The Macroeconomic Effects of the Federal Reserve's Unconventional Monetary Policies

Finance and Economics Discussion Series

Divisions of Research & Statistics and Monetary Aﬀairs

Federal Reserve Board, Washington, D.C.

The Macroeconomic Eﬀects of the Federal Reserve’s

Unconventional Monetary Policies

Eric M. Engen, Thomas Laubach, and David Reifschneider

2015-005

Please cite this pap er as:

Engen, Eric M., Thomas Laubach, and David Reifschneider (2015). “The Macroeconomic

Eﬀects of the Federal Reserve’s Unconventional Monetary Policies,” Finance and Economics

Discussion Series 2015-005. Washington: Board of Governors of the Federal Reserve System,

http://dx.doi.org/10.17016/FEDS.2015.005.

NOTE: Staﬀ working papers in the Finance and Economics Discussion Series (FEDS) are preliminary

materials circulated to stimulate discussion and critical comment. The analysis and conclusions set forth

are those of the authors and do not indicate concurrence by other members of the research staﬀ or the

Board of Governors. References in publications to the Finance and Economics Discussion Series (other than

acknowledgement) should be cleared with the author(s) to protect the tentative character of these papers.

Page1of54



TheMacroeconomicEffectsoftheFederalReserve’s

UnconventionalMonetaryPolicies*



EricEngen,ThomasLaubach,andDaveReifschneider

FederalReserveBoard

January14,2015



Abstract

Afterreachingtheeffectivelowerboundforthefederalfundsrateinlate2008,theFederalReserve

turnedtotwounconventionalpolicytools—quantitativeeasingandincreasinglyexplicitandforward‐

leaningguidanceforthefuturepathofthefederalfundsrate—inordertoprovideadditionalmonetary

policyaccommodation.Weuse

surveydatafromtheBlueChipEconomicIndicatorstoinferchangesin

private‐sectorperceptionsoftheimplicitinterestraterulethattheFederalReservewouldusefollowing

liftofffromtheeffectivelowerbound.Usingourestimatesofthechangesovertimeinprivate

expectationsfortheimplicitpolicyrule,

andestimatesoftheeffectsoftheFede ral Reserve’s

quantitativeeasingprogramsontermpremiumsderivedfromotherstudies,wesimulatetheFRB/US

modeltoassesstheactualeconomicstimulusprovided byunconventionalpolicysinceearly2009.Our

analysissuggeststhatthenetstimulustorealactivityandinflation waslimited

bythegradualnatureof

thechangesinpolicyexpectationsandtermpremiumeffects,aswellasbyapersistentbeliefonthe

partofthepublicthatthepaceofreco verywouldbemuchfasterthanprovedtobethecase.Our

analysisimpliesthatthepeakunemploymenteffect—subtracting

1¼percentagepointsfromthe

unemploymentraterelativetowhatwouldhaveoccurredintheabsenceoftheunconventionalpolicy

actions—doesnotoccuruntilearly201 5,whilethepeakinflationeffect—adding½percentagepointto

theinflationrate—isnotanticipateduntilearly2016.

 

JELclassification:E5



Keywords:federalfundsrate,

forwardguidance,large‐scaleassetpurchases,monetarypolicyreaction

function,zerolowerbound



____________________________

*Theviewsexpressedinthispaperarethoseoftheauthorsanddonotnecessarilyreflecttheviewsof

othersattheFederalReserveBoardorintheFederalReserveSystem.WethankJim

Clouse,BillEnglish,

RefetGürkaynak,MikeKiley,DavidLebow,EdNelson,andseminarparticipantsattheEuropeanCentral

BankandtheBankofFranceforhelpfulcomments. 

Page2of54



I. Introduction



Bytheendof2008,theFederalReserve’sconventionalmonetarypolicytool,thefederalfundsrate,was

atitseffectivelowerboundastheeconomywasinthemidstofafinancialcrisisanddeeprecession.In

thesecircumstances,theFederalOpenMarketCommittee

(FOMC)turnedtotwounconventionalpolicy 

tools—quantitativeeasingprogramsandincreasinglyexplicitand forward‐leaningguidanceforthe

futurepathofthefederalfundsrate—inordertoprovideadditionalmonetarypolicyaccommodationto

helpendtherecessionandstrengthentheeconomicrecovery.

Theseunconventionalpolicyactions

wereintendedtoputdownwardpressureonreallonger‐terminterestratesandmoregenerallyto

improveoverallfinancialconditions,includingbolsteringpricesforcorporateequitiesandresidential

properties.Morefavorablefinancialconditionswould,inturn,helpboostaggregatedemandandcheck

undesirabledisinflationarypressuresby

providingincreasedsupportforconsumerspending,

construction,businessinvestment,andnetexports.



AsizablenumberofstudieshaveinvestigatedthefinancialmarketeffectsoftheFOMC’sunconventional

actions,especiallywithregardtotheFederalReserve’sassetpurchases,andfoundnoticeableeffects,

onbalance,onlong‐terminterestrates.

Bycontrast,therehavebeenrelativelyfewstudiesofthe

effectsoftheseactionsonrealactivityandinflation,andthisworkhasfocusedalmostexclusivelyon

macroeconomiceffectsarisingonlyfromthereductionsintermpremiumscausedbylarge‐scaleasset

purchases.

Forthesereasons,relativelylittleisknownaboutthemacroeconomicstimulusprovidedby

theeffectsoftheFOMC’spost‐crisisforwardguidanceandassetpurchasesonexpectationsforthe

futurepathofshort‐terminterestrates,norhowthoseexpectationaleffectsmayhaveinteractedwith

termpremiumshifts.Acentral

themeofthispaperisthatthetwotypesofpolicyactionsarehighly

interdependent,andthattheirmacroeconomiceffectsthusneedtobeevaluatedjointly.Intryingto

gaugethemacroeconomiceffectsofunconventionalmonetarypoliciessincelate2008,wetakeinto

accountthemarkedevolutionovertimeofthe

public’sexpectationsforbothfuturemonetar ypolicy

andtheoveralleconomy;thesefactors,asouranalysiswillshow,haveanimportantbearingonthe

actualmonetarypolicystimulusprovidedbyforwardguidanceandquantitativeeasinginrecentyears.



Thenextsectionofourpaperprovidesasummaryoftheunconventional

monetarypolicyactionstaken

bytheFederalReservesincelate2008.Thethirdsectionpresentsourmethodologyandassessmentof

howprivate‐sectorexpectationsaboutmonetarypolicyevolvedfromearly2009throughlate2013.In

particular,weusesurveydatafromtheBlueChipEconomicIndicatorstoinferthegradualchanges

that

havetakenplacesinceearly2009inprivate‐sectorperceptionsoftheFOMC’simplicitpolicyrule—that

is,thewayinwhichshort‐terminterestrateswouldbeadjustedinresponsetomovementsinreal

activityandinflation.ThefourthsectiondiscussestheotherchanneloftheFOMC’sunconventional

policy

actions,theestimatedeffectsoftheFederalReserve’squantitativeeasingprogramsonthe

trajectoriesofthetermpremiumsembeddedinlonger‐terminterestrates.Withtheseestimatesof

expectationaleffectsandtermpremiumshiftsinhand,weareinapositiontoscoretheactualstimulus



Althoughforwardguidanceforthefederalfundsratehadbeenusedinthepast,policymakersstillconsideredthe

natureofitsrecentusetobeunconventionalornontraditional;seeBernanke(2012),forexample.

ThesestudiesincludeBauerandRudebusch(2012),D’Amico,English,Lopez‐SalidoandNelson(2013),D’Amico

andKing(2013),Gagnonetal(2011),HamiltonandWu(2012),JoyceandTong(2013),KrishnamurthyandVissing‐

Jorgensen(2011),LiandWei(2013),MeaningandZhu(2011),Neely(2010),Rosa(2012),Rogers,Scottiand

Wright

(2014),andSwanson(2011).

ThesestudiesincludeChen,Curdia,andFerrero(2011),Chungetal(2012),BaumeisterandBenati(2013),Gertler

andKaradi(2013),andWealeandWieladek(2014).

Page3of54



torealactivityandinflationprovidedbyunconventionalpolicysinceearly2009,conditionalonsome

modelofthemacroeconomy.ForthispurposeweusetheFRB/USmodel,whoseproperties are

discussedinthefifthsection.Thesixthsectionthensummarizesresultsfromcounterfactualsimulations

ofFRB/US

designedtoscoretheactualmacroeconomicstimulusprovidedbytheFOMC’spost‐crisis

forwardguidanceandassetpurchases;animportantfeatureofthisexerciseistheuseofalternative

specificationsofthemodelinordertoaddressimportantuncertaintiesaboutthecurrentdynamicsof

theeconomy.Weconcludewithsomeobservations

aboutthepotentialimplicationsofouranalysisfor

theefficacyofmonetarypolicyintheeventtheeconomyweretoagainexperienceadeep and

prolongedslump.



Severalkeyfindingsemergefromouranalysis.Intheyearsfollowingtherecessionandfinancialcrisis,

wefindthatprivate‐sectorforecastersgradually

cametoperceivethattheFOMCinthefuturewould

pursueasignificantlymoreaccommodativepolicythantheyhadanticipatedatthestartofthecrisis,in

thesensethattheyrevisedupmarkedlyovertimetheirestimateoftheFOMC’sresponsivenessto

economicslack.Thischangeinpublicperceptionsof

theFOMC’simplicitpolicyrule—whichwe

estimateputconsiderabledownwardpressureonreallong‐terminterestratesovertime,overand

aboveanyeffectsassociatedwithchangesintheunderlyingoutlookforrealactivityandinf lation—

presumablyoccurredinresponsetotheFederalReserve’squantitativeeasingandincreasinglyforceful

forwardguidance

forthefederalfundsrate.Togetherwiththedownwardpressureontermpremiums

associatedwithassetpurchases,theseexpectationaleffectsappeartohaveeasedfinancialconditions

appreciablyrelativetowhattheyotherwisewouldhavebeen,therebyprovidingappreciablesupportto

theeconomicrecoveryovertime.However,ouranalysisalsoshows

thatthenetstimulustorealactivity

andinflationwaslimitedbythegradualnatureofthechangesinpolicyexpectationsandtermpremium

effects,aswellasbyapersistentbeliefonthepartofthepublicthatthepaceofrecoverywouldbe

muchfasterthanprovedtobe

thecase.Partlyforthesereasons,andpartlybecauseoftheinherent

lagsinthemonetarytransmissionmechanism,ouranalysisimpliesthatthemacroeconomiceffectsof

theFOMC’spastunconventionalpolicyactionshaveprobablyyettomanifestthemselvesinfull.In

particular,weestimatethatthepeakunemploymenteffect—subtracting1.2

percentagepointsfromthe

unemploymentraterelativetowhatwouldhaveoccurredintheabsenceoftheunconventionalpolicy

actions—doesnotoccuruntilearly201 5,whilethepeakinflationeffect—adding0.5percentagepointto

theinflationrate—isnotanticipateduntilearly2016.



