion of he price level/ppp GDP per capia in a member counry from he euro area average. The coefficien γ in he above equaions capures he persisence of dispersions, wih a lower value of he coefficien in absolue erms corresponding o slower adjusmen. The esimaion resuls imply ha persisence of inflaion and growh dispersions has increased under EMU. Alhough he above framework does no idenify he sources of persisence in inflaion and growh dispersions, aside from convergence facors, anoher reason could be differen degrees of srucural reforms among euroarea members during EMU. In Persisence of Inflaion and Growh Dispersions / Inflaion Growh Inflaion Growh Wih lagged dependen variable Shor-erm, γ Long-erm, γ ρ Adjused R / Lower coefficien in absolue erms means higher persisence.

10 8 paricular, ECB (25) finds ha services prices and differences in wage developmens have been major sources of inflaion persisence, and he degree of srucural reform in boh services secors and labor markes has differed noiceably among euro-area members over he pas decade. D. Shock ransmission and relaive imporance of channels The relaive imporance of he ransmission channels was sudied using a general equilibrium model. 2 Equaions () (4) represen he core of he model. y β + β yˆ 6 + ε () = y + β 2 y+ + β 3hg + β 4rg + β 5zg y where, y is oupu gap, hg is real house prices gap, rg is real ineres rae gap, zg is real exchange rae gap, and ŷ sands for he curren period euro area oupu gap in a counry model/he lagged oupu gap of a counry in he euro area model. Poenial oupu, equilibrium ineres rae, equilibrium real exchange rae, and equilibrium real house prices are also defined in he model, which allows consisen esimaes of he gaps wihin he model (for full model specificaion see he Appendix). π π = α π + ( α ) π + + α 2 y + α 3Δz + α 4π + ε (2) ˆ where, π is inflaion, z is logarihm of real exchange rae, πˆ denoes he curren period euro area inflaion in a counry model/he lagged inflaion of a counry in he euro area model, and Δ is firs difference operaor. hg = hg + γ rg ρ + ε (3) hg where, hg is real house prices gap. rs rs = δ rs + ( δ)[ re + π + δ 2 ( π + π ) + δ 3 y ] + ε (4) where, rs is nominal ineres rae, re is real equilibrium ineres rae, and π is inflaion arge. Equaion () is an aggregae demand funcion, which has a lagged erm o capure persisence in he daa and a forward looking componen as in Gali and Gerler (999). Aggregae demand depends also on he real ineres rae gap, capuring he ineres rae channel, he real 2 The model was esimaed using Bayesian echniques for Ausria, France, Germany, Ialy, Neherlands, Porugal, Spain, and he euro area (weighed average of he sample counries) over hree periods: (i) he full sample, 98-26; (ii) pre-emu period, ; and (iii) EMU period, While formally he above counries renounced heir moneary policy in 999, he ineres raes have converged and exchange rae was no used as a policy ool since 996. Hence, for esimaion purpose he EMU period sars in 996.

11 9 house price gap, reflecing cyclical effecs from asse prices on aggregae demand, exernal demand. Equaion (2) is a sandard open economy Phillips curve in which inflaion is driven by demand condiions, exchange rae developmens, and exernal shocks. Equaion (3) defines real house price gap as a funcion of real ineres raes. Real ineres rae changes affec house prices by changing he opporuniy cos of capial invesed in housing, he cos of servicing morgage credi, and he presen value of fuure household earnings. Finally, equaion (4) is a moneary policy reacion funcion, in which he cenral bank cares abou inflaion and oupu gap. The impac of various shocks on dispersions was assessed by simulaing he esimaed models for each counry wih ha for he euro area ogeher. The ineracion of area-wide and counry-specific shocks akes place hrough he demand and supply equaions in he model of he euro area and ha of a member counry. Area-wide demand and supply shocks are assumed o affec each counry conemporaneously, while counry-specific shocks are assumed o affec euro-area demand and inflaion wih a lag. Euro Area Demand (lagged demand in member counry, oher variables) Supply (lagged inflaion in member counry, oher variables) Oher equaions The simulaions wih he esimaed models sugges ha: Member Counry Demand (demand in euro area, oher variables) Supply (inflaion in euro area, oher variables) Oher equaions The EMU-relaed changes in ineres raes have conribued do growh divergences, accouning on average for abou 25 percen of hem (Figures 6). The impac of house prices differs across he counries reviewed and explains abou 5 percen of growh dispersions (Figures 7 8). E. Cross-counry consumpion and income smoohing The degree of cross-counry risk sharing and he role of he financial sysem over ime was esimaed using a panel regression feauring cross-counry correlaions of GNP/consumpion condiional on oupu. The idea is ha absence of correlaion beween GNP/consumpion and oupu suggess risk sharing, e.g., via credi or capial markes. Thus, Δ, (5) x i, Δx = α + β ( Δyi, Δy ) + ε i, where he deviaion of per capia GNP/consumpion growh of counry i from he euro area ( x Δx ) is regressed on he deviaion of counry i per capia oupu growh from he Δ i,

