expandedkim

WORKINGPAPERSERIESExchangeRateConstraintsandMoneyControlinKoreaGuyhanKimWorkingPaper1995-011Ahttp://research.
stlouisfed.
org/wp/1995/95-011.
pdfFEDERALRESERVEBANKOFST.
LOUISResearchDivision411LocustStreetSt.
Louis,MO63102TheviewsexpressedarethoseoftheindividualauthorsanddonotnecessarilyreflectofficialpositionsoftheFederalReserveBankofSt.
Louis,theFederalReserveSystem,ortheBoardofGovernors.
FederalReserveBankofSt.
LouisWorkingPapersarepreliminarymaterialscirculatedtostimulatediscussionandcriticalcomment.
ReferencesinpublicationstoFederalReserveBankofSt.
LouisWorkingPapers(otherthananacknowledgmentthatthewriterhashadaccesstounpublishedmaterial)shouldbeclearedwiththeauthororauthors.
PhotocourtesyofTheGatewayArch,St.
Louis,MO.
www.
gatewayarch.
comEXCHANGERATECONSTRAINTSANDMONEYCONTROLINKOREAJuly1995JELCLASSIFICATION:E51,E58,F31,F32KEYWORDS:Capitalmobility,Moneycontrol,Offsetcoefficient,SterilizationpoliciesGyuhanKimVisitingScholar,FederalReserveBankofSt.
LouisandSeniorEconomistInstituteforMonetary&EconomicStudiesTheBankofKorea110,3-Ga,Namdaemun-Ro,Jung-GuSeoul,KoreaTheauthorwishestothankBillDewald,CletusCoughlin,BillGavin,PatriciaPollard,ChrisNeelyandMikePakkofortheirhelpfulcommentsandsuggestionsandRichardTaylorforresearchassistance.
AbstractThispaperappliestwotraditionalempiricalapproachestoinvestigatehowsuccessfullyKoreamanagedtocontrolmoneysupplyinresponsetothechangesinforeignreservessince1980.
Oneistoestimateoffsetcoefficientsandtheotheristoestimatesterilizationcoefficients.
Theestimationresultsareasfollows.
Reflectingstrictandeffectivecapitalcontrols,thedomesticmonetarypolicieswerepartiallyoffsetmainlybyshort-termcapitalflows.
TheBankofKoreapursuedsterilizationpoliciesveryactively.
Particularly,sterilizationwasmoreactiveduringthelate1980sofhugecurrentaccountsurplusthanduringtheearly1990sofasurgeincapitalflows.
Therewereover-sterilizationduringtheearly1980swhenforeignreserveskeptdropping,whichimpliedthatmonetarypoliciesweredirectedtoworseningthebalanceofpaymentsdeficitratherthanimprovingit.
I.
IntroductionInsmallopeneconomieswithexchangerateconstraintsandhighcapitalmobility,thescopeofmonetarycontrolisgreatlylimitedbytheoffsettingchangesinforeignexchangereservesmainlybycapitalflows,unlessthecentralbanksactivelysterilizethem.
Duringthe1980safteralongperiodoffixedexchangeratesystem,Koreaadoptedmultiplepeggedexchangeratesystemwhichallowedexchangeratestochangefrequentlybutmostlyunderthegovernmentcontrol.
Korea,therefore,hadtocontinuestrictcapitalcontrolsinordertoensuremonetarycontrol.
Theyexperiencedandarenowexperiencingbigpressureofmoneygrowthfromtheforeignsector:ahugecurrentaccountsurplusinthelate1980sandasurgeincapitalinflowsinthe1990s.
Koreaisnowspeedingupderegulatingcapitaltransactions,anditwilldamagethecontrollabilityofmoneysupply,unlesstheymovetowardfreefloatingexchangeratesystem.
ThisarticledealswithanissueofhowsuccessfullyKoreamanagedtocontrolmoneysupplyunderexchangerateconstraintssince1980.
First,theeffectivenessofcapitalcontrolsisexploredbyestimatingoffsetcoefficients.
Secondly,thesterilizationpoliciesoftheBankofKoreaareinvestigatedbyestimatingsterilizationcoefficients.
Thispaperproceedsasfollows.
Insectiontwo,ageneraldiscussionispresentedonthecontrollabilityofmoneysupplyunderexchangerateconstraints.
Insectionthree,theKoreanmonetaryexperiencesaredescribedinsomedetail.
Insectionfour,somepiecesofrelevantempiricalstudiesareundertaken,andthefinalsectionfiveisreservedforevaluationandprospectsasaconclusion.
1ILGeneralDiscussionTheswingsofmoneysupplythroughtheforeignsectorareunavoidableunderexchangerateconstraints,becauseexchangeratesdonotmovefreelytoremovethebalanceofpaymentsimbalance.
Thisgivesmuchtroubletothecentralbankofasmallopeneconomyinkeepingmoneysupplyundercontrol.
Particularly,capitalmarketlinkagesamongcountrieslimitsubstantiallythecontrollabilityofthenationalmonetaryauthoritiesoverthemoneysupplyandthedomesticinterestratesthroughoffsettingcapitalflows(Frenken,J.
A.
&M.
L.
Mussa[l981]).
Forexample,adecreaseinthedomesticcredittocombattheinflationarypressuremayincreasedomesticinterestratesorenforcefirmstofinanceabroad,andthuswillinduceacapitalinflowandcorrespondingincreaseinforeignexchangereservesthatwillrapidlyincreasethemoneysupplyback.
TheunfavorableeffectsoftheseoffsettingcapitalflowsonmoneycontrolwereempiricallyfoundinWestGermanyduringtheperiodoffixedexchangeratesystembetween1960and1972(Porter[1972],Kouri&Porter{1974],Neuman[1978],Obstfeld[l980]),andinsomeSouthAmericancountriessuchasColumbia,Mexico,Venezueladuringthe1970swhenthesecountriesranthecrawlingpegexchangeratesystem(Cumby&Obstfeld[1981],Kamas[1986],Rennhack&Bonangelino[1988]).