II. TheFederalReserve’sUnconventionalMonetaryPolicies



After

thefederalfundsratereacheditseffectivelowerbound,theFOMCturnedtotheunconventional

strategyofusingbothquantitativeeasing(QE)alongwithincreasinglyexplicitandforward‐leaning

guidanceaboutthefederalfundsratetoprovideadditionalpolicyaccommodation.

Inthefirstpartof

thissection,weprovideabriefsummaryofhowbothapproachesevolvedovertime.Afterthat,we

reviewthegeneralimplicationsoftheseactionsforfinancialmarketexpectationsandotherfactors.





OtherunconventionalactionstakenbytheFederalReserveduringthisperiodincludedseveralspecialcredit

programsthatwereintendedtoimprovethefunctioningoffinancialmarkets,alongwiththeintroductionof

macroeconomic“stresstests”forthelargestbankstoverifythattheyhadsufficientcapital.Bothprograms

probablyhelpedto

supporteconomicactivityduring2009andearly2010tosomedegreebyenhancingtheflowof

creditandbyincreasingthepublic’sconfidenceinthebankingsystemandthebroadereconomy.Wehavenot

triedtoincorporatetheseprogramsintoouranalysis,however,becauseofthedifficultiesinaccessingtheireffects



usingmacroeconometricmodels.

Page4of54



Summaryofunconventionalpolicyactionssincelate2008



TheFOMC’sQEprogramsweremostlycomprisedoflarge‐scaleassetpurchases(LSAPs)oflonger‐term

Treasuryandagencymortgage‐backedsecurities(MBS),butalsoincludedthematurityextension

program(MEP);thevariousstagesoftheseQEpolicy

actionsaresummarizedinTable1.Cumulatively,

theFederalReserve’sholdingsofTreasurynotesandbondsalongwithagencyMBSandagencydebt

rosefromaround$500billionpriortothefinancialcrisistoover$4trillionwhenthemost‐recentLSAP

programconcludedinOctober2014.Thecounterpartonthe

liabilitysideoftheFederalReserve’s

balancesheetwasasubstantialincreaseinitsshort‐termliabilities(primarilybankreserves).

TheLSAPs

andMEPsmateriallylengthenedtheaveragedurationofthesecuritiesheldbytheFederalReserve,and

thusreducedtheaveragedurationofTreasuryandagencysecuritiesheldbythepublic,relativetowhat

otherwisewouldhaveoccurred.



AssummarizedinTable2,theFOMCalsobeganprovidingforwardguidanceforthefuturepathofthe

federalfundsrateinlate2008.However,theinitialguidanceonlyadvisedthat“weakeconomic

conditionsarelikelytowarrantexceptionallylowlevelsofthefederalfundsrateforsometime;”such



limitedguidancewasnotmarkedlydifferentfromthatemployedbytheFOMCin2003and2004.

Althoughthewords“sometime”werereplacedby“anextendedperiod”inearly2009,neitherphrase

providedtheprivatesectorwithmuchspecificityabouteitherthelikelydateofliftoffofthefederal

funds

rateortheeconomicconditionsthatwouldtriggerit.Asaresult,thepublicmighthave

reasonablyinterpretedtheguidanceasconsistentwiththeFOMC’saveragehistoricalbehavior,given

theseverityoftherecessionandtheprojectionsoffuturerealactivityandinflationmadeatthetime,

ratherthanassignaling

anyintentionoftheFOMCtodepartfromitsusualpolicyreactionfunction.



Thespecificityofforwardguidancechangedwith thereleaseoftheAugust2011statementbythe

FOMC,inwhichtheCommitteenotedthateconomicconditionswouldlikely“warrantexceptionallylow

levelsofthefederalfundsrateat

leastthroughmid‐2013”.

TheFOMCcontinuedtoissuesimilar

calendar‐basedguidanceforroughlythenextyearandahalf,althoughtheciteddateswererevisedto

“late‐2014”attheJanuary2012meetingandto“mid‐2015”attheSeptember2012meeting.Suchdate‐

basedguidancehelpedfinancialmarketparticipantsandothers

comparetheirexpectationsforthe

liftoffofthefederalfundsratedirectlytothoseoftheFOMC,andsohadthepotentialtoprompt

revisionsinthepublic’sexpectationsregardingthetimingofliftoffandpossiblythelonger‐runconduct

ofmonetarypolicyaswell.Thatsaid,suchdate‐basedforward

guidancewasprobablydifficultforthe



TheFederalReservedidnotholdanyagency‐relateddebtandMBSasoflate2008andheldmorethan$200

billioninshort‐termTreasurybillsjustbeforetherecession.Bylate2014,theFederalReservedidnotholdany

Treasurybillsandhadabout$1.75trillioninagency‐

relatedsecuritiesandalmost$2.5trillioninTreasurynotes

andbonds.TheFederalReservehadacquiredabout19percentoftotalfederalTreasurydebtheldbythepublic

whenthemost‐recentLSAPprogramconcluded,onlysomewhatlargerthanthe15percentshareitheldjustprior

tothelastrecession.

TheFederalReserve’sbalancesheetexpandedsignificantlybecauseoftheLSAPprograms,

buttotalfederaldebtheldbythepublicalsoincreasedsubstantiallyfromabout35percentofGDPin2007to

almost75percentofGDPin2014.

TheaveragedurationofthesecuritiesheldintheFederalReserve’sportfolioincreasedfrom1.6yearsjustprior

tothemost‐recentrecessionto6.9yearsneartheendof2014.

Thisguidanceappearedtohavebeenasurprisetofinancialmarketparticipants.TheeventstudybyFemia,

Friedman,andSack(2013),forexample,foundthatbothlonger‐terminterestratesandtheforeignexchangevalue

ofthedollardeclined,whileabroadindexofU.S.equityvaluesrose,immediatelyfollowing

thisannouncementof

calendar‐basedforwardguidancebytheFOMC.

Page5of54



publictointerpretbecauseanydifferencesbetweentheFOMC’sliftoffguidanceandprivateforecastsof

liftoffcouldhavereflected,atleastinpart,differencesinforecastsforeconomicactivityandinflation.



TheFOMCeventuallyshiftedawayfromcalendar‐basedguidanceandadoptedamorestate‐contingent



approachattheDecember2012meeting.Inthatmeeting’sstatement,theFOMCannouncedthatit

wouldkeepthefederalfundsrateatitseffectivelowerbound“atleastaslongastheunemployment

rateremainsabove6½percent,inflationbe tweenoneandtwoyearsaheadisprojectedtobeno

more

thanahalfpercentagepointabovetheCommittee’s2percentlonger‐rungoal,andlonger‐terminflation

expectationscontinuetobewellanchored.”SubsequentFOMCstatements releasedthroughlate2013

reiteratedthisguidanceaboutthethresholdsthatwouldneedtobecrossedbeforetheCommittee

wouldconsiderraisinginterestrates,

whileaddingthatthedecisiontotightenwouldalsotakeaccount

ofbroaderlabormarketconditionsandfinancialdevelopments.Anothernoteworthychangeinforward

guidancewasmadeattheDecember2013meeting,whentheCommitteeemphasizedthatthefederal

fundsratewaslikelytoremainatitseffectivelowerbound

wellpastthetimethattheunemployment

ratefellbelow6½percent,especiallyifinflationweretocontinuetorunbelowitslonger‐runtarget

levelof2percentandlonger‐runinflationexpectationsremainedstable.





Ofcourse,theforwardguidanceissuedinFOMCstatementswasnottheonlywaytheFederalReserve

mayhaveinfluencedpolicyexpectationsthroughitscommunications.Amongothercommunications

thatprobablyinfluencedtheviewsoffinancialmarketparticipantsandothersintheprivatesectorover

thepastfewyearswerecomments

aboutthefactorslikelytoinf luence future policydecisionsmadeby

individualFOMCparticipantsinspeeches,testimony,pressconferences,andotherforums;theFOMC’s

regularly‐publishedSummaryofEconomicProjectionswithforecastsofthefederalfundsrate,economic

activity,andinflationmadebyindividualFOMCparticipantswithoutattributionsinceJanuary2012;

and

theannualstatementsissuedbytheFOMCsinceJanuary2012aboutitslonger‐rungoalsandpolicy

strategy.Inthisregard,severalspeechesbyYellen(2012a,b,c)areworthhighlightingbecausethey

emphasizedthattheFOMC’sguidanceatthetimeaboutthelikelydateofliftoffwasbroadlyconsistent

with

theprescriptionsofpoliciesthatweremoreaccommodativethanastandardTaylor‐typepolicy

rule.Indeed,Bernanke(2012)notedthatchangestoprivate‐sectorforecastsappearedtoreflect“a

growingappreciationofhowforcefultheFOMCintendstobeinsupportingasustainablerecovery.”



Potentialeffectsofunconventionalpolicyactions



BoththeQEprogramsandtheforwardguidanceoftheFOMCwereintendedtofurtherhelpsupport

economicactivityandcheckundesirabledisinflationarypressuresafterthefederalfundsratehadbeen

loweredtoitseffectivelowerbound.TheQEpolicieswerethoughttoputdownwardpressureonthe

generallevelof

longer‐terminterestratesprimarilybecausetheyreducedtheaveragedurationofthe

TreasurysecuritiesandagencyMBSheldbythepublic,whichinturnwouldcausethepricechargedby

investorsforlendingoverthelongerrun—thetermpremium—todecline .

Asaresult,borrowingcosts



Asaconvenientcutoffpointforouranalysis,weconsidertheeffectsofFOMCunconventionalpolicyactions

takenthroughtheendof2013only.Ourfocusisonthemacroeconomiceffectsofprovidingadditional

accommodationthroughunconventionalpolicies.WiththeannouncementaftertheDecember2013FOMC

meetingofthefirst

reductioninthepaceofassetpurchases—implying,amongotherthings,thatdownward

pressureontermpremiumswouldstabilizeasthestockoftheFederalReserve’slong‐termsecuritiesholdings

leveledout—weconsideritunlikelythatsubsequentcommunicationshavesoughttoprovideadditional

accommodation.

SeeKrishnamurthyandVissing‐Jorgenson(2011),D’Amico,English,Lopez‐Salido,andNelson(2012),andLiand

Wei(2013)fordiscussionsofthismechanism.

Page6of54



forresidentialmortgages,autoloans,businessborrowing,andothertypesoflonger‐termlendingwould

decline,allelseequal;inaddition,arbitrageeffectsimplythatcorporateequitypricesandhomevalues

wouldriseandtherealforeignexchangevalueoftheU.S.dollarwoulddecline.



itself,forwardguidanceaboutfuturemonetarypolicyneed notprovidemonetarypolicystimulusto

theeconomy,especiallyifitmerelyconfirmsthepublic’sexistingexpectationsforthecentralbank’s

likelyresponsetomovementsin economicactivityandinflationovertime.