12 euro area ( Δ yi, Δy ), and he coefficien β measures uninsured risk over ime (β = means perfec risk-sharing). To assess he conribuion of he financial sysem, he ime varying coefficien β is specified as a funcion of ime and he dispersion of financial developmen in EMU counries: 3 β = β + β + β F F ) (6) ( 2 i, Subsiuing (6) ino (5) resuls in: Δxi, Δx = α + β ( Δy i, Δy ) + β( Δyi, Δy ) + β 2 ( Fi, F )( Δyi, Δy ) + εi,. (7) The coefficien β measures he average uninsured risk, β indicaes how risk-sharing evolves over ime, and β 2 capures he effec of financial sysem. A negaive β 2 coefficien lowers he degree of co-movemen beween consumpion/gnp wih oupu, reducing he amoun of unshared risk. Equaions (5) and (8) were esimaed for he euro area excluding Ireland and Luxemburg over he period The empirical resuls sugges ha he financial sysem has played a role in income (=GNP) bu no in consumpion risk sharing. The coefficien on he ineracion of he financial sysem developmen wih GDP (coefficien β 2 ) is significan for GNP, bu insignifican for consumpion. Also, he conribuion of he financial sysem o risks sharing does no seem o have changed significanly over ime, as he rend coefficiens in boh he GNP and consumpion equaions are insignifican. Regression Resuls / GNP Consumpion β.94.7 (.3) (.23) β -.5. (.) (.) β (.8) (.24) Sandard errors in parenheses. Esimaion period The GDP share of credi of deposi money banks o he privae secor was used as a proxy for financial sysem developmen. The advanage of his indicaor is ha i does no consider credi issued o governmens, bu a shorcoming is ha i capures only he role of he banking sysem and no of oher financial insiuions or he securiies marke. However, given he dominan role of he banks in he euro area, using his indicaor may no resul in a large bias. In fuure work he following alernaive indicaors could be used as a cross-check: (i) liquid liabiliies, comprising currency and ineres-bearing liabiliies of bank and non-bank financial inermediaries; (ii) sock marke capializaion; and, (iii) he common componen of he hree measures from a principle componen regression.

13 Unshared Risk / (In percen) Gross Naional Income 95 Consumpion / 5-year moving average of he coefficien β in equaion (5). This finding suggess ha risk sharing via financial markes is beer developed for invesmen (which accouns for much of he difference beween GNP and consumpion) han for household consumpion. 4 One reason may be he sill limied inegraion of reail banking in Europe. While he share of cross-border holdings of equiies by euro-area residens doubled beween 997 and 25, he median share of oal asses of branches of euro-area banks ha are locaed ouside home counries remained pracically unchanged a below 3 percen of all euro-area banking asses; he same figure for subsidiaries increased marginally o around 3 percen in 25, from around 9 percen in 2 (see ECB, 27). Overall, i is well known ha reail banking is appreciably less well inegraed han many oher financial aciviies (among ohers, see Decressin e al., 27). Indicaors of Financial Inegraion The Degree of Cross-Border Holdings of Equiy Issued by Euro Area Residens (in percen) Dispersion of Toal Asses of Euro Area Banks' Branches Across Euro Area Counries (in percen of oal assess of he euro area banking secor) Dispersion of Toal Asses of Euro Area Banks' Subsidiaries Across Euro Area Counries (in percen of oal assess of he euro area banking secor) Inra-euro area Exra-euro area Source: ECB. Maximum Minimum Median 4 Auxiliary regressions confirmed his and are available upon reques.

