speeches · July 11, 2002

Speech

William Poole · President

Fed Policy to the Bond Yield MidwestRegionoftheNationalAssociationofStateTreasurers MissouriHistoryMuseum St.Louis,Missouri July12,2002 When I learned that this meeting Louisfortheircomments,especiallytheBank’s was to be held in this beautiful Researchdirector,BobRasche,butIretainfull museumoftheMissouriHistorical responsibilityforerrors. Society, I immediately fell into a reflective mood. What I have to say does not remotely compare in importance with the story LONG-RUN CONSIDERATIONS of Lewis and Clark, or other important events in Totackletheissuesinvolvedwithmyques- Missouri history. But I am going to look at some tions,I’mgoingtoexploresomeratherabstract unusual features of our current economy in the considerations,andthenapplythemtoourcurrent context of what history can tell us. circumstances.I’llstartwithlong-runconsidera- Mytopicisoneparticularaspectofrecent tions,whichwecanthinkofasapplyingtodata behavioroffinancialmarkets—thefactthatlong- averagedover,say,fivetotenyears. terminterestrateshavechangedlittleduringa Manyyearsago,thedistinguishedeconomist periodwhenshort-termratesfelltolevelsnot IrvingFisherprovidedasignificantinsightinto seensincetheearly1960s.Morespecifically,the thebehaviorofinterestratesbypartitioning federalfundsrate,whichtheFederalReserve observednominalratesonconventional,non- targets,fellfrom6percentatthebeginningof indexedbondsintotwocomponents:oneisthe 2001to1percentinDecember2001,wherethat realrateofinterestandthesecondtheanticipated ratesitstoday.Butthe10-yearTreasurybondrate, rateofinflation.Theformerrepresentstherate whichwasatadover5percentwhentheFedfirst startedeasingpolicyinJanuaryoflastyear,has ofinterestthatwouldbeobservedinanoninfla- fluctuatedbetweenroughly4percentand5per- tionaryenvironment.Thelatteristhepremium centduringmostofthisperiod.Theratewasclose thatthemarketaddstointerestratestoreflect to5percentinMayoflastyear,anditapproached anticipatedinflationoverthelifeofthebondcon- 4percentaftertheterroristattacks,butthoserates tract.Thepremiumis,ofcourse,negative—adis- weretemporary.Whatexplainsthishistorically count—inadeflationaryenvironment. unusualbehavior,wherethelongrateseemsso Theusefulnessofthisdecompositiondepends littleinfluencedbyFedpolicy?Howdowego inpartonwhetherthetwocomponentsmove fromFedpolicytothebondyield?Everyonewith independentlyofeachother.Basedondecades responsibilityforraisingfundsinthecapital ofresearch,economistsnowviewthisindepend- marketshastobeconcernedwiththesequestions. enceconditionasareasonablecharacterization BeforeIattempttoaddresstheseissues,I ofthelong-runbehaviorofinterestrates,though wanttoemphasizethattheviewsIexpresshere notnecessarilyappropriatefortheanalysisof aremineanddonotnecessarilyreflectofficial interestratefluctuationsoveranyshortperiod. positionsoftheFederalReserveSystem.Ithank Forthisdiscussion,Iwillassociatetheantic- mycolleaguesattheFederalReserveBankofSt. ipatedinflationcomponentwiththatpartof 1 FINANCIALMARKETS interestrateswhichtheFederalReserve,ormore tivetoTreasurybondstendtobehigherinthe generally,acentralbank,caninfluenceinthe riskierenvironmentcreatedbyhigherinflation. longrunthroughtheeffectsofmonetarypolicy Tosummarizetheeffectsofhigherinflation ontherateofinflation.Iwillassociatethereal oninterestrates,thehighertheaveragerateof rateofinterestcomponentwiththatpartofinterest inflationthehigherwillbetheexpectedrateof rateswhich,giventheindependenceofthetwo inflation,whichgetsbidintonominalinterest interestratecomponents,theFederalReserve ratesasinvestorsprotectthemselvesfromerosion haslittleandonlyindirectinfluenceoninthe