Inflaion Targeing Through Shor Term Ineres Rae: Ausralian Eperience Nisar Ahmed Samiul Parvez Ahmed Absrac: The aricle is based on empirical daa from Ausralian Economy. I made an aemp o evaluae he success of Reserve Bank of Ausralia (RBA) in conrolling inflaion by aggressively using cash rae. The main objecive of he research is o find ou he relaionship of RBA s moneary policy wih Taylor Rule; wheher i follows he Taylor Rule mechanically or uses a differen form (e.g. forward looking) or depends on oher facors as well. Oher models have been used for esimaion and he resuls are analyzed subsequenly. The evidences from he es resuls sugges he successful implemenaion of ineres rae rule by he Reserve Bank o sabilize he inflaionary pressure in erms of forward looking perspecive. 1.0 INTRODUCTION The moneary policy has evolved in he las few decades owards a new concep inflaion argeing. Many counries are already pracicing inflaion argeing as a ool for moneary policy. New Zealand was he firs counry ha inroduced inflaion argeing in 1990. Till hen i dispersed o oher counries, i.e. Canada, UK, Sweden, Ausralia, Finland ec. The basic idea of inflaion argeing is ha, cenral bank adjuss he shor erm ineres rae in such a way ha ensures he projecions of inflaion and oher variables saisfy he arge crierion. Though he objecive of inflaion argeing is o sabilize he inflaion around he arge, i also akes ino accoun he imporance of oupu gap by allocaing some weigh ono i. The inflaion argeing is based on he following characerisics: Numerical Inflaion Targe: Cenral bank usually specifies he arge as a range raher han poin esimaion. Inflaion-Forecasing-Targeing: The decision making process is saed as inflaion-forecasing-argeing. Transparency & Accounabiliy: The overall process mus be ransparen and cenral bank mus be accounable for he consequences. Usually, in he inflaion argeing regime, cenral bank makes public announcemens abou any changes in policy indicaing boh curren inflaion and epeced fuure inflaion saes and seps needed o keep inflaion in rack. The desired level of arge inflaion acually achieved gradually over a period han in one insance. The sabiliy is linked wih he concep Price Sabiliy, and his price sabiliy does no refer o zero inflaion raher i is closer o a low level (i.e. 2%) annual rae of price changes. The arge price level need no remain consan indefiniely, bu could be allowed o drif upward in a predeermined way over ime (Goodhar and Vinals, 1994; Svensson, 1996). Assisan Professor, American Inernaional Universiy, Bangladesh Lecurer, American Inernaional Universiy, Bangladesh 1
The inflaion argeing ignores he unanicipaed shocks and does no ry o offse; hus in he long run i migh resul in large variaion in price level. In conras, under sric price level argeing, he shocks are offse by couner policy implemenaions ha keeps he long run variances a minimal hough i creaes large swings in shor run. 1.1 The Moneary Policy Rule The cenral bank in an economy ses he shor erm ineres rae in response o economic f aciviy. The r = + gy + h( ) + r (see Appendi A for noaion) rule can be used o se he shor erm ineres rae once he parameers are deermined given real GDP and inflaion. Taylor s basic rule suggess ha: Easing money marke condiions in response o lower inflaion or declines in producion Tighening money marke condiions in response o higher inflaion or increase in producion In boh he cases he g and 1+h would be posiive As he size of he parameers is no saed; oo high or low value could resul in poor economic performance. Even hough moneary policy rule has been proven o be effecive based on empirical evidence, sill here are some argumens wheher policy rule or discreion will dominae he moneary decisions. Generally, policy rules are algebraic epressions enailing fied seings for he policy insrumens; whereas discreion is he seings for he insrumens of policy and are deermined from scrach each period wih no aemp o follow a reasonably well defined coningency plan for he fuure (Taylor, 1993). The general idea abou policy rule is he fied seings of he insrumens of moneary policy. Bu Taylor has presened his view for policy rule in differen way: The rules are responsive, calling for changes in he money supply, he moneary base, or he shor erm ineres rae in response o changes of he price level or real income. 