Duringthe1970stherewerelotsofresearchontheseoffsettingcapitalflowstoseehowfeasiblemonetarycontrolcanbeunderfixedexchangeratesystemandhighcapitalmobility.
Anassertionwasmadeonthebasisofthemonetaryapproachtothebalanceofpaymentsthatmonetarycontrolistotallyinfeasibleintheshortrunaswellasinthelongrunbecauseof2completeoffsettingcapitalflowsrespondingtothedomesticinterestratemovements(Mundell[1961],Swoboda[l973]).
Accordingtotheirviewanyattempttosterilizecapitalflowsthroughdomesticcreditmeasurescannotsucceed,eventemporarilybecauseitinducesanotheroffsettingmassivecapitalflowswithouttheinitialmoneyeffectsofmonetarypolicymeasures.
'Therewasadifferentviewthatmoneycontrolissomewhatfeasible,becausecapitalflowscannotoffsetdomesticmonetarypolicycompletelyinarealcasewherebondsdenominatedindifferentcurrenciesarenotperfectsubstitutes(Branson[1970],Obstfeld[1980]).
Theyhadadifferentviewonthecapitalmovements:capitalmovementsarenotsimplycontinuingflowsofforeignexchangebutlargelyonce-and-for-allstockadjustmentsofinvestorsportfoliosmadeinresponsetochangeininterestratedifferentials.
Theyinsistthatthesterilizationpoliciesofacentralbankmayinsulatedomesticmoneysupplyfromchangesinforeignexchangereservesatleastintheshortrunormediumrun.
But,inthelongrun,sterilizationcannotsustainamoneysupplythatdiffersfromtheequilibriumlevelofmoneydemand.
Fromthepracticalandtheoreticalprospective,wecanconcludethatwhetherasmallcountrycancontrolmoneysupplyproperlyunderexchangerateconstraintsdependsnotonlyonhoweffectivelyacountrycancontrolcapitalflowstopreventthemfromoffsettingchangesindomesticcreditbutalsoonhowactivelyacentralbanksterilizeschangesinforeignexchangereservesinaworldofhighlyintegratedcapitalmarkets.
~Somecriticsweremaderespecttothisassertion.
(1)ifwetakeoutsidemoneyintoconsideration,completeoffsetisnotpossibleunderperfectcapitalmobility(Fratianni[1977]),(2)completeoffsetbycapitalflowsarenotsufficientconditionforanemasculatedmonetarycontrol(Kanniainen[1977]).
3(Graph1):ExchangeRateMovementsinKoreaQuarterlyDataIndexCurrentAccount(MillionsU.
S.
$)160150-140-130120110-100-90-80-70-60IIIIIII198081828384858687888990919293941)Index,intermsofforeigncurrencies,calculatedbasedonexchangeratesJapan,England,Germany,Canada,Singapore,Taiwan)(1985=100).
2)Index,intermsofU.
S.
dollars(1985=100).
andCPIsofsevenmajortradingpartners(U.
S.
,~50004000CurrentAccountaNbminalExchangeRateS-4000-3000-2000-1000-0--1000--2000--3000—S.
RealeffectiveExchangeRate*III.
TheKoreanExperienceForeignExchangeandCapitalControlsInKorea,upuntil1990theexchangerateswerenotdeterminedinthemarketbutcontrolledbytheauthorities.
2Duringthistimetheexchangerateswerecontrolledtosatisfypurchasingpowerparityconditionbyandlargelesttheexportinggoodsshouldbeuncompetitiveintheworldmarketbytheunfavorableexchangeratesmovements.
Thiswasconsiderednecessarybecauseexportgrowthwasessentialforthecontinuedeconomicgrowth.
Astheforeignexchangemarketgrewrapidlyinsize,anewexchangeratesystem,so-calledmarket-averageexchangeratesystem,wasadoptedinMarch1990toallowmarketforcestodeterminethewon/dollarrateinthedomesticforeignexchangemarketwithinabandsetbytheauthorities.
3Reflectingtheserealities,themovementsofthenominalandrealexchangewerecloselyrelatedtocurrentaccountpositionasshowngraphl.
~Duringtheperiodofcurrentaccount2Fromearly1970suntilJanuary1980,KoreamaintainedapegtotheU.
S.
dollar.
Subsequently,itwasreplacedbythemultiple-basketpeggedsysteminwhichtheexchangeratesweresetbytheauthoritiesdailyonthebasisofabasketofcurrenciescomprisingatrade-weightedbasketandSDRbasketaswellasofjudgementsofauthoritiesontheothereconomicfactorssuchasdomesticandforeignpricetrends,andthebalanceofpaymentsposition.
Underthissystemtheexchangerateswerefloatingbutmanagedbytheauthoritiestokeepthematthedesiredlevel.
~Underthenewsystem,thedailyexchangerateisdeterminedbytheweightedaverageofwon-dollartransactionsconductedonthepreviousbusinessdayamongforeignexchangebankswithabandsetbytheauthorities.
Thisbandwaswidenedfrom0.
4%aboveorbelowtheprevailingrateto0.
6%inSeptember1991andwasagainenlargedto0.
8%inJuly1992andwasfurtherexpandedto1.
0%inOctober1993inordertoenhancetheroleoftheexchangerateasapricevariableandboostmarkettransactions.
ItisempiricallyassertedthatthechangesinnominalexchangerateshadcleareffectsonthetradebalanceastheoryindicatesinKorea.
4deficit(1980-1985),boththenominalandrealexchangeratescontinuedtorise(wondepreciated),whiletheydropped(wonappreciated)considerablyduringtheperiodofsurplus(1986-1989).
