speeches · August 5, 1999

Speech

William Poole · President

Synching, Not Sinking, the Markets MeetingofthePhiladelphiaCouncilforBusinessEconomics FederalReserveBankofPhiladelphia Philadelphia,Pennsylvania August6,1999 Ihavelongbeenfascinatedwiththestudyof ofpolicy“perfection”isthatthemarketsandthe market responses to policy actions. Prices centralbankhavethesameinformationset, in speculative markets respond sensitively incompletethoughitmaybe.Informationis to events, including policy actions and incompletebecausethefutureisuncertainand hints about policy actions, of all kinds. The mar- thedisciplineofeconomicsisunsureaboutmany kets I am referring to include the equity, bond, importantrelationships. commodityandforeignexchangemarkets.Policy I’mdeliberatelyusingtheword“perfect” actionsincludethosebytheFederalReserve,the ratherthantheword“optimal”becauseIwant executive branch of the federal government, the toavoidtheimplicationthatthesubjectathand Congress, courts and many other governmental isanoptimalcontrolissue,whichwouldinvolve bodies. Although today I will discuss this topic suchmattersasapossibletrade-offbetween in the context of Federal Reserve policy actions, employmentstabilityandpricestability.Iam I mention the wide array of markets and govern- assumingthatsocietyhassomehowmadeits mental units to emphasize that the subject is a choicesaccordingtotheobjectivesofmonetary very broad one indeed. policyandhowtheyaretobepursuedwithin Myinterestinmarketresponsestopolicy theconstraintsofhowtheeconomyworks.The actionshasonlybeenheightenedbymystill-new terminologyisabitawkward;informationis jobasSt.LouisFedpresident.Ihavefoundmyself inherentlyimperfect,butmytopicconcernsdiffer- puzzlingovermarketreactionstoFedpolicyand encesininformationheldbythecentralbankand speculationaboutFedpolicy.Iamcertainly themarkets.Bypolicy“perfection,”Imeanthat acutelyawareofthefactthatwhatFedofficials thedifferencesininformationhavedisappeared— saycanmovemarkets.Theissueformeishowto boththecentralbankandmarketshavethesame understandmarketresponsesatadeeperlevel, imperfectsetofinformation.I’llkeep“perfection” becauseIthinkimprovedunderstandingwill strengthenmonetarypolicyandreducemarket inquotationmarkstohelpmakeclearthatthe volatility. issueistheequalityofimperfectinformation Myquestionisthis:Whatwouldweexpect betweenthecentralbankandthemarkets. toobserveinspeculativemarketsifmonetary Theassumptionthatmonetarypolicyis policywereworking“perfectly”?Idonotmean “perfect”impliescertainthingsabouthowmarkets “perfectly”insomeutopiansense,butwithinthe shouldbehave.Iwillconcentrateonthebehavior contextoftheworldofincompleteinformation offinancialmarkets,andwilltakeasgivensthe inwhichweactuallylive.Weknowthatpolicy policygoalssoIcanconcentrateonthe“perfec- problemsmayarisewhenthecentralbankand tion”ofinformationbetweenthemarketsandthe themarketshavedifferentinformationsets,or centralbank.Thistopicisalargeone;mypur- havedifferentperceptionsaboutthestateofthe poseistointroducemymodel,explainhowthis economyorthedirectionofpolicy.Myconcept modelcanhelpusunderstandhowpolicyhas 1 MONETARYPOLICYANDINFLATION evolved,suggestwaysinwhichdeparturesfrom Reserveactionsorhadpossibleimplicationsfor “perfection”indicateproblemswithmyunder- futureFederalReserveactions: lyingmodelandwaystoimprovepolicy.I’llonly • Sunday,August15,1971.PresidentNixon