1 Policymakers do no need o follow he policy rules mechanically. Policy rules do no provide enire eplanaions of he moneary issues. Such as, wheher a rise in he price is emporary or permanen requires analyzing several measures; i.e. consumer price inde, producer price inde, employmen cos inde, epecaion on inflaion, erm srucure of ineres raes, surveys ec. The quaniaive mehods aid o analyze hese issues more precisely bu only algebraic epression is no sufficien o eplain hese measures. Therefore, policy rule and discreion should be complemen o each oher no subsiue. The concep of policy rule is also changing. Economiss sugges ha policy rules should no be inerpreed as fied seings for he policy insrumens. Fleibiliy (i.e. feedback or responsiveness) should be par of he policy rules ha is policy rule also incorporaes judgmens. In ligh of view, a policy rule would include a nominal income rule in which he cenral bank akes acion o keep nominal income on arge, bu i would no include pure discreionary policy (Taylor, 1993). A simple eample of Taylor Rule ha how i works is as follows: n r = ( r + ) + λ ( ) + γ (see Appendi A for noaion) 1 Discreion versus policy rules in pracice John B. Taylor, 1993 2
ineres rae 12 11 10 9 Figure: I: Simple Taylor Rule Simple Taylor Rule Taylor Rule Ineres rae for 4.5% real rae 10.75% Targe Inflaion=Acual Inflaion=2.5% 9.5% 8 7 6 (2.5,7) 5 4 3 2 1 0 0 1 2 3 4 5 6 Assumpions: Long Run Real Ineres Rae r n = 4.5%, Targe Inflaion Rae = 2.5%, γ =1.5, γ =0.5. For simpliciy i is assumed ha oupu gap, =0 7 1.2 IMPLICATIONS OF THE PARAMETERS λ & γ X Inflaion Parameer: If, λ>1 λ >1 means ha i is he condiion for he ineres rae rule where inflaion and economic aciviy will be sabilizing, (given γ >0). I implies ha, if he inflaion rises above he arge level, hen he cenral bank mus increase he nominal ineres rae more han in line wih he increase in inflaion; so ha i will raise he real ineres rae. Therefore, i will dampen he real economic aciviy in he shor run, and subsequenly i will resul in decrease in inflaion rae. Therefore, he rade off beween inflaion conrol and oupu growh eiss in he shor run in his policy regime. Inflaion Parameer: If, λ <1 λ <1 means ha i is he condiion for he ineres rae rule where inflaion and economic aciviy will be desabilizing or i will accommodae he shocks of he economy. In his insance, hough he cenral bank increases he nominal rae in response o an epeced rise in inflaion, bu he increase in nominal rae is no enough o sop he real ineres rae from declining. In his, accommodaive regime, self fulfilling burs of inflaion and oupu may be possible (Bernanke and Woodford, 1996; Clarida e al., 1997). Oupu Deviaion Parameer: γ γ >0: Similar logic applies here as well. γ >0 means sabilizing condiion, (given λ >1). γ <0: Desabilizing condiion. γ =0: There is no oupu gap. Benchmark: λ =1 & γ =0 o evaluae differences in he esimaed policy rules. 1.3 FORWARD LOOKING MODEL Pure forward looking model can be defined as he one in which he bank s acion a any ime is condiioned solely upon hose aspecs of he sae of he world ha are relevan for forecasing he subsequen evoluion of he bank s arge variables--inflaion rae & aggregae economic aciviy (M.Woodford, 2000). Oher auhors also described forecased inflaion 3
argeing in oher ways. According o Lars E. O. Svensson (1999) inflaion forecasing argeing is an economic model which is used o generae condiional forecas pahs for he arge variables from he presen onwards associaed wih alernaive feasible policies. The cenral bank hus selecs he bes applicable opion depending on is crierion, i.e. arge fuure inflaion. Clarida, Gali, & Garler (2000) developed simple forward looking model based on Taylor s backward looking reacion funcion (1993). The applicabiliy of his rule depends on he degree of auonomy of a cenral bank over is economy. They assumed ha he domesic moneary policy is independen of any eernal consrains. The main operaing insrumen is he shor erm nominal ineres rae because by varying he nominal ineres rae he real ineres can be changed. Bu hen again, he relaionship beween shor erm ineres rae and he real economic aciviy is shor run as well. In he medium run, he oupu acually comes o is naural level and he policy resuls in only nominal changes in prices. Sill, he policy s effec on he shor run oupu in he economy acually limis cenral bank s policy managemen. As an eample, if cenral bank inends o reduce inflaion in he shor run hen i has o increase nominal shor erm ineres rae. As a resul his will reduce he oupu as well, depending on he nominal sickiness. Thus, he cenral bank has o consider hese issues while aking is policy decision. I is assumed ha in each period he cenral bank has a arge for he nominal shor erm ineres rae. 2.0 EMPIRICAL RESEARCH The empirical research is based on Ausralian economy. I mainly focuses on how Reserve Bank of Ausralia (RBA) eercises is moneary policies and how i conrols he inflaion and how i achieves he inflaion sabilizing. An overview of Ausralian moneary policy and RBA s operaions are discussed in he following secion. 