Speech

Understanding the Fed DyerCountyChamberofCommerceAnnualMembershipLuncheon Dyersburg,Tennessee August31,2006 PublishedintheFederalReserveBankofSt.LouisReview,January/February2007,89(1),pp.3-13 People often ask me questions about Themodelofhowtheeconomyworksis the Fed, sometimes out of simple complicated,andIcouldnotpossiblybeginto curiosity and sometimes out of a real presentithere.ButIwillsaythatourunderstand- need to know for business reasons. ingofhowtheeconomyworksisbasedoneco- Portfoliomanagers,forexample,havearealneed nomictheoryandanenormousbodyofempirical to know. My remarks reflect my effort to provide researchthatteststhetheory.Ourunderstanding rather systematically some answers to common isoftenqualitative,andweknowthatwemust questions. And I will also answer questions that attachstandarderrorstoournumericalpredic- ought to be put to me, but usually are not. There tions.Anactiveresearchprogramwithinthe is no reason why the Federal Reserve should be FederalReserveandbyacademicandbusiness a mysterious organization—we ought to be economistscontinuouslyrefinesthetheoryand responsive to your concerns. ourempiricalunderstanding. Obviously,Iwanttoemphasizethattheviews Letmeuseananalogy:Hurricaneforecasting Iexpressherearemineanddonotnecessarily hascomealongway,but,asanyonewhowatches reflectofficialpositionsoftheFederalReserve theweathernewsknows,theforecastsarenot System.Isuspectthateachofusinvolvedin perfectlyreliable.Shipcaptainshavetomake FederalReservepolicywouldanswertheques- tionssomewhatdifferentlyandemphasizediffer- policydecisionsonwhatcoursestoset,taking entthings.Inanyevent,I’lloffermyanswers.I intoaccounttheforecastsandwhatisknown thankmycolleaguesattheFederalReserveBankof aboutforecastaccuracy.Economicpolicymakers St.Louisfortheircomments,especiallyRobertH. havetomakethesamesortsofdecisionsbased Rasche,seniorvicepresidentanddirectorof onincompleteknowledge. research,whoprovidedspecialassistance. So,givenpolicyobjectives,andgivenaview abouthowpolicydecisionsaffecttheeconomy, thecentralbankcaninprinciplespecifyapolicy THE ROLE OF ECONOMIC rule,orresponsefunction,thatguidespolicy.To SCIENCE IN MONETARY POLICY achieveagoodresult,thegeneralpublicandmar- ketparticipantsneedtounderstandtheobjectives Averygeneralquestionconcernsthebasison andtheresponsefunctionsothattheprivate whichpolicydecisionsaremade.Itisimportant economycandetermineitsactivitieswithfull torecognizethateconomistshavedevelopeda knowledgeofhowthecentralbankwillact.Of formaltheoryofmonetarypolicyoverthepast 60yearsorsoandthatthistheoryreallydoes course,uncertaintyisaninherentcharacteristic guideourthinking.Thetheoryhastwological oftheeconomicworld.Whatshouldbepredict- parts.Thefirstisaclearsetofobjectives.The ablearethecentralbank’sresponsestothenever- secondisaspecificationofhowpolicy,inpursuit endingsequenceofsurprisesthatcharacterize oftheseobjectives,affectstheeconomy. theeconomicenvironment. 1 FEDERALRESERVE Monetary Policy Objectives feelingthatpricestabilityisamoreimportant nationalobjectivethaneithermaximumsustain- CongresssetsthemissionoftheFederal ablegrowthorahighlevelofemployment,but Reserveintheconductofmonetarypolicy.Origin- ratheronthereasonedconclusionthattheobjec- ally,theFed’smissionwasspecifiedintheFederal tiveofpricestabilityisanessentialprerequisite ReserveActsignedintolawbyPresidentWoodrow fortheirachievement”(McChesneyMartin,1959, WilsoninDecember1913.TheFed’scurrentman- p.5).Inhis1979confirmationhearing,Chairman date,setformallyinanamendmenttotheFederal PaulVolcker(U.S.Congress,1979)madethisstate- ReserveActin1977andreaffirmedin2000, ment:“Ibelievethatultimatelytheonlysound requirestheFederalReservetopursuethreeobjec- foundationforcontinuinggrowthandprosperity tivesthroughitsconductofmonetarypolicy.They oftheAmericaneconomyismuchgreaterprice are“maximumemployment,stablepricesand stability.”Earlyinhistenure,ChairmanAlan moderatelong-terminterestrates”(Bernanke, Greenspan(1988)concurredinthisview:“The 2006a).Economistsrecognizethatlong-term Committeecontinuedtofocusonmaintaining