ofthepurchasingpowerofthecurrency.Secondly, longrun. higherinflationisinvariablymoreerraticinflation, WhatIcalltheanticipatedinflationcompo- whichmakesinflationmoredifficulttopredict. nenthastraditionallybeenassociatedwith Thatuncertaintyinturnmakestheeconomyand expectedinflation.Acentralbankthatwantsto manybusinessesmorerisky.Asaconsequence, achievemoderatelevelsofmarketinterestrates investorsbidhigherriskpremiaintonominal overthelongrunshouldseektoproducean interestrates. environmentoflowinflation.Successinmain- Whatdeterminestherealrateofinterest? taininglowinflationisessentialifexpectations Thisisthecomponentofthenominalmarketrate, oflowinflationaretobecomeentrenchedin averagedoveraperiodofyears,thatislargely marketthinking. independentofFederalReserveactions. Amorecompleteanalysisofanticipated Acrosstimeandacrosseconomiestheaverage inflationrecognizestheeffectofinflationon valueoftherealrateofinterestdependsonthe riskpremia.Manystudieshavefoundthathigher entirestructureoftheeconomy:marketstructure, ratesofinflationalsogeneratemorevolatileinfla- taxstructure,productivitytrends,andthelike. tion.Thusitisreasonabletoattributeaninflation Ofparticularimportanceisthefactthatmore riskpremiumtothedifferencebetween,say,the rapidlygrowingeconomies,reflectingpublic 10-yearTreasurybondrateandthecorresponding policiesconducivetoproductivebusinessinvest- 10-yearrisk-freerealrateofinterest.Thereare, mentandentrepreneurialactivity,generatehigher therefore,twoaspectsoftheanticipatedinfla- realratesofinterest.Thelogiciseasytounder- tioncomponentofinterestratesthatareaffected stand:Whentheexpectedrealreturntobusiness byFederalReservepolicy.Establishingalow- investmentinnewtechnologyandnewmarkets expected-inflationenvironmentwillproduce ishigh,theexpectedrealreturnonbonds,which lowerlong-terminterestratesbothbecause competeforinvestors’funds,willalsohavetobe expectedinflationislowerandbecauseinflation high. riskisreduced.Animplicationofthisobservation isthatthecentralbankcanreducetheaverage levelofmarketinterestratesbyincreasingthe SHORT-RUN CONSIDERATIONS credibilityofitslow-inflationobjective,which willtherebyreducetheinflationriskpremium. Wecanthinkofthelong-runrelationshipsas Thenextstepintheanalysisistonotethat determiningthelevelsaroundwhichshort-run higherinflationvolatilityisassociatedwithhigher fluctuationsininterestratesoccur.Overperiods volatilityoftheeconomyingeneral.Economic measuredfromafewweekstoseveralyears, boomsandbuststendtobemoreextremewhen FederalReservepolicyactionscanbeexpected inflationisvariable.Asaconsequence,riskier toimpactinterestratesacrossthematurityspec- firmsandindustriestendtobemoreriskythe trum.Attheveryshortestmaturity,theFOMC highertherateofinflation.Thehigherriskshows setsatargetforthefederalfundsrate—theso- upininterestratespreadsbetweenmoreandless called“intended”federalfundsrate.Through riskybondsofcomparablematurity.Forexample, openmarketoperations,theTradingDeskatthe thespreadsoflowerratedcorporatebondsrela- NewYorkFedkeepsthedailyaveragefundsrate— 2 FedPolicytotheBondYield whattheFederalReservecallsthe“effective” toturningpointsineconomicactivity—thecycle federalfundsrate—veryclosetotheFOMC’s peaksandtroughsdesignatedbytheNational intendedrate.Clearly,atthismaturitytheFederal BureauofEconomicResearch.Therelationship Reservehasalmosttotalcontroloverthereal betweenpeaksandtroughsofshort-terminterest rateofinterestintheshortrun.Forexample,if ratesandNBERcyclepeaksandtroughsissome- theFOMCweretoadjusttheintendedrateeither whatvariable.Shortratessometimesleadand upordownatitsnextmeeting,therealrateof sometimeslagthecycleturningpoints,although interestwouldchangebythesameamountbecause usuallynotbymorethanafewmonths.Themost therewouldbenoimmediateresponseofthe recentcyclewasnotunusualinthisregard;using