2.1 OVERVIEW OF AUSTRALIAN MONETARY POLICY The objecives of moneary policy in Ausralia are formally developed in he Reserve Bank Ac (1959): 2 I. The sabiliy of currency of Ausralia II. Mainenance of full employmen in Ausralia III. The economic prosperiy and welfare of he people of Ausralia The firs objecive is relaed o inflaion argeing. The sabiliy of currency means he sabiliy of price ha is he equaliy in purchasing power Inflaion over he Long Run pariy. In order o achieve his objecive RBA sared o 10 arge inflaion. Ausralia sared is informal inflaion argeing as early as 1989. Reserve Bank of Ausralia (RBA) sared achieve low inflaion arge during 1989 (Sevens, 1999). In 1993, Bernie Fraser, hen governor of RBA, firs announced he inflaion 2 Percenage 8 6 4 2 0-2 86Q1 Formally Targe Inroduced in Mid-1993 87Q1 88Q1 89Q1 90Q1 91Q1 92Q1 93Q1 94Q1 Time Period 95Q1 96Q1 97Q1 98Q1 99Q1 00Q1 01Q1 02Q1 Moneary Policy Reserve Bank of Ausralia, hp://www.rba.gov.au/educaion/moneary _policy.hml 4
argeing as 2-3%. The inflaion arge was formalized in a join governmen-rba Saemen on he conduc of Moneary Policy in 1996. 3 The moneary policy in Ausralia is responsible for shor erm rade-offs beween inflaion and aciviy or employmen keeping in mind he policy objecive ha is meeing he arge on average over he course of an economic cycle. The rade-offs can be described as: A policy relaaion helps o simulae he economic aciviies and ulimaely i will increase inflaion and accommodae economic growh. In conras, ighening of policy will resul in slower economic growh and inflaion in he following period. Hence, i is eviden ha in he shor erm, policy makes a choice beween inflaion conrol and economic growh. 2.1.1 Implemenaion of Moneary Policy The moneary policy insrumen used by RBA is he shor erm ineres rae; in his case i is Cash Rae a marke ineres rae on overnigh funds charged by financial inermediaries. RBA operaes hrough he open marke operaions o eercise he changes. The cash rae is deermined in he money marke hrough he demand and supply equilibrium of he overnigh funds. The cash rae iself is highly relaed o oher shor erm ineres raes in he marke and hereby i fies all he shor erm ineres raes (money marke raes and key raes of banks & financial inermediaries) in overall economy. Thus, cash rae can be very powerful in conrolling he aggregae economy. The closeness of cash rae and oher shor erm raes (i.e. 90 days T- Bill) is given in he above diagram. Percenage 20 18 16 14 12 10 8 6 4 2 0 Shor Term Ineres Raes Cash Rae 86Q1 87Q1 88Q1 89Q1 90Q1 91Q1 92Q1 93Q1 94Q1 95Q1 96Q1 97Q1 98Q1 99Q1 00Q1 Time Period 90 Days T-Bill 2.1.2 RBA s Performance over he Inflaion Targeing Period Ausralia sared is inflaion argeing when he inflaion was already low. In 1993 he inflaion was jus over 2% afer he high inflaion of lae 80s. Is Inflaion & Cash Rae objecive was o lock he 20 inflaion a low level and consequenly form a framework 15 for epecaions. One of he 10 imporan feaures of RBA s policy is ha i ignores he unavoidable emporary flucuaions in inflaion. rcenage Pe 5 0 1986-5 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Cash Rae i Inflaion dp 3 Opimal Moneary Policy Rules and Inflaion Targes: Are Ausralia, Canada, and New Zealand Differen from he U.S.?-----Sean Collins and Pierre L. Siklos, March 2004. 5
The economic oucomes of he wo periods, period of inflaion argeing and previous period, have been given below 4 : Comparison of Inflaion during Targeing and Previous Period 1980-1992 1993-June 2000 Inflaion (Sandard Deviaion) 7.0 (2.4) 2.1 (0.6) GDP Growh (Sandard Deviaion) 2.8 (2.7) 4.4 (0.8) The main poins are: The inflaion was low during argeing period relaive o is previous period. Over his period he average inflaion were jus over 2%. The variabiliy of inflaion in argeing period was lower han he previous period as well. Bu ineresingly, he inflaion argeing did no hamper he oupu growh of he economy. The oupu growh averaged over 4% per annum since 1993. 2.2 OBJECTIVES OF THE RESEARCH The wo main objecives demonsraed in his research are as follows: 1. Idenificaion of he eisence of a relaionship beween Cash Rae wih inflaion and oupu gap by using simple Taylor Rule. 2. Idenificaion of he eisence of a relaionship beween Cash Rae and forecass of fuure inflaion and oupu gap along he lines suggesed by he Forward Looking Model. 