Long run demand for money in India: A co-integration approach

Transcription

1 MPRA Munich Personal RePEc Archive Long run demand for money in India: A co-inegraion approach Sahadudheen I Pondicherry Universiy 202 Online a hp://mpra.ub.uni-muenchen.de/65563/ MPRA Paper No , posed 23. July 205 4:33 UTC

2 LONG RUN DEMAND FOR MONEY IN INDIA: A COINTEGRATION APPROACH (970 TO 200) Sahadudheen I M.Phil scholar Deparmen of Economics Pondicherry Universiy, India Shd4frnds@gmail.com

3 ABSTRACT Demand for money plays a pivoal role in deermining he welfare implicaions of moneary policy acions in an economy. This sudy esimaed he demand for money in India and invesigaed various deerminans of demand for money for he period 970 o The sudy uilized Johansen-juselius coinegraion analysis o es for he exisence of a long run relaionship beween he variables and an Error Correcion mehod is hen used. The sudy concluded ha he income and price has a posiive effec on he demand for money. On he oher hand, ineres rae and exchange rae has a negaive. The income elasiciy is.98 and showing significan, implying ha in India, a one percen economic growh requires around.98 percen increase in he naion s money supply. KEYWORDS: Demand for money, India, uni roo, coinegraion, ECM JEL: E4 INTRODUCTION Demand for money is an imporan concep in he hisory of economic hough and can be defined as he desired holdings of money balances in he form of cash or bank. According o Mankiw he demand for money reflecs he degree of willingness o possess money by economic eniies. I plays a pivoal role in deermining he welfare implicaions of moneary policy acions in an economy. Since he early 970s a sizable chunk of empirical lieraure in moneary economics has been concerned wih he demand for money, because money demand is considered as an imporan indicaor of growh of a paricular economy. The demand for money funcion creaes a background o review he effeciveness of moneary policies, as an imporan issue in erms of he overall macroeconomic sabiliy. Money demand is an imporan indicaor of growh of a paricular economy. The increasing money demand mosly indicaes a counry's improved economic siuaion, as opposed o he falling demand which is normally a sign of deerioraing economic climae. LITERATURE OF THE STUDY I is essenial o review he lieraure in he relevan field boh heory and pracice as o arrive a reasonable and meaningful conclusion. Wih his background on mind some of he earlier sudies on demand for money in naional and inernaional level are reviewed.

4 THEORETICAL LITERATURE REVIEW There are diverse specrums of money demand heories which address a broad range of hypoheses. These heories bring forward he relaionship beween he quaniy of money demanded and a se of economic variables. Classical Theory Money demand heories dae back o he quaniy heory of money. Fisher (9) provided he famous classical equaion in his classical work he purchasing power of money. He examined he link beween he quaniy of money (M) and he oal amoun of spending on final goods and services, aggregae nominal income /oal spending (PxY). If he velociy of money (average number of imes per year ha a dollar is spend) is V=PxY/M, Muliply boh sides of he equaion by M VxM/ =PxY/ x M/ MV=PY Where, M= quaniy of money, V=Velociy, P= Price level, Y= aggregae oupu or income Cambridge Approach The Cambridge approach is associaed wih he neo-classical economiss, Pigou (97) and Marshall (923). The Cambridge approach sressed he demand for money as public demand for money holdings, especially he demand for real balances, which was an imporan facor in deermining he equilibrium price level consisen wih a given quaniy of money. Divide boh side of he exchange equaion (MV=PY) by V MxV/V=PxY/V M=PY/V, M=/V xpy, M=k xpy I can be wrien as, Here we can replace /V in he equaion by k where, k is a consan