inflationrateor,Iwouldpredictundertoday’s monthlyaveragedata,the3-monthTreasurybill conditions,theexpectedinflationrate. ratereachedapeakinNovember2000,four TheFederalReservehasnodirectinfluence monthsbeforethecyclepeakinMarch2001. overinterestratesatlongermaturities,ineither Historically,long-termratesalsoturnwithin theshortorlongrun.Nowwearegettingcloser afewmonthsofcyclepeaksandtroughs.But toaddressingtheissueIposedatthebeginning. recentexperiencehasbeensomewhatdifferent. Wheredoesthelongratecomefrom?Aninvestor Foronething,graduallydeclininginflation hastheoptionofperiodicallyrollingovershort- broughtratesdownonaverageduringthecyclical termbondsorsimplyholdingalong-termbond. expansionfrom1982to1990,andagainduring Althoughthedifferentstrategiescarrydifferent theexpansionfrom1991to2001.Morestrikingly, risks,abstractingfromthoseconsiderations,we thepeakinthe10-yearTreasuryrate,using expectthattensuccessiveone-yearinvestments monthlyaveragedata,wasinJanuary2000,14 willhaveaboutthesamereturnasoneten-year monthsbeforethecyclepeak.FollowingJanuary investment.Thusthekeytounderstandingthe 2000,longratesdeclinedsignificantlyoverthe relationshipbetweenlongbondyieldsandshort courseoftheyearandthen,asIhavenoted,have bondyieldsistounderstandinvestorexpectations fluctuatedinarelativelynarrowrangefromlate offutureshortyields. 2000tothecurrentday. Instudyinghistorytoprovideinsightsinto Inshort,wehavetoexplainnotonlywhy today’seconomy,perhapsthesinglemostimpor- thelongratedidnotfall,onaverage,during tantobservationisthat,aftertheearly1960s, 2001,whentheFedwasaggressivelycuttingthe changesininflationexpectationswereoften intendedfederalfundsrate,butalsowhythelong importantinchanginginvestorexpectationsabout ratebegantofallsomuchaheadoftheMarch thelikelycourseofshort-terminterestrates.Over 2001cyclepeak,contrarytotypicalexperience thepastthreeorfouryears,however,alltheevi- historically. dencewehaveindicatesthatinflationexpecta- Giventhatlongratesreflectmarketexpecta- tionshavechangedlittle.Itisessentialtokeep tionsoffutureshortrates,thekeytoanswering thisimportantpointinmindwhenstudyingpast thesequestionsistounderstandhowmarket interestratepatterns. expectationswerechangingduringthisperiod. Itiswell-knownthatinterestrateshave GiventhattheFedcontrolstheshortendofthe exhibitedastrongprocyclicalpatternhistorically. yieldcurve,wehavetocirclebacktoaskwhat Bothshort-termandlong-terminterestratestend themarketmightreasonablybelieveaboutmone- toriseineconomicexpansionsand,absentrising tarypolicy. inflation,tendtofallineconomiccontractions. Forstarters,inanenvironmentinwhichlong- Theamplitudeofthecyclicalfluctuationsin runinflationaryexpectationsarewellanchored, short-termratesissubstantiallylargerthanthat theFedneednotbe,andisnot,hypersensitive oflong-termrates. toinflationconcerns.Ifinflationconcernsare Wecanbemorepreciseaboutcyclicalinterest notactiveintheshortrun,whatshouldbethe ratepatternsbycomparingthebehaviorofrates guidingprincipleofFedpolicy?Inmyview,in 3 FINANCIALMARKETS thisenvironmenttheFederalReservecancredibly Weknowthattobethecasefromobservingthe attemptcountercyclicalmonetarypolicy.That behaviorofindexedTreasuryyields,whichpro- is,theFedhasroomtoadjustpolicyinaneffort videadirectmarketreadingontherealrateof toreduceunwantedfluctuationsinemployment interest.Thisobservationfitsinwithmyearlier andoutput. commentthattherealrateofinterestisrelated TheFedhasroomtoact,butdoesithavethe totherateofeconomicgrowth. knowledgetoact?Ithasbeenwelldocumented Still,themarketwouldnothavebiddown thatforecasters,includingFedforecasters,have longratesintheabsenceofananticipationthat greatdifficultypredictingtheturningpointsof