Dcevloovershooting多恩布什超调模型.docx

Dcevloovershooting多恩布什超调模型.docx

《Dcevloovershooting多恩布什超调模型.docx》由会员分享，可在线阅读，更多相关《Dcevloovershooting多恩布什超调模型.docx（29页珍藏版）》请在冰点文库上搜索。

Dcevloovershooting多恩布什超调模型

|_

~

吾尝终日而思矣，不如须臾之所学也；吾尝而望矣，不如登高之博见也。

－－《荀子·劝学》

Dornbusch'sOvershootingModelAfterTwenty-FiveYears

SecondAnnualIMFResearchConference

Mundell-FlemingLecture

KennethRogoff1

EconomicCounselorandDirectoroftheIMFResearchDepartment

November30,2001（revisedJanuary22,2002）

I.Introduction

ItisagreathonortopaytributeheretooneofthemostinfluentialpaperswritteninthefieldofInternationalEconomicssinceWorldWarII.RudigerDornbusch'smasterpiece,"ExpectationsandExchangeRateDynamics"waspublishedtwenty-fiveyearsagointheJournalofPoliticalEconomy,in1976.The"overshooting"paper-aseveryonecallsit-marksthebirthofmoderninternationalmacroeconomics.ThereislittlequestionthatDornbusch'srationalexpectationsreformulationoftheMundell-Flemingmodelextendedthelatter'slifeforanothertwenty-fiveyears,keepingitintheforefrontofpracticalpolicyanalysis.

Thislectureisdividedintothreeparts.First,Iwilltrytoconveytothereaderasenseofwhy"ExpectationsandExchangeRateDynamics"hasbeensoinfluential.Mygoalhereisnotsomuchtoofferacomprehensiveliteraturesurvey,thoughofcoursetherehastobesomeofthat.Rather,Ihopethereaderwillgainanappreciationofthepaper'senormousstatureinthefieldandwhysomuchexcitementhasalwayssurroundedit.Tothatend,IhavealsoincludedsomematerialonlifeinDornbusch'sMITclassroom.Thesecondpartofthelectureisamoredetacheddiscussionoftheempiricalevidenceforandagainstthemodel,andathumbnailsketchofthemodelitself.Thefinalsectiontouchesoncompetingnotionsofovershooting.

II.TheOvershootingModelinPerspective

OneofthefirstwordsthatcomestomindindescribingDornbusch'sovershootingpaperis"elegant".Policyeconomistsareunderstandablycynicalaboutacademics'preoccupationwiththeoreticalelegance.ButDornbusch'sworkisaperfectillustrationofwhythesearchforabstractbeautycansometimesyieldalargepracticalpayoff.ItispreciselythebeautyandclarityofDornbusch'sanalysisthathasmadeitsoflexibleanduseful.Likegreatliterature,Dornbusch（1976）canbeappreciatedatmanylevels.Policymakerscanappreciateitsinsightswithoutreferencetoextensivemathematics;graduatestudentsandadvancedresearchersfoundwithinitarichlodeofsubtleties.

Asecondwordtodescribetheworkis"pathbreaking".Iwilloffersomequantitativeevidencelater,butsufficetosayherethatliterallyscoresofPh.D.theses（includingmyown）havebuiltuponDornbusch（1976）.ItisnothyperboletosaythatDornbusch'snewviewoffloatingexchangeratesreinvigoratedafieldthatwasonitswaytobecomingmoribund,usingonlydated,discreditedmodelsandmethods.Dornbusch（1976）inspiredfreshthinkingandbroughtinfreshfacesintothefield.Inpreparingthislecture,Ire-readMauriceObstfeld'ssuperbinauguralMundell-Fleminglecturefromlastyear（IMFStaffPapers,Vol.47,2001）.Obstfeld'spaperspansthewholemodernhistoryofinternationalmacroeconomics,fromMeadeto"NewOpenEconomyMacroeconomics",butthemainemphasisisonBobMundell'spapers.I,andperhapsmanyotherreaders,foundObstfeld'sdiscussionenlighteninginpartbecausewedonothavethesameintimateknowledgeofMundell'spapersthatwedoofDornbusch（1976）.Mundell'sprofoundlyoriginalideasare,ofcourse,atthecoreofmanythingswedoinmoderninternationalfinance,andhewastheteacherofmanyimportantfiguresinthefieldincludingMichaelMussa,JacobFrenkel,andRudigerDornbusch.MundellisacreativegiantwhowasthinkingaboutasinglecurrencyinEuropebackwhenintergalactictradeseemedlikeamorerealistictopicforresearch.ButthemethodsandmodelsinMundell'spapersarenowbadlydated,andarenotalwayseasytodigestfortoday'sreader（evenifatthetimetheyseemedapictureofclaritycomparedtotheexistingstateoftheart,Meade（1951））.OneoftheremarkablefeaturesofDornbusch'spaperisthattoday'sgraduatestudentscanstilleasilyreaditintheoriginaland,asIwilldocument,manystilldo.

Thereadershouldunderstandthatasnovelastheovershootingmodelwas,Dornbuschwashardlywritinginavacuum.JoAnnaGray（1976）,StanleyFischer（1977）,andNedPhelpsandJohnTaylor（1977）wereallworkingonclosedeconomysticky-pricerationalexpectationsmodelsataroundthesametime.StanleyBlack（1973）hadalreadyintroducedrationalexpectationstointernationalmacroeconomics.Dornbusch'sChicagoclassmateMichaelMussa（mypredecessorasEconomicCounsellorattheFund）wasalsoworkingactivelyintheareainthetime,thoughhedelayedpublicationofhismainpieceonthetopicuntilMussa（1982）.TherewereotherswhowerefishinginthesamewatersasDornbuschataroundthesametime,（e.g.,HansGenbergandHenrykKierzkowski,1979）.ButtheeleganceandclarityofDornbusch'smodel,anditsobviousandimmediatepolicyrelevance,putshispaperinaseparateclassfromtheotherinternationalmacroeconomicspapersofitstime.