Rather,significant

economicstimuluswouldrequirethatfinancialmarketparticipantsandothersrespondtotheguidance 

bymarkedlyalteringtheirbeliefsaboutfuturemonetarypolicyinamoreaccommodativedirection.At

issue,then,iswhetherthepublicinterpretedtheguidanceprovidedbytheFOMCsincelate2008asa

credible

signalthat,astheeconomyrecovered,theFederalReserveintendedtokeepthefederalfunds 

ratelowerforlongerthanwouldotherwisebeexpectedgiventheFOMC’shistoricalresponsesto

movementsinresourceutilizationandinflation.

IfsuchashiftinpublicperceptionsoftheFederal

Reserve’simplicitpolicyreactionfunctiondidinfactoccur,thenitshouldhavecausedexpectationsfor

thefuturepathofshort‐terminterestratestoshif tdownandexpectationsforfutureinflationtomove

up,therebyputtingdownwardpressureon

reallonger‐runinterestratesin the nearterm evenasthe

nominalfederalfundsrateremainedstuckatitseffectivelowerbound.Allowing forarbitrageand

spillovereffectstootherassetmarkets,thischangeinexpectationscouldhavehelpedtoincreasethe

valueofcorporateequities,realestate,andother

formsofwealth,lowertherealforeignexchange

valueofthedollar,andimprovefinancialconditionsmoregeneral ly.



TheFederalReserve’sassetpurchasesalsocouldhavepromptedshiftsinpolicyexpectationsthat

stimulatedtheeconomybeyondwhatlowertermpremiumsbythemselves directlyprovided.

12

Given

theunprecedentedmagnitudeofitsquantitativeeasi ngprograms, suchassetpurchasesmayhave

concretelyanddramaticallydemonstratedtheFOMC’sdesiretoprovideadditionalaccommodation, 

despiteaccompanyingconcernsbysomeabouttheeffectivenessandpotentialcostsofexpandingthe

FederalReserve’sbalancesheet.Ifso,theQEprogramspotentiallyboosted

thecredibilityandhence

effectivenessoftheFOMC’sforwardguidancebyhelpingtoconvincefinancialmarketparticipants and

othersthattheFOMCreallywouldbepersistentlymoreaggressivethanpreviouslythoughtinusingall

itspolicytoolstobringtheeconomybacktowardfullemployment,whileperhapsbeingmoretolerant

amodesttemporaryovershootininflationaswell.Suchshiftsinpolicyexpectationscausedbythe

QEactionsareextremelydifficult,ifnotimpossible,todisentanglefromchanges inexpectations

associatedwithforwardguidancealone,giventhatbothtypesofunconventionalpolicieswereused

simultaneouslyandprobablyweremutuallyreinforcing.

Forthisreason,weestimatethe

macroeconomiceffectsofthecombina tionofthesetwounconventionalmonetarypolicies.





Ofcourse,suchconfirmationsmightreduceuncertaintyabouttheFOMC’sfuturebehavior,andifbydoingso,

financialmarketparticipants,firms,andhouseholdsbecamelessconcernedaboutpotentialriskstotheoutlook,

thentheclarifyingofforwardguidancecouldconceivablyboosteconomicactivitytosomedegree.

SeeCampbell,Evans,Fisher,andJustiniano(2012)foradiscussionofthedistinctionbetweenforwardguidance

thatchangesprivateexpectationsbypubliclycommittingtheFOMCtodeviatefromitsusualpolicyreaction

function—whattheycall“Odyssean”forwardguidance—fromforwardguidancethatmaysimplyprovidea

forecastoffuturepolicythat

onlyconfirmsthepublic’sexpectationsthattheFOMCwillfollowitstypicalpolicy

reactionfunction—whattheycall“Delphic”forwardguidance.

Inparticular,thispossibilityhasbeennotedbyBauerandRudebusch(2011),Bhattarai,Eggertsson,andGafarov

(2014),andWoodford(2012).

Page7of54



III. MeasuringShiftsinMonetaryPolicyRateExpectations



Thissectiondescribesourprocedureforestimatingchangesinprivate‐sectorexpectationsaboutthe

wayinwhichmonetarypolicymakerswerelikelytorespondtofuturemovementsineconomicactivity,

inflation,andfinancialconditions.Toconstructourestimateswe

usesurveydatafromtheBlueChip

EconomicIndicators.Theco‐movementsoftheeconomicprojectionsreportedintheBlueChipsurvey,

togetherwiththewaysinwhich theseforecastshaverevisedovertime,allowustoinferhowprivate‐

sectorforecasters’beliefsabouttheFOMC’simplicitpolicyreactionfunctionchanged

sincelate2008.In

thisregard,ourmethodologydiffersimportantlyfrompreviousworksuchasCampbelletal.(2012),in

thatthetypicalapproachhasbeentousesurveydatatoinferexpectationsoftemporarydepartures

from“standard”centralbankbehavior(i.e.,shockstoafixedrule),ratherthansystematic

changesto

beliefsabouthowthecentralbankwouldrespondtoeconomicconditions.



Forourpurpose,themostinformativesurveyresultsarethosepublishedeveryMarchandOctober,

whichincludenotonlyquarterlyforecastsofeconomicactivity,inflation,andinterestratesforthe

currentandcomingyear,whicharecollected

everymonth,butalsoforecastsofannualconditions 

extendingsixyearsintothefuture.Importantly,theMarchandOctobersurveysalsoreportforecasters’

assessmentsofconditionsinthelongerrun,therebyallowing ustoinferanychangesovertimein

privateassessmentsofthenaturalrateofunemployment, theequilibriumreal

rateofinterest,andthe

FOMC’seffectiveinflationtarget,whicharekeyparametersintheperceivedpolicyreactionfunction.



Figure1summarizestheevolutionsince2007oftheBlueChipconsensusprojectionsfortheannual

averageofthe3‐monthTreasurybillrate,theyear‐over‐yearrateofGDPpriceinflation,andthe

averageannualunemploymentrate,togetherwithanimputedconsensusforecastfortheannualoutput

gap.

Aftertheeffectivelowerboundforthefederalfundsratewashitinlate2008,privateforecasters

forsometimeprojectedthattheTreasurybillratewouldbeginrisingatarelativelyrapidpacewithina

fewquarters.Startingin2011,however,theprojectedliftoffdatewaspushedfurtherinto

thefuture;in

addition,theexpectedpaceoftighteningafterliftoffdeclinedappreciablywhilethenominallevelof

short‐termratesexpectedtoprevailinthelongrunfellnoticeably.Presumably,theserevisionsinthe

expectedpathofshort‐terminterestratesareexplainedatleastinpartbymarkedrevisions

overtimein

theoutlookforeconomicactivity,asprivateforecastersgraduallycametorealizethatthepaceof

recoveryinthelabormarketandthebroadereconomywouldbemarkedlyslowerthaninitially

anticipatedintheimmediateaftermathoftherecession.Incontrast,revisionsintheoutlookfor

inflation

publishedafterMarch2009appearlesssubstantial,withinflationalwaysprojectedtoconverge

relativelyquicklybackuptoaround2percent.Giventherecentaveragehistoricalwedgebetween

inflationmeasuredusingGDPpricesandinflationmeasuredusingthePCEpriceindex,theout‐yearBlue



AlthoughtheBlueChipEconomicIndicatorsreportsupdatedprojectionseachmonth,theforecasthorizonforall

buttheMarchandOctoberreleasesextendsonlytothecomingyear.Asaresult,BlueChipprojectionsreleasedin

theothertenmonthsoftheyearareoflimitedvalueinidentifyingshifts

inprivateforecasters’perceptionsofthe

FOMC’simplicitreactionfunction,particularlyasalmostallforecastsreleasedafterearly2009showedshort‐term

interestratesattheireffectivelowerboundthroughmostifnotallofthefollowingyear.

AlthoughtheBlueChipsurveydoesnotcollectforecastsoftheoutputgap,theonespublishedinMarchand

Octoberdoreportprojectionsforthelong‐runleveloftheunemploymentrate—thatis,thenaturalrate.Private

forecasters’implicitprojectionsfortheoutputgapmadeattimetcanthus

beapproximatedas2





∗







,

where



∗

denotestheconsensusprojectionofthenaturalrate,





istheprojectedunemploymentratejyears

ahead,and2istheassumedvalueofthecoefficientinOkun’sLaw.

Page8of54



Chipforecastsappeartobefairlywellalignedwiththelonger‐terminflationobjectiveof2percent

(definedintermsofthePCEpriceindex)thattheFOMCannouncedinearly2012.





Inourmodel‐basedevaluationofthequantitativeeffectsofunconventionalpoliciesonrealactivityand

inflationlaterinthepaper,changesinpublicperceptionsoftheFOMC’sreactionfunction—p resumedto

beadirectconsequenceoftheFederalReserve’sforwardguidanceandassetpurchases—areoneofthe

twometricsof

the“impulse”fromunconventionalmonetarypolicy(theotherbeingQE‐relatedshifts in

termpremiums).Togaugethemagnitudeofchangesintheperceivedpolicyruleovertime,weusethe

BlueChipforecaststoinferprivate‐sectorbeliefsabouttheFOMC’slikelyfutureresponsestochangesin

economicconditions.

ThehistoricalbehavioroftheFederalReserveandothercentralbankscanbe

approximatedquitecloselybysimplereactionfunctions,evenifthepolicymakersattheseinstitutions

donotliterallyemployaruletosetthepolicyrate.

Consistentwiththatevidence,itisreasonableto

assumethat,attimeperiodt,privateforecastersandothersbasetheirj‐step‐aheadexpectationsfor

short‐terminterestratesonaTaylor‐typeruleofthefollowingform:



(1) 











∗

















∗















Here,



denotestheprojectednominalshort‐terminterestrate,



∗

istheprojectedlevelofthereal

short‐terminterestrateconsistentwithfullemploymentoverthelongerrun(i.e.,theequilibriumreal

rate),istheprojectedinflationrate,

∗

istheperceivedinflationtargetoftheFOMC,and



isthe

projectedoutputgap.Asnotedabove,theextendedBlueChip surveyprovidesdataonallthetermsin

thisexpressionexceptthecoefficientstobeestimated—αandβ—withthecaveatthatprojectionsof

the3‐monthTreasurybill

ratemustbeusedinplaceofforecastsoftheexpectedfederalfundsrate;we

thinkthisisareasonablesubstitution,giventhehighcorrelationbetweenthetwoseries.

17





Aswritten,equation(1)allowsforshiftsovertimein publicperceptionsoftheFOMC’sresponsiveness

tomovementsinrealactivityandinflatio n.Toprovideabenchmarkofprivateforecasters’beliefsabout

thereactionfunctionpriortothefinancialcrisis,weestimatetheparametersαandβovertheperiod

fromthe

early1990s,whenthelong‐horizonBlueChipforecastsfirstbecomeavailable,totheeveofthe

recessionin2007,treatingbothparametersasfixedcoefficientsinthecontextofapanelregression

involvingpooledBlueChipconsensusprojections ofv a ryinghorizonspublishedatdifferenttimes.