shortrates,controlledbytheFed,wouldbefalling. businesscyclesorevenrecognizingthemsoon Infact,themarketexpectedthattheFedwould aftertheyoccur.Hence,thebestthatcanbereason- respondtotheweakeningeconomy;longrates ablyexpectedisthattheFOMCwouldbeableto camedownduring2000inanticipationofthe initiatepolicyactionsseveralmonthsinadvance actionthattheFOMCsubsequentlytook.The ofcycleturningpoints,ortoadjustpolicyonthe timingoftheFed’sJanuary2001ratecuttookthe basisofaccumulatingevidencetohelpreducethe marketbysurprise,butnotthefactofthecut. magnitudeofarecessiononceoneisobservedas Moreover,oncetheratecutsbegan,theoddson havingstarted. arevivalofeconomicactivityrose,whichIbelieve iswhybondratesdidnotfallastheFOMCcut Themostrecentcycleisausefulexampleof theintendedfundsraterepeatedlyoverthecourse exactlythisprocess.Thebusinesscyclepeakis oftheyear. datedinMarch2001.TheFOMCstartedlower- Allduringthecourseof2001,uptothetime ingtheintendedfundsrateatthebeginningof oftheterroristattacksinSeptember,currentdata January2001,atwo-monthleadonthecycleturn- cameingenerallyweakerthanexpectedbutfore- ingpoint.AtthepreviousmeetinginDecember, casterskeptexpectingthattheeconomicrecovery theFOMChadindicateditsconcernthatthe wasjustaroundthecorner.TheFedresponded economymightbeweakeningwiththislanguage totheweakerdatabycuttingthefundsrateaggres- initspolicystatement: sively,andthebondmarketrespondedtothose “Againstthebackgroundofitslong-rungoals cutsandtheexpectationofeconomicrevivalby ofpricestabilityandsustainableeconomic holdinglongratesinarelativelynarrowrange. growthandoftheinformationcurrentlyavail- Moreover,therewereanumberofinstances able,theCommitteeconsequentlybelievesthat inwhichdatareleasessuggestedthattheecon- therisksareweightedmainlytowardcondi- omymightseearevivalfairlyquickly,andthese tionsthatmaygenerateeconomicweakness tendedtokeeplongratesfromfollowingthe intheforeseeablefuture.” declinesinthefederalfundsrate.Letmecitejust AswillbeclearifyoureadtheFOMC’spub- oneexampleofmany.OnFriday,April27,2001, lishedminutesoverthecourseoflastyear,the the10-yearTreasurybondyieldjumpedby14 Committeedidnotforeseetheextentofthe basispoints,alargechangeforasingleday.The downturn.Butoverthecourseoftheyear,the marketwasrespondingtothereleaseoftheGDP Committeedidsensethecontinuingweakness estimateforthefirstquarter,whichshowedgrowth anddidrespondreadilytoincominginformation ata2percentannualrate.Thatwasanincrease suggestingthattheexpectedrevivalofactivity fromthe1percentrateinthefourthquarterand wasnotoccurring. doubletheincreasethatthemarkethadbeen Myinterpretationoftheseeventsisthatin expecting.Inreportingonmarketactivity,the 2000,especiallytowardtheendoftheyear,the WallStreetJournalsaidthat,“manyalreadyhad bondmarketsensedthattheeconomywasweak- beenwageringthattheFed’saggressivemonetary ening.Thedeclineinthenominalyieldwas easingthisyearwouldspurgrowthandsparka almostentirelyduetoadeclineintherealyield. reboundinstocksbeforelong…Now,analysts 4 FedPolicytotheBondYield say,theTreasurymarketcouldfaceapainful Then,whentheFOMCacts,orfailstoact,asthe periodinwhichyieldscontinuetoratchethigher, Committeethinksmakesmostsensegiventhe theFedeaseslessandpeoplepullmoneyoutof informationavailable,littleifanymarketresponse bondsinanticipationofacontinuedresurgence willbeobserved.Themarkethasthesameinfor- instockprices”(April30,2001,p.C15). mationtheFeddoes,toacloseapproximation, Theviewthateconomicrevivalwasjust anddrawsthesameconclusionsfromthatinfor- aroundthecornerremainedintoearlySeptember. mation,uptotheinevitableprofessionaldiffer- However,whentheterroristattacksoccurred, encesofopinion. theoutlooksuddenlylookedmuchworse.The Itisworthemphasizingthattheprocesscan Fedcuttheintendedfundsratesharplyfurther, workthiswaybecausetheFedistransparent