5 Keynesian Demand for Money Keynes (936) buil on he Cambridge approach o provide a more rigorous analysis of money demand, focusing on he moives of holding money. Keynes s liquidiy preference heory emphasizes he role of ineres raes in he demand for money. He disinguished hree moives for holding money: he ransacions moive, he precauionary moive and he speculaive moive. All he hree moives influence a paricular person s holdings of money. Keynes argued ha he demand for money for ransacions and precauionary moives depends on he level of income, while speculaive demand for money depends on ineres raes. From Keynes perspecive, he demand for real money balances (Md) is a posiive funcion of real income (Y) and a negaive funcion of ineres rae (r) as depiced by he liquidiy preference funcion given in equaion. Md= f(y,r); fy>0,fr<0 Invenory Theory of Boumol Baumol (952) and Tobin (956) independenly developed similar demand for money models, which demonsraed ha even money balances held for ransacions purposes are sensiive o he level of ineres raes. The Baumol - Tobin model analyses he coss and benefis of holding money for ransacions purposes. The benefi is convenience and he cos of his convenience is he ineres forgone. A person can hus hold a porfolio of moneary asses and non-moneary asses.if r is he difference in he reurn beween moneary and non-moneary asses and b as he cos of ransferring non-moneary asses ino moneary asses, such as a brokerage fee, hen a person minimizes he sum of brokerage coss and ineres income forgone. This leads o a well-known square-roo formula given in equaion below. Md= (by/2r) This saes ha he demand for real money balances m is direcly proporional o ransacions coss b and real income Y, and inversely proporional o he ineres rae r. In he ideal world of he Baumol-Tobin model, he elasiciy of money demand in response o income and ineres rae mus be 0.5 and -0.5 respecively.

6 Friedman Theory In 956, Friedman developed he modern quaniy heory of money. He applied he heory of porfolio choice and posulaed ha he demand for money mus be influenced by he same facors ha influence he demand for any asse. Thus according o Friedman, he demand for money funcion is given by equaion below. (m/p) d = f(y p ;r b -r m ;r e -r m ;π e -r m ) where (M/P) d is he demand for real money balances, Y p is a measure of wealh or permanen income, r m is he expeced reurn on money, r b is he expeced reurn on bonds, r e is he expeced reurn on equiy, and π e is he expeced inflaion rae. EMPIRICAL LITERATURE REVIEW The empirical lieraure on demand for money is exensively available. The empirical invesigaion may shed some ligh on which specificaion is more likely o be beer. A number of researchers have been esimaed India's money demand funcion. Among hese he firs sudy was conduced by Moosa in 992. His sudy explicily considers he saionariy of, and coinegraion relaionships among, he variables of he money demand funcion. He used hree ypes of money supply, cash, M, and M2 o perform coinegraion ess on real money balances, shor-erm ineres raes, and indusrial producion over he period 972Q o 990Q4. Resuls indicaed ha for all hree ypes of money supply, he money balance had a coinegraing relaionship wih oupu and ineres raes. Bahmani-Oskooee and Rehman (2005) analyzed he money demand funcions for India and six oher Asian counries during he period of 972Q-2000Q4. Using he ARDL approach hey performed coinegraion ess on real money supplies, indusrial producion, inflaion raes, and exchange raes (in erms of US dollar). For India, coinegraing relaionships were deeced when money supply was defined as M, so hey concluded ha M is he appropriae money supply definiion o use in seing moneary policy. Das and Mandal (2000) considered only he M3 money supply in saing ha India's money demand funcion is sable. They used monhly daa for he period of April 98 o March 998 o perform coinegraion ess and deeced coinegraing vecors among money balance, indusrial producion, shor-erm ineres raes, wholesale prices, share prices, and real effecive exchange raes. Their posiion, herefore, was ha long-erm money demand relevan o M3 is sable.