beabletotouchuponafewoftheseareas. closedthegoldwindowandannounced Herearethekeyquestionsmyanalysis comprehensivewageandpricecontrols. addresses:Undera“perfect”monetarypolicy, Thefollowingday,the20-yearTreasury wouldweobservelargemarketresponsestopolicy bondyieldfellby18basispointsandthe actions?And,wouldweobservelargemarket DowJonesIndustrialsroseby3.8percent. responsestoinnovations,orsurprises,ineco- • Saturday,October6,1979.TheFederal nomicdata?Iwillarguethatundera“perfect” Reserveintroducedadramaticsetofnew monetarypolicy,weshouldnotexpecttoobserve policies.Whenthemarketsreopenedthe anymarketreactionstoFedpolicyactions.Every followingTuesday(Mondaywasthe Fedpolicyactionshouldbecompletelyantici- ColumbusDayholiday),the30-year patedbythetimeitoccurs,andthereforeshould Treasurybondyieldroseby25basispoints, beanoneventinthemarkets. andtheDowJonesIndustrialssankby4.5 I’llproceedbyfirstlistingafewexamplesof percent. dramaticpolicychangesthatdidhavemajor • Sunday,September22,1985.TheG5coun- effectsonmarketprices.ThenI’llbrieflydiscuss triesannouncedthePlazaAgreement, theefficientmarketsmodelthatservesasthebase- whichcalledforthesecountriestopursue lineexplanationofpricedeterminationinspecu- policiestodepreciatetheforeignexchange lativemarkets.Becausemytopicconcernsthe valueoftheU.S.dollar.Thetrade-weighted interactionsofthemarketsandthepolicymakers, dollarindexfellby3.5percentthenextday, I’llnextdiscussFederalReserveobjectivesand the30-yearTreasurybondyieldrose6basis policyimplementation.Puttingthemarketand points,andtheDowIndustrialsrose1.4 Fedpolicytogether,myclaimisthatpolicy“per- percent. fection”requirestheFedandthemarketstoreact thesamewaytoarrivinginformation.Inthiscase, Thesethreeexamplesreflectdramaticpolicy themarketandtheFedareinsynch;themarket changes.Icouldofferhundredsofothersmaller anticipatesFedpolicyactionsandisnotsurprised andmoreroutineexamples.Ihaverefrained bythem.Largechangesinmarketpricestopolicy fromlistingexamplesofmarketresponsesto actionsobviouslyindicatemarketsurprises.In nonpolicyinformation,suchastheemployment suchcases,policychangescansinkthemarkets, reportthatarrivesonthefirstFridayofevery orsendthemintoorbit. month,butIamsurethateveryonehereisfamil- Beforeproceeding,Iwanttoemphasizethat iar—perhapsalltoofamiliar—withsuchmarket theviewsIexpressherearemineanddonot responses.Marketsreacttodatareleasesinpart necessarilyreflectofficialpositionsoftheFederal becauseofthebeliefthatthecentralbankwill respondtotheinformation. ReserveSystem.Ithankmycolleaguesatthe Whatdowemakeofthesemarketreactions FederalReserveBankofSt.Louisfortheircom- tobothpolicyactionsanddatareleases?Froma ments,whileretainingfullcreditforerrors. policyperspective,arethesemarketresponses simplyanunavoidablesideeffectofpolicy actions?Shouldthemonetaryauthoritiesattempt MARKET REACTIONS TO toavoidlargemarketreactions,orarethemarket POLICY ACTIONS reactionsanessentialpartoftheprocessbywhich Let’sstartbyexaminingafewdramaticexam- monetarypolicyeffectsaretransmittedtothe plesofpolicychangesthatreflectedFederal economy? 