interestratesincorporateapremiumforexpected theeconomicexpansionandonprogresstoward inflation.Thus,theobjectivesofpricestability pricestability,whichwasseenasanecessarycon- andlowlong-terminterestratesareessentially ditionforlong-termsustainedeconomicgrowth.” thesameobjective. InJuly2006,ChairmanBenS.Bernanke(2006b) This“dualmandate,”so-calledbecauseofits acknowledgedthefollowing:“Theachievement emphasisonbothemploymentandpricestability ofpricestabilityisoneoftheobjectivesthatmake objectives,differsfromthatofothercentralbanks, uptheCongress’smandatetotheFederalReserve. especiallythe“inflationtargeting”centralbanks. Moreover,inthelongrun,pricestabilityiscritical Inflationtargeters,includingtheBankofCanada, toachievingmaximumemploymentandmoder- theBankofEngland,andtheReserveBankof atelong-terminterestrates,theotherpartsofthe NewZealand,amongothers,operateunderan congressionalmandate.” agreementwiththeirrespectivegovernmentsthat Ibelievethatwecangoastepbeyondthese definespricestabilityasthesingleobjectiveand statements.Inmyview,thegoalofpricestability specifiesaquantitativedefinitionoftheinflation mustbetheprimarygoalforthreereasons.First, objective.Inasimilarfashion,theEuropean inthelongrun,employmentandeconomicgrowth CentralBank(ECB)isgivenapricestabilityman- aremaximizedinanenvironmentofpricestability. dateundertheMaastrichtTreaty,thoughthetreaty Second,onlyinanenvironmentofpricestability doesnotgiveanumericvalueorrangetotheECB. andmarketconfidencethatthecentralbankwill TheECBhasinterpreteditsmandateasprevent- continuetomaintainpricestabilitywillthecentral ingtheinflationratefromexceeding2percent bankbeinapositiontoactdeliberatelytooffset perannumovera“mediumterm”horizon. manytypesofdisturbancesthatwouldotherwise Today,thereisgeneralagreementamong createfluctuationsinemploymentandoutput. professionaleconomistsandcentralbankers TheFederalReservedoesnothavethepowerto aroundtheworldthat,inthelongrun,monetary completelyoffsetallsuchdisturbances,butit policycannotachieveatradeoffbetweeninflation cancushiontheireffectsandtherebyimprove andemployment.SuccessiveFedchairmenhave economicstability.Finally,pricestabilityisa emphasizedthatpricestabilityisnotonlyaman- goalinitsownrightsimplybecausepriceinsta- datedobjectiveofmonetarypolicybutalsothe bilitycreatesarbitraryandunfairredistributions meansbywhichmonetarypolicycontributesto ofincomeandwealth. achievingtheothertwoobjectives.Theviewgoes Ihaveoftennotedthatmyownpersonal backatleasttoChairmanWilliamMcChesney preferenceistodefine“pricestability”asacon- Martin:“Myinterestinamonetarypolicydirected ditioninwhichtherateofinflation,properly towardadollarofstablevalueisnotbasedonthe measured,isonaveragezero.Iinsertthequalifier 2 UnderstandingtheFed “properlymeasured”topointoutthatactualprice relationshipimpliesthattheFOMCadjuststhe indicesmayhavestatisticalproblemssuchthat targetfederalfundsrateupaseithertheobserved zeromeasuredinflationonaparticularpriceindex inflationrateexceedsthetargetlevelofinflationor mightnotinfactreflectatruestateofzeroinfla- realGDPexceedspotentialrealGDP.Conversely, tion.Althoughmyownpreferenceisforzeroinfla- undertheTaylorrule,theFOMCreducesthetarget tionproperlymeasured,Ibelievethatacentral federalfundsratewheninflationfallsbelowits bankconsensusonsomeothernumericalgoalof targetand/orrealGDPfallsshortofpotentialreal reasonablylowinflationismoreimportantthan GDP.Thustherelationshipincorporatesthepri- theexactnumberchosen.Thus,Ifindthatrecent macyofalong-runinflationobjectivewhileincor- discussionofa“comfortzone”of1to2percent poratingshort-runstabilizationefforts. inflationmeasuredbythepriceindexforpersonal Figure1showstheactualvalueofthefederal consumptionexpenditures(PCE),excludingthe fundsratetargetonFOMCmeetingdatesstarting volatilefoodandenergycomponents,isperfectly in1987aswellasacomputedvaluebasedon consistentwithmyownthinking. Taylor’soriginalformulaandtheinformation Notethatthecongressionalmandatetothe availabletotheFOMCatthetimeofeachmeet- FederalReservedoesnotincludeanyofnumerous ing.2Theinflationrateinthefigureisthetotal