TheBlueChipalsoreportsconsensusforecastsfortheoverallCPI,whichmightatfirstglancebethoughtofas

moresimilarinnaturetooverallPCEprices—theFOMC’spreferredmeasureofinflationoverthelongerterm—

thantheGDPpricemeasureweuseinouranalysis.However,insettingmonetary

policy,theFOMChas

traditionallysoughttolookthroughtemporarypricefluctuationsandinsteadfocusonunderlyinginflationtrends

by,amongotherthings,closelymonitoringthegenerallymorepersistentmovementsincorePCEinflationaswell

asmeasuresoflong‐terminflationexpectations.Giventhepronouncedshort‐runeffectsof

transitoryfoodand

energypricefluctuationsontheoverallCPI,researchershavetendedtouseeithercore(CPIorPCE)inflationor

GDPpriceinflationwhenestimatingtheFOMC’sreactionfunction.TheBlueChipsurveydoesnotreportforecasts

forcoreinflation.

Inparticular,thisbehaviorhasbeennotedbyTaylor(1993,1999)andTaylorandWilliams(2011).

Inouranalysis,thepredictedvalueoftheequilibriumrealinterestrateattimetisassumedtoequalthelong‐

runBlueChipforecastforthenominalTreasurybillratelessthelong‐runprojectionforGDPpriceinflation.

Similarly,theperceivedvalueoftheFOMC’sinflationtargetis

assumedtoequalthelong‐runprojectionforGDP

inflation.

Page9of54



Interestingly,theestimatedparametersareveryclosetothosesuggestedbytheoriginalTaylor(1993)

rule.



Bycontrast,onecouldplausiblyexpecttheFederalReserve’sunconventionalpoliciesoverthepastfew

yearstohavegraduallyalteredpublicperceptionsofthefutureresponsivenessoftheFOMCtochanges

ineconomicconditions.Time‐varyingperceptionsoftheFOMC’sresponsetoeconomicconditions

couldtakedifferentforms;theycoul d

alteroneorbothoftheresponsecoefficientsαandβ,orthey

couldhavetemporarilyreducedtheinterceptoftheperceivedpolicy rulebelowthelonger‐run

equilibriumrealrate



∗

.GiventheemphasisplacedonexpectedlabormarketconditionsbytheFOMC

initscommunications,wefocushereontimevariationintheeconomicslackparameterwhilekeepingα

fixedandsettingtherule’sintercepttothesurvey‐impliedlong‐runequilibriumrealrate.Inparticular,

onewouldexpectthevalue

ofβconsistentwiththeBlueChipprojectionstoriseovertime.





Accordingly,wedonotpoolresults fromdifferentsurveysinourregressionanalysisforthepost‐

recessionperiodbutinsteadestimateseparatecoefficientsforeachMarchandOctobersurveyreleased

from2009through2013.Giventhatthesurveysreportannualprojectionsforonlythecurrentyear

throughsixyearsahead

(plusanassessmentoflong‐runconditions),thenumberofobservations

employedineachindividualregressionisnecessarilysmall.Toimprovethedegreesoffreedomofthis

exercise,wefixthevalueofαinalltheregressionsat0.5,itsvalueinthestandardTaylorruleand,as

discussedabove,consistent

withtheaveragebehavioroftheFOMCperceivedbyprivateforecastersin

theyearspriortothefinancialcrisis.Duetothecomparativelyminordeviationsofinflationfromthe2

percentobjective,andthefactthatthesedeviationswereexpectedtocloserelativelyquickly,allowing

fortimevariationinthe

coefficientαwouldinanycasenothavehadlargeeffectsontheexpected

futurepathofshort‐terminterestrates

Inlightofthelimitednumberofobservations,wealsodonot

considertemporarydeviationsoftheinterceptfromthelong‐runequilibriumrealrate.



Table3summarizestheresultsfromthisregressionanalysis.Whenannualprojectionsatallpublished

forecasthorizonsareincludedintheregression(thefirstcolumn),

theestimatedcoefficien tonthe

outputgaprisesmarkedlyovertime,fromlessthan0.3inthe2009and2010surveystomorethan0.9

intheOctober2013survey.However,theshiftintheparameterestimatesovertime is biased

downwardsincethenear‐termannualprojectionsofthe

Treasurybillrateinmanyofthesurveysare

constrainedbythezerolowerbound(ZLB),whichcausestheactualnear‐termprojectionsofshort‐term

interestratestodeviatefromwhattheunconstraine d policyrulewouldotherwiseprescribebyan



Forexample,inapooledsampleofallMarchandOctoberBlueChipsurveystakenfrom1992through2007,the

estimatedcoefficientsforαandβare.517(.184)and.424(.063),respectively.(Standarderrorsinparentheses.)

Inthisregression,dummyvariablesareincludedforthezero‐step‐aheadforecaststocontrol

forthefactthatthe

observationpartlyreflectsactualdata.

Inprinciple,perceivedchangestomonetarypolicyintentionscouldmanifestthemselvesasanticipatedfuture

shockstoanunchangedFOMCimplicitpolicyrule,inplaceoforinconjunctionwithchangesintherule’s

coefficients.InouranalysisoftheBlueChipprojections,however,werestrictourselvestocoefficientshiftsfor

reasonsofparsimonyandbecausesuchshiftsbettercapturethesystematicchangesinpolicymanifestedinBlue

Chipprojectionsthatextendsixyearsintothefuture.Incontrast,Campbelletal(2012)useBlueChipquarterly

forecastsforthecurrentandcomingyeartoestimateshockstoanunchangedrule,possibly

becausetheiruseof

shorter‐horizonprojectionswouldnotpermitestimationof time‐varyingpolicyrules.



Forexample,iftheinflationcoefficientintheperceivedpolicyrulehasremainedabout0.5,thenthemarginal

effectoftheprojectedinflationgapontheTreasurybillrateisonlyontheorderof25basispointsonaveragein

thefirstyearofprojectionsmadesinceearly2009,

withtheeffectdiminishingrapidlywiththeextensionofthe

forecasthorizonthereafter.

Page10of54



increasingamountovertime.Tomitigatethisproblem,thesecondcolumnofthetablereportsresults

whentheregressionsonlyincludeforecastsinwhichtheprojectedTreasurybillrateisgreaterthan0.3

percentagepoint,acut‐offpointconsistentwithafederalfundsrateabove

itscurrenttargetrangeof0

to25basispoints.(Inpractice,thisscreeninginvolvesdroppingthecurrent‐yearand,inthelater

surveys,theone‐year‐aheadprojections.)Inthiscase,theincreaseovertimeintheestimatedvalueof

βimplicitintheconsensusBlueChipprojectionsiseven

morepronounced,climbingfromabout0.15in

theMarch2009surveyto1.60intheOctober2013survey.Basedontheseresults,onecouldjudgethe

effectivenessoftheFederalReserve’sunconventionalpolicyactionsininfluencingpolicyexpectationsas

considerable,particularlyaftertheFOMCbeganissuingmoreexplicitforwardguidancein

August2011.



Figure2illustratestheimplicationsfortheprojectedtrajectoryofshort‐terminterestratesofthese

inferredshiftsintheexpectedpolicyreactionfunction.Tobegin,theblacklinesshowtheevolutionof

theactualBlueChipprojectionsoftheTreasurybillratepublishedoverthe2009‐2013

period,whilethe

dashedredlinesshowwhatwouldhavebeenprescribedbythes tandardTaylor(1993)ruleconditional

ontheaccompanyingBlueChipprojectionsforeconomicslack,inflation,theequilibriumrealrate,and

theFOMC’slong‐runinflationtarget.Asthefigureshows,privateforecastersinearly2009anticipated

that

short‐terminterestrateswouldbegintorisemuchearlierandfasterthanpredictedbytheTaylor

(1993)rule;thisprojectedrapidtighteningmayhavereflectedforecasters’beliefthattheFOMCwould

bereluctanttokeepthefundsrateatalevelwellbelowanythingseensincethe1930s,even

ifsucha

lowratewasconsistentwithperceptionsoftheFOMC’spastbehaviorasapproximatedbytheTaylor

rule.ThispatternpersisteduntilOctober2011,whentheBlueChipprojectionsbecamereasonably

well‐alignedwiththestandardTaylorrule.Afterthisdate,privateforecastersappearedtoanticipate

thattheFOMC

wouldbemorepatientthanthestandardTaylorruleimpliedforpolicyastheeconomy

recovered,toadegreethatincreaseda ppreciablyovertime.Theestimatedchangesinpolicy

expectationsunderlyingtheBlueChipprojectionsisreasonablywellapproximatedbythebluedashed

line,whichshowsthepolicyprescriptionsfrom

asimpleruleinwhichtheperceivedvalueofβrevises

overtimeasreportedinthesecondcolumnofTable3,resulting—asonewouldexpect—inmuchbetter

trackingperformance.



Nonetheless,therulesestimatedwithannualdatahavedifficultyaccountingforprivateforecasters’

expectationsforthetimingofliftoffasreportedin

both2009surveysandintheMarch2010andthe

March2011surveys.Thesediscrepancies,plustheemphasisplacedinFOMCcommunicationsonthe

conditionsthatwouldwarrantkeepingthefederalfundsratenearzero,suggesttakingacloserlookat

BlueChipprojectionsofunemploymentandinflationatthe

timeofliftoff,asisdoneinTable4.Thefirst

fewcolumnsofthetablesummarizetheconsensusforecastsforeconomicconditionsinthequarter

duringwhichprivateforecastersimplicitlyexpectedthefederalfundsratetofirstriseaboveitseffective

lowerbound.

Intheimmediateaftermathofthefina ncialcrisis,privateforecastersanticipatedthat

liftoffwouldoccurinlate2009eventhoughtheunemploymentratewouldstillbeabove9percentand

GDPpriceinflationwouldbeonly1percent.Astimepassed,however,privateforecasterspushedthe



AlthoughtheBlueChipEconomicIndicatorssurveydoesnotpublishfundsrateprojections,forecastsofthe3‐

monthTreasurybillratecanbeusedtoinfertheexpectedliftoffdate.Inparticular,thebillrateshouldequalthe

expectedaveragevalueofthe(overnight)federalfundsrateoverthenext

91days,abstractingfromtax

considerationsandtransitorysafe‐haveneffects.Giventhattaxeffectsareessentiallyzerowhentheinterestrates

areverylow,andgiventhatsafe‐havenconcernsarepresumedtobeapproximatelynormalwhentheFederal

Reservebeginsraisingthefederalfundsratetarget,weassumethat

iftheprojectedT‐billratefirstrisesabove0.3

percentinquartert,thenliftoffisexpectedtooccurinquartert+1.Asdiscussedintheappendix,thisassumption

appearsreasonablyconsistentwithfinancialmarketparticipants’forecastsoftheliftoffdateforthefederalfunds

rate.