2 Synching,NotSinking,theMarkets THE EFFICIENT MARKETS pricestabilitywill,ifanything,yieldloweraverage unemploymentthanwillprevailathigherinfla- PARADIGM tion.Pricestabilitywillcontributetomaximum Theefficientmarketsparadigmsurelyhasto sustainableeconomicgrowth. bethestartingpointinunderstandingspeculative So,theFed’sprimarygoalispricestability markets.Accordingtothisview,speculative anditssecondarygoalisstabilityofoutputand marketsrespondefficientlyasmarketparticipants employmenttotheextentpossible.However,my assessallrelevantinformation—absolutelyevery- argumentdoesnotdependonthespecification thingthatmightinfluencemarketprices—and ofthemonetarypolicygoals—substituteyourown bidmarketpricesupordownaccordingly.Given ifyoudonotlikemyformulation.Ofcourse,the thatinvestorsbidcurrentpricestolevelsatwhich marketsmustassessnotonlytheFed’sgoals,but risk-adjustedexpectedratesofreturnareequalized alsoitsskillinachievingthosegoals.Ialsotake acrossvariousinvestmentalternatives,eachnew asgiventhepolicyprocedurestheFedusesto pieceofinformationmaymovemarketprices. reachitsgoals.ThepointisthattheFed’spolicy Theefficientmarketsmodelisnotperfect,butit goalsandproceduresarekeypiecesofinformation certainlygoesfarinexplainingthebehaviorof tothemarketsbecausethisknowledgehelps speculativeprices. marketparticipantspredicthowtheFedwill So,marketsrespondtotheflowofallsortsof respondtonewinformation. information,includingthatfromthecentralbank. Howwouldweexpectmarketstobehaveif FederalReservepolicyactions,andstatements FederalReservepolicywere“perfect?”Ifpolicy byFederalReserveofficials—especiallythe were“perfect,”wewouldcertainlynotexpect chairman—affectmarketexpectationsaboutthe thefederalfundsrate—theFed’sshort-runpolicy futureand,therefore,currentmarketprices. instrument—toremainforeverconstantatan Theinformationthatmatters,ofcourse,is unchangedlevel.Thefedfundsratewouldhave newinformation.Everythingpredictablehas tobehighersometimesandlowersometimesto alreadybeenbidintomarketprices;onlythe beconsistentwiththepolicyobjectives.How reportscomingacrossthewiresthatchangethe wouldtheFeddecidewhenandbyhowmuchto probabilitiesoffutureoutcomesaffectcurrent changethefederalfundsrate?Well,asnewinfor- marketprices.Thispointiswellunderstoodby mationarrived,theFedwouldprocessthatinfor- mostmarketparticipantsandmostofthefinancial mationtodecideonadjustmentsofthefederal press. fundsrate.FOMCmembersareconstantlyexam- iningtheflowofincominginformationonthe stateoftheeconomyandworkingtodecidewhat MONETARY POLICY ACTIONS policyactionsmaybenecessarytokeeptheecon- NowIwanttodiscusstheinteractionsof omyonthedesiredtrack. speculativemarketswithmonetarypolicyactions. Ofcourse,FOMCmembersmayhavedifferent Marketparticipantsaretryingtoforecastthe interpretationsoftheincomingflowofdataand future,andsotheyarenaturallyinterestedinwhat theappropriatepolicyresponses.Theimplications theFedistryingtodo.Letmemaketheassump- forpolicyofaparticulareventarerarelyperfectly tion—whichIthinkisaccurate,butwillnotargue clear.Thefactisthateconomicsprovidestremen- here—thatthegoaloftheFederalReserveisto dousguidance,butdoesnotprovidecalculations keeptherateofinflationlowandsteady,agoal outtotheseconddecimalplace.Indeed,some- thatI’llcall“pricestability.”Also,insofaras timeseventheappropriatedirectionofpolicy possiblegiventhepricestabilitygoal,theFed actionisunclear.Nevertheless,itishelpfulto wantsitspolicyactionstocontributetothesta- thinkaboutpolicythisway:Thereisinprinciple bilityofemploymentandoutput.Ibelievethat somecorrectpolicyresponsetoeachpieceof 3 MONETARYPOLICYANDINFLATION informationthatcomesalong,andthattheaim