objectivesthatfromtimetotimehavebeenadvo- CPI.Through2000,thegapbetweenrealGDP catedbysupportersofvariousinterests:forexam- andpotentialrealGDPisthevaluemeasuredby ple,stableexchangerates,stableassetprices,or thestaffoftheBoardofGovernorsatthetimeof housinginvestment.Clarityofobjectivesisan eachFOMCmeeting. importantattributeofmonetarypolicytodayand After2000,thestaffassumptionsaboutthe contributesgreatlytoitssuccess. GDPgaparenotyetpubliclyavailable,sothe dottedlineinthegraphforthisperiodiscom- Systematic Policy putedwiththeGDPgapasconstructedbythe ThedualnatureoftheFedmandateiswell CongressionalBudgetOffice.Alsobeginningin summarizedinthe“Taylorrule.”In1993,Stanford 2000,theFOMCchangeditsinflationobjective economistJohnTaylorproposedasimpleformula intwoways.First,theCommitteeemphasized relatingthefederalfundsrateto(i)along-run theinflationrateasmeasuredbychangesinthe inflationtargetand(ii)short-rundeviationsof PCEinflationrateratherthantheCPIinflation inflationfromthattargetandshort-rundeviations rate.Second,theCommitteeemphasizedthecore ofrealgrossdomesticproduct(GDP)(Taylor,1993) PCEindex,whichexcludesthevolatilefoodand fromameasureof“potentialrealGDP.”1Taylor energycomponents.Hence,itislikelythatafter suggestedthathissimplerelationshipcharacter- 2000theTaylorformuladoesnotaccuratelyreflect izedinbroadoutlinetheactualbehaviorofthe theinformationusedbytheFOMCasinputtoits federalfundsrateintheearlyyearsofthe deliberations. GreenspanFOMC.Theessenceofthisrelationship Notethatthetargetfundsratepredictedby isthat,inthelong-run,theFOMCseekstokeep theTaylorformulagenerallytrackstheactual thefederalfundsrateroughlyconsistentwitha fundsratethrough2000,thoughtherearesizable levelthatisbelievedtoproduceatargetlevelof andpersistentdeviationsofthefundsratefrom inflation.Taylorassumedatargetrateofinflation thevaluespredictedbytheformula.Nevertheless of2percentperyearmeasuredbythetotalcon- severaloftheseepisodesareconsistentwitha sumerpriceindex(CPI).Intheshortrun,the systematicmonetarypolicy.First,in1989,the 1 Taylorcomparedthevaluesofhisformulaagainsttheobservedhistoryofthefundsratefrom1987through1992. 2 Inthefigure,theTaylorformulaisevaluatedonthebasisof“realtime”informationthatcouldhavebeenusedinreachingapolicydecision. SinceMay17,1989,allchangesintheintendedfundsratehavebeen25basispointsormultiplesthereof.SinceApril18,1994,allchanges intheintendedfundsratehavebeenvotedonbytheFOMCeitherataregularlyscheduledmeetingoronanintermeetingconferencecall. 3 FEDERALRESERVE Figure 1 GreenspanYears: Federal Funds Rate and Taylor Rule (CPIp*=2.0,r*=2.0)a=1.5,b=0.5 Percent 12 Board’s Output Gap:CPI,1987:09–2000:10 FederalFundsRate 10 CBOOutputGap:CPI,2000:11–2006:06 8 6 4 2 0 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 FOMC Meeting Dates FOMCincreasedthetargetfundsratemorequickly turesfromtherulemakegoodsensewheninfor- thanpredictedbytheformula,suggestingthatthe mationbeyondthatincorporatedintheruleis Committeerespondedmorevigorouslytorising available.Forexample,policyisforwardlooking, inflationthanincorporatedintheTaylorspecifica- whichmeansthatfromtimetotimetheeconomic tion.Second,during1990-91,theFOMCreduced outlookchangessufficientlythatitmakessense thefundsratemorequicklythanpredictedbythe fortheFOMCtosetafundsratetargeteitherabove formula,suggestingastrongerresponsetothe orbelowthelevelcalledforintheTaylorrule, recessionthanincorporatedintheTaylorspeci- whichreliesonobservedrecentdataratherthan fication.Third,betweenlateSeptember1992and oneconomicforecastsoffuturedata.Othercir- February1994,thetargetfundsratewasheldat cumstances—anobviousexampleisSeptember11, alowerlevel(3percent)thanpredictedbythe 2001—callforapolicyresponse.Theseresponses Taylorspecification.Itwasduringthisperiodthat canbeandgenerallyareunderstoodbythemarket. theFOMCexpressedconcernabout“financial Thus,suchresponsescanbeeverybitassystem- headwinds”thatwererestrainingtherecovery aticastheresponsesspecifiedintheTaylorrule. fromthe1990-91recession.Finally,inthefallof 1998,theFOMCloweredthefundsratewhenthe