SinceKoreaopeneditsstockmarketpartiallyin1992,theyaregreatlyaffectedbythecapitalflows.
Undertheseexchangerateconstraints,Koreahadtoputstrictcontrolsonthecapitalflowsinordertoenhancemonetarycontrol.
5Untilearly1990swhenthecapitalmarketliberalizationproceededrapidly,mostofthecapitaltransactionstakingplaceoutofpurelyfinancialmotiveswereprohibitedandonlythosewhichhadsomeconnectionswithrealsectorwerepermittedwithlimits.
Reflectingthesecontrolsoncapitaltransactions,mostofthecapitaltransactionsduring1980swerereal-sector-relatedonessuchasloans,tradecredit,anddirectinvestmentasshowntable1.
However,since1990capitaltransactionsoutoffinancialmotivesintheformofequitiesanddebentureissuedsurgedenoughtogivetroubleincontrollingmoneystock.
Itisalsoremarkablethatboththecapitalcontrolsanditsrelaxations6havebeenpursuedasoneoftheeffortstoeffectivelycontrolmoneysupply.
Forexample,duringthelatterpartof1980swhenpersistentcurrentaccountsurpluswasabigpressureformoneygrowth,thecapitaloutflowswerederegulated,whilecapitalinflowwerekeptunderstrictcontrol,andviceversa.
5.
.
.
Theotheraimsofcapitalcontrolsare(I)tochannelforeignsavingtothestrategicallyimportantsectors,and(ii)toprotecttheunderdevelopeddomesticfinancialsectorsfromoutsidecompetition,(iii)tomakeefficientuseoflimitedforeigncurrencyfunds.
6Sincetheearly1980sanumberofmeasureshavebeentakentoremoveoreasethesecontrols.
Thegovernmentsubstantiallyloosenedtheforeignexchangeconcentrationsystemandtheceilingsoftheexchangepositionofforeignexchangebankshasbeenraisedseveraltimes.
Italsoabolishedorrelaxedtheregulationsoncapitaltransactionssuchasforeigninvestmentsandoverseasportfolioinvestment.
Noticeably,theKoreansecuritiesmarketwerepartiallyopenedinJanuary,1992byallowingforeignsecuritiescompaniestoownupto10%ofthepaid-incapitaloflargeKoreansecuritiescompanies.
5Table1:CapitalFlowTrendinKorea(milliondollars)1980-19841985-19891990-1994inflowoutflowinflowoutflowinflowoutflowloans!
)1754280771240223629425810583directinvestmentdomestic57412531031954524660abroad--34710168627217portfolioinvestmentdomestic799165217514374000512233abroad--10121083942010321importcredit59803579137106071047150410144407short-term58259562386923269670146377141147long-term154416751828137740333260exportcredit817690219165106352449420462short-term328730991545219625822876long-term48895922762084392191217586total2)8689475301992641090422339732058831)publicloans+commercialloans+bankloans2)Theotheritemswhichhavenetbalancesuchasadvancesunderred-clauseL/Careexcluded6ForeignReserveChangesandMoneyControlDuetothewideopeningofKoreaneconomytotheworld,combinedwithrelativelyinflexibleexchangerate,theforeignreserveshavebeenadominantsourceofthereservebaseinKoreaasshowningraph2.
Duringtheearly1980swhenKoreacontinuedtohavebalanceofpaymentdeficit,domesticmonetarypolicycouldenjoydegreeoffreedom,becauseforeignsectorwasanetreservedemandsector.
But,sincethelate1980sforeignreservesdevelopedhardshipsinkeepingmoneygrowthatthetargetrate,becauseitswitchedtoanetsupplysector.
7Ahugecurrentaccountsurplusbetween1986and1989wasabigpressureformoneygrowth,whichthreatenedtoaccelerateinflationandhadlongservedtoabsorbliquiditysincethen.
Theactivesterilizationpolicieswerepursuedtokeepmoneygrowthundercontrol.
Domesticcreditwasgreatlycontractedthroughasubstantialreductioninpolicyloansandatightercontroloverbankloanstolargefirms.
Sinceitwasnotenough,salesofmonetarystabilizationbondsbytheBankofKoreaexplodedandreserverequirementsratioswereincreasedmorethandoubletoabsorbtheliquidity.
Inadditiontothesemonetaryreactions,thegovernmentallowedthewontoappreciateagainstthedollar,restrictedsomecapitalinflows,andliberalizedsomecapitaloutflows.
Astherestrictionsoncapitalflowswereremarkablyliftedintheearly1990s,ieapartialopeningofstockmarketinJanuary1992,thesurgeincapitalinflowscontinuouslyexpands7Since1957Koreahasestablishedmonetarytargetingasthebasicmonetarypolicyregimeinordertoefficientlytackletheinflationarypressures.
TheannualtargetratesofmoneygrowtharesetmainlyonthebasisofrealGNPgrowthrateandtargetinflationrate.
From1979onwards,acertainpoint(recentlyacertainrange)ofannualgrowthrateofConsistingofcurrencyincirculationplustotaldepositsofbankinginstitutionshasbeenamainmonetarytarget.
7(Graph2):ReserveBaseanditsForeignSourceQuarterlyDataMillionsofWon30000-25000-20000-15000-10000-5000-0-IIIIIIIIIReserveBase/SS.
~IForeignSourceES——0~—I19808182838485868788899091929394foreignreservesinspiteofrecentconsecutivecurrentaccountdeficitasshowntable2.
Sincethishugecapitalinflowisputtingtremendouspressureforwonappreciation,thepurchaseofforeignexchangebythecentralbankisrequiredtokeepexchangerateatproperlevel.
Thisinterventioninthemarketnecessitatesthecentralbanktopracticesterilizationpoliciestokeepmoneygrowthattargetlevel.