Page11of54



projecteddatefortheonsetoftighteningfurtherandfurtherintothefuture,presumablyinresponseto

adisappointinglyslowpaceofrecovery,continuedmoderateinflation,asequenceoflarge‐scaleasset

purchaseprograms,andincreasinglyaggressiveforwardguidance.Critically,theyreviseddownsharply

theirexpectations

fortherateofunemploymentthatwouldprevailatlif toff toabout6½percentbylate

2013,alevelbroadlyconsistentwiththeFOMC’sannounce dunemploymentthreshold,whilegradually

revisinguptheirexpectationsfortheaccompanyingrateofinflationtoabout2percent.





Holdingthevalueofαconstantat0.5,andassumingthattheFOMCraisesthefederalfundsrateto50

basispointsatliftoff,wecanestimatethevalueofβconsistentwiththeprojectionsofeconomic

conditionsatliftoff(t=LO)inaspecificBlueChipforecastas:



(2) 



0.5  

∗





0.5









∗





2









∗





⁄



ThefinalcolumnofTable4reportstheresultsfromthiscalibrationexerciseusingtheprojectedvalues

fortheunemploymentrateandinflationatliftoffreportedinthesecondandthirdcolumnsofthetable

andthelong‐runBlueChipforecastsfortheunemploymentrate,inflation,and

therealfundsrate

reportedinthefourththroughsixthcolumns.Acomparisonoftheseresultswiththefinalcolumnof

Table3revealssimilarrevisionsinβovertime.Forexample,theimplicitoutputgapcoefficientatliftoff

climbsfrom0.25intheMarch2009surveyto1.6intheOctober

2013survey,achangealmostidentical

tothatobtainedusingannualprojections.Moreover,mostofthecumulativerevisiontoβoccursafter

early2011,asisthecasewiththeannualestimates.Finally,Figure2illustratesthat,fromtheOctober

2011surveyon,theprescriptionsoftheestimatedliftoffrules(dashed

greenlines)arequitesimilarto

onesgeneratedusingtheestimatedan nualrules(dashedbluelines).Priortothatsurvey,however,it

appearsthatprivateforecastersanticipatedthat thebehavioroftheFOMCwouldbenoticeablymore

accommodativeatthetimeofliftoffthanitwouldbeinthelongerrun.

(Asinthecaseoftheannual‐

basedestimates,theshiftsintheexpectedsensitivityofthefederalfundsratetoslackatthetimeof

liftoffwouldnotbemateriallyalteredifwehadinsteadassumedanaccompanyingdeclineinthe

coefficientontheinflationgap,giventhatprivate

forecastersfromOctober2011onconsistently

expectedinflationtobe closetoitslong‐runlevelatthetimeofliftoff.)



Sofar,wehavemadenoallowanceinouranalysisforthepossibilitythatprivateforecastersmayhave

expectedtheFOMCtobehaveinertially,inthatwehavenot

includedthelaggedTreasurybillrate inour

regressionsasfollows:



(3) 















1









∗

















∗

















TheBlueChipforecastsreleasedduring2013,aftertheinitialannouncementofthresholds,mayinfactbemore

consistentwiththeFOMC’sforwardguidancethanimpliedbytheestimatesreportedinTable4becausethelatter

areaffectedbymodestquarterlyimputationerrorsbeginningwiththeOctober2011release.Beginning

withthat

release,privateforecastersanticipatedthatliftoffwouldoccursometimeintheyearfollowingthecomingyear.

BecausetheBlueChipsurveyprovidesquarterlyforecastsforthecurrentandcomingyearonly,theexact

quarterlytimingofliftoffanditsaccompanyingconditionsintheOctober2011releaseandbeyondmust

be

imputedusingannualprojections.Theseimputationsaresomewhatimprecise,particularlywithregardstothe

federalfundsrate,whichissubjecttothenon‐lineareffectivelowerboundandtendstoberaisedindiscretesteps

bytheFOMC.Asaresult,theliftoffdateactuallyanticipatedbyprivateforecasters

couldbeaquarterortwolater

thanreportedinthetable,implyingthattheaccompanyingunemploymentratecouldbeasmuchas¼percentage

pointlower.Theestimatedinflationrateatthetimeofliftoffwouldbelittlechanged,however.

Page12of54



Forforecastspublishedfrom2009through2011,thereappearstobelittleevidenceofinertia.For

forecastspublishedfrom2012on,however,thestoryissomewhatdifferent.Ifλisassumed

toequal0.8onaquarterlyfrequency,andβiscalibratedasshownintheright‐most

columnofTable3,

thentheresultingruledescribes theBlueChipforecastsaboutaswellasthenon‐inertialestimatedrule.

ThetrackingperformanceofthiscalibratedruleisillustratedbytheorangelineinFigure2.



Thereareseveralcaveatstotheresultspresentedhere.First,ouranalysisassumes

thatprivate

forecastersbasetheirprojectionsofshort‐terminterestratesonasimplepolicyruleinvolvingonlythe

outputgap,inflation,andassessmentsoftheequilibriumrealrateandtheFOMC’sinflationtarget.In

practice,however,theirprojectionsmightincorporateotherinfluencesaswell,suchasthecyclical

growth

rateofrealGDP,thehealthoffinancialinstitutions,oraperceptionthateconomy’sequilibrium

realrate(theinterceptintherule)istime‐varying.Second,ourapproximationfortheprojectedoutput 

gapimplicitintheBlueChipsurveycouldmisstateprivateforecasters’actualviewsaboutthelikely

trajectoryforfuture

resourceutilization,aswouldoccurifforecastersseethenaturalrateof

unemploymentasvaryingovertheprojectionhorizo nandthereforenotequivalenttotheirprojections

oftheunemploymentrateinthelongerrun,oriftheyanticipatevariabilityintheOkun’sLaw

relationshipperhaps,forexample,relatedtounusual

projectedmovementsintrendlaborforce

participationandtrendpr oductivity.Third,evenifourresultsdoaccuratelygaugetheevolutionofthe

monetarypolicyviewsofprivateforecasters,thoseviewsmaynotconformcloselytotheviewsof

households,firms,andthefinancialmarketsmorebroadly.(Asdiscussedintheappendix,

however,

forecastsmadebyfinancialmarketparticipantsdoseembroadlyconsistentwiththosepublishedin the

BlueChipEconomicIndicators.)Nonetheless, weviewtheresultsaspointingstronglytoamarkedshift

intheperceptionsofprivateforecastersoftheFOMC’swillingnesstofollowahighlyaccommodative

monetarypolicy.



IV.

SpecifyingtheEffectsofFederalReser veQuantitativeEasingonTermPremiums



Inadditiontoestimatingshiftsinperceptionsoftheimplicitpolicyreactionfunction,ouranalysisalso

needstotakeintoaccounttheothermonetary“impulse”fromunconventionalpolicy—thechangesin

termpremiumsassociatedwiththeFederalReserve’squantitativeeasing

programs.Forestimatesof

theseeffects,wedrawontheconsiderablebodyofworkthathasemergedinrecentyearsconcerning

thefinancialmarketeffectsoflarge‐scaleassetpurchases.Usingavarietyoftechniques,anumberof

studiesfindonbalancethattheFOMC’sassetpurchaseshavesucceededinreducing

yieldsonlonger‐

termTreasurysecuritiesandagencyMBSbyanappreciableamount,andsomehavefoundnoticeable

reductionsincorporatebondyields.

Inmanyofthesestudies,however,thereportedredu ctionsin

yieldsmayreflectamixofchangedexpectationsaboutthefuturepathofshort‐terminterestratesand

policy‐drivendeclinesin termpremiums.



Forouranalysis,weusethetermpremiumestimatesreportedbyIhrigetal.(2012)becausethose

estimatesarederivedusingamethodologythatallowstheeffectsofshiftsintermpremiumsonlong‐

termTreasuryyieldsduetochangesinquantitativeeasingpoliciestobeisolatedfromotherinfluences,

includingtheeffectsofchangesinexpectationsforthepathofthefederalfundsrate.Indeed,their

resultshaveseveraladvantagesoverotherresearchinthisarea.Toestimatethetermpremiumeffects



ThesestudiesincludeBauerandRudebusch(2012),D’Amico,English,Lopez‐SalidoandNelson(2013),D’Amico

andKing(2013),Gagnonetal(2011),HamiltonandWu(2012),JoyceandTong(2013),KrishnamurthyandVissing‐

Jorgensen(2011),LiandWei(2013),MeaningandZhu(2011),Neely(2010),Rosa(2012),Rogers,Scottiand

Wright

(2014),andSwanson(2011);alongwithKiley(2012)andGilchrist,López‐Salido,andZakrajšek(2014).

Page13of54



ofthequantitativeeasingprograms,Ihrigetalrelyheavilyontheempiricalarbitrage‐freeterm

structuremodeldevelopedbyLiandWei(2013).Thismodelpresumestheexistenceofpref erred‐

habitatinvestorsofthesortdescribedbyVayanosandVila(2009),whichinturnimplies

thattheslope

oftheyieldcurveshouldrespondtosu pplyfactorssuchaschangesinthesizeandmaturitystructureof

theoutstandingstockofTreasuryandagencysecuritiessuppliedtothepublic.





Ihrigetal.adjustthemethodologyoriginallyemployedbyLiandWeitoscorethetermpremiumeffects

ofLSAP1andLSAP2byusingtheFOMC’snormalizationplansannouncedinmid‐2011,whichlaidoutthe

relationshipbetweenthetimingofliftoffofthefederalfundsratefromthelower

boundontheone

hand,andtheendofreinvestingmaturingassetsfromtheSOMAportfolioandthetimingofassetsales

ontheother.TheycombinethisinformationwithBlueChipexpectationsforthetimingofliftoffto

arriveatmarketexpectationsforthepathofthebalancesheet.

Ihrigetal.alsoemploytheLi‐Wei

methodologytoscoretheeffectsofthematurityextensionprogramsthattheFederalReserveinitiated

inSeptember2011andco mpletedinlate2012:ByreducingtheaveragedurationoftheTreasury

securitiesheldbythepublic,MEP1andMEP2putfurthermodest

downwardpressureonterm

premiumsevenasthesizeoftheFederalReserve’sportfolioremainedunchanged.

26





Figure3plotstheIhrigetal.estimatesofthetermpremiumeffectsofthequantitativeeasingprograms.

Theblacklineindicates thatthefirstroundofassetpurchasesisestimatedtohaveloweredtheterm

premiumon10‐yearTreasuryyieldsroughly40basispointsonaverageatthe

startoftheprogramin

early2009.