2.3 METHODOLOGY 2.3.1 Daa The daa are quarerly ime series seasonally adjused daa. The daa have been colleced from he daa base called Daa Sream. For he model esimaion and he parameers esimaion, STATA program is used. Cash Rae is used as shor erm ineres rae. Cash Rae is he marke ineres rae on overnigh funds ha is direcly conrolled by Reserve Bank of Ausralia (RBA). Real Gross Domesic Produc (GDP) is used o measure oupu. To measure he oupu gap, he log of real GDP was de-rended using a quadraic rend. The real GDP is measured in 1997-98 prices. Consumer Price Inde (CPI) is used o measure he inflaion. The baseline inflaion measure is he rae of change of CPI beween wo subsequen quarers. The base year for CPI is 1990. The inflaion arge is inerpreed in erms of a measure of he overall CPI, a change o which he Treasurer and RBA agreed (1998). 5 2.3.2 Assumpions & Raionales The main operaing insrumen of moneary policy, used by RBA, is a shor erm ineres rae (cash rae). RBA can affec real aciviy in he shor run by changing he cash rae. RBA is assumed o have a arge for nominal shor erm ineres rae wihin each operaing period based on he sae of he economy. RBA is assumed o make one year-ahead forecas in inflaion as i seems good indicaor of he medium erm rend in inflaion. 4 Ausralia s Eperience wih Inflaion Targeing---Ian J. Macfarlane (2000) 5 Moneary Policy Reserve Bank of Ausralia, hp://www.rba.gov.au/educaion/moneary _policy.hml 6
RBA has is oal auonomy o conduc is operaion. The domesic moneary policy is assumed o be free of any eernal consrains. Echange rae shocks are ignored in he seing of policy as hey are epeced o have only emporary impac on inflaion. 2.3.3 Time Horizon The daa have been colleced for 71 quarers saring from 1985:1 o 2002:3. Afer calculaing he inflaion, 4 quarers daa were los. Therefore, Taylor Rule has been formed during 1986:1 o 2002:3. For dynamic model, wih one lag period in cash rae, assumed ha he RBA se cash rae considering las quarer s figure. Thus he cash rae was lagged by one quarer. In forward looking model i is assumed ha he RBA looks ahead one year for inflaion forecas and one quarer ahead of oupu gap. In addiion, in he model, he RBA is assumed o se is arge by adjusing o las one year s acual rae. 2.3.4 Implemenaion of Taylor Rule In secion 3.1, simple Taylor Rule has been esimaed and ploed agains he wo policy rules (1 & 2). The simple Taylor Rule akes he following form: [Noaions: Appendi A] n r = ( r + ) + λ ( ) + γ Taylor Rule 3.5 + 1.5λ + 0.5 ; =1986:Q1 o 2002:Q3..Policy Rule 1 3.5 + 1.5λ + 1.0 ; =1986:Q1 o 2002:Q3.Policy Rule 2 2.3.5 Model Esimaion In secion 3.2.1 saic model is esimaed. The saic model akes he following form: > = α + α + α ; Where α = r n + λ r 0 1 2 0 In secion 3.2.2 he dynamic model wih one quarer lag in cash rae is esimaed. The following model is used: > ρr + ( 1 ρ) α + (1 ρ) α + (1 ρ α r = 1 0 1 ) 2 In secion 3.2.3 he forward looking model is esimaed by using he following form: r 1 ρ){ α + γ E + γ E ρr, where α = r n + γ = ( + 1, k + 1, q} + 1 Finally, in secion 3.2.4, he backward looking model is esimaed by: r 1 ρ){ α + γ + γ ρr.. (7) = ( + 1, k + 1, q} + 1 Where, α = r n + γ and k = q = -ve [Noaion: Appendi A] 3.0 DATA ANALYSIS AND FINDINGS In he beginning of his secion, daa has been uilized o depic he simple Taylor Rule and is implicaions have been discussed. In he following secions four models (saic, dynamic, forward looking, & backward looking) have been esimaed and heir implicaions been discussed. 3.1 SIMPLE BASE LINE POLICY RULE: TAYLOR RULE Taylor Rule has been implemened in Ausralian economy aking he acual oupu gap and inflaion for period 1986:1 o 2002:3. γ is assumed o be 1.5 (sabilizing condiion) in boh Policy Rule 1 and 2, and he γ is assumed o be 0.5 in Policy Rule 1 and 1.0 in Policy Rule 2. r n = ( r + ) + λ ( ) + γ.base Rule 7
> r n = ( r + λ ) + γ + γ > r = (4.5 + 2 1.5 2) + 1.5 + 0. 5 Here, 2, γ = 1.5, γ = = 0.5, =1986:Q1 o 2002:Q3, Consan=3.5 calculaed aking r n =4.5%, is he sample average. > r = 3.5 + 1.5 + 0. 5...Policy Rule 1 Again, > r = (4.5 + 2 1.5 2) + 1.5 + 1. 0 Here, 2, γ = 1.5, γ = 1.0, =1986:Q1 o 2002:Q3 = r n =4.5%, is he sample average > r = 3.5 + 1.5 + 0. 5 > r = 3.5 + 1.5 + 1. 