14 2 The poenial welfare gains from furher financial inegraion among EMU members are subsanial for each euro-area counry. A comparison of he volailiy of he individual euroarea members privae consumpion growh wih he volailiy of euro-area oupu growh suggess a high poenial for addiional risk sharing. Euroarea oupu is less volaile han consumpion in each member sae. Specifically, a. percenage poins, he sandard deviaion of euro-area oupu growh is far lower han he maximum sandard deviaion of privae consumpion growh of around 3 percenage poins and lower han he minimum sandard deviaion of privae consumpion growh of.2 percenage poins Euro Area: Volailiy of Member Saes Privae Consumpion and Euro Area Oupu (Sandard deviaion over 99-26, in percenage poins) Ausria Belgium Finland France Germany Greece Ireland Ialy Luxemburg Neherlands Porugal Spain Euro area GDP IV. CONCLUSIONS Euro-area members share common shocks and heir conribuion o growh and inflaion has increased since EMU bu dispersions remain. The resuls of his paper sugges ha inflaion and growh in EMU counries are o a large exen driven by common shocks, he imporance of which has increased over ime. This suggess ha he common moneary policy may have conribued significanly o he business-cycle synchronizaion and sabilizaion, as migh have he beer synchronizaion of fiscal policy. In paricular, while common shocks explained around 3 percen of growh and inflaion in euro-area members before he inroducion of he euro, heir conribuion increased o around 6 percen afer ha. A he same ime, common shocks increasingly rigger common responses. Accordingly, heir conribuion o dispersions has declined. This is imporan no leas because i suggess ha he poenial for moneary policy o efficienly and effecively address common shocks has increased. The remaining dispersions are predominanly driven by counry-specific shocks. Several facors have conribued o he idiosyncraic shocks ha drive remaining dispersions. One facor has been income and price level convergence. This facor could accoun for over 3 percen of he remaining inflaion dispersions and has gained imporance in growh dispersions under EMU. I will persis bu is force will diminish. Anoher se of facors, which is no necessarily orhogonal o he firs one, has been EMU-relaed changes in he ineres raes and house price developmens. This can accoun for 4 percen of growh divergences and may largely be of a one-off naure. The persisence of idiosyncraic shocks raises he imporance of faciliaing adjusmen o shocks. The funcioning of he labor and produc markes could be improved o foser a beer

15 3 operaion of he compeiiveness channel and higher produciviy growh. The laer would faciliae faser adjusmen during downurns, given downward nominal wage rigidiy. Fiscal policy in he member saes could absorb idiosyncraic shocks by allowing auomaic sabilizers o work (see Eichengreen and Wyplosz, 998). Finally, a fully inegraed financial sysem could serve as a powerful insurance mechanism agains asymmeric shocks, by allowing relaively sable consumpion funded via privae raher han public borrowing and governmen inervenion despie flucuaions in domesic oupu. Furher inegraion of he euro-area financial sysem could significanly enhance income smoohing and reduce dispersions. While financial secor inegraion has acceleraed, i has no achieved is full poenial. For example, as shown in he lieraure and confirmed in his sudy, he conribuion of he financial secor o income smoohing could be increased significanly some sudies sugges by 2 percenage poins or more, if is level of inegraion reaches ha of he Unied Saes. Furher inegraion of he European capial markes can play an imporan role in his respec. In addiion, inegraion of reail banking can also conribue o risk sharing, as he resuling flow of cross-counry ineres paymens will help counries smooh idiosyncraic shocks and incomes, lowering consumpion growh dispersions in he euro area.