motesmaximumsustainableeconomicgrowth. oftheFOMCistodialinthatresponseatitsnext Wecannothopeorexpecttoavoidallsurprises, meeting.Theappropriateresponse,forexample, forthenatureofourworldisthatthefutureis maybetokeepthefundsratesteady.Indeed,Iam unpredictable.Theseunpredictableeventsinclude convincedthatoneofthegreatestbenefitsahigh naturaldisturbances,suchasearthquakesand degreeofmarketconfidenceintheFedaffordsis floods,politicaldisturbancesathomeandabroad, theFed’sabilitytosittightandwaituntilit’squite manychangesintastesandtechnology,andso clearwhichFedpolicyactionsareappropriate. forthandsoon.Marketsrespondefficientlyto Nowlet’sreturntothemarketresponses. thesedisturbancesmostofthetime.Ouraimis ThemarketsandtheFedarebothrespondingto formonetarypolicytooffsetshocks,whenpossi- thesameflowofinformation.Atthispoint,let’s ble,topreventthemfrompushingtheeconomy assumethatthemarketsandtheFedgetthesame awayfrompricestability.Ofcourse,wealsowant informationatthesametime—neitherhasan toavoidintroducingmonetarypolicydisturbances informationaladvantage.Itiseasytoseethenature persethatadverselyaffectpricestability. oftheexpectationalequilibrium.Themarketsand Thejobofthecentralbankistomaintaina theFedhaveacommonresponsetonewinforma- clearfocusonpricestabilityandtoconveythat tion.ThemarketsknowtheFed’spolicyobjectives focustothemarkets.Thecentralbankandthe andthepolicyadjustmentsthatareappropriate marketscanthenrespondinidenticalfashionto giveneachpieceofnewinformation.TheFOMC theflowofincominginformation,reachingthe meetseverysixweeks,butbythetimeofeach sameconclusionsastoimplicationsoftheinfor- meeting,themarketsknowfullwellwhatpolicy mationformonetarypolicyadjustments.Inthis adjustment,ifany,isnecessaryanddesirable. model,themarketsarenottakenbysurpriseby TheFedadjustspolicyaccordingtothemarket policyactions,fortheyhavealreadyfiguredout forecast,andnooneissurprised.Fedactionisa whatneedstobedone. noneventinthemarkets. Ithinkthisidealizedpicturetakesusalong ThepredictabilityofFedpolicyactionsunder waytowardunderstandinghowmonetarypolicy theseconditionsisthecentralinsightofthe andthemarketsshouldinteractwhenpolicyison analysis.GiventhattheFOMCactsasthemarket asuccessfultrack.Inthisenvironment,theFed forecasts,theFOMC’sactionsarenotthemselves andthemarketsaresynched.Withcompletesyn- informationandthereforeelicitnoadjustmentof chronization,themarketsandtheFedhaveacom- marketprices.Thoseadjustmentshavealready monunderstandingoftheobjectivesofmonetary takenplaceduringtheperiodbetweenFOMC policyandacommoninterpretationofthesignifi- meetingsasthemarketsrespondtothesteady canceofeachpieceofincominginformation. flowofnewinformationintheformoftheemploy- Supposethefederalfundsfuturesmarketdoes, mentreport,housingstarts,productivity,employ- infact,accuratelyforecastdecisionsatFOMC mentcostindex,etc.,etc.,etc. meetings.IsthatanindicationthattheFedis simplyfollowingthemarkets,andnotexercising itsproperleadershiprole?Obviously,Ithinknot. MarketpricesthatanticipatewhattheFOMCis SYNCHING MARKETS AND goingtodoarenotonlyconsistentwithpolicy POLICY “perfection,”inthesensethatIhavebeendis- Thisidealizedpictureofthemarketsandthe cussing,butalsoactuallynecessaryforpolicy Fedrespondingthesamewaytothesamedatais, “perfection.”Regularityandpredictabilityare Ibelieve,themodelweshouldallbestrivingto