0.Policy Rule 2 3.2 IMPLICATION OF POLICY RULE VS CASH RATE Policy Rule 1: Policy Rule 2: 3.5 + 1.5λ + 0.5 ; =1986:Q1 o 2002:Q3 3.5 + 1.5λ + 1.0 ; =1986:Q1 o 2002:Q3 According o he policy rule, afer eperiencing high inflaion during he 80s, Reserve Bank should have increased he Cash Rae high enough o increase he real rae ha would ulimaely pull down he demand; as a resul economic growh would be slowed down and subsequenly he inflaion would fall. The rade-off beween inflaion sabiliy and economic growh always been an issue bu sill in inflaionary siuaion economic slow-down can be considered as viable macro-objecive; because in he long run i would creae a sound basis for growh. In he policy rule 1(a) and 2(a), he γ is assumed o be 1.5 and hereby based on ha, he desired level of cash rae been calculaed while considering he γ as 0.5 (in rule 1a) or 1.0 (in rule 2a). Under policy rule 1(a), in 2 nd quarer of 1988, according o he policy rule 1(a), he Cash Rae should have been 14.62% (poin A in figure 1) whereas he acual cash rae was 12.29%--way below he arge ineres rae. In conras, in 1990 (quarer 1) he cash rae was roughly he same (poin D in he figure 1) as required by policy rule 1(a) ha is Reserve Bank was aggressively conrolling he inflaion by seing higher cash rae. Furher down he line, during early-90s he cash rae was way above he policy rules. Despie he decreasing rend, he cash rae superceded he rae suggesed by he policy rule. Reserve Bank ook more aggressive moneary policy han suggesed by Taylor Rule. The responsiveness of he policy aken by Reserve Bank was higher. Objecive of Reserve Bank of Ausralia is o keep he inflaion rae beween 2 o 3% on an average, over he business cycle. 8
3.1.1 Figure (a) Cash Rae and Policy Rule 1 and 2: 1986:Q1 o 2002:Q3 Policy Rule1 Gpie=1.5, G-=0.5 Policy Rule2 Gpie=1.5, G-=1 Cash Rae 25 17.01% 14.62% 12.29% Percenage 20 15 10 A B Policy Rule 1 D Cash Rae Policy Rule 2 Gap beween Policy Rules & Cash Rae 5 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Time Period Eperienced High Inflaion Gradual Inflaion Sabilizaion Increase in Inflaion Lower Inflaion Period Oil Price Shock Phase 1 Phase 2 Phase 3 Phase 4 Phase 5 Informal Inflaion Targeing Sared Formal (eplici) Inflaion Targeing Sared Sared o overcome shor erm oil price shock 9
3.3 Chronological Eplanaions 6 Phase 1: Eperienced High Inflaion and Se High Cash Rae Major rises in he ineres rae occurred during 1981/82, 1985 and 1988/89, which were designed o reduce he inflaionary boom and resuled in slower growh in demand. Marke eperienced high inflaion during 1988 and 1989; money marke ineres rae increased by 7% since he early monhs of 1988 ill 1990 and subsequenly marke slowed down a bi due o reduced spending. The 3 rd /4 h quarer in 1989, he cash rae reached o is highes level around 18%. Phase 2: Gradual Reducion of Cash Rae o Reacivae he Economy Saring from he 1990 he Reserve Bank realized ha he economy was slowed down because of such high cash rae se in early period. Even in he 1 s quarer 1990, he cash rae was a one of is highes levels, 17.0133%. The objecive was o slowdown he economy in order o reduce he inflaion. The effec of his move was working ou hroughou he economy by slowing down he economic aciviies. Laer, Reserve Bank ook he oher way round booss he economy by reducing cash rae. In January 1990, Reserve Bank reduced he ineres and his reducion in he cash rae was epeced o flow hrough he securiy yields and ineres raes charged by he inermediaries. Bu economy was sill in inflaionary pressure. In February, 1990 Reserve Bank furher reduced he cash rae by 0.5% as he aggregae demand was decreasing. Inflaion remained a high level and in order o move o he righ direcion Reserve Bank hough o reduce he cash rae furher down. Thus, i kep on reducing he cash rae in Augus/Ocober/December roughly by one percenage poin respecively, aiming o reduce he inflaion in he medium run while avoiding ecessive slowdown in he economy. During ha ime, reail rade was fla, invesmen reduced, and employmen was also falling, impors declined, and credi growh also declined resuled from earlier higher rae. Reserve Bank mainained he rend of cash rae reducion during 1991 (May/Sepember/November) by 1 percenage poin. In lae 1991 he inflaion came down o around 1.5% which is lower han heir arge inflaion rae: 2--3%. In spie of he lower inflaion rae, Reserve Bank gradually kep on reducing he ineres rae because he economy didn recover he earlier higher ineres rae effec as i was epeced. The unemploymen rae was fairly high. The early period of 1993 followed he similar rend ha is he reducion in cash rae bu a a lower level. The economic aciviies were sronger han he earlier period. The Ausralian economy abou o ener ino he period of susained low inflaionary growh. Phase: 3: Periods of Cash Rae Increase: Forward Looking Approach In augus 1994, Reserve Bank increased cash rae by 0.75%, in Ocober i raised he cash rae by anoher 1%; hese increases were firs in almos over he las four years. Their policy was forward looking and herefore he major reason for he rae increase was o avoid overheaing of he economy and consequenly prolonging he period of susained growh. I was necessary o susain non-inflaionary growh as he economy was growing robusly. The economic recovery was already achieved and he oal oupu was abou 10% higher han is low poin in mid-1991. The pace of aciviy was acceleraed in he previous year and unemploymen rae decreased. Household spending increased han he forecased figure and aggregae credi was increased wih susained epansion in economic aciviy. The reducion in cash rae was o avoid any overheaing in economy. 6 Media Release by Reserve Bank of Ausralia. 10