Thereafter,thetermstructuremodelpredictsthatthedownwardpressureontheterm

premiumunderthefirstLSAPprogramwouldbeexpectedtodiminishsteadilyastheaveragematurity

oftheFederalReserve’sholdingsdeclined,thescale oftheeconomyincreased,andthe

contemporaneousBlueChipprojectionsforthedate

ofliftoffofthefederalfundsrate—whichwould

signalthestartofactivebalance‐sheetnormalization—drewnearer.Asindicatedbythegreenline,the



UsingmonthlydatafromMarch1994throughJuly2007onvolumesandyieldsofTreasurysecuritiesandagency

MBS,LiandWeifindstrongempiricalsupportforthistheoreticalproposition.Theyalsousetheirestimatedmodel

toinferthetermpremiumeffectsofchangesinthesizeandcompositionof

publicholdingsofTreasurysecurities

andagencyMBSofthesortthatresultedfromthefirsttwoLSAPprograms,andconcludethattheywerequite

large.

WhereasLiandWei’sestimatesoftheeffectsofthefirsttwoassetpurchaseprogramswerebasedonthe

assumptionthattheFOMCwouldbeginsellingassetstwoyearsaftertheendofeachprogramandconcludesuch

salesoveraspanofthreeyears,Ihrigetal.tiethe

beginningofsalestotheexpecteddateofliftoffbasedonBlue

Chipsurveyexpectations,whichatthetimewasexpectedtooccursoonerthantwoyearsaftertheconclusionof

eachprogram,therebyreducingthetermpremiumeffects.

AnassumptionmaintainedthroughoutouranalysisisthattheTreasurydidnotchangeitsstrategyforthe

maturitycompositionofitsissuanceofnewsecuritiesinresponsetotheFOMC’sassetpurchases.TheTreasury’s

debtmanagementstrategyhadbeenpersistentlyincreasingtheaveragematurityofnewly‐issuedTreasury

securitiessince

around2002.AlthoughtheTreasury’sstrategyhadtheoppositeeffectontheaveragedurationof

outstandingTreasurysecuritiesastheFOMC’sQEpolicies,theTreasuryhadbeguntopursueitsstrategylong

beforethefinancialcrisisandcontinuedwithitevenwiththeFOMC’sdecisiontoexpandthesizeandduration

of

theFederalReserve’ssecurityholdings,whichwasconsistentwiththeTreasury’susualpracticewhenfacedwith

thelargeincreasesinfederalgovernmentborrowingduringthisperiod.Accordingly,themarginaleffectof

quantitativeeasingwasstilltoreducetermpremiumsrelativetowhattheyotherwisewouldhavebeen.(See

Greenwood,

Hansen,Rudolph,andSummers(2014)foramorecriticalviewoftheTreasury’sdebtmanagement

practicewhiletheFederalReservewaspursuingitsQEprograms.)

Aportionofthiseffectoccursearlyin2009becausetheFederalReservemadeitsfirstannouncementabout

assetpurchasesattheveryendof2008.Thefullscopeofthefirstassetpurchaseprogramwasnotknownuntil

afterthereleaseoftheFOMCstatementinlateMarch2009.

Page14of54



projectedtrajectoryoftermpremiumeffectsareestimatedtoshiftdownslightlywiththeFOMC’s

announcementthatpaymentsontheFederalReserve’sholdingsofsecurities(particularlyMBS)would

bereinvested,slowingtherateofdeclineinthesizeandaveragedurationofitsportfolio.

Commencementof

asecondroundofassetpurchasesinNovember2010isestimatedtoprompta

furtherdownwardrevisionintheprojectedpath,whichalsoincorporatestheeffectsoftheassets

purchasedunderLSAP1(solidredline).Importantly,theinitialrevisionintermpremiumeffectsatthe

startofLSAP2isfollowedby

furtheronesasBlueChipforecastsforthedateofliftoffarepushedfurther

offintothefuture(thereddashedlines).Finally,theIhrigetalresultsindicatethatthematurity

extensionprogramsfirstincreasedandthenmaintainedthecumulativedownwardpressureonterm

premiumstoabout60basis

throughmid‐2012.



ThethreepurplelinesshowtheextensionoftheIhrigetal.methodologytoincludetheeffectsofthe

thirdroundofassetpurchasesthatbeganinSeptember2012.Theimplementationofthisprogramis

estimatedtohavegreatlymagnifiedthedownwardpressureontermpremiums,bringingit

closeto120

basispointsatthestartoftheprogram.Marketparticipantsinitiallyexpecteditssizetobeabout$1.2

trillion,substantiallysmallerthanthefirstassetpurchaseprogram;nonethelessitseffectswere

substantiallylarger,inpartbecausepurchasesofTreasurysecuritieswereconcentratedinlonger

maturitiesthanthey

hadbeeninLSAP1,andinpartbecausetheexpansionoftheSOMAholdingswas

expectedtolastlongerthantheoriginalpurchasesin2009.Asshownbythefirstdashedline,the

projectedlonger‐runtrajectoryofthecumulativeeffectsfromallassetpurchaseprogramsshifteddown

inresponseto

FOMCcommunicationsindicatingthatitwouldprobablyeschewactivesalesofitsMBS 

holdingsonceitbeganshrinkingthesizeofitsbalancesheet,incontrasttotheCommittee’ s previous

guidance.Finally,marketexpectationsgraduallyshiftedtowardtheprogram’sultimatesizeof$1.5

trillion.



Becausedurationeffectsshouldvaryacross

securitiesofdifferentmaturities,theresultsreportedin

Figure3applyonlytothetermpremiumsembeddedin10‐yearTreasuryyields.Simulationsofthe

Li‐Weitermstructuremodelpredictthattheeffectsofthevariousassetpurchaseprogramsontheterm

premiumsembeddedin30‐yearTreasuryyieldsshould

beonaverageonly35percentorsoofthese

estimatedfor10‐yearsecurities;thecorrespondingfigurefor5‐yearTreasurynotesisabout80

percent.





V. EstimatingtheMacroeconomicEffectsofUnconven tionalPoliciesUsingFRB/US



HavingquantifiedtheeffectsoftheFederalReserve’sunconventionalpolicyactionsonperceptionsof

itsimplicitpolicyruleandontermpremiums,wenowturntotheissueoftheeffectsoftheseactionson

aggregateeconomicactivityandinflation

overthepastfewyears.Thissectiondiscussesourchoiceofa

structuralmacroeconometricmodelforsimulatingtheresponseoftheeconomytounconventional

policymeasures,andthenpresentsanddiscussesresultsforsomeillustrativepolicyactions.



TheFRB/USmodel



Toquantifythemacroeconomiceffects ofunconventionalpolicies,weneedamacroeconomic

modelthatisstructuralinthesensethatitcanappropriatelycontrolfortheeffectsofpolicy

measuresthroughexpectations.Infact,asouranalysiswillhighlight,expectationsaboutthefuture

courseofpolicyarecriticallyimportantfortheeffectivenessoftheunconventionalpolicyactions



WethankMinWeiforprovidinguswiththesefigures.

Page15of54



studiedhere.WhilethisgeneralpointhasbeenappreciatedsinceLucas’(1976)paper,thisissueis

ofparticularrelevanceinthecircumstancesofthepastfewyears,whenthepublichashadtorely

disproportionatelyonpolicyannouncementsforformingexpectationsbecausetherewasnopast

recordofunconventionalpolicyactionsattheZLBtogoby.



Asmentionedearlier,onlyafewstudieshaveexaminedthemacroeconomiceffectsof

unconventionalpoliciesinstructuralmodels.Thesestudiesaremostlyconductedusingdynamic

stochasticgeneralequilibrium(DS GE)modelsbuiltaroundarepresentativehouseholdandassuming

someformoffinancialfriction,suchasinChenetal.(2012)andGertlerandKaradi(2013).Inour

studyweusetheFederalReserveBoard’sFRB/USmodel,whichdepartsfromtheseothermodelsin

certaindimensionsthatwewillbrieflydiscusshere.MoredetailedinformationabouttheFRB/US

modelingeneralisavailableelsewhere.



ThreepropertiesofFRB/USinparticularareworthhighlighting.First,theFRB/USmodelfeaturesa

broadarrayoffinancialassetswhosepricescanserveaspotentialtransmissionmechanismsfor

monetarypolicy:Besidestheshort‐terminterestratethatisthe(conventional)toolofmonetary

policy,therearelong‐termTreasurysecuritiesatthreedifferentmaturities,residentialmortgages,

corporatebonds,corporateequity,andtherealtrade‐weightedexchangerate.Thepricesofthese

variousassetsarelinkedtoeachotherviaarbitrageconditionsthatin cludetermandriskpremiums

whicharethemselvesmodeledasendogenousvariables,dependinginmanyinstanceson

expectationsaboutthefuturecyclicalpositionoftheeconomy.InthebaselineversionofFRB/US

usedinthispaper,termpremiumsareaffectedonlybyunconventionalmonetarypolicymeasures,

whereasriskpremiumsonotherassetsareendogenous.Thebroadarrayoffinancialassetsin

FRB/US,andinparticularthecentralroleplayedbyspecificmeasuresoflong‐terminterestrates,

makesthemodelveryusefulforstudyingtheeffectsofunconventionalmonetarypolicy,especially

assetpurchases.Thatsaid,thereliabilityofitssimulationresultsdependsimportantlyonthe

empirical“reasonableness”ofitspredictionsforQE‐relatedspillovereffectsonprivateborrowing

rates,corporateequityprices,andtheexchangerate—somethingthatwewillexamineshortly.

Fromsomeperspectives,theapproachtakeninFRB/USmayhavecertainlimitations:Practicallyall

spendingdecisionsdependonlonger‐termrealinterestrates,therebyassigningequallystrong

effectsonaggregatedemandtochangesintermpremiumsandinexpectedfutureshortrates;and

premiumsaremodelledinreduced‐formratherthanbeingderivedexplicitlyfromfinancial

frictions.





Asecondfeatureisthatthereisnorepr esentativeagentintheFRB/USmode l;insteadthereare

bothliquidity‐constrainedandunconstrainedhouseholds,wheretheformerspendalltheirincome

eachquarterandthelatterconsumeandinvestbasedontheirassessmentoftheirlifetime

resources.Inmakingthisassessment,unconstrainedhouseholdsdiscountfuturelaborandtransfer

incomeataratesubstantiallyhigherthanthediscountrateonfutureincomefromnon‐human

wealth,reflectinguninsurableindividualincomerisk.

Notably,aggregatefutureincomeisvalued



Documentationofthemodelisavailableatwww.federalreserve.gov/econresdata/frbus/us‐models‐about.htm,

includingacompletelistingofmodelequationsandcoefficients,papersdiscussingdifferentaspectsofFRB/US,

andillustrativesimulationprograms.

Kiley(2012)presentsevidencethattheaggregatedemandeffectsofchangesintermpremiumsmaybeweaker

thanthoseofexpectedfutureshortrates.

