Demand for Money in Dollarized, In- transition Economy: The Case of Vietnam. Watanabe Shinichi & Pham Thai Binh. Outline

Transcription

1 Demand for Money in Dollarized, In- ransiion Economy: The Case of Vienam Waanabe Shinichi & Pham Thai Binh 2 Ouline Inroducion Theoreical and Empirical Framework Models and Analyses Conclusions 3

2 Inroducion Simple Correlaion beween Growh rae of M2 and Inflaion (Quarerly daa) M2 1/93-4/4 3/97-4/ Puzzle: he growh rae of broad money was negaively correlaed wih inflaion 4 Simple Correlaion: Growh rae of Broad Money and Inflaion 1/93-4/4 Inflaion M2.69 M2VND.693 FCD /97-4/4 Inflaion Puzzle: However, M2VND is posiively correlaed wih inflaion. 5

3 %/year Figure 1: Annual Changes :1 1994:3 1996:1 1997:3 1999:1 2:3 22:1 23:3 INF M2 M2VND FCD 6 % 7 Figure 2: Annual Changes in Price, Broad Money, Broad Money of Vienam Dong and Foreign Currency Deposis (1997:3-24:4) :3 1998:3 1999:3 2:3 21:3 22:3 23:3 24:3 INF M2 M2VND FCD 7

4 Simple Correlaion: Growh rae of Broad Money and Inflaion M2 M2VND FCD 1/93-4/4 Inflaion /97-4/4 Inflaion Puzzle: However, M2VND is posiively correlaed wih inflaion. The negaive correlaion beween M2 and inflaion is caused by he negaive correlaion beween FCD and inflaion 8 Inroducion (con) Moneary policy sill lack money demand model, wha is a key ool o use money supply for conrolling inflaion. In view ha Vienam is highly dollarized, he above observaions indicae ha, in order o solve he puzzle, we should look a M2 componens: M2VND & FCD, insead of looking a M2 as a whole. Only a few researches abou money demand in highly dollarized economies. However, here is no empirical research generaing he money demand model by aking fully ino accoun he dollarizaion siuaion in financial sysem in Vienam. 9

5 Inroducion (con) Objecive: - Aims o esimae he money demand funcion of Vienam and sudies he characerisics of he economy. - Arising from he puzzle of negaive correlaion beween growh rae of broad money and inflaion, he paper develops and esimaes he money demand model by formulaing demand for real M2VND and demand for real FCD. - Conducs some analyses o research empirically he economy. Quarerly daa from 1993:1-24:4. Coinegraion, error correcion model, impulse response and variance decomposiion. 1 II. Theoreical and Empirical Framework Many ried o heorize and empirical research he money demand: Fisher, Pigou, Keynes, Tobin, Friedmand Money as a par among a porfolio of asses including money wih oher financial and real asses. he sandard money demand equaion: M = P f ( S, OC ) M/P: Real money Y/P: real economic aciviies OC: opporuniy cos of holding money. 11

6 II. Theoreical and Empirical Framework (con) In highly dollarized economies: Branson and Henderson (1985) shows ha i is necessary o separae broad money ino domesic and foreign money : m m * + + = f ( y, i, p, i *, e ) + = f ( y, i, p, i *, e ) General Empirical Research abou Money Demand A large sream of empirical researches have been dedicaed o demand for money, especially in developed counries. One imporan facor fuelling he empirical researches on money demand is big advancemens in ime series analysis in he pas decade. To ge he meaningful resuls, he wo mos imporan poins in modeling demand for money: i) Framework of model: boh coinegraion and error correcion model proved o be useful ools in money demand research Coinegraion popularly o capure he long run equilibrium Error-correcion model reflec he shor run dynamic of money demand 13

7 ii) Selecion and represenaion of variables. - Money: M, M1, M2, M3 or broader moneary aggregaes or componens of heirs; - Economic ransacions: GDP, GNP or Indusrial Index - Opporuniy coss: mos imporan facor in geing meaningful resuls -Own rae of money: M, M1: zero; and broad money: shor erm deposi ineres rae. - Subsiued asses o money: - Domesic financial asses: governmen bonds, bills, - Foreign financial asses: ineres raes of FCD or expeced rae of depreciaion - Real asses: inflaion 14 Empirical Research abou Money Demand in Dollarized Economy Dollarizaion is a siuaion where one or some foreign currencies serves one or more funcions of domesic currency (medium of exchange, accouning and sore of value). Currency subsiuion is he siuaion where foreign currency is used only as means of paymen. Asse subsiuion is siuaion where foreign currencies served as sores of value. 15