Table2:BalanceofPaymentsTrendinKorea(net,milliondollars)1981-851986-8919901991199219931994currentaccountcapitalaccountlong-termshort-term-11,16233,686-2,179-8,728-4,528385-4,7787,980-12,9163,8824,2278,3436,8799,0868,511-13,9135484,1867,2338,9006,134-5319973,334411,110-2,0212,952overallaccount-7,60521,530-274-3,7414,8986,5422,802EveniftheBankofKoreastruggledtokeepmoneygrowthundercontrolagainsttheexternalpressureasdescribedabove,itisundeniablethatthemovementsofforeignexchangereserveshadsignificanteffectsonmoneygrowthandthusinflationasshowningraph3.
Duringtheearly1980sKoreacouldsustainlowinflationrateexcept1980and1981whenKoreahadthesecondoilshockthankstorelativelylowerrateofmonetarybasegrowth.
Butafter1986therateofmoneygrowth(M2)rosesubstantiallyduetothecontinuedhugecurrentaccountsurplus,whichputKoreaundertremendousinflationarypressureevenuntilearly1990s.
Therecentincreaseintherateofmonetarybaseisbelievedtobeassociatedwithasurgeincapitalflows.
8(Graph3):MoneyGrowthandInflationinKoreaMnualDataPercent50-40-30-20-10-0M2198081I~I1I~I~I8283848586878889IIII9091929394\RB\.
1-10~SInflation(CPI)IV.
EmpiricalFindingsInthissectiontwopiecesofempiricalstudiesarepresentedtoinvestigatehoweffectivelyKoreamanagedtocontrolmoneystockinresponsetobigchangesinforeignreservessince1980:bothoffsetcoefficientsandsterilizationcoefficientsareestimated.
(1)offsettingcapitalflowsOneoftheempiricalapproachestomeasurethecontrollabilityofthemoneystockistoestimatetheoffsetcoefficient:thefractionofagivenchangeinthedomesticcreditthatisoffsetbycapitalflows.
Traditionally,theoffsetcoefficientwasestimatedinthecontextofmonetaristorportfoliobalancemodelineitheroftwoways.
Thereducedcapital-flowequationisestimateddirectly;alternatively,thestructuralassetdemandequationsareestimated,andthen,theoffsetcoefficientiscalculatedfromtheestimatedinterest-sensitivities.
8Inthispaper,theoffsetcoefficientsareestimatedinareduced-formmodelofcapitalflowsdevelopedfromageneralequilibriummodeloftheKoreanfinancialmarkets.
Accommodatingthetheoreticallyderivedcapital-flowequationasmuchaspossible9,thefollowingcapital-flowequationissetupforestimation.
Modelderivationandsourcesofdata8Thereduced-formestimatestendtobebiasedtoward-1ifthecentralbankssystematicallyattempttosterilizereserveflows(KouriandPorter1974,Obstfeld1982)orunlessspeculatingcapitalflowsareproperlycapturedinthecapitalflowequation(Obstfeld1982,Laskar1982).
Thestructuralestimatesarebiasedtoward0ifprivateagentsinternalizethegovernmentbudgetconstraint(Pasula1994).
~Amongtheexplanatoryvariablesinthederivedcapitalflowsequation,thechangeincapitaltransactionsmotivatedbythecapitalmarketliberalizationiscapturedbyadummyvariablerepresentingpartialstockmarketopeningin1992.
1andthecapitalflowsaccompaniedbytradeisreflectedthecurrentaccount.
9usedinregressionsareprovidedinAppendices.
TC~=a0+a,L~.
NDA~+a2CA~+a3LI(i*+E)1+a4LXY~+a5LXW~+a6DUMMY+e~Here,TCiscapitalinflow,NDAisdomesticsourceofmonetarybase,CAiscurrentaccountsurplus,1*isforeignexchangerate,Eisexpectedrateofdepreciation,Yisrealincome,Wisnominalwealth,DUMMYisadummyvariablerepresentingtheperiodafterJanuary,1992,and~isadisturbanceterm.
Inordertofindoutsourcesofoffsettingcapitalflowsinmoredetail,someofthecombinationsofshort-termcapitalflows,long-termcapitalflows,anderror&omissionsareusedasdependentvariablesinthecapital-flowequation.
TheestimationresultsareprovidedinTable~Someimplicationscanbedrawnfromtheseestimationresults.
First,inspiteof10TheHausman'stestforexogeneityisappliedinordertodetecttheproblemofsimultaneousbiaswhichmaycomefromtheendogeneityofthedomesticcreditvariablei~NDAt.
ToimplementtheHausmantest,addto(1)thevariable~NDAt,whichisthepredictedvaluefromanOLSregressionof~NDAtontheinstrumentalvariables,includingtheotherregressorsin(1).
ThehypothesisofnosimultaneousbiasisequivalenttothehypothesisthatthecoefficientofANDAtiszero.
TheresultofOLSestimation(witht-statisticsinparentheses)isTCI-32.
72-0.
351ANDAt-0.
496CAe-159.
6A(I*+E)t+0.
452Ayt+0.
616AWt-0.
045ANAt+(1.
359)(0.
214)(3.
182)(0.
234)(3.
190)(5.
887)(1.
278)1102.
1DUMMYR2=0.
708,DW=1.
801(2.
140)ThehypothesisthatthecoefficientofANAtiszeroisnotrejected.
SoOLSorGLSregressionmethodsareappliedforestimation.
Asshownintabletheestimatedoffsettingcoefficientusing2SLSmethodmakeslittledifference.
10Table3:CapitalFlowEquation(1980:1to1994:4)constantZ~NDA~.
NDA-CAB~(i*+E)z~Y~WDummyR2DWz~RRSTCLTCTC1TC2OLS2SLS-0.
243(3.
532)-0.
053(0.
721)-0.