Credibility of the Inflation Objective Taylorspecificationpredictedthattheratewould beheldconstant.Atthistime,concernabout AcriticalingredientintheTaylorspecification financialstabilityfiguredstronglyinpolicydelib- asadescriptionofmonetarypolicyisthelong- erationsinthewakeoftheAsianfinancialcrisis, runtargetrateofinflation.Inpractice,financial theRussianDefault,andthecollapseofLongTerm marketparticipantsandthepublicingeneral CapitalManagement(LTCM). cannotadequatelyunderstandtheFed’smonetary TheFOMC,andcertainlyJohnTaylorhimself, policy—thatis,thestrategicthinkingthatguides viewtheTaylorruleasageneralguideline.Depar- thesequenceofindividualpolicyactions—with- 4 UnderstandingtheFed outagoodunderstandingofwhattheFOMCcon- economyfromevidenceofthecurrentstateofthe siderstobeanacceptablelong-runaveragerate economyasindicatedinregulardatareleases, ofinflation.Whenmonetarypolicymakersartic- andotheractivities,suchaspoliticaleventsand ulatetheirgoalforlong-runinflationandpursue policyactionsthatinfluenceeconomicactivity crediblepoliciestoachievethatgoal,theyprovide andmarketprices.Theevidenceforsuchforward- thebasisfor“anchoring”theinflationexpectations lookingexpectationsiswidespread.Futurescon- thatguideconsumptionbehaviorofhouseholds tractsincommodities,interestrates,andforeign andinvestmentdecisionsoffirms.Inflationexpec- exchangeareactivelytradedinlargevolumeon tationsalsodeterminetheinflationpremiumsin organizedexchanges.Surveysofforecastsof nominalinterestratesthatarerequiredtoequili- forthcomingdatareleasesappearregularly(e.g., bratefinancialmarkets. TheTicker,whichappearseveryMondayinthe Evidencesuggests,particularlyintheU.S. WallStreetJournal,andtheCalendarofReleases, economy,thattheactualinflationexperienceis whichappearseachFridayonthecoverpageof drivenbyinflationexpectations,resourceutiliza- U.S.FinancialDatapublishedbytheFederal tion—usuallymeasuredbyagaptermasinthe ReserveBankofSt.Louis).3Pricesinfinancial Taylorrule—and“supplyshocks”suchaschanges marketsrespondtodifferencesbetweenthe intheworldmarketpricesofenergyandother observedinformationontheeconomicsituation commodities.Ofthese,themostsignificantfactor andtheexpectationsaboutsuchinformation— historicallyhasbeentheinfluenceofinflation thatis,marketsrespondto“news.” expectations.Hence,whentheanchorforinfla- Therearemanyclaimsbutnoconvincing tionexpectationsbeginstodrag,actualinflation documentedevidenceoflaggedresponsesto becomesvolatile,andtheresultingdistortionsto “news.”Economictheorysuggeststhatmarket economicactivityandconditionsinfinancial pricesofassetsshouldbehavelikerandom marketsproducesignificantdisruptionsinthe walks—thattheaccumulatedinformationata economy.Themostrecentsevereexamplefrom pointintimeshouldhavenopredictivepower oureconomichistoryofinflationexpectations gettingoutofcontroloccurredin1977-79. forfuturechangesinprices.Putanotherway,new Itisaterriblethingifmonetarypolicymakers informationisquicklyreflectedinmarketprices, losetheircredibilitythattheywillmaintainlow leavingnoremainingpredictablechangeinmarket andstablelong-runinflation.Oncecredibilityis pricesthatwouldpermitaninvestortoexpect impaired,itcanonlybere-establishedthe“old anabove-normalreturnfrombuyingorselling fashionedway”—policymakershavetoearnit! theasset. Restoringcredibilitytakestimeinthefaceofsub- Ahugebodyofempiricalresearchisbroadly stantialpersistenceintheactualinflationprocess. consistentwiththishypothesis.Thus,market Ittookwelloveradecadetocompletelyrestore commentarythattoday’smarketadjustmentsare lowinflationintheUnitedStatesaftertheGreat causedby,ordueto,continuingconcernsover Inflationofthe1970s,andintheprocessthe implicationsofoldinformationareofdoubtful economyexperiencedtheworstrecession,in validity.Traders,financialjournalists,andeco- 1981-82,sincetheGreatDepression. nomicpunditsseemtobelievethattheyhaveto attributemarketadjustmentstosomething,even whenthereisnoevidencetosupporttheasserted INTERPRETING NEW reasons. Ananalystwhoispresumedtobeanexpert INFORMATION oughtoftentosaythatpricechangesappearinex- Financialmarketparticipantsformexpecta- plicableorrandom.Butfewexpertsinfactsay tionswithrespecttotheprospectivestateofthe suchathing.Therefore,Irecommendthatthe 3 ThelatterdataarefromasurveyconductedbyThomsonFinancial. 