importantpolicygoals.Marketsuccessinantici- achieve.Wewantpricestabilitypreciselybecause patingFOMCactionsindicatesFedsuccessin webelievethatavoidinginflationsurprisesadds designingpoliciestoachievegoalssocietyaccepts, totheefficiencyofthemarketeconomyandpro- andinconveyingthosepoliciestothepublic.I 4 Synching,NotSinking,theMarkets putthepointthiswaybecausenocentralbank ofthesignificanceofreadilyobservabledata.The inademocraticcountrycanlongpursuegoals morecompletetheconvergenceofviewsbetween notacceptedbythesocietyatlarge.Themarkets themarketandtheFed,thebettertheeconomy andtheFedcannotconvergeonacommonunder- willwork.Convergencereducesmarketvolatil- standingofthedirectionofmonetarypolicyif ityandexpectationalerrors,whichcanleadto theFeddoesnotpursueitsgoalsinaconsistent resourcemisallocation.Firms,forexample,may fashionovertime. makeinvestmentsthatprovetobeunprofitable WhatdowemakeofcasesinwhichFedpolicy becausetheirexpectationswerewrong. actionscreatelargemarketresponses?Clearly, TheFedhasnotreachedthepointthatits policymakerscansinkthemarkets,orsendthem policyactionselicitzeroresponseinthemarkets. intoorbit,whenasurprisepolicyactionboosts Butletmeofferahypothesis,whichIhavenot marketprices.Changesinpolicydirectionarenot yetinvestigatedbuthopetobeableto.Everyone necessarilyundesirable;policyoughttochange agreesthat,inrecentyears,economicoutcomes— ifithasdriftedoffcourse.That,Ibelieve,wasthe intermsofbothinflationandunemployment— casewiththe1979changeinFedpolicy.Whatis havebeenbetterthaninthepast.Myhypothesis unfortunateaboutsuchacase,however,isthat isthatmonetarypolicyhasbeenmoreregularand policydriftsoffcourseinthefirstplace.Sinking, predictablethanitusedtobe.Thishypothesis ororbiting,themarketscertainlycanreflectsome canbetestedbyexaminingwhetherFedpolicy sortofpolicyfailure,eitherbecauseanundesir- actionsaccountforasmallerfractionofthevari- ablepolicyisbeingcorrectedorbecausepolicyis anceofinterestratechangesinrecentyearsthan takingoffinanunforeseenandundesirabledirec- inthepast.Puttheotherwayaround,myhypothe- tion.Ibelievethatthe1971policyturntoward sisisthatnonpolicyevents,suchasdatareleases, comprehensivewageandpricecontrolswasan accountforalargerfractionoftotalinterestrate exampleofpolicytakingawrongturn.Iinclude variancenowthantheydidinthepast. thiscaseinmylistofexamplesbecausetheintro- Mymodelofsynchingthemarketsandpolicy ductionofwageandpricecontrolsdidhave isincompleteinsomeimportantrespects.Two monetarypolicyimplications;manyobservers issuesparticularlyconcernme. thoughtthecontrolswouldtakecareofinflation First,thepureversionofthemodelrequires andpermitmonetarypolicybothtobemore thattheFedandthemarketshavethesameinfor- expansionaryanddrivedownunemployment. mationabouttheeconomy.Ithinkthat,relative Sinkingmayalsoreflectsomesortofmarket tothemarkets,theFedactuallyhassuperior error.Ihavedeeprespectformarketjudgments informationinsomecasesandinferiorinforma- butdonotbelievethattheyareinvariablycorrect. tioninothercases.Whatisclearisthatfullsyn- Sometimesmarketswakeuptoerrors,andprices chronizationwiththemarketsrequiresthatthe adjustrapidly.Verylittleisknownaboutthissub- Fedpaycarefulattentiontobothcollectingand ject,butthe1987stockmarketcrashiscertainly conveyinginformation.Transparencyandclarity anexampleofamarketerror.Eitherthecrash arenecessaryingredientsforpolicysuccess. wasanerror,themarketadvancepriortothe