297(3.
194)-0.
377(4.
545)-51.
87(0.
559)41.
78(0.
421)-10.
09(0.
084)-107.
1(1.
004)-32.
72(0.
236)-0.
259(3.
663)-0.
107(1.
410)-0.
366(3.
977)-0.
437(5.
374)-0.
396(2.
224)-0.
104(2.
020)-0.
385(6.
954)-0.
490(7.
272)-0.
523(8.
132)-0.
495(5.
937)0.
2580.
7420.
7080.
70798.
22(1.
075)-258.
8(2.
648)-160.
6(1.
351)-70.
05(0.
623)-159.
6(1.
289)0.
025(1.
739)0.
019(1.
255)0.
044(2.
369)0.
032(1.
815)0.
045(2.
158)1.
8861.
9191.
8041.
8010.
073(2.
524)-0.
017(0.
560)0.
056(1.
481)0.
053(1.
514)0.
062(1.
371)-367.
9(1.
693)1499.
7(6.
446)1130.
8(4.
000)1090.
3(4.
164)1102.
1(3.
209)1)t-staticsareinparentheses.
TC1=STC+LTC,TC2=TC1+ERR(Graph4):DomesticandForeignInterestRateDifferentialQuarterlyData1)Corporatebondyieldrate(3Yearmaturity).
2)LIBO(3months)+expectedrateofdepreciation.
Percent19808182838485868788899091929394strictcapitalcontrolsinKorea,monetarypolicieswerepartiallyoffsetbycapitalflows,mainlytheshort-termcapitaltransactions.
Becausetheregulationsontheportfoliotransactionswereextremelystrict,theseoffsettingcapitalflowswerelargelyassociatedwithborrowingandrepaymentsbyfirmsornon-bankfinancialinstitutionsratherthaninterest-sensitivecapitalflowslikeportfolioinvestment.
Short-termborrowingaremainlyconnectedwithimporttransaction,whilelong-termborrowingarecomposedofbankloans,commercialloans,tradecredits,debenturesissued.
Theseoffsettingcapitaltransactionsweremadebylargefirmsinordertosubstitutedomesticsourcesoffundforforeignsourceswheneverloanavailabilityranoutdomesticallyintheeventoftightmonetarypolicies.
Moreover,theycouldtakecostadvantageoverdomesticloansbyborrowingmoneyabroad,becauseforeigninterestratesstayedmostlyfarlowerthandomesticinterestratesasshowningraph3.
Ascapitalmarketliberalizationproceedsrapidlyduring1990s,theportionofinterest-sensitivecapitalflowswhichoffsetmonetarypolicyissuspectedtorise.
Secondly,capitalflowsinKoreaaremoreassociatedwithmonetarypoliciesorcurrentaccountthanportfoliochoicebehavior.
Particularly,short-termcapitalflowswerelargelymotivatedbymonetaryconditions,andtradetransactions,whilelong-termcapitaltransactionswerelargelymadetooffsetcurrentaccountimbalances.
Thiscanbewellexplainedbytheestimationresultsthattheexplanatoryvariablessuchasforeigninterestrate,domesticincomeandwealtharemuchlesssignificantthanothervariableslikechangesindomesticsourceofmoneysupplyandcurrentaccount.
Thisreflectsthefactthattherewerestringentrestrictionsonportfolioadjustmentsinfinancialmarkets.
Thirdly,thefrequentchangesinreserverequirementratioduring1980swerenotdetectedtocauseoffsettingcapitalflows.
Thisimpliesthatreserverequirementpolicieswerenotsoeffectiveastoinfluencetheliquidityconditionsin11themarket.
(2)sterilizationpoliciesTheabilitytoneutralizechangesinforeignreservesoriginatinginpersistentbalanceofpaymentimbalancesorforeignexchangemarketinterventionisaprerequisiteforcontrollingmoneysupplyinnon-reserveeconomies.
Thus,manycentralbanksareundertakingactivesterilizationpoliciestomaintaintheircontrollabilityofmoneystock.
Sterilizationpoliciesaresystematicattemptsbythecentralbankstoneutralizethemonetaryimpactofchangesintheirforeignreservesbyoppositemovementsintheirholdingsofdomesticassets.
Thebasicwaywemeasurethedegreeofsterilizationbyacentralbankisbyestimatingamoney-supplyreactionfunctionwhichprovidesaformalstatementofthebehaviorofthemonetaryauthorities.
Thedegreeofsterilizationismeasuredbythesterilizationcoefficient-coefficientofnetforeignassetinthereactionfunction.
FollowingCumbyandObstfeld(198l),themoneysupplyreactionfunctionfortheBankofKoreaismodeledasafunctionofchangeinnetforeignassets,governmentdeficit,realeffectiveexchangerateandproductiongrowth.
~~SincetheBankofKoreaviewsthetransmissionofmonetarypolicyasdependingonM2ratherthanbankreserves,~NDA,isNFAandBUDDEFTarescaledbythepredetermined,requiredreserveratio.
Aftermultiplyingthroughbythepredetermined,requiredreserveratio(q~1),Iobtaintheexpressionforthemoneysupplyreactionfunctioninthenextpage(HerringandMarston(1977)).
123L~NDA~=a0+a1ExNFA1+a2BUDDEFT~+a3q~1REER~+a4q11GIPI~++cii=!
Here,NFAisnetforeignassetoftheBankofKorea,BUDDEFTisgovernmentdeficit,REERisrealeffectiveexchangerate,andGIPIisindustrialproductiongrowth,Siisseasonaldummies,qisreserverequirementratio,andcisadisturbanceterm.
Thecoefficienta1isthesterilizationcoefficient,whichmeasurestheextenttowhichtheBankofKoreaattemptstoneutralizethemoneycreationresultingfromchangeinitsforeignreservesthroughcountervailingdomesticmonetarymeasures.