5 FEDERALRESERVE Figure 2 Actual and Predicted Changes in EurodollarYields,3-Month Horizon 3-MonthActualEurodollarYield Change(basis points) 300 CorrelationofPredictedChanges Perfect Forecast Line with ActualChanges=0.65 200 100 0 March1995 –100 –200 March 2001 Blue = Dec1994–June2003 Black=Sept2003–June2006 –300 –300 –200 –100 0 100 200 300 3-Month-AheadPredictedEurodollarYield Change(basis points) waryobserverbeskepticalaboutpurportedexpla- adjustmentsdoexplainsomestockmarketfluctu- nationsforpricechangesandshouldalways ations.Iacceptthatfact.Indeed,thetransmission checkanexplanationagainstbehaviorinother mechanismlinkingmonetarypolicydecisionsto markets.Inmyexperience,explanationsforstock changesintheinflationrateandemployment marketfluctuationsareespeciallysuspect.In requiresucheffects.But,Ireallydon’twanttobe recentyears,Ihaveoftenreadclaimsthatthe heldresponsibleforstockmarketfluctuations stockmarketwentdownbecauseofinterestrate thatoccurredforotherreasonsoraresimply fears,onlytofindthatinterestratesinthebond inexplicable!So,nexttimeyoureadthatthestock marketwereunchangedorwentdowninsteadof marketwentdownbecauseof“interestratefears,” up.Giventhatsomanyinstitutionsdealinboth pleasedotakeaquicklookatthebondtablesto theequityandbondmarkets,itmakesabsolutely seeifinterestratesactuallychangedthatday.4 nosensethatinterestratefearscoulddrivedown stockpricesandhavenoeffectoninterestrates. Newinformationdrivesbothmarketadjust- Clearly,FederalReservepolicyadjustments mentsandpolicychanges.Policydecisionsordi- andmarketexpectationsaboutfuturepolicy narilycannotbesetlonginadvancebecausethe 4 ThosewhoreportonthestockmarketdonotconfinetheirmisleadingcausalstatementstotheFed.Inmid-August2006,theDowJones averagecloseddownover90points,attributedbyatleastonereportto“oilpricefears.”Lookingattradingintheoilmarkets,Septemberoil futuresontheNYMEXexchangecloseduponly4cents,butOctoberfuturescloseddown19cents. 6 UnderstandingtheFed Figure 3 Actual and Predicted Changes in EurodollarYields, 6-Month Horizon 6-MonthActualEurodollarYield Change(basis points) 300 CorrelationofPredictedChanges Perfect Forecast Line with ActualChanges=0.54 200 100 June1995 –100 Dec 2000 –200 Blue = Dec1994–June2003 June2001 Black=Dec 2003–June 2006 –300 –300 –200 –100 0 100 200 300 6-Month-AheadPredictedEurodollarYield Change(basis points) FOMCmustbeopentochangingitspolicystance federalfundsrate.Considertheforecastingrecord inresponsetonewinformation,whichisinher- onthree-monthandsix-monthhorizons.Figure2 entlyunpredictable.Togainasenseoftheimpact focusesonthedifferencebetweentheyieldona ofnewinformationoninterestrates,I’llanalyze three-montheurodollarfuturescontractwiththree datafromtheeurodollarfuturesmarket.Eurodollar monthstomaturityandtheactualeurodollar depositsarenotfederalfunds,thoughchanges deposityield.Thisdifferenceisplottedonthe inthetwoyieldsarehighlycorrelated.Tostudy horizontalaxisagainsttheactualchangeinthe marketexpectationsaboutthefederalfundsrate yieldoneurodollardepositsoverthecorrespond- overhorizonsoffourormoremonths,itisbest ingthreemonthstothematuritydateonthever- touseeurodollarfuturesbecausethesecontracts ticalaxis.Thus,thethree-monthactualchange areactivelytradedoverfarfuturehorizons, intheeurodollardepositrateisplottedagainst whereastradingoffederalfundsfuturesisrela- theforecastedchangeintheeurodollardeposit tivelythinonhorizonsofmorethanafewmonths. yieldoverthesametimeperiod.Theobservations Pricesintheeurodollarfuturesmarketpro- aretakeneverythreemonthsstartinginDecember videdirectinformationonmarketexpectations 1994.5Thesameexerciseisrepeatedatasix- offutureFOMCpolicyactionssettingthetarget monthhorizoninFigure3. 5 Hence,theobservationsarenonoverlapping. 