7 Parvez Azim, Nisar Ahmed,Sami Ullah,Bedi-uz-Zaman.Muhammad Zakaria (200) esimaed he demand for money in Pakisan for he period 973 o 2007 using Auoregressive Disribued Lag (ARDL) approach o coinegraion analysis. The empirical resuls shows ha here is a unique coinegraed long-run relaionship among M2 moneary aggregae, income, inflaion and exchange rae. The income elasiciy and inflaion coefficiens are posiive while he exchange rae elasiciy is negaive. The resuls show ha income and inflaion variables are posiively associaed wih money demand while exchange rae negaively affecs money demand. AL-Abdulrazag Bashier and Abdullah Dahlan (20) made an aemp o examine he money demand funcion and is sabiliy in Jordan over he period by using Johansen-Juselius Coinegraion es and VAR. Their empirical findings sress he exisence of a posiive relaionship beween money aggregaes and he level of income while he relaionship is negaive for exchange rae. DATA AND METHODOLOGY In order o esimae he demand for money in India, he following daa are used. The daa used in his sudy are cumulaed from various secondary sources. The variable such as Broad money (M3), nominal Gross domesic produc, wholesale price index (WPI), call money rae, `-$ bilaeral exchange rae are colleced from various Reserve Bank of India bullein. The daa colleced over a period of o The WPI is colleced on he basis of consan prices, whereas nominal GDP is on consan prices. To invesigae he above issue he sudy uses he 40 observaions. In order o esimae he demand for money funcion in India, we considered five variables, namely M3 (Nominal money), nominal GDP, WPI, call money rae and bilaeral exchange rae beween rupee and dollar. The saisical and ime series properies of each and every variable are examined using he convenional uni roo es and employed coinegraion and error correcion mehod. ECONOMETRIC MODEL There is a diverse specrum of money demand heories emphasizing he ransacions, speculaive, precauionary consideraions. All he heories share common imporan variable. The general agreemen in he lieraure is ha a money demand equaion should conain a scale variable o he level of ransacions in he economy and a variable represening he opporuniy cos of holding money. In he conex of an open economy, a variable such as

8 exchange rae can be included in he money demand equaion o reflec he impac of currency depreciaion on money demand. Specificaion of he Model The general specificaion begins wih he following funcional relaionship for he demand for money: M =f(s,oc,x) The demand for nominal balances M is a funcion of he chosen scale variable(s) o represen he economic aciviy and he opporuniy cos of holding money (OC) and exchange rae. Alhough here are several funcional forms of specifying money demand funcion, here is general consensus ha he log linear version is he mos appropriae funcional form because i performs beer han he oher forms because he log linear form allows for inerpreaion of coefficiens of variables in logarihms as elasiciies. We sar wih a sandard money demand funcion in which nominal money balances are expressed as a funcion of nominal income, price level, ineres rae and exchange rae. We expec he esimae of income is expeced o be posiive; an esimae of price level, ineres rae and exchange rae are expeced o be negaive. For esimaion purposes, we use he logarihmic ransformaion of annual daa for he period 970:7 2009:0. We specify he following money demand equaion, where all variables are expressed in logarihmic forms, U is a random error erm, and is a annual ime index. ln (M) =α+ 0 lny + lnp + 2 ln r + 3 lnx +u () M= Nominal money Y= Nominal gross domesic produc ( base year prices) P= wholesale price index ( base year prices) r= (+Call money rae) X= rupee- dollar bilaeral exchange rae U= error erm

9 ECONOMETRIC METHODOLOGY AND EMPIRICAL RESULTS Uni Roo Tes The firs sep of he sraegy of our empirical analysis involves deermining he order of inegraion. Mos ime series are rended and herefore in mos cases are nonsaionary. The problem wih nonsaionary or rended daa is ha he sandard OLS regression procedure can easily lead o incorrec conclusion. A series of Augmened Dickey-Fuller uni roo es is performed o deermine he order of inegraion of he variables. Table () shows he ADF es resuls for boh a he level and he firs difference on inercep and inercep and rend. Table () Inercep only Inercep and rend Variables Level Firs difference Level Firs difference Prob: value Prob: value Prob: value Prob: value ln M (0) (0) (0) (0) ln Y.0000(0) (0) (0) (0) ln P 0.86(0) 0.003(0) (0) (0) ln r 0.484(0) (0) (0) (0) Ln x (0) (0) 0.783(0) (0) (Numbers in parenhesis are he number of lags) The repored resul in able () reveals ha he hypohesis of a uni roo can be rejeced in all variables in levels. However, he hypohesis of a uni roo is rejeced in firs differences a 0.05 level of significan which indicaes ha all variables are inegraed of degree one, I(). Tha means all he variables achieve saionariy only afer firs difference. The esimaion of he equaion by direc OLS gives he following inegraion equaion.