8 Empirical Research abou Money Demand in Highly Dollarized Economy Rodriguez and Turner (23) invesigaed he money demand in Mexico by coinegraion echnique: M ln( ) P f e d = α o + α 1 ln Y1 + α 2 ln i + α 3i + α 4 e + α 5π + α 6 ln( ) 1 M P + u em ln( M * ) em M * f f e = α o + α1 ln Y1 + α 2 ( i π ) + α 3 ( i π ) + α 4 e + α 5 ln( ) 1 + u d : he difference beween he inflaion raes of he domesic and foreign economies. 16 Empirical Research abou Money Demand in Highly Dollarized Economy Porqueras, Serrano and Somuano (1999) abou 13 Lain America counries during : log( 1 Ri R i m ) = log( m f i d i ) = d γ + γ 1log( yi) + λ 2Ii + λ 3Ei + ε m f, m d are real foreign (domesic) money E i is he exchange rae depreciaion I d i is ineres rae of domesic deposis; y is real naional oupu 17

9 Aarle Van Bas and Budina Nina (1995) invesigaed he money demand in six ransiional economies in Easern Europe during (quickly reforms, huge shocks in price and sharp increase in dollarizaion): Following Cagan Model (hyperinflaion): ln( m ) = c + e e e δ D χπ + γe + ηd E + ε em ln( P * ) = ln m = d + * * e e e χ D απ + δe + φd E + D is he dummy variable represening he financial liberalizaion in ransiional economies during 199s. e is expeced inflaion rae E e is expeced rae of depreciaion 18 ε Empirical Research abou Money Demand in Highly Dollarized Economy Considering ha Vienam is highly dollarized, his paper invesigaes he money demand in Vienam by separaing he broad money ino broad money of domesic currency and foreign currency deposis and esimae he money demand for each of hese moneary aggregaes. The explaining variables will be carefully chosen based on he analysis of financial operaions in Vienam. 19

10 III. Models and Analyses RM 2D = M 2ofVND Ln ( CPI NGDP RGDP = Ln( GDPdeflaor *1 ) *1) RFCD = FCD Ln ( CPI *1 ) DEPO : 6 monh VND deposi nominal ineres rae INF_A = (lncpi lncpi -1 )*4 f RF = DEPOF + E rae of foreign reurns = 6 monh USD deposi nominal ineres rae plus depreciaion rae DIFFER = RF DEPO differences beween raes of reurns of USD and VND deposis 2 III. Models and Analyses (con) RM 2D RFCD = o + β 1 RGDP + β 2DEPO + β 3INF _ A + β 4 β RF + (+) (+) (-) (-) = + γ 1RGDP + γ 2 (+) (+) γ DIFFER + Expeced Signs: - RGDP: ransacion effec (+) - DEPO: own rae of reurn on money (+) - INF_A: subsiued real asse (-) - RF: subsiued foreign financial asses (-) - DIFFER: speculaion beween VND and USD (+) ε ε Quarerly daa from 1993:1-24:4 - daa availabiliy - inflaion is excepionally high during 1991 and

11 Uni Roo Tes RM2D RFCD RGDP INF_A RF DEPO DIFFER Lags ADF saisics s Diferences RM2D RFCD RGDP RGDP INF_A RF DEPO DIFFER Lags ADF saisics -3.36** -2.98** *** -3.37** -4.29*** -5.19*** -6.69*** (**), (***): significan a 5%, and 1% levels Tess use lags lengh from 1 o 1. The repored resul is chosen wih minimum AIC or Schwarz crierion. All are I(1), excep RGDP is I(2) (48 observaions). However a general knowledge is ha GDP should be I(1). Based on general knowledge o 22 consider RGDP as I(1). Coinegraion Tes If a linear combinaion of some non-saionary daa series is saionary, i is called coinegraed. considered as long-run equilibrium relaionship among he variables. Tess begins wih eigh lags and repeaed by reducing consecuively one lag a a ime unil he lag lengh reaches one. Daa are seasonally adjused: no seasonal dummy A dummy variable in RFCD equaion o reflec he srong VND depreciaion during Asian crises. DUM=1 during 1997:1-2 and 1997:4-1998:4 DUM= for he ohers. 23