7 FEDERALRESERVE Ineachofthetwographs,thediagonalline 50basispointsinthefirsthalfof1995(atthe fromthelowerlefttotheupperrightrepresentsa JanuaryFOMCmeeting). lineofperfectforecasts—ifallthepointslayalong AnotherlargemissoccurredinDecember thisline,theeurodollarfuturesmarketwould 2000.Atthattime,thefuturesmarketforecasts neverhavemadeanyerrorsinforecastingthe wereforadeclineintheeurodollaryieldof35 changeintheeurodollaryieldoverthesucceed- basispointsoverthefollowingthreemonthsand ingthree(six)months.Itisevidentfromthefigures atotalof67basispointsoverthesix-month thatthefuturesmarketsfallshortofsuchperfec- period.Instead,theFOMCactedaggressivelyto tion.Infact,overtheentireperiod,thecorrelation lowerthefundsratetargetstartinginJanuaryand betweentheforecastchangesandtheactual continuingthroughMay2001byatotalof250 changesonathree-monthhorizonisonly0.65;on basispoints.TheFOMCactedaggressivelyas thesix-monthhorizonthecorrelationfallsto0.54. incominginformationpointedtogrowingweak- Asanaside,notethateconomistsandstatisti- nessineconomicactivity.BoththeFOMCandthe ciansusuallymeasureforecastingaccuracybythe marketsweresurprisedbyincominginformation squareofthecorrelationcoefficient,orR2.Thus, indicatingthattheeconomywasweakening acorrelationof0.65isanR2of0.42,whichmeans quicklyandsignificantly. thattheforecastsembeddedintheeurodollar Itisrarethatasingledatareportisdecisive futuresmarketexplain42percentofthevariance fortheFOMC.Theeconomicoutlookisdeter- offluctuationsintheactualeurodollaryield.Thus, minedbynumerouspiecesofinformation.Impor- unpredictableeventsevenoverathree-month tantdatasuchastheinflationandtheemployment horizonareresponsibleformorethanhalfofthe reportsarecross-checkedagainstotherinforma- varianceoftheeurodollaryield.Overasix-month tion.TheFOMCisawareofthepossibilityofdata horizon,theR2is0.29,whichmeansthatunpre- revisionsandshort-runanomalies. dictableeventsareresponsibleformorethan70 Mykeypointisthatmarketcommentary percentofthevarianceoverasix-monthhorizon. indicatingthattheFOMCisunpredictableisoff Theoverallcorrelationsmasksomeinteresting base.Typically,theFOMCcannotbepredictable details.Thepointsplottedinblackineachfigure becausenewinformationdrivingpolicyadjust- representobservationssincethemiddleof2003, mentsisnotpredictable.Allofuswouldliketo whentheFOMCstartedproviding“forward beabletopredictthefuture.WeintheFeddothe guidance”regardingfuturepolicyactions.Note bestwecan,butthemarketsshouldnotcomplain thatthesepointsscatterfairlytightlyaroundthe thattheFOMClacksclairvoyance!Whatthe lineofperfectforecasts—marketswerenotpartic- FOMCstrivestodoistorespondsystematically ularlysurprisedbytheevolutionofpolicyactions tothenewinformation.Thereisconsiderable overthisperiod.Incontrast,considertheforecasts evidencethatthemarketdoessuccessfullypredict forthefirsthalfof1995.Thethree-monthand FOMCresponsestotheavailableinformationat six-monthfuturesmarketforecastsinDecember thetimeofregularlyscheduledmeetings.6 1994wereforlargepositivechangesintheeuro- dollardepositrateoverthesucceedingthreeand sixmonths.Intheevent,theeurodollardeposit FOMC PROCESSES ratefellabitoverthesehorizons.TheDecember 1994futurespriceswereobservedshortlyafter TheBoardofGovernorsandtheReserveBanks theFOMCincreasedthefundsratetargetby75 arefortunatetohavehighlyprofessional,nonpo- basispointsinNovember1994,andthemarket liticalstaffsofeconomists.Theroleofthestaffis hadexpectedfurthersubstantialincreases.Infact, toprovideanalysisofcurrenteconomiccondi- theFOMCincreasedthefundsratetargetbyonly tions,forecastsoftheevolutionoftheeconomy 6 See,forexample,Poole(2005). 8 UnderstandingtheFed overahorizonofacoupleofyears,andassess- policyaction(settingofthefundsratetarget)given mentsoftheriskstothoseforecasts.Suchinfor- policygoalsandtheoutlookfortheeconomy. mationisavaluableandvaluedinputintopolicy Unanimousdecisions,whiledesirable,arenot discussions.Imyselfdonotfinallymakeupmy required,andmembersarefreetodissentfrom mindonapolicypositionuntilI’veheardboth theconsensusviewiftheyfeelstronglythatan staffpresentationsandtheviewsofotherFOMC alternativepolicyactionispreferable.Indeed,I participants.Moreaccurately,IgointoeachFOMC believethatitismyobligationundertheFederal