Themarginalpropensityofhouseholdstoconsumeoutofdifferenttypesofincomecanvary,dependingon

whichgroupofhouseholdsreceivestheincome.Forexample,transferincomeisdisproportionatelyreceived

Page16of54



bydifferentdiscountfactorsdependingonitssourceandtheaverageageoftherecipient

household.Asaresult,theeffectiveplanninghorizonfortheaveragehouseholdinFRB/USisclose

tothefive‐yearperiodadvocatedbyFriedman(1957)ratherthanthemuchlongertimehorizonof

householdsembeddedinatypi calDSGEmodel.Thismodelingchoicereduces,inourview

realistically,theabilitytoaffectcurrenteconomicactivitythroughpoliciesthatraiseexpectationsof

futurerealactivityandincomes(forexample,asanalyzedinWerning(2012)).Theabsenceofa

representativehouseholdalsomeansthatwecannotevaluateunconventionalpoliciesaccordingto

theirwelfareimplications;instead,wesimplyreporttheirestimatedeffectsonrealeconomic

activityandinflation.



Finally,priceandwageinflationdynamicsfollowaNewKeynesianPhillipscurvespecificationinthe

presenceofnonzerotrendinflation.

WhilethisspecificationisverysimilartothoseusedinDSGE

models,theestimateddegreeofinertiainpriceandwageinflationishigherthaninmostDSGE

modelsbasedonthesticky‐priceparadigm.Thisfeatureinteractswithhouseholds’highdiscount

rateoffutureincomediscussedbeforetodampentheeffectsofannouncedfuturemonetarypolicy

changes,suchasforwardguidance.Inparticular,thecombinationofthesetwofeaturesexplains

why,asourresultswillshow,theFRB/USmodeldoesnotexhibitthe“forwardguidancepuzzle”that

besetsstandardDSGEmodels(DelNegroetal.,2013).

AsshowninAppendix2,inasituation

whenthenominalshortrateisheldconstantforanextendedperiodoftime,forexampl ebecause

thezerolowerboundisbinding,acredibleannouncementtodayoflowershort‐termratesinthe

futureraisescurrentoutputandinflationinacanonicalDSGEmodel, andtheseincreasesarelarger

thefurtherintothefutureistheannouncedaction.Givenconstantnominalinterestrates,the

increaseininflationreducescurrentandexpectedfuturerealinterestrates,whichinturnhasstrong

aggregatedemandeffectsbecauseofthehigherinterestratesensitivityduetohouseholds’longer

planninghorizons.Bothofthesechannelsaresubstantiallyand,inourview,plausiblyattenuatedin

FRB/US.



Monetarypolicyismodeledasasimpleruleforthefederalfundsratesubjecttothezerolower

boundonnominalinterestrates;importantlyforouranalysis,theparametersofthepolicyruleused

insimulationscanbemodifiedasdesiredtobeconsistentwithprivatebeliefsaboutthelikelyfuture

behavioroftheFOMC.Inourcounterfactualsimulations,thepronouncedweaknessofthe

economysincelate2008meansthattheZLBmarkedlyandpersistentlyconstrainsactualand

expectedmonetarypolicy.Moreover,theimpositionoftheZLBmakesthemodelhighlynon‐linear,

therebyprofoundlyinfluencingthesimulateddynamicsoftheeconomy.





InourbaselineversionoftheFRB/USmodel,allprivateagentsareassumedtohaverational

expectations—thatis,beli efswhichareconsistentwiththedynamicsofthefullmodel,conditionalon

theanticipatedbehaviorofpolicymakers.Thepublicdoesnotnecessarilyhaveperfectforesightabout



byretireeswhoarewell‐advancedintheirlifecycles.Inaddition,aggregationacrossagegroupsleadsto

averagepropensitiestospendthatvaryacrossdifferenttypesofaggregateincome,reflectingvariationsinthe

distributionofincomeacrossgroups.

ThespecificationofthePhillipscurveusedinFRB/USisbasedonthemodeldevelopedbyCogleyandSbordone

(2008);thismodelallowsfortimevariationintheunderlyingtrendininflation,reflecting(forthemostpart)

changesinthecentralbank’sinflationobjective.InFRB/US,thepost‐1979trend

isbasedonmeasuresofexpected

long‐runinflationasreportedinitiallyintheHoeySurveyandlatertheSurveyofProfessionalForecasters.

SeeChung(2015)foracomparisonofthemacroeconomiceffectsofforwardguidanceintheFRB/USmodel

comparedtotwostandardDSGEmodels.

Tosolvethemodel,weusedthesolutionmethoddescribedbyBrayton(2012).

Page17of54



theshocksthatwillhittheeconomyinthefuture,however,northeFOMC’sresponsetothoseshocks.

Whileweviewanexaminationoftheeffectsofunconventionalmonetarypolicyinarangeofalternative

modelsasaworthwhileundertaking,itisbeyondthescopeofthispaper.Instead,wewillreportresults

oftheseeffectsforarangeofalternativeparameterizationsofFRB/US,suchasalternativeassumptions

abouttheformationofexpectationsbyvariousagents,aboutinflationdynamics,andabouttheinterest

elasticityofaggregatedemand.



Illustrativesimulations



Beforeturningtoestimatesoftheeffectsoftheunconventionalpolicyactionsactuallyundertakenby

theFederalReserve,weillustratethemacroeconomiceffectsofforwardguidanceandassetpurchases

inahypotheticalscenarioinwhichtheU.S.economyishit withasequenceofverylargeandpersistent



contractionaryshocks.Inordertoillustratesomekeyfactorsthatinfluencethenetstimulusprovided

byunconventionalpolicy,andsoneedtobetakenintoaccountwhenwescoretheactualeffectiveness

oftheFOMC’sactionsinthenextsection,werunthisexerciseunderalternativeassumptionsabout

boththe

public’sunderstandingofthepersistenceoftheeconomicdownturnandthespeedatwhich

policymakersrespond.



Thescenarioiscalibratedsuchthat,underabaselineassumptionofthefederalfundsratefollowingthe

standardTaylorruleandwithoutanyassetpurchases,thefederalfundsrateisatitseffectivelower

boundforfiveyears.Oncetheeconomicslumpstarts,thepublicisassumedtohaveperfectforesight

abouttheshocksthatwillbuffettheeconomyovertime,aswellasthemonetarypolicyresponseto

thoseshocks.AsindicatedbytheblacklinesinFigure4,themagnitudeoftheseshocks

causesthe

unemploymentratetoroughlydoublewithinthefirsttwoyears,from5¼percentpriortotheslumpto

apeakof10¼percent,andtodeclinethereaftergradually,whileinflationdropsbymorethan1

percentagepoint.Despitetheassociateddeclineininflationexpectedoverthenext10

years,thereal

10‐yearyieldfallsbyabout2½percentagepoints,reflectingthelowerpathoftherealfederalfundsrate

overthattimehorizon.Becausepolicymakersdonotengageinlarge‐scaleassetpurchasesinthis

scenario,termpremiumsareassumedtobeunaffectedalthoughriskpremiumsoncorporate

bondsand

equityincreaseendogenously.



AstheredlinesplottedinFigure4illustrate,theseadverseeffectswouldbenoticeablymitigatedby

raisingtheoutputgapresponsecoefficientfrom0.5to1,whichwoulddelayliftofffromtheeffective

lowerboundbytwoquarters.Thereafter,thefederalfundspathis

littlechangedcomparedtothe

standardTaylorrulesimulation,butgiventhehigherpathforinflation,thereal10‐yearTreasuryyield

runsabout¼percentagepointlowerforseveralyearsinarow.Similar, albeitslightlysmaller,effects

fromthischangeinpolicyruleobtaininamoderateslumpscenario,

inwhichthefederalfundsrate

underthestandardTaylorruleisconstrainedforonlythreeyearsbytheeffectivelowerbound(theblue

andgreenlines).



Themarginalunemploymentrateandinflationeffectsofvariouspolicyrulechangesunderthesetwo

scenariosareshowninFigure5,expressedas

deviationsfromoutcomesunderthestandardTaylorrule.

Inadditiontotherulewithanoutputgapcoefficientequalto1(thereddashedlines),thefigurealso

showstheeffectsofraisingtheoutputgapcoefficientto1.5,ofraisingitto1.0butinadditionswitching

toaninertial

rulewithacoefficientof0.8onthelaggedfederalfundsrate,orofimplementinga

thresholdstrategywherebythefederalfundsratestaysattheeffectivelowerbounduntilthe

unemploymentratefallsbelow6½percentandthereafterfollowsthestandardTaylorrule.The

macroeconomiceffectsofadoptingeither

oftheserulesarelargerinthehighlypersistentslump

Page18of54



scenario(theleftpanels)thaninthemoderatelypersistentslumpscenario(therightpanels),andare

thelargestfortheinertialrule,followedbytherulewithanoutputgapresponseof1.



ThefirsttwocolumnsofTable5provideinformationaboutthechannelsinFRB/USthroughwhichthe

changeintheperceivedpolicyruleistransmittedtorealeconomicactivityandinflation.Inthehighly

persistentslumpscenario(theupperrowsofthetable),long‐termrealTreasuryyieldsdeclinemostly



becauseofanincreaseinexpectedinflationratherthanareducti onintheexpectednominalpathof

futureshort‐terminterestrates;theinflationeffectisespeciallypronouncedinthecaseofachangeto

aninertialpolicyrule(column2).Endogenousdeclinesinriskpremiumsleadtolargerreductions

inreal

corporateyieldsaswellasanotableincreaseinequityvaluations.Financialmar keteffectsareslightly

smallerinthemoderatelypersistentslumpscenario,pointingtotheimportanceofexpectationsabout

thepersistenceofadverseeconomiccircumstances;differencesfromthehighly‐persistentscenario

wouldbemoreappreciableiftheslump

wasevenmoretransitory.



Figure6highlightstheimportanceofinformationalassumptionsfortheestimatedmacroeconomic

effectsofthesepolicies.Theleft‐handpanelspresenttheresponsesofunemploymentandinflation

underthreeoftherulesthatwerereportedinFigure5,againexpressedrelativetooutcomesunderthe

standardTaylorrulewithoutasset

purchases,inthecontextofthehighlypersistentslumpscenario.As

mentionedbefore,thepublichasacompleteunderstandingofhowtheslumpwilldevelopovertime,

andmonetarypolicyrespondsimmediately tothecrisis.Incontrast,themiddleandright‐handpanels

ofthechartshowthemarginaleffectiveness

ofthesamepoli cyinitiativesunderalternativeassumptions

forprivateagents’knowledgeofthefutureandthetimingofpolicyresponses.Specifically,themiddle

panelsofFigure6reportthemarginaleffectivenessofthesameshiftsinpolicyruleswhenagents

initiallyexpectthecontractioninrealactivitytobemuch

lesssevereandpersistentthanactuallyturns

outtobethecase,andsoonlygraduallycometounderstandthattheeconomyhasenteredaprolonged

ZLBepisode.Theright‐handpanelsofthefigureshowthestimulusprovidedwhenthepublicfully

understandsthenatureoftheshockshitting

theeconomybutthecentralbankshifts toadifferent

policyruleonlythreeyearsaftertheonsetofthecrisis,andthepublicdoesnotanticipateanychangein

policybeforehand.Ascanbeseen,eithertypeofdelaynotonlyshiftsthetimingofstimulusbut

actuallyreducesits

magnitude—animportantconsiderationinanyscoringoftheefficacyoftheFOMC’s

unconventionalpolicyactions,giventhegradualnatureoftherevisionstoboththeBlueChip

projectionsandpolicyexpectationsobservedfromearly2009throughlate2013.