16 4 REFERENCES Darvas, Z., A. K. Rose, and G. Szapáry, 25, Fiscal Divergence and Business Cycle Synchronizaion: Irresponsibiliy is Idiosyncraic, NBER Working Paper No. 58. Decressin, Jörg, Hamid Faruqee, and Wim Foneyne (eds.), 27, "Inegraing Europe's Financial Markes," Inernaional Moneary Fund, forhcoming. Eichengreen, B. and C. Wyplosz, 998, The Sabiliy Pac: More han a Minor Nuisance? Economic Policy, Vol. 3, No. 26, pp EC, 26, Adjusmen Dynamics in he Euro Area: Experiences and Challenges. EC EU Economy Review 26. ECB, 25, Moneary Policy and Inflaion Differenials in a Heerogeneous Currency Area, ECB Monhly Bullein, May 25. ECB, 27, Financial Inegraion in Europe, ECB. Galí, J. and Gerler, M., 999, Inflaion Dynamics: A Srucural Economeric Analysis, Journal of Moneary Economics, 44(2), Giannone, D., and L. Reichlin, 26, Trends and Cycles in he Euro Area: How Much Heerogeneiy and Should we Worry abou I? ECB Working Paper No Hoeller, P., C. Giorno, and C. de la Maisonneuve, 24, One Money, One Cycle? Making Moneary Union a Smooher Ride, OECD Economics Deparmen Working Papers, No. 4. Kalemli-Ozcan, S., B. E. Sørensen, and O. Yosha, 24, Asymmeric Shocks and Risk Sharing in a Moneary Union: Updaed Evidence and Policy Implicaions for Europe, CEPR Discussion Paper Series No Marinheiro, C. F., 23, Oupu Smoohing in EMU and OECD: Can We Forego Governmen Conribuion? A Risk Sharing Approach, CESIFO Working Paper No. 5. Obsfeld, M. and G. Peri, 998, Regional Non-Adjusmen and Fiscal Policy, Economic Policy, Vol 3, No. 26, pp Rogers, J., H., 27, Moneary Union, Price Level Convergence, and Inflaion: How Close is Europe o he USA? Journal of Moneary Economics, 54(3), Sandra Eickmeier, 26, Comovemens and Heerogeneiy in he Euro Area Analyzed in a Non-saionary Dynamic Facor Model, Deusche Bundesbank Discussion Paper No. 3. Van den Noord, 24, Modeling Cyclical Divergence in he Euro Area: The Housing Channel, OECD Economics Deparmen Working Papers No. 4.

17 5 Figure. Euro Area: Growh and Inflaion Dispersions Variance / Inflaion Oupu Inflaion Growh Panel esimaes wih ime-varying slope /, 2/ Variance explained by cyclical / Variance explained by rend / Source: IMF, World Economic Oulook. / Excluding Ireland and Luxemburg. 2/ Esimaed equaion: x where -growh/inflaion of each, i x = α i + β ( x, i x ) + ε, i x, i member sae; x -Euro area growh/inflaion; -persisence parameer. β

18 6 Figure 2. Euro Area: Uni Labor Coss and Curren Accoun ULC - Deviaions from Euro Area Trend (In percen, index, in USD) AUT IRL FIN FRA DEU NLD BEL LUX GRC ESP ITA PRT 8 FIN Curren Accoun and ULC 5 Curren accoun balance, in percen of GDP, average, AUT IRL DEU FRA NLD BEL GRC LUX ESP ITA ULC - Deviaions from Euro Area Trend (in percen, index, in USD) Sources: IFS and WEO daabase.

19 7 Figure 3. Euro Area: Conribuion of Common Shocks o Inflaion and Growh, Pre-EMU and EMU (In p percen) ) Growh Inflaion EMU Pre-EMU Source: IMF Saff calculaions.

20 8 Figure 4a. Euro Area: Conribuion of Counry-specific Shocks o Growh Dispersions, Pre- EMU and EMU (In percen) Ausria Belgium Finland France Germany Greece Source: IMF Saff calculaions. EMU Pre-EMU

21 9 Figure 4b. Euro Area: Conribuion of Counry-specific Shocks o Growh Dispersions, Pre- EMU and EMU (Concluded) (In percen) Ireland Ialy Neherlands Porugal Spain Source: IMF Saff calculaions. EMU Pre-EMU

22 2 Figure 5a. Euro Area: Conribuion of Counry-specific Shocks o Inflaion Dispersions, Pre- EMU and EMU (In percen) Ausria Belgium Finland France Germany Greece Source: IMF Saff calculaions. EMU Pre-EMU

23 2 Figure 5b. Euro Area: Conribuion of Counry-specific Shocks o Inflaion Dispersions, Pre- EMU and EMU (Concluded) (In percen) Ireland Ialy Neherlands Porugal Spain Source: IMF Saff calculaions. EMU Pre-EMU

24 22 7 Figure 6. Euro Area: Ineres Raes and Growh Dispersions 7 IRL Cumulaive GDP growh, in percen, NDL FIN BEL AUT FRA DEU ITA PRT GRC ESP Real ineres rae decline, percenage poins, Growh Dispersions Due o Ineres Raes (In percen) AUT FRA DEU ITA NDL PRT ESP Source: IMF Saff calculaions.