12 Coinegraion Tes Equaion RM2D H HA r = r * Trace Saisics r 1 r 2 r 3 r 2 r 3 r r 4 r 5.16 Maximum Eigenvalue r = r 1 r 2 r 3 r 4 r 1 r 2 r 3 r 4 r * RFCD 48.7** ** % Criical Value RM2D RFCD Probabiliy (%) RM2D RFCD Coinegraion Tes RM2D = 2.76*RGDP *DEPO 3.16*INF_A -1.94*RF (8.5) (1.8) (-3.9) (-3.2) RFCD = * RGDP * DIFFER (4.74) (7.89) High income elasiciy: he economy is in he moneizaion process. (China.doc) DEPO and INF have srong impacs on money demand. RF coefficien is lower han DEPO coefficien: domesic ineres rae is sill more imporan han rae of foreign reurn in deermining demand for RM2D DEPO parly absorbed RF effec. Speculaion effec (DIFFER) is srong: asse subsiuion is really a maer in Vienam. Srong VND depreciaion shif VND deposis and real asses o USD, decreasing demands for goods and reducing inflaion leading o a srong increase in RFCD. 25

13 Coinegraion ess I is ineresing o see ha he long-erm money demand is exising in Vienam. Previously, i was well believed ha i is difficul o find a model for long run demand for money in Vienam. The coinagraion ess idenified only one coinegraion for real broad money of domesic currency and one unique coinegraion for foreign currency deposis. 26 Error Correcion Model ECM provides informaion abou how adjusmen is aking place o resore he equilibrium in he long run money demand in response o he shor run deviaion. Based on he limied number of observaions, he maximum number of lags should no exceed four. Based on he minimum AIC, The number of lags in RM2D equaion is chosen o be 4 and lags lengh in RFCD equaion is chosen o be 1. 27

14 RM2D Error Correcion Model Variable Coefficien Sandard Error -Saisic C *** ECT(-1) *** RM2D(-1) *** RM2D(-2) *** RM2D(-3) ** RM2D(-4) ** RGDP(-2) *** RGDP(-3) ** DEPO(-1) *** INF_A *** INF_A(-4) *** RF ** RF(-4) ** R2 =.876 Adjused R2 =.826 DW = (***) and (**) significan a he 1 and 5% levels. 28 RM2D Error Correcion Model ECT is negaive and very significan (.1% level), validaing he long run relaionship in he coinegraion es. The adjusmen akes abou 15 quarers. This resul is in line wih he radiional view ha money is largely passive in socialis economies (Aarle and Budina, 1995). In shor run, ineres rae has he sronges effecs on RM2D. The coefficien of DEPO in shor run model (1.13) is smaller han ha in he long run model (4.261), hus reflecing lags in moneary ransmission mechanism (needs some periods o pass he full effec of he increase in ineres rae on o money demand). 29

15 RFCD Error Correcion Model Variable Coefficien Sandard Error -Saisic Prob. C ***.5 ECT(-1) **.219 DUM ***.1 RFCD(-1) ***.6 DIFFER(-1) ***.52 R2 =.575 Adjused R2 =.534 DW = (***) and (**) significan a he 1 and 5% levels. 3 RFCD Error Correcion Model ECT is negaive and significan a 1% level, validaing he long run relaionship in he coinegraion es. In shor run, RGDP doesn play an imporan role in deermining demand for money. 31

16 Impulse Responses Impulse response maps he effecs of a variable s shock on all variables in he vecor error correcion model. Based on sandard macro economic heory, ordering: RF, DEPO, RM2D, RGDP, INF_A 32 RF Innovaion Domesic ineres rae increases immediaely around.47 poins higher han is pre-shock level reflecing he uncovered ineres rae pariy ineres rae policy aims o keep a higher ineres rae o preven a currency fligh o USD RF DEPO RM 2D RGDP IN F_A

17 Responses of RM2D o oher Innovaions RGDP shock leads o increase in RM2D Inflaion and rae of foreign reurn influence on RM2D as srong as GDP does bu in opposie direcion. Increase in ineres rae or a conracionary moneary policy cause he money aggregae o decrease bu his effec lass only around 3 quarers RF DEPO RGDP INF_A Responses of RFCD and RGDP o DIFFER Innovaion Order: DIFFER, RFCD, RGDP RFCD RGDP RFCD a 2nd quarer: overshooing effec A he 3rd quarer, people will buy USD o proec money value as hey are averse o he uncerainies in he 35 fuure