10 M= y p r -0.44x (2) ( ) (.64363) ( ) ( ) ( ) (0.000) (0.000) (0.000) (0.840) (0.0248) Adj R 2 = F= DW= The esimaed parameers of equaion are in accordance wih economic heory. Ineres rae and exchange rae have negaive parameers while nominal income and price level has posiive elasiciy. All coefficiens are saisically significan a 0.05 % level excep ineres rae. Here we have high R 2 and -values, bu U is no whie noise. Ofen i is idenified wih R 2 > D-W saisic. All he variables give he expeced resul, bu he nonsaionariy of variable biased he previous esimaion, and he low value of DW compared o R 2 can be inerpreed as sign of spurious regression. Selecion of Lag Lengh The crierion for selecing he lag lengh consis an imporan sep. There are differen ess ha would indicae he opimal number of lags. The sudy uilizes he SC crierion o ensure sufficien power of he Johansen procedure. Table (2) Lag AIC SC * * (VAR lag order selecion crieria included observaion 37) Coinegraion The nex sep in our empirical analysis is o es for coinegraion. Since he variables are considered o be I(), he coinegraion mehod is appropriae o esimae he long run demand for money. The concep of coinegraion is ha non-saionary ime series are coinegraed if a linear combinaion of hese variables is saionary. The coinegraion requires he error erm in he long-run relaion o be saionary. Suppose here are wo variable Y ad X and boh Y and X follows I () process, Sill he linear combinaion U =Y - αx is I (0). If so, boh Y and X are said o be coinegraed and a is he coinegraing

11 parameer. The maximum likelihood approach o es for coinegraion is based on he following sysem of equaions p x = π x + π x + ε i= The number of independen coinegraing vecor is equal o he rank of marix π, If rank of π = 0; hen π is a null marix and equaion urns ou o be a VAR model, whereas If rank of π =, here is one coinegraing vecor and π x - is an error correcion erm. Johansen suggess ha i can be done by esing he significance of characerizes roos of π. i i Suppose ha π is a 3x3 marix and he ordered characerisics roos are λ > λ 2 > λ 3 If rank of π = 0 hen λ i = 0; hence, ln(- λ i ) = 0 whereas, If rank of π = uniy hen 0 < λ < and ln( λ ) will be negaive and he res ln(- λ 2 ) = ln(- λ 3 ) = 0 Johansen suggess wo es saisics o es he null hypohesis ha numbers of characerisics roos are insignificanly differen from uniy. λ λ race max n ( r) = T ln( ˆ λi ) i= r+ ( r, r + ) = T ln( ˆ λ λ i = esimaed characerisic roos or Eigen values T = he number of usable observaions λ race es he null hypohesis r = 0 agains he alernaive of r > 0 λ max es he null hypohesis r = 0 agains he alernaive of r = The heory assers ha here exiss a linear combinaion of his non-saionary ha is saionary. Solving for he error erm, we can rewrie he relaion as e =α- 0 lny - lnp - 2 ln r - 3 lnx (3) Since {e } mus be saionary, i follows ha he linear combinaion of inegraed variables given by he righ hand side of mus also be saionary. r+ )

12 Johannsen Coinegraion Resul Table (3) Sample adjused Unresriced coinegraion Rank es (Trace) Null Eigen Trace saisics 5 percen Porb.** hypohesis Value criical value r=0 * r * r r r Unresriced coinegraion Rank es (Maximum Eigenvalue) Null Eigen Max-Eigenvalue 5 percen Porb.** hypohesis) Value saisics criical value r=0* r r r r (* denoes he rejecion of he hypohesis a he 0.05 level. And ** are Mackinnon-Hauge- Michelis (999) p-values.) The above able shows ha he null hypohesis of no coinegraion is rejeced a he convenional level (0.05) and he sudy conclude ha here exiss a relaionship among he proposed variables in he long run. Trace es indicaes ha here are wo coinegraiong vecor is here, whereas Eigen value es indicaes ha here is a leas one linear combinaion in he long run.