meetingwithaviewontheappropriatepolicy ReserveActtodissentwhenIbelievestrongly actiongivenmyassessmentoftheeconomicout- thatanalternativepolicycoursewouldbebetter. look,butItrytobeasopenasIcantohavingmy Historically,dissentswerenotunusual;though, viewalteredbydiscussionattheFOMCmeeting. intherecentyears,theyhavebeenrelativelyrare. TherearecertainlyinstanceswhenIhavechanged theviewItookintothemeetingasaconsequence Communication ofthediscussion. Oncepolicyactionhasbeenset,itisabso- lutelynecessarythatcommunicationtothepublic Distinction Between Members and aboutthepolicyactionnotbegarbled.Hence, Participants theChairmanistheonlyparticipantwhospeaks Exceptwhentherearevacanciesinoffices, officiallyfortheCommittee.Hepresentsofficial thereare19principals,orparticipants,ateach FederalReservepositionsthroughtestimony FOMCmeeting—the7membersoftheBoardof beforecongressionalcommitteesandinpublic Governorsandthe12ReserveBankpresidents. speeches.TheminutesofFOMCmeetings,cur- Alloftheseparticipantsarefullyengagedinpre- rentlyreleasedthreeweeksaftereachmeeting,also sentingtheirviews.Theybringinformationfrom representtheofficialpositionoftheCommittee. theirbusinessandacademiccontacts,comment ParticipantsotherthantheChairmanexpress onstaffpresentations,discussanalyticalissues theirownviewsinspeeches.Thesespeechesoften relevanttounderstandingtheeconomyandpolicy mayseemtoreflecta“partyline”butarerarely issues,andpresenttheirviewsastothemost centrallycoordinatedinanyway.Inmyexperi- appropriatepolicyaction.However,atanypartic- ence,theonlytimetherehasbeenarealeffortto ularmeetingthereareonly12votingmembers coordinatepubliccommentsbytheparticipants oftheCommittee.Eachofthe7membersofthe wasinthelatesummerof1998.Atthattime, BoardofGovernorsisapermanentmemberof financialmarketswereskittishasaresultofthe theFOMC.ThepresidentoftheFederalReserve RussiandefaultandfinancialtroublesofLTCM. BankofNewYorkisalsoapermanentmember. Irecallaninformalgatheringinthelatesummer TheremainingFederalReserveBankpresidents of1998withChairmanGreenspanandacouple rotateasmembersoftheFOMC.Aspresidentof ofotherFOMCmemberswhentheChairmanmade theFederalReserveBankofSt.Louis,Iamina arequestthatwesayverylittlegiventherather rotationgroupwiththepresidentsoftheFederal tensestateofthemarketsastheLTCMsituation ReserveBanksofAtlantaandDallas.Thisstruc- unfolded. tureoftheFOMCdatesfrom1942.Intheearly Seemingcoordinationatothertimesisacon- yearsoftheFOMC,notalltheprincipalswere sequenceofthefactthatFOMCparticipants allowedtoparticipate,orevenattend,meetings ordinarilyseethingsquitesimilarly.Butmost atwhichtheywerenotvotingmembersofthe participantsarenotshyaboutexpressingtheir Committee. differences.Differencesareregisteredinaformal ThepurposeofFOMCmeetingsistoreacha waythroughdiscountratedecisionsoftheReserve consensusamongtheparticipantsand,particu- Banksandinformallythroughpositionsstated larly,amongthemembersabouttheappropriate inspeeches.Anyonecanobtainanexcellentfeel 9 FEDERALRESERVE forwhatgoesonatanFOMCmeetingbyreading discussionatthemeetingitself.Ihavealready transcriptsthatarereleasedwithalagoffiveyears documentedthemostimportantreasonforthis andaremadeavailableonthewebsiteofthe policyofmine.Inherentlyunpredictableinforma- BoardofGovernors. tioncanarriverightuptothedayofthemeeting. Itwouldnotbesoundpracticeforpolicymakers Dealing with the Press tolockthemselvesintodecisionsimperviousto newinformation. DifferentFOMCparticipantshavedifferent AnotherreasonwhyIordinarilydonothave attitudesandcomfortlevelsindealingwiththe press.Imyselfgivemanyspeechesandalmost asettledpolicypositionweeksbeforeanFOMC alwaystalkwiththepressaftermyspeeches. meetingisthis:Ifolloweconomicreportscon- ItrytobeasclearasIcan,butfromtimetotime tinuouslybetweenFOMCmeetings,butdonot IrealizeafterthefactthatIhavenotexpressed ordinarilydigdeeplyintothemuntiltheperiod myselfasclearlyasIwouldhaveliked.When ofintensepreparationtheweekbeforeanFOMC readingpressaccounts,beawarethatFOMCmem- meeting.Thus,myviewsonincominginformation bersmisspeakfromtimetotimeandpressinter- areoftententativeandincomplete;Iknowthat