Second,thereareconsiderabledifferencesin crashwasanerror,orbothwereerrors.Noeco- professionalopinionabouthowtheeconomy nomicdataorpolicychangesarrivedatthetime works.ThedebatesinsidetheFedandoutside ofthecrashtojustifyanadjustmentthatlarge. theFedaresimilar.ThemarketsandtheFedwill Inanyevent,thekeypointremains.Alarge neverbecompletelyinsynchbecausetherewill marketresponsetoaFedpolicyactionisevidence alwaysbesomethingforeconomiststoargue thatthemarketsandtheFedarenotinsynch. about,andtheFedandthemarketswillnotnec- EitherthemarketortheFed,orboth,musthave essarilycometothesamejudgments.Still,itis beenoperatingonthebasisofdifferentinforma- importantnottolosesightofthefactthatthere tion,whichmayincludedifferentassessments isanenormouscommonbaseofunderstanding 5 MONETARYPOLICYANDINFLATION betweentheFedandthemarkets,andthatthis priceswillnotrespondtothembecausetheyare commonbasehasalottodowithpolicysuccess. fullyanticipated. Ibelievethatapolicyagendadesignedto heightenthedegreetowhichtheFedandthe CLOSING COMMENT marketsareinsynchisanambitiousandworthy objective.WeintheFedneedtoworkontwo I’llclosebyre-emphasizingmymaintheme. fronts,inmyopinion.Oneisthepolicyfront WhenthemarketsandtheFedareinsynch,both itself,makingsurethatpolicyactionsareas willhaveacommonreactiontoincomingdata, appropriatelytimedandscaledaspossible.The andthemarketswillcorrectlyanticipateFed secondisonthedisclosurefront,makingsurethat policyactions.Anenvironmentinwhichmarkets knowledgeinsideandoutsidetheFedconverges correctlyanticipateFedactionsimpliesasituation tothemaximumpossibleextent. inwhichFedpolicyiswidelyunderstood,regular, Progressonbothfrontswillrequirecontinu- andpredictable.ThefactthatFedpolicyactions ingresearch.Itiscleartomethatnewinsightsinto sometimestakethemarketsbysurpriseshows theconvergence,orlackthereof,ofinformation thatwehavenotreached“perfection”yet. betweenmarketsandtheFedwillplayacentral Still,itisimportanttorecognizethattheFed roleinthisresearch.Myinsighttodayiscom- hasmadetremendousprogressoverthelast20 pletelyconsistentwith—indeedisimpliedby— yearsorsoinpursuingaconsistentpolicy rationalexpectationsmacromodels.WhatIhad designedtoestablishpricestabilityasthenorm notdonebeforeinmyownmindisrelatethese fortheeconomy.TheFedandthemarketsare abstractmodelstothedailyebbandflowofmar- mostlyinsynch;surprisesintheincomingdata— ketreactionstonewinformation.Theconclusion whetheronprices,employment,GDP,activityin Ihavebeendiscussing—that,withfullconver- economiesabroad,andsoforth—aresurprisesto genceofinformation,Fedpolicyactionswillnot bothmarketsandtheFedandbothreadthesur- affectmarketpricesbecausethemarkethas prisesprettymuchthesameway.Ifthemarket alreadypredictedthem—initiallysurprisedme. andFedreadingsbecomeidentical,wecanexpect ButthemoreIthinkaboutthematter,themore thatFedpolicyadjustmentswillconveynonew compellingtheconclusionis.Ihopeyouagree. informationtothemarket,andthereforemarket 6

Cite this document

APA

William Poole (1999, August 5). Speech. Speeches, Federal Reserve. https://whenthefedspeaks.com/doc/speech_19990806_poole

BibTeX

@misc{wtfs_speech_19990806_poole,
  author = {William Poole},
  title = {Speech},
  year = {1999},
  month = {Aug},
  howpublished = {Speeches, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/speech_19990806_poole},
  note = {Retrieved via When the Fed Speaks corpus}
}