18 Variance Decomposiion Variance Decomposiion of Rae of Foreign Reurn (RF) RM2D RGDP DEPO INF_A RF Variance decomposiion splis variance of forecas errors for one variable ino conribuions of shocks of oher variables in VECM, hus saing how imporan each innovaion is in effecing he variables in VECM. Mos of variances of he forecas errors of RF are explained by shocks in iself. Tha can be explained by he fac ha he exchange rae is srongly conrolled by cenral bank, herefore ineres rae, oupu, money demand has small role in influencing exchange rae. However, inflaion is sill an imporan facor affecing exchange rae, parly reflecing he purchasing power pariy RM2D RGDP DEPO Variance Decomposiion of VND Ineres Rae (DEPO) INF_A RF GDP shock does no affec VND ineres rae in shor run as well as in longer periods. RF has very imporan effec on domesic rae in shor run as well as long run. This parly reflecs he uncovered ineres pariy condiion. Offseing RF shock is a very imporan mission of VND ineres rae. Overall, he ineres rae policy aims a offseing he RF shock in shor run. In a longer run, ineres rae policy aims a mainaining exchange rae and price sabiliy and shock from money demand. However, he weakness of ineres policy is ha i has no role in sabilizing oupu. 37

19 RM2D RGDP DEPO Variance Decomposiion of Real Broad Money (RM2D) INF_A RF Shock of VND ineres rae only affecs RM2D in shor run. In long run, i has no considerable effecs on RM2D RF shock plays he second mos imporan aribues o he variance of RM2D and accouns for around % variance. This confirms RF innovaion as he mos imporan facor affecing RM2D variance, excep RM2D shocks iself. From 6h quarers, each shock from RGDP, INF_A and RF accouns for around 1% variances of forecas errors, showing he relaive imporance of hese variable s innovaions in affecing RMD Variance Decomposiion of INF_A RM2D RGDP DEPO INF_A RF A firs quarer, around 29.4 and 11.8% of is variance come from shock of domesic rae and RM2D, respecively. Afer 1 quarers, he conribuion of RM2D increases o 24.9%, while ha of domesic rae decrease slighly o 22%. Moneary policy is responsible for up o 4-45% of he inflaion variances. 39

20 Variance Decomposiion of Real Oupu (RGDP) 8 6 Variance Decomposiion of Real Broad Money of Foreign Curency (RFCD) 4 2 LM2VNDSA LGDPSA VND_SA INF_SA USD_SA 4 2 RFCD RGDP DIFFER Almos 9% of oupu variances is explained by iself. RM2D accouns for 11% of forecas errors of oupu a he 1s quarer bu his conribuion has decreased quickly. Like a classical economy, RM2D only affec oupu in shor periods and has no effec in longer periods. 4 Conclusions In order o solve he puzzle of negaive correlaion beween M2 and INF, he paper formulaes demand for money by esimaing demand for RM2D and RFCD A number of ineresing and very useful analyses of Vienam economy are also found. The semi elasiciy of DIFFER is very large, implicaing asse subsiuion is a really maer in Vienam financial sysem. The income elasiciy in boh models is around wo, explained by he fac ha Vienam is in he moneizaion process. The shor run model evidenced ha oupu has no role in deermining demand for RFCD in shor run. 41

21 Conclusions (con) Impulse response shows ha an unanicipaed shock of RF increases VND ineres rae (uncover ineres pariy condiion). Variance decomposiion affirms ha he ineres rae policy aims a offseing he shock from RF in he shor run. In he long run, he policy arges mainaining exchange rae and conrolling price. However, he weakness of ineres policy is ha i has no role in sabilizing oupu, his can be due o he rigid ineres rae policy of cenral bank. Ineres rae, oupu, RM2D has no role in influencing exchange rae. Inflaion is he only imporan facor affecing exchange rae, reflecing he purchasing power pariy. 42 Conclusions (con) Oupu is affeced considerably by RM2D only in he firs wo quarers and his effec will be decreased afer ha. The moneary policy has very significan role in conrolling inflaion, suggesing he curren arge of moneary policy in Vienam of simulaing economic growh by a long-lasing expansionary moneary policy seems o be improper. Imporan meaning o Vienam. The resuls may help o improve he effeciveness of moneary policy by applying financial programming o conrol inflaion beer. Realizes he possibiliy of seing up economeric models o use in analyze Vienam economy. 43

22 Thank You!!! 44