Figure7showsthemarginaleffectsofannouncing a$1.5trillionLSAP

programattheonsetofthecrisis,

conditioningontheinterestrateruleinplaceatthetimeoftheannouncement.Basedonsimulationsof

theLi‐Weimodel,suchaprogramwouldbeexpectedtoreducethetermpremiumembeddedinthe10‐

yearTreasuryyieldabout60basis

pointsinitially,withtheeffectfadingawaythereafteratroughlythe

averagepaceshowninFigure3.Columns3to5ofTable5illustrateforthreeoftherulesshownin

Figure7that,inFRB/US,suchareductioninpremiumsdirectlyincreasesthedownwardpressureon

Treasuryyieldsand

indirectlyinfluencesotherassetpricesthrougharbitrageeffects,therebyeasingthe

costofborrowingforhouseholdsandfirms,checkingtherecession‐drivendeclinesincorporateequity

valuationsandhouseholdwealth,andfurtherloweringtherealforeignexchangevalueofthedollar.



ThesimulationresultsarebroadlyconsistentwiththeresultsreportedinTables3and4.Inparticular,inthe

middlecolumnofTable3,theestimatedcoefficientontheoutputgaprisesabove1intheMarch2013BlueChip

survey,whentheunemploymentrateatthetimeofliftoff

isexpectedtobe6.8percent.Bycomparison,inthe

simulationinfigure5withβequalto1,theunemploymentrateatthetimeofliftoffinthehighlypersistentslump

scenariois6.4percent.

Page19of54



Formostoftherules,themarginaleffectsoftheLSAPprogramareverysimilar.Thestrongereconomic

conditionsinducedbytheassetpurchasescallforthastrongerinterestrateresponseafterliftoffunder

themore‐aggressivepolicyrules,illustratingthatinFRB/US,fundsratepolicy

andassetpurchasesactto

someextentassubstitutes.



TheanticipatedreductioninthepathofTreasurytermpremiumsassociatedwithassetpurchases

influencesmacroeconomicconditionsviatheformer’seffectsonavarietyoffinancialmarketvariables.

Exceptforthesizeofthetermpremiumeffect,whichwetakefromthe

analysisofIhrigetal.(2013),the

financialmarketresponsesshowninTable5areendogenouslygeneratedbyFRB/US.Arethemodel’s

predictionsfortheresponseoffinancialmarketconditionstoashiftinTreasurytermpremiumsinline

withtheempiricalevidence?InAppendix2,wesummarizesomeof

thefindingsofeventstudies

followingQE‐relatedevents.Althoughthereissomevarietyinresultsacrossstudies,agoodcentral

estimateisthataQE‐inducedreductionin10‐yearTreasuryyieldsby20basispointsisassociatedwitha

reductioninthe30‐yearmortgagerateofabout25

basispoints(reflectingthesizeableshareofMBS

purchasesintheFederalReserve’sQEprograms),areductioninBBBcorporatebondsofabout15basis

points,anincreaseinequitypricesofabout1¾percent,andareductionintherealexchangerateof

about¾percent.Therelativemagnitudes

ofthemodel‐generatedresponsesshownincolumns3to5

ofTable5areveryclosetotheseestimates,suggestingthatthemodelprovidesagoodapproximation

ofthisstageofthetransmissionchannel.



ThedifferencesbetweentheleftandrightpanelsinFigure7illustratethattheeffectsof

LSAPsdepend

importantlyonhowlongthefederalfundsrateisexpectedtoremainattheeffectivelowerbound,and

hencehowfarintothefuturehighershort‐terminterestrateswillbegintooffsettheeffectsofthe

LSAPs.Accordingly,ifagentsatthelaunchofaQEprogram

aresubstantiallyoverestimatingthepaceat

whichtheeconomywillrecoverandsoprojecttooearlyadateforliftoffofthefundsrate,thedirect

downwardpressureonlong‐terminterestratescausedbyassetpurchasesinitiallywillbesubstantially

offsetbyasimultaneousupwardrevisiontotheexpectedpath

offutureshort‐termrates.Similar

considerationsattendthestimulusprovidedbyanannouncedchangeintheimplicitpolicyrule.Iflabor

andproductmarketgapsarealreadyexpectedtoclosequicklyevenwithouttheadoptionofamore

accommodativerule,thenincreasingthecoefficientonslackoradoptinga

moreinertialapproachto

policysettingwillnotmarkedlyaltertheprojectedfuturepathofthefundsrate,incontrasttowhat

wouldoccurifslackwasexpectedtoremainsizeableformanymoreyears.



VI. TheEffectsoftheFOMC’sUnconventionalPolicyActions—SimulationResults



Thissectionpresentsourquantitative

assessmentsoftheactualstimulustorealactivityandinflation

providedbytheFOMC’sforwardguidanceandassetpurchasessinceearly2009,basedon

counterfactualsimulationsofFRB/US.Webeginwithadiscussionofthedesignofourcounterfactual

simulationanalysis,whichamongotherthingshastodealwithcomplications arising

fromgradual

learningonthepartofthepublicaboutfuturemonetarypolicyaswellasthelikelymagnitudeand

persistenceofthepost‐crisiseconomicslump.Usingthebaselineversionofthemodel,wefindthatthe

FOMC’sunconventionalpolicyactionsapparentlyprovidedonlyasmallboosttothe

realeconomy

duringtherecessionandtheinitialrecoveryperiodbutthattheseeffectshavesincebecomemore

substantial;thebaselinecounterfactualsimulationsalsosuggestthatinflationeffectstodatehavebeen

relativelymodest.Considerableuncertaintyattendstheseassessments,however,assimulations

generatedusingplausiblealternativespecificationsofthemodelyield

awiderangeofestimated

unemploymentandinflationeffects;inaddition,our analysisdoesnottakeintoaccountthepossibility

Page20of54



thattheFOMC’sunconventionalactionsmayhaveledtomarkedimprovementsinhouseholdand 

businessconfidence,therebysignificantlyboostingtheeconomybeyondwhatourmodel(s)predict.



Simulationdesign



Toassesstheconsequencesofannouncingasystematicchangeinmonetarypolicy,onetypicallystarts

withabaseline

embodyingsomereferencerulecalibratedtomatchtheFOMC’srecentbehavior,and

thensimulatestheeffectofswitchingtoanewpolicyruleundertheassumptionthatagentshave

rationalexpectationsandthepolicyannouncementiscompletelycredible.Astheprecedingsimulations

demonstrate,however,thestimulusprovidedbyunconventionalpolicy

atanypointintimedepended

notonwhattheFOMCwouldactuallydofromthatpointforward,butonwhatthepublicatthetime

expectedtheFederalReservetodointhefuture—animportantconsiderationforourefficacy

assessmentsgiventhattheperceivedpolicyruleand theexpectedtrajectory

oftermpremiumeffects

evolvedgraduallyfromearly2009tolate2013.Controllingforthepaceoflearningisalsoimportant

withregardstochangesovertimeinperceptionsofthefundamentalshocksactingtoshapethe

economicoutlook,giventhattheillustrativesimulations alsoshowedthattheactualstimulus

imparted

byforwardguidanceandassetpurchasespartiallydependsonhowlongthefundsrateisexpectedto

remainconstrainedbythezerolowerbound.



Tocontrolfortheselearningeffects,weemployasequenceofbaselinesofpastandprojected

conditions,eachofwhichrepresentsasnapshotofpast

economicdevelopmentsandexpectationsfor

thefuturetakenataparticularpointintime.Thisapproachenablesustocontrolfortheactualtiming

oftheshiftsinexpectationsassociatedwithrevisionsinboththeperceivedvalueofβintheimplicit

fundsrateruleandtheexpectedtrajectoryofQE‐related

termpremiumeffects.Inaddition ,itallowsus

toparsetheforcesdrivingchangesovertimeintheeconomicoutlook—asreflectedinrevisionstothe

extendedBlueChipprojections—intothoseassociatedwithperceivedinnovationsinpolicy,andthose

associatedwithchangingassessmentsofthefundamentalshockstoaggregatedemandand

other

factorshittingtheeconomy,bothnowandinthepast.



Forcomputationalconvenience,weassumethatexpectationsaboutfuturepolicyandotherfactorsare

updatedtwiceayear,onceinthefirstquartertotakeonboardontheinformationprovidedbythe

MarchBlueChipsurveys,andoncein

thethirdquartertoincorporatetheresultsfromtheOctober

surveys.Eachupdateofexpectationsrequiresthecreationofanewbaseline;thus,weconstructa

2009Q1baseline,a2009Q3baseline,a2010Q1baseline,andsoontoourfinal2013Q3baseline.The

expectedtrajectoryofQE‐relatedterm premium

effectsincorporatedintoeachbaselinematchesthe

correspondingpathestimatedbyIhrigetalandreportedinFigure3;forexample,thepathincorporated

intothe2009Q3baselineisthesameastheLSAP1path,whil e thepath incorporatedintothe2010Q3

baselinematchestheinitialestimatedtrajectoryofLSAP2effects.

Withregardstorevisionsinthe

policyruleacrossbaselines,ourbase‐caseassumptionisthattheperceivedvaluesofλandαarealways

zeroand0.5,respectively,butthattheperceivedvalueofβguidingfuturemonetarypolicyratchetsup

markedlyasonemovesfromthe2009Q1baselinetothe2013Q3

baseline.Asdiscussedearlier,it

appearsthatprivateforecasterspersistentlyanticipatedpolicyatthetimeofliftofftobemore

accommodativethanthatwhichwouldbefollowedoverthelongerrun. Accordingly,weassumethat



ThefulldimensionsofthefirstLSAPprogramwerenotannounceduntilaftertheMarch18,2009FOMC

meeting,althoughsomeinformationabouttheFederalReserve’sintentionstobuyMBSwasavailableinlate2008.

Forthisreason,weassumethatagentsin2009Q1expectedassetpurchasestolowerterm

premiumsbyroughly

10basispointsinthefirsthalfof2009,withthedownwardpressureerodingawayrapidlythereafter.

Page21of54



agentsatthetimeofeachsurveyexpectedβinthefuturetoequalthevaluesreportedinTable4