The inflaion iself was in arge level 2%. A he end of he year (December) Reserve Bank furher increased he cash rae by 1% as he growh was sronger han epecaion. The implied consumer price inde was increasing a 2% rae. Phase 4 Periods of Relaively Lower Inflaion The 1997 was a period of economic growh wih low inflaion. In 1997, RBA reduced he ineres rae wo imes (May, June) and pulling i down o 5% from 6%. The forecased inflaion was less han 2%, hus RBA wen for easing of moneary policy. Acually, he economy was freed from he inflaionary pressure; he underlying consumer price inflaion was 1.7% (lower han heir arge) in mid year. In December 1998, RBA furher reduced he cash rae by 25 basis poins as he Ausralian economy was growing faser han he Bank or mos oher forcasers had epeced. Therefore, RBA increased he cash rae by 25 basis poins in November, 1999. Phase 5: Periods of Oil Price Increase, Flucuaing Economy & Effec of Sep 11 2000: Increase in Cash Rae: Ausralia s economy had sronger GDP growh and employmen growh which was higher han epecaion during 2000. CPI rose o 3.2 % over he year o June quarer due mainly o inernaional oil price increase and also domesic overheaing of economy. Epecaions of inflaions increased as well. The oil price increase seemed o be a emporary shock; herefore RBA had o deviae from is inflaion argeing. In 2000, RBA increased he cash rae o 6.25% from 5%. 2001: Decrease in Cash Rae: The picure changed in 2001. Inernaional economic condiions weakened due o he erroris aack on he USA on Sepember 11. Through he rading parners, i affeced Ausralian economy as well. Oil price declined sharply. There was downward pressure on inflaion. Equiy prices declined, corporae profi reduced and demand for credi by business reduced as well. Thus, RBA reduced he cash rae si imes hroughou he year and pulled down he cash rae o 4.25% in December 2001. 2002: Increase in Cash Rae The economic climae changed markedly in 2002. The economic condiions were beer boh in Ausralia and res of he world. There were upurn among Ausralia s main rading parners in he Eas Asia region (ecep Japan). Businesses were epanding heir invesmens, household spending increased and employmen increased as well. Inflaion reached 2.9% ha is close o he upper arge level. Thus, RBA increased he cash rae o 4.75% in mid 2002. 3.2 ESTIMATION OF FOUR MODELS In his secion four models have been esimaed and he resuls are discussed: 1. The Saic Model 2. Dynamic Model wih 1 Lag Period in Ineres Rae (Cash Rae) 3. Forward Looking Model 4. Pure Backward Looking Model 11
3.2.1 Saic Model The saic model considered he acual inflaion and oupu gap in each period o idenify he policy rule, hence calculaed heγ, γ and he consan. Table I: Reacion Funcion: Saic Model Parameers Value P Value Implicaion γ γ 1.3114 > 1 P=0.000 Sabilizing Condiion 0.2647 > 0 P=0.133 Saisically no significan α 0 3.532263 P=0.000 Saisically significan Eplanaions of he esimaion of he Saic Model The γ is significanly higher han one, γ =1.31. I means ha a rise in inflaion of one percenage poin induces he Reserve Bank o raise real raes by 31 basis poins. The coefficien γ is posiive bu saisically no significan. The consan is posiive and saisically significan. 3.2.2 Dynamic Model: Wih 1 Lag Period in Ineres Rae In dynamic model he ineres rae seing is assumed o be done wih a smoohing adjusmen. Acually in real siuaion he Reserve Bank does no wan o change he ineres oo differen from is previous period s rae, because of fear of disrupion of capial markes, loss of credibiliy from sudden large policy reversals ec. In his secion i is assumed ha Reserve Bank adjuss he acual rae o he arge in he following way: Table II: Reacion Funcion: Dynamic Model wih 1 Lag in Cash Rae Parameers Value Tes Value Implicaions γ γ 1.0478 > 1 P=0.8725 No significanly higher han 1 1.2841 > 0 =2.86 (H 0 : =0) Saisically significan ρ 0.8758647 SE: 0.0369322 Saisically significan α 0 3.1332943 =1.84 [H 0 : (1-ρ)α 0 =0] Saisically no significan The lag in cash rae is one quarer, i -1. We assume ha Reserve Bank of Ausralia looks behind 1 quarer of cash rae in order o define is policy. Eplanaions of he esimaion of he Dynamic Model wih 1 lag period in r Generally, ineres rae rule implies ha he sign of he response of he real rae arge o changes in epeced inflaion and oupu gap depends on wheher γ is greaer or less han one and on he sign of γ respecively. The resul depics ha a rise in he inflaion of one percen induces he Reserve Bank o raise real raes by 5 basis poins. The saisical resul saed ha γ is no higher han one, he null hypohesis could no be rejeced in his case ha γ is equals o one. Therefore, here γ could be equals o or lower han one. The coefficien γ is posiive and saisically significan. I means ha one percen rise in he 12