13 Table (4) The coinegraion equaion is depiced in above able which reveals ha he income and price has a posiive effec on he demand for money. On he oher hand, ineres rae and exchange rae has a negaive. The income elasiciy is.98 and showing significan, implying ha in India, a one percen economic growh requires around.98 percen increase in he naion s money supply. Ineres rae and exchange rae carries expeced negaive and significan coefficien. The Dynamic Shor Run Relaionship (VECM) By specifying he long run demand for money in an error correcion model, he shor run as well as he long run effecs of all righ hand side variables in equaion are esimaed in one sep, which is a major advanage ha error correcion modeling has in comparison o oher esimaion. The dynamic relaionship includes he lagged value of he residual from he coinegraing regression (ε - ) in addiion o he firs difference of variables which appear in he righ hand side of he long run relaionship (y, p, r and x). The inclusion of he variables from he long run relaionship would capure shor run dynamics. To sar, we define he error correcion erm by e =α- 0 lny - lnp - 2 ln r - 3 lnx (4) 0,, 2 and 3 are coinegraing coefficien e = he error from a regression of M on y, p, r and x. The ECM simply defined as X X r r p p y y m m X r p y a m X X r p y a m r X r p y a m p X r p y a m y X r p y a m m ε λ ε λ ε λ ε λ ε λ + = + = + = + = + = ) ( ) ( ) ( ) ( ) ( Normalized coinegraion coefficiens lnm lny lnp lnr lnx (0.5342) (0.0896) (0.075) (0.060)

14 Where, he elemens of ε s are whie noise errors and λ s are speed of adjusmen parameers and α and are shor run parameers. All he variable in he ECM are saionary, and herefore, he ECM has no problem of spurious regression. Table (5) Error correcion D(M) D(Y) D(P) D(r) D(X) Coin Eq Sandard error ( ) (0.0677) ( ) ( ) (0.262) saisics [ ] [ ] [ ] [ ] [-2.865] The above able shows he speed of adjusmen coefficiens, which reveals ha only hree variables are adjusing. The adjusmen coefficien on coinegraion equaion for he price is posiive, bu quie rapid 9% per year. The adjusmen coefficien for ineres rae is showing negaive, as i should no be, bu boh adjusing coefficien are showing significan. Similarly adjusmen coefficien for exchange rae is showing negaive, as i should be. Bu he esimaed error correcion model enjoys a very low goodness of fi (R 2 = , adj R 2 = ). The empirical sudy is performed by using PC version of Eviews 6.0. CONCLUSION The sudy used five variables exracing 40 annual observaions from 970 o Since all he variables have uni roo a levels he sudy uilizes Johansen-juselius coinegraion analysis o es for he exisence of a long run relaionship beween he variables. The coinegraing regression so far considers only he long-run propery of he model, and does no deal wih he shor-run dynamics explicily. For his, he error correcion from he long rum money demand is hen used as a dynamic model o esimae he demand money demand. The race es indicaes ha here are wo coinegraiong vecor is here, whereas Eigen value es indicaes ha here is a leas one linear combinaion in he long run. The sudy concluded ha he income and price has a posiive effec on he demand for money. On he oher hand, ineres rae and exchange rae has a negaive. The income elasiciy is.98 and showing significan, implying ha in India, a one percen economic growh requires around.98 percen increase in he naion s money supply. Ineres rae and exchange rae carries expeced negaive and significan coefficien.

15 REFERENCES. Al-Abdulrazag Basheer and Abdullah Dahlan(20), The Money Demand Funcion for Jordan: An Empirical Invesigaion, Inernaional Journal of Business and Social Science March 5, 20, Vol Bahmani-Oskooee, M and Rehman Hafez (2005) Sabiliy of he Money Demand Funcion in Asian Developing Counries, Applied Economics, 37, Damodar N Gujarahi, Basic Economerics, fourh ediion 4. Das, S and K Mandal (2000) Modeling Money Demand in India: Tesing Weak, Srong & Super Erogeneiy. Indian Economic Review, 35, Dimiorios Aseriou, Applied Economerics, A Modern Approach Using Eviews And Microfi, Palgrave Macmillan Publicaion, Juan J. Dolado a, Jesús Gonzalo and Francesc Marmol Coinegraion, February, Liang Xueping, Analysis of money demand 8. Moosa, I. A. (992). The Demand for Money in India: A Coinegraion Approach, The Indian Economic Journal 40(): Parvez Azim, Nisar Ahmed,Sami Ullah,Bedi-uz-Zaman.Muhammad Zakaria Demand for Money in Pakisan: an Ardle approach, Global Journal of Managemen and Business Research Page 76 Vol. 0 Issue 9,December 200