Ifa1is0,thenthereisnosterilization,becausethebalanceofpaymentsleadtoaproportionateincreaseinthemoneysupply.
Ifa1is-1,thencompletesterilizationispracticed.
Valuesofa1between-1and0indicatepartialsterilization.
12First,themoney-supplyreactionfunctionisestimatedovertheperiodfrom1980:1to1994:1applyingGLSand2SLSmethod.
Then,itisestimatedseparatelyduringtwoperiodsofbalanceofpaymentdeficitandsurplustocomparethesterilizationbehavioroftheBankofKoreabetweenthem.
Theestimationresultsareprovidedintable4~1312Apositivevalueofa1wouldbeanevidencethatmonetarypolicyaimedatexternalbalanceratherthaninternalone.
13TheHausman'stestforexogenetyisalsoappliedtodetecttheproblemofsimultaneousbiasintheequation.
Thehypothesisofnosimultaneousbiasisnotrejected.
Asshownintable4,theestimatedsterilizationcoefficientsusing2SLSmethodmakesasmalldifference.
ANDA=258.
0-0.
756ANFA+0,307BUDDEFT+46.
77q,1REER,-920.
8qGIPI,+0.
196AN~A-(0.
627)(7.
502)(4.
400)(2.
523)(2.
512)(1.
497)1203.
1Si-196.
1S2-11.
58S3R2=0.
790,DW=2.
237,p=0.
565(1.
985)(0.
647)(0.
020)13Table4:MoneySupplyReactionFunction(1980:1to1994:4)periodconstant~NFA(1)z~NFA(2)BUDDEFTREER(T—l)GIPR(T-2)SlS2S3R2pD.
W80:1-94:4GLS441.
5-0.
640-0.
7010.
28246.
40-104.
6-1406-263-1690.
780-0.
572.
23(1.
11)(9.
82)(10.
40)(4.
11)(2.
48)(2.
80)(2.
34)(0.
87)(0.
28)2SLS267.
9-0.
5580.
30645.
92-942.
5-1192-161-50-0.
572.
09(0.
57)(6.
36)(3.
67)(2.
09)(1.
99)(1.
75)(0.
44)(0.
07)SURPLUS337.
4-0.
664-0.
7310.
30649.
07-910.
6-1080-148-150.
755-0.
572.
07(0.
81)(9.
05)(9.
70)(4.
27)(2.
48)(2.
37)(1.
72)(0.
47)(0.
02)DEFICIT-223.
9-2.
802-2.
860.
35544.
1-1188.
2-1643-75-2340.
5322.
27(0.
41)(4.
28)(4.
16)(3.
67)(1.
11)(2.
32)(2.
45)(0.
16)(0.
37)1)t-staticsareinparentheses.
2)Thecolunmsof~NFA(1)and~NFA(2)representestimatedcoefficientsofL~.
NFAwhenthedependentvariableisAB-I~TNFAandi~B-~NFA-E~RRrespectively.
(Graph5):TheBankofKorea:ChangesinDomesticandForeignAssetsQuarterlyDataBillionsofWon19808182838485868788899091929394*ForeigncurrencydepositsatdomesticbanksbytheBankofKoreaisincluded.
Allthecoefficientshavetheexpectedsigns.
TheestimationresultsprovidestrongevidencethattheBankofKoreapursuedapolicyofsystematicsterilizationaseasilyseeningraph4.
Duringtheperiodofbalanceofpaymentsdeficitmostlyfrom1980to1985,over-sterilizationpolicieswerepursued.
Theywereintendedtosupporttheeconomicgrowthbyarapidincreaseindomesticcrediteventhoughtightmonetarypolicieswerenecessarytocombatthebalanceofpaymentimbalance.
Duringthesurplusperiod,activesterilizationpolicieswerepracticedtoreducetheinflationpressureacceleratedbythesurgeinmoneysupplyfromtheforeignsector.
Inadditiontothis,theextendtowhichtheBankofKoreaattemptedtosterilizetheincreaseinforeignreservesiscomparedbetweenduringtheperiodofcurrentaccountsurplusandtherecentperiodofcapitalaccountsurplususingadummyvariable.
zXNDA1=285.
7-0.
516ANFA1-0.
219A(NFAD)1+0.
341BUDDEFT1+(0.
687)(5.
013)(1.
942)(4.
786)28.
51q11REER11-811.
86q~.
.
1GIPI1,2-771.
9Si+31.
13S2+191.
1S3(1.
316)(2.
161)(1.
177)(0.
100)(0.
297)R2=0.
780,DW=2.
228,p=-0.
570TheaboveresultshowsthattheBankofKoreapursuedmoreactivesterilizationpoliciesduringtheperiodofcurrentaccountsurplusinthelate1980sthanrecentperiodofcapitalaccountsurplus.
Thisresultcanbeexplainedbythefactsthatduringrecentperiod(1)sterilizationpolicyinstrumentssuchasmonetarystabilizationbondwhichweremostfrequentlyusedareseverelylimitedduetotheburdenofinterestpayments(2)exchangeratesarebecomingmoreflexible14enoughtoabsorbtheexternalmonetarydisturbances(3)changesinforeignreservesaremorefrequentsothatitishardtoneutralizethemeffectivelyandconsistently.
15V.
ConclusionUndertheexchangerateconstraints,themoreeffectivethecapitalcontrolsandthemoreactivethesterilizationpoliciesare,themoreeffectivelyacentralbankcancontrolmoneysupply.
ThispaperappliedtwotraditionalempiricalapproachestoinvestigatehoweffectivelyKoreamanagedtocontrolmoneysupplyinresponsetothechangesinforeignreservessince1980s.