--------------------------------------------------------------------------------

A.StillaUsefulPolicyTool

AwordaboutNewOpenEconomyMacroeconomics,whichObstfeldsurveyedlastyear;certainlythisliteraturehascometodominatetheacademicliteratureoninternationalmacroeconomicpolicy.2Superficially,ofcourse,mostofthenewergenerationmodelsappearquitedifferentfromDornbusch'smodel,notleastbecausetheyintroducerigorousmicrofoundationsforconsumerandinvestorbehavior.Atthesametime,however,theycanbeviewedasdirectdescendants.Formally,NewOpenEconomyMacroeconomicsattemptstomarrytheempiricalsensibilityofthesticky-priceDornbuschmodelwiththeelegantbutunrealistic"intertemporalapproachtothecurrentaccount".3

Butevenwiththeinevitableonslaughtofmoremodernapproaches,theDornbuschmodelisstillverymuchalivetodayonitsown,preciselybecauseitissoclear,simpleandelegant.Let'sbehonest.Ifoneisinapinchandneedsaquickresponsetoaquestionabouthowmonetarypolicymightaffecttheexchangerate,mostofuswillstillwanttocheckanyansweragainstDornbusch'smodel.

Dornbusch'svariantoftheMundell-Flemingpaperisnotjustaboutovershooting.Thegeneralapproachhasbeenappliedtoahostofdifferentproblems,includingthe"Dutchdisease,"thechoiceofexchangerateregime,commoditypricevolatility,andtheanalysisofdisinflationindevelopingcountries.Itisaframeworkforthinkingaboutinternationalmonetarypolicy,notsimplyamodelforunderstandingexchangerates.Butwhatsoldthepapertopolicymakers,whatstillsellsthepapertograduatestudents,isovershooting.OnehastorealizethatatthetimeDornbuschwaswriting,theworldhadjustmadethetransitionfromfixedtoflexibleexchangerates,andnoonereallyunderstoodwhatwasgoingon.ContrarytoFriedman's（1953）rosydepictionoflifeunderfloating,exchangeratechangesdidnotturnouttosmoothlymirrorinternationalinflationdifferentials.Instead,theywereanorderofmagnitudemorevolatile,farmorevolatilethanmostexpertshadguessedtheywouldbe.AlongcomesDornbuschwholaysoutanincrediblysimpletheorythatshowedhow,withstickyprices,instabilityinmonetarypolicy-andmonetarypolicywasparticularlyunstableduringthemid-1970s-couldbetheculprit,andtoafargreaterdegreethananyonehadimagined.Dornbusch'sexplanationshockedanddelightedresearchersbecauseheshowedhowovershootingdidnotnecessarilygrowoutofmyopiaorherdbehaviorinmarkets.Rather,exchangeratevolatilitywasneededtotemporarilyequilibratethesysteminresponsetomonetaryshocks,becauseunderlyingnationalpricesadjustsoslowly.Itwasthisideathattookthepaperfrombeingamere"A"toan"A++".Asweshallsee,Dornbusch'sconjectureaboutwhyexchangeratesovershoothasprovenofrelativelylimitedvalueempirically,althoughaplausiblecasecanbemadethatitcapturestheeffectsofmajorturningpointsinmonetarypolicy.Butthetruestrengthofthemodelliesinthatithighlightshow,intoday'smoderneconomies,oneneedstothinkabouttheinteractionofsluggishlyadjustinggoodsmarketsandhyperactiveassetmarkets.Thisbroaderinsightcertainlystillliesatthecoreofmodernthinkingaboutexchangerates,evenifthedetailsofourmodelstodaydifferquiteabit.

PaulSamuelsononceremarkedthatthereareveryfewideasineconomicsthatareboth（a）trueand（b）,notobvious.Dornbusch'sovershootingpaperiscertainlyoneofthoserareideas.Now,ofcourse,unlessoneissteepedinrecenteconomictheory,littleofwhatappearsintoday'sprofessionaleconomicsjournalswillseemobvious.However,thatisonlybecauseittakesconstanttrainingandretoolingtobeabletofollowtheassumptionsinthelatestpapers.Onceyoucanunderstandtheassumptions,whatfollowsisusuallynotsosurprising.Butthisiscertainlynotthecasewiththe"overshooting"result,asIwillnowbrieflyillustrate.

B.Overshooting:

TheBasicIdea

Sincethislectureisaimedatabroadaudience,itisnotmyintentiontoinvoketoomanymathematicalformulas,thoughtherewillbeafew.Asmallnumberofequationsisnecessaryifonlytoimpressuponthereaderhowsimpletheconceptreallyis.Thereadercaneasilyskipoverthem.

Tworelationshipslieattheheartoftheovershootingresult.Thefirst,equation

（1）below,isthe"uncoveredinterestparity"condition.Itsaysthatthehomeinterestrateonbonds,i,mustequaltheforeigninterestratei*,plustheexpectedrateofdepreciationoftheexchangerate,Et（et+1-et）,whereeisthelogarithmofth