oupu gap induces he Reserve Bank o increase he nominal raes by 28 basis poins, aking inflaion as consan. Bu mos imporanly, as he γ is no saisically higher han one (or could be lower han one), i can be deduced from he empirical evidence ha RBA migh no follow his backward looking model while implemening heir moneary policy. 3.2.3 Forward Looking Model Reserve Bank of Ausralia believes ha he inflaion argeing has o be forward looking, as evidenced by he operaion of moneary policy since is incepion. However, a he same ime i believes ha he formulaion allows for boh forecasing and lags in he effecs of moneary policy in he economy. 7 Therefore, he policy rule followed by he Reserve Bank includes boh epecaions and also he pas inflaion eperiences. Thus, in he following forward looking model, boh epecaions and lagged inflaions have been esimaed and discussed. The Reserve Bank has arge for shor erm nominal ineres rae, ha is cash rae (he marke ineres rae on overnigh funds) in Ausralia, r. In he baseline case, he assumpion has been made ha he arge depends on boh epeced inflaion and oupu. Table III: Reacion Funcion: Forward Looking Model γ γ ρ α 0 Base Line 2.1374115 0.48431173 0.6074498 0.01132639 (C) p=0.0507 =0.74 (H 0 : =0) (SE: 0.0668521) =0.01 Adding: Lagged Inflaion a (D) 1.7293043 p=0.323 es _b[dp]=(1-_b[l.i]) 0.47082143 =0.83 (H 0 : =0) 0.5466824 (SE: 0.1457823) [H 0 : (1-ρ)α 0 =0] 0.63336941 =0.29 [H 0 : (1-ρ)α 0 =0] Esimae of for baseline specificaion, =3.95. The insrumens are -1, -2, -3, -4 and -1, -2, -3, -4 and i -4. Esimae of assumes ha long-run equilibrium real ineres rae is equal o sample average of r n =4.5. We assume ha Reserve Bank of Ausralia looks ahead 4 quarers for inflaion and 1 quarer for oupu gap ha is k=4, q=1. a -4, Eplanaions of he esimaion of he Forward Looking Model The op line of he Table III repors he resuls for he baseline specificaion. The mos imporan esimaion is he coefficien on he inflaion gap, γ : 2.14. The resul depics ha a rise in he epeced annual inflaion of one percen induces he Reserve Bank o raise real raes by 114 basis poins. The general rule saes ha in order o sabilise he inflaionary pressure he γ should be greaer han one, bu in his case he γ is high above one. The predicion of he Reserve Bank ha i increases he real rae in case of inflaionary pressure is saisically significan as long as γ is significanly greaer han one. The esimaion of he oupu gap is also ineresing. The coefficien γ is posiive. One percen rise in he oupu 7 Abou Moneary Policy Reserve Bank of Ausralia, hp://www.rba.gov.au/ Moneary Policy/abou_moneary_policy.hml. pg. 2. 13
gap induces he Reserve Bank o increase he nominal raes by 48 basis poins, aking inflaion consan. The ineres rae smoohing parameer ρ is saisically significan as well. The long run inflaion arge =3.95. The sample average real rae was aken as he esimaion of he long run real rae ha is 4.5 (appro.). Then he following epression was used o calculae he : n = ( r α)/( γ 1) The esimaed value, 3.95, acually is higher han he Reserve Bank s arge of 2-3%. Alernaive o he baseline specificaion has been considered as well. The resuls are menioned in he second row of able III. The lagged inflaion for four quarers has been included in he reacion funcion along wih he epeced inflaion and oupu. Afer esimaion wih lagged inflaion he γ becomes 1.73 bu i is no saisically significan. The ineres rae smoohing parameer ρ is saisically significan. And he coefficien γ is roughly unchanged. 