Oneistoestimateoffsetcoefficientswhichmeasuretheeffectivenessofcapitalcontrolsandtheotheristoestimatesterilizationcoefficientswhichinformhowactivelythemonetarypoliciesaredirectedtooffsetthechangesinforeignreservestokeepmoneysupplyundercontrol.
Theestimationresultsareasfollows.
SinceKoreaundertookcapitalcontrolsstrictlyandeffectively,onlysomeofcapitaltransactionssuchastradecreditsandloanswereabletooffsetdomesticmonetarypoliciestosomedegree.
Moreover,theBankofKoreapursuedsterilizationpoliciesveryactively.
Particularly,sterilizationwasstrongerandalmostcompleteduringlate1980sofhugecurrentaccountsurplusthanearlyl990sofasurgeincapitalflows.
Therewereover-sterilizationsduringearly1980swhenforeignreserveskeptdropping,whichimpliedthatmonetarypoliciesweredirectedtoworseningthebalanceofpaymentsdeficitratherthanimprovingit.
Since1990stheinterestratesandcapitaltransactionshasbeenderegulatedveryrapidlyinKorea,itisessentialforKoreatomovequicklytowardmoreflexibleexchangeratesystemtomaintaincontrollingmoneysupplytoadesirableextent.
16AppendixIDerivationofreduced-formcapitalflowequationSupposethattherearefourfinancialassetsinasmallopeneconomy;money,governmentbond,foreignfinancialassets,andotherdomesticfinancialassetsownedbyforeigners.
Andthiseconomyiscomposedofthreesectorsi.
e,government,privatesector,andforeignsector.
Thegovernmentsectorincludesthecentralbank,whilethebankingsectorisapartoftheprivatesector.
Thiseconomyundertakescapitalcontrolwhichlimitsthekindandamountoffinancialassetsthatprivatesectorobtainsandsuppliesabroad.
Thus,itisassumedthattheprivatesectorisallowedtoownsomeportion(oc0)oftheamountofforeignassetsthatotherwisetheywouldhave,andtosupplysomeportion(~35)ofdomesticfinancialassetsthatotherwisetheywouldsupply.
Besides,foreignersarenotallowedtoowndomesticcurrencyandgovernmentbond,butcanownequityandclaimsonloanssuchasbankloans,tradecredits,commercialloans,andsoon.
Thenthegeneralequilibriumportfoliobalancemodelofthiseconomyisexpressedasfollows.
ThedemandforreservebaseisgivenbyM~=pL(I,y,W)(1)Thedemandforgovernmentbond(BG)isBGd=F(I,y,W)(2)17Therealizeddemandforforeignfinancialassets(B*)isB*d=aJ(I,I*+E,y,W)(3)Therealizedsupplyofdomesticfinancialassetstoforeigners(BF)isBFs=~3H(I,f+E,y)(4)Here,Misreservebase,Iisdomesticinterestrate,pisdomesticprice,1*isforeigninterestrate,Bisexpectedrateofdepreciation,yisdomesticrealincome,andWisdomesticnominalwealth.
Inaddition,wehavefouridentityequationstocompletethemodel.
=NFA+NDA(5)ANFA=TC+CA(6)TC=ABFs-AB*d-k(TB)(7)W=M+B0+B*(8)Here,NFAisnetforeignassetofthecentralbank,NDAisnetdomesticassetofthecentralbank,TCiscapitalinflow,CAiscurrentaccountsurplusandkwhichisafunctionoftradeaccountsurplus(TB)istrade-connectedcapitalflowssuchastradecredits,advancesunderred-clauseL/C,andsoon.
Assuminginfinityelasticityofforeigndemandfordomesticfinancialassets,wegettothefollowingreduced-formcapitalinflowequationthroughcomparativestatic18analysiswithequilibriumconditionsineachassetmarket.
11~-aJ.
~iXi=ANDA+CA+i+EL~-~3H.
+aJ.
L~-~H~+czJ~L~-~H1+aJ~-L~+aJ~~J+LiXW+(A~H-AaJ)--kL~-13H+aJ.
L~-~H~+aJ~L,-I3H~+aJ~~H-aJ.
~H.
—aJ.
(RH.
*-aJ.
*)(~H.
—aJ.
)TC=IILINDA+I1CA+Ii~J~'11LI(i*+E)L~-~H.
+aJ.
L~—~H~+aJ1L~—~H,+aJ~(L+aJ)(13H-aJ.
)~J+LL.
-[3'3'11+J]Ay-[WW+J]LIW+I(LI~H-AaJ)L1-~H.
+~xJ.
~'L1-~+aJ.
WL1-+aJ.
L.
-____kL~-~H~+aJ~Thecapitalflowequationderivedaboveshowsthatcapitaltransactionsaremotivatedbydomesticmonetarypolicy(ANDA),currentaccountsituation(CA),portfoliobehavioroftheprivatesector(A(I*+E)4y4W),capitalmarketliberalization(L43H-LsaJ),andtrade.
Appendix2NotesonthedataThisappendixdescribesthedataseriesunderlingtheestimatespresentedinthispaper.
Weemploythefollowingabbreviations:19MBBOK=MonthlyBulletinoftheBankofKoreaIFS=InternationalFinanceStatisticsFFK=FlowsofFundsinKorea(1)capitalflowequationSTC:surplusonshort-termcapitalaccount,inmillionsofdollars.
Source:MBBOKLTC:surplusonlong-termcapitalaccount,inmillionsofdollars.
Source:MBBOKERR:errorsandomissions,inmillionsofdollars.
Source:MBBOKCA:currentaccountsurplus,inmillionsofdollars.
Source:MBBOKB:high-poweredmoney,inbillionsofwon.
Source:MBBOKNFA:netforeignassetsoftheBankofKorea,inbillionsofwonincludingDOME.
Source:MBBOKDOME:depositoftheBankofKoreaatdomesticbanksinforeigncurrencywhichiscountedasoneofthenetdomesticassetsoftheBankofKorea,inbillionsofwon.