25 Figure 7. Euro Area: House Prices and Growh Dispersions House Prices and Growh (Cumulaive change, , in percen) IRL Growh 2 ESP AUT DEU FIN PRT BEL ITA FRA NLD House prices 3 Growh Dispersions Due o House Prices (In percen) AUT FRA DEU ITA NDL PRT ESP Sources: BIS and IMF Saff calculaions.

26 24 Figure 8. Euro Area: Impulse Response of Oupu o House Price Shock ( percenage poins shock o house prices).3.25 Ausria.3.25 France Germany.3.25 Neherlands Ialy.3.25 Spain Source: Fund saff esimaes.

27 25 APPENDIX DYNAMIC GENERAL EQUILIBRIUM MODEL The model consiss of he following equaions. Oupu Aggregae demand: y y = β y + β y + β hg + β rg + β zg + β yˆ ε () where, y is oupu gap, hg is real house prices gap, rg is real ineres rae gap, zg is real exchange rae gap, and ŷ sands for he curren period euro area oupu gap in a counry model/he lagged oupu gap of a counry in he euro area model. Poenial growh g g = λ g _ s + ( λ) g + ε (2) where, g is poenial growh, and g_s is seady sae growh. Poenial oupu lye lye = lye + g + ε (3) where, lye is logarihm of poenial oupu. GDP ly = lye + y (4) where, ly is logarihm of GDP. Phillips curve: π π = α π + α π + α y + α Δz + α ˆ ( ) π + ε (5) where, π is inflaion, z is logarihm of real exchange rae, πˆ denoes he curren period euro area inflaion in a counry model/he lagged inflaion of a counry in he euro area model, and Δ is firs difference operaor. House prices Real house price gap hg hg = ρ hg + γ rg + ε (6) where, hg is real house prices gap. Growh of real house prices gh gh = τ gh _ s + ( τ ) gh + ε (7) where, gh is growh of real house prices, and gh_s is equilibrium growh of real house prices. Equilibrium real house prices he he = he + gh + ε (8) where, he is logarihm of real equilibrium house prices. Real house prices h = he + hg (9) where, is logarihm of real house prices.

28 26 Exchange rae Real exchange rae gap zg = z ze () where, zg is real exchange rae gap, z is logarihm of he real exchange rae, and ze is logarihm of real equilibrium exchange rae. Equilibrium real exchange rae ze ze = + ε () ze Uncovered Ineres Pariy z z = z + + ( r rf ) + ε (2) where, r is domesic ineres rae, and rf is foreign ineres rae. Ineres raes Real ineres rae r = rs π + (3) where, r is real ineres rae, and rs is nominal ineres rae. Equilibrium real ineres rae re re = ϑ re _ s + ( ϑ) re + ε (4) Where, re is real equilibrium ineres rae, and re_s is seady sae real ineres rae. Real ineres rae gap rg rg = r re + ε (5) Moneary policy reacion funcion rs rs = δ rs + ( δ)[ re + π + δ 2 ( π + π ) + δ 3 y ] + ε (6) where, rs is nominal ineres rae, re is real equilibrium ineres rae, and π is inflaion arge. Observable variables: Real GDP, inflaion, real effecive exchange rae, nominal ineres raes, real house prices, governmen balance, and foreign ineres raes. Unobservable variables: Poenial oupu, oupu gap, real equilibrium ineres rae, real ineres rae gap, real equilibrium exchange rae, real exchanger rae gap, real equilibrium house price, real house price gap. Uni roo variables: Real GDP, real equilibrium house prices, real house prices, real exchange rae, and real equilibrium exchange rae Sample: EMU7 (Ausria, France, Germany, Ialy, Neherlands, Porugal, and Spain). Sample period: Frequency: Quarerly, annual.