3.2.4 Simple Backward Looking Model In his secion, backward looking model has been esimaed in order o find ou wheher he Reserve Bank follows he backward model for is policy rule or no. Two ime horizons are considered here by assuming: A. Reserve Bank follows las quarer s inflaion & oupu gap o ake is curren policy B. Reserve Bank follows las year s inflaion and las wo quarers oupu gap o ake is curren policy Table IV: Reacion Funcion: Pure Backward Looking Model γ γ ρ α A. k=q=-1; (l.i) B. k=-4, q=-2; (l4.i) 0.5805325 p=0.2820 0.6358965 p=0.0723 1.6490642 =2.76; (H 0 : =0) 0.9016162 (SE: 0.0418369) 4.2876505;=1.86 [H 0 : (1-ρ)α 0 =0] 1.4124478 0.5176188 4.6353185;=3.79 =4.61; (H 0 : =0) (SE: 0.1084923) [H 0 : (1-ρ)α 0 =0] The lags: A. -1 and -1, and i -1. B. -2 and -4, and i -4. Eplanaions of he esimaion of Backward Looking Model The γ in boh he cases is lower han one depicing he desabilizing condiion as oucome of moneary policy. Thus, i can be concluded from he above resul agains is moneary policy evidence, ha Reserve Bank does no follow pure backward looking model for is policy decision. 4.0 CONCLUSION Among he differen models, he forward looking model (secion 3.2.3, Cases: C & D), γ has, which is saisically higher han one. Thus, from he empirical evidence and in comparison wih he es resuls, i is eviden ha Reserve Bank of Ausralia successfully implemened ineres rae rule o sabilize he inflaionary pressure in erms of forward looking perspecive. The es values also sugges ha he value of should be higher han one. In conclusion, i can be deduced ha Reserve Bank of Ausralia has been fairly responsive o inflaion rae and his also confirms wih he empirical resul ha hey changed he cash rae aggressively o conrol he inflaion rae during is inflaion argeing period. γ 14
REFERENCES 1. John B. Taylor, (1998), An Hisorical Analysis of Moneary Policy Rules, NBER Working Paper Series 6768, Ocober 1998. 2. Richard Clarida, Jordi Gali, Mark Gerler. (2000). Moneary Policy Rules and Macroeconomic Sabiliy: Evidence and Some Theory. The Quarerly Journal of Economics, February 2000. 3. Richard Clarida, Jordi Gali, Mark Gerler. (1998). Moneary policy Rules in Pracice: Some Inernaional Evidence. Elsevier Science B.V. 1998. 4. Michael Woodford. The Taylor Rule and Opimal Moneary Policy. Working Paper, Princeon Universiy, Princeon. 5. Rober G. King, (2000), The New IS-LM Model: Language, Logic, and Limis, Federal Reserve Bank of Richmond Economic Quarerly, Vol. 86/3, Summer 2000. 6. Jordi Gali, (2001), New Perspecives on Moneary Policy, Inflaion, and he Business Cycle, Prepared for World Congress of he Economeric Sociey, Seale, Augus 11-16, 2000. 7. Lars E. O. Svensson, (2002), Inflaion Targeing: Should i be Modeled as an Insrumen Rule or a Targeing Rule?, NBER Working Paper 8925, May 2002. 8. Michael Woodford, (2000), Pifalls of Forward-Looking Moneary Policy, The American Economic Review, Vol.90, No.2, 100-1004, May 2000. Journal of he American Economic Associaion. 9. Ben S. Bernanke, Frederic S. Mishkin, (1997), Inflaion Targeing: A New Framework for Moneary policy?, The Journal of Economic Perspecives, Vol. 11, No.2, 97-116, Spring 1997. Journal of he American Economic Associaion. 10. Sean Collins, Pierre L. Siklos, (2004), Opimal Moneary Policy Rules and Inflaion Targes: Are Ausralia, Canada, and New Zealand Differen from he U.S.?, March 2004. 15
APPENDIX: A: NOTATIONS TAYLOR RULE: A. r = + gy + h( ) + r f r: Shor erm ineres rae, : Inflaion rae, : Targe Inflaion rae, y: % change in real oupu (Y) from rend, r f : Cenral bank s esimae of he equilibrium real rae of ineres, h: Amoun by which he cenral bank raises he e-pos real ineres rae (r-) in response o an increase in inflaion, g & (1+h): response coefficiens. n B. r = ( r + ) + λ ( ) + γ r n : Long Run Real Ineres Rae, : Targe Inflaion rae, λ : Inflaion Parameer, γ : Oupu Deviaion Parameer, : Inflaion a period, : Oupu gap in period FORWARD LOOKING MODEL: A.a.1: r = r + β E{ Ω } ) + γe{ Ω } (, k, q r : Targe Rae for nominal ineres rae in period, ß: Inflaion Parameer,,k : The percenage change in he price level beween periods and +k;,q: Average oupu gap beween period and +q, E: Epecaion operaor, : Targe Inflaion, r: The Desired nominal rae when boh inflaion and oupu a heir arge level, Ω : The informaion se a ime, γ: Oupu gap parameer. A.a.2: r + v = ρr 1 + ( 1 ρ) r r : Acual Fund Rae se by Reserve Bank; r : Targe Rae for nominal ineres rae in period ; ρ:degree of Smoohing of ineres rae changes; r -1: Ineres rae in period -1 (previous period). A.a.3: rr = r E {, k Ω } and rr = r rr : Real Rae Targe a period ; rr: Long run equilibrium real rae; E: Epecaion operaor; : Targe Inflaion; Ω : The informaion se a ime ;,k : The percenage change in he price level beween periods and +k. BACKWARD LOOKING MODEL r = 1 ){ rr ( β 1) + β, k + γ, q} + ρ( L) r 1 ( ρ + ε Where k = q = negaive r : Targe rae; ρ: Degree of Smoohing of ineres rae changes; r -1: Ineres rae in period -1 (previous period); ß: Inflaion Parameer;,k : The percenage change in he price level beween periods and +k;,q: Average oupu gap beween period and +q; γ: Oupu gap parameer 16