Source:MBBOK.
NDA:netdomesticassetsoftheBankofKorea,inbillionsofwon.
CalculatedasAB-ANFA.
Source:MBBOKq:reserverequirementratio(quarterlyaverage).
Source:MBBOKL:totaldepositsofdepositmoneybanks,inbillionsofwon(quarterlyaverage).
Source:MBBOK20ARR:changeinreserverequirements.
Calculatedas(q1-q~.
.
1)*L1~11*:LondonInterbankOfferRatesonUSdollardeposits(3monthmaturity).
Source:IFSB:expecteddepreciationrateofwon.
Generatedfromthefollowingestimatedequation.
Source:MBBOKB=0.
004-0.
512CA/B1R2=0.
526,DW=1.
828(2.
29)(7.
89)y:realGrossNationalProductinbillionsofwon(1985=100)Source:MBBOKW:nominalwealthinbillionsofwon.
Theseriesisthesumofgovernmentbonds,money(M2)andforeignassetownedbyprivatesectorincludingbankingsector.
Source:FFK(2)monetarypolicyreactionfunctionBUDDEFT:governmentbudgetdeficit,inbillionsofwon.
Source:MBBOKGIPI:growthrateofindustrialproductionduringthepreviousquarter.
CalculatedusingindustrialproductionindexinKorea(l990=lOO).
Source:MBBOKREER:realeffectiveexchangerateindexinKorea(l985=100).
CalculatedbasedonexchangeratesandCPIsofsevenmajortradingpartners(U.
S.
,Japan,England,Germany,Canada,Singapore,Taiwan).
21REFERENCEAmas,Linda,"TheBalanceofPaymentsOffsettoMonetaryPolicy:Monetarist,PortfolioBalance,andQuinicineEstimatesforMexicoandVenezuela,"JournalofMoney,Credit,andBanking(November1986),PP.
467-481.
Fratianni,Michele,"ANoteontheTheoryofOffsettingCapitalFlows,'JournalofMonetaryEconomics3(1977),PP.
133-138.
Frenkel,JacobA.
andMichaelL.
Mussa,"MonetaryandFiscalPoliciesinanOpenEconomy,"ABAPapersandProceedings(May1981),PP.
253-258.
Frenkel,JeffryA.
,"LiberalizationofKorea'sForeignExchangeMarkets,"PacificBasinWorkingPaperNo.
PB92-08(March1992),FederalReserveBankofSanFrancisco.
Gros,Daniel,"TheEffectivenessofCapitalControls,"IMFStaffPapers,PP.
621-642.
Hagen,JurgenVon,"MonetaryTargetingwithExchangeRateConstraints:TheBundesbankinthel980s,"FederalReserveBankofST.
LouisReview(September/October1989),PP.
53-69.
Henderson,DaleW.
,"ModelingtheInterdependenceofNationalMoneyandCapitalMarkets,"AmericanEconomicReview(February1977),PP.
190-199.
Herring,RichardJ.
andRichardC.
Marston,"SterilizationPolicy:theTrade-offbetweenMonetaryAutomyandControloverForeignExchangeReserves,"EuropeanEconomicReview10(1977),pp.
325-343.
Kanniainen,Vesa,"OnOffsettingCapitalFlowsandMonetaryAutonomyofaSmallOpen22Economy,"Economica(May1984),PP.
177-186.
Kouri,PenttiJ.
K.
andPorter,MichaelG.
,"InternationalCapitalFlowsandPortfolioEquilibrium,"JournalofPoliticalEconomy(December1974),PP.
21-39.
Laskar,DanielM.
"Short-runIndependenceofMonetaryPolicyunderaPegged-Exchange-RatesSystem:AnEconometricApproach,"JournalofInternationalMoneyandFinance(April1982),PP.
57-79.
Lee,ChungSikandEungJinKim,"MonetaryManagementPoliciesandProceduresinKorea,"unpublished(December1994)Lindner,DeborahJ.
,"ForeignExchangePolicy,MonetaryPolicy,andCapitalMarketLiberalizationinKorea,"InternationalFinanceDiscussionPapersNo.
435(August1992),BoardofGovernorsoftheFederalReserveSystem.
Moreno,Ramon,"ExchangeRatePolicyandInsulationfromExternalShocks:TheExperiencesofTaiwanandKorea-1970-1990,"PacificBasinWorkingPaperNo.
PB93-05(March1993),FederalReserveBankofSanFrancisco.
Obstfeld,Maurice,"CanWeSterilizeTheoryandEvidence,"ABAPapersandProceedings(May182),PP.
45-50.
SterilizationandtheOffsettingCapitalMovements:EvidencefromWestGermany,1960-1970,"NBERWorkingPaperNo.
494(June1980).
Obstfeld,MauriceandRobertE.
Cumby,"CapitalMobilityandtheScopeforSterilization:Mexicointhe1970s,"NBERWorkingPaperNo.
770(September1981).
22Pasula,Kit,"Sterilizatoin,RicardianEquivalenceandStructuralandReduced-FormEstimatesoftheOffsetCoefficient,"JournalofMacroeconomics(Fall1994),PP.
683-699.
Porter,MichaelG.
,"CapitalFlowsasanOffsettoMonetarypolicy:TheGermanExperience,"IMFStaffPaper(July1972),PP.
395-424.
Roubini,Nouriel,"OffsetandSterilizationunderFixedExchangeRateswithanOptimizingCentralBank,"NBERWorkingPaperNo.
2777(November1988).
TheBankofKorea,FinancialLiberalizationandInternationalizationinKorea(May1994).
FinancialSystemandMonetaryPolicyinKorea(October1989).
ForeignExchangeSysteminKorea(March1991)MonetaryPolicyanditsInstrumentsinKorea(April1993).
23