pretationsarenotalwaysasintended.Wealltry I’llbereviewingallavailableinformationinthe tobeascarefulaspossiblebutarenotinfallible. intensepreparationperiod.Ibelievethatitwould Bewaryofheadlines.Reporterswilltellyouthat notbehelpfultothemarketsformetoconvey theyaresometimesfrustratedwiththeheadlines viewsthatIknowaretentativeandincomplete; thatappearovertheirstories.Beawarealsothat thus,Itrynottospeculateaboutthesignificance wireservicereportsaresometimesdesignedto ofnewinformationforthepolicydecisiontobe movemarkets.Considerationsbehindpolicy debatedatthenextFOMCmeeting.Myresponsi- adjustmentsareoftencomplex—becautious bilityistoconveyaccurateinformationand, aboutsimpleinterpretations. equally,nottobeasourceofmisleadingorinac- Ibelievethatoneofmyresponsibilitiesisto curateinterpretationsofincominginformation. communicatetoawideraudiencethroughthe press,doingthebestIcantobeaccurateandto Bottom Line conveybothpolicyfundamentalsandpolicy TheFederalReservehastheresponsibilityto nuances.Awell-informedpublicisessentialto provideleadership.Theidealsituationiswhen aneffectivemonetarypolicy.Inmyexperience, themarketcanreasonablypredictwhattheFed pressreportsaregenerally,butnotalways,accu- rate.ThefinancialjournalistswithwhomIinteract isgoingtodobecausetheFedhasprovidedthe aregenuinelyinterestedingettingthestoryright. leadershiptomakeclearitsobjectivesandhow Theydoaskprobingquestions,astheyshould, itpursuesthoseobjectives.TheFedisnotand butdonottrytoimposetheirownpersonalslant oughtnottobeviewedasanadversaryofthe tothereportstheypublish.Inaccuraciesaregen- markets.Policyactionsandstatementsdohave erallyaconsequenceofthecomplexityofthesub- marketeffects.Thoseareunavoidable,buttheFed jectandtheneed,onmypartandthejournalists’ strivestomakepolicyasclearasitcansothat part,tomakereportssimpleenoughforawide whatisreallysurprisingthemarketsisnotFed audiencetounderstand. actionsbutthearrivalofnewinformationthat Whentalkingwiththepress,oneofthepoints surprisestheFedandmarketstogether. ItrytoconveyisthatIdonotcometoafirmcon- I’vetriedtoanswerquestionsthatarecom- clusionaboutmypolicypositionuntiljustbefore monlyputtomeandtoprovideageneralframe- theFOMCmeeting;asIhavesaid,eventhen,I workforbetterunderstandingoftheFederal domybesttomaintainanopenmind,whichcan Reserve.AndIhopethatIhaveprovokedsome bechangedbythestaffpresentationandgeneral additionalquestions. 10 UnderstandingtheFed REFERENCES Bernanke,BenS.“TheBenefitsofPriceStability.” RemarksattheCenterforEconomicPolicyStudies, WoodrowWilsonSchoolofPublicandInternational Affairs,PrincetonUniversity,Princeton,NJ, February24,2006a;www.federalreserve.gov/ boardDocs/Speeches/2006/200602242/default.htm. Bernanke,BenS.TestimonybeforetheCommitteeon Banking,HousingandUrbanAffairs,U.S.Senate, July19,2006b;www.federalreserve.gov/boarddocs/ hh/2006/july/testimony.htm. Greenspan,Alan.TestimonybeforetheCommittee onBanking,FinanceandUrbanAffairs,United StatesHouseofRepresentatives,February28,1988; reprintedintheFederalReserveBulletin,April1988, p.227. McChesneyMartin,William.LettertoSenatorPaul Douglas,December9,1959. Poole,William.“HowPredictableIsFedPolicy?” SpeechdeliveredattheUniversityofWashington, Seattle,WA,October4,2005;www.stlouisfed.org/ news/speeches/2005/10_04_05.htm. Taylor,JohnB.“DiscretionversusPolicyRulesin Practice.”Carnegie-RochesterConferenceSerieson PublicPolicy,December1993,39,pp.195-214. UnitedStatesCongress.HearingsontheNomination ofPaulW.VolckertobeChairman,Boardof GovernorsoftheFederalReserveSystem,before theCommitteeonBanking,HousingandUrban Affairs,UnitedStatesSenate,Ninety-sixthCongress, FirstSession,July30,1979,p.20. 11

William Poole (2006, August 30). Speech. Speeches, Federal Reserve. https://whenthefedspeaks.com/doc/speech_20060831_poole

@misc{wtfs_speech_20060831_poole,
  author = {William Poole},
  title = {Speech},
  year = {2006},
  month = {Aug},
  howpublished = {Speeches, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/speech_20060831_poole},
  note = {Retrieved via When the Fed Speaks corpus}
}