andbondratesfelltowhatturnedouttobetheir aboutitsobjectivesandmethodsofanalysis.If lowsfortheyearasforecastersreviseddowntheir themarketdidnotunderstandwhattheFedis employmentandoutputforecasts. doingandwhy,itwouldoftencometoadifferent Intheevent,theeconomydidnotsinksharply. judgmentthantheFedonthebasisofinforma- PromptactionbytheFedandtheresilienceof tionavailable.Monetarypolicywoulditselfbea theU.S.economycarriedusthrough.Asdata sourceofuncertainty,addingrisktothemarket. arrivedinOctoberandNovember,indicating Letmenowlookahead.Inrecentweeks, thathousingactivityremainedverystrongand muchoftheeconomicdatahasbeenonthedis- thatcarsaleswererespondingvigorouslytothe appointingside.Thatisespeciallytrueofthe autocompanyincentives,theoutlookturned employmentreports.Inflation,however,continues brighter.Hereitishelpfultolookatweeklyaver- tobewellcontrolled.So,whereistheeconomy agedata.Onthatbasis,the10-yearTreasuryrate headed? reacheditslowof4.30percentinearlyNovember. Icannotofferaguessthatisanybetter Theflowofstrongereconomicdataledthebond informedthantheconsensusofprofessionalfore- markettoadjustquickly;insixweeks,the10-year casters,whostudythesemattersforaliving.The Treasuryratewasupbyabout85basispoints. prevailingviewisthattheeconomicexpansion Myinterpretationofthisperiodisbasedon willcontinueandthatitspacewillpickupfrom extensiveandongoingresearchattheSt.Louis thatofrecentmonths.Thatexpectationisreflected FedontheinteractionsofthemarketsandFed inthecurrentleveloflong-terminterestrates. policy.Icansummarizewhatwehavelearned Althoughthe10-yearTreasuryrateisdownmore thisway:Inanenvironmentinwhichmarket than50basispointsfromitslevelinMarch,much participantsunderstandhowtheFederalReserve ofthatdeclineshould,inmyview,beinterpreted interpretsincomingdataontheeconomy,the asevidencethatthemarketbelievesthattheodds marketcanforecastfutureFedpolicyactionswith arelowerthanbeforethattherecoverywillpro- someprecision.Infact,themarketcanforecast ceedsorapidlythattheFedwillberequiredto theseactionswithaboutasmuchprecisionas tightenpolicyrelativelyquickly.Theimportant theFedcanforecastitsownactions! point,though,isthatthemarketbelievesthatthe ThemarketandtheFedbothfacethesame recoverywillcontinue. uncertaintiesabouthowtheeconomywillevolve. IfyoufollowtheflowofdataascloselyasI Insuchcircumstances,adjustmentsinmarket do,yourealizethattheforecastsflowingfromthe ratescananddooccurinadvanceofthepolicy dataarealwayssubjecttorevision.Whatisnote- actionbytheFOMC.TheFOMCmeetsabout worthyaboutthecurrentstateofmonetarypolicy everysixweeks,buttheflowofinformation isthatuncertaintyoverpolicyitselfhasbeen occurscontinuously.Themarketrespondsday reduceddramaticallyinrecentyears.Thatis,the byday—indeed,hourbyhour—totheflowof wayinwhichtheFedwillreacttochangingcir- information,accumulatingitssignificanceforFed cumstancesisnotinmuchdoubt.Whatisindoubt actionrightuptothetimeofeachFOMCmeeting. ishowcircumstanceswillchange.Theworldis 5 FINANCIALMARKETS fullofsurprises,asIamsureyourexperience willconfirm. Iknowthatwecomplainaboutallthisuncer- tainly,butwouldn’ttheworldbeadullplace withoutit?PerhapsIshouldlistentomyown wordsandrejoiceintheendlessfascinationofthe dynamicworldwelivein.Infact,Idorejoiceas Ifindmyjobendlesslyexcitingandinteresting. 6

Cite this document

APA

William Poole (2002, July 11). Speech. Speeches, Federal Reserve. https://whenthefedspeaks.com/doc/speech_20020712_poole

BibTeX

@misc{wtfs_speech_20020712_poole,
  author = {William Poole},
  title = {Speech},
  year = {2002},
  month = {Jul},
  howpublished = {Speeches, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/speech_20020712_poole},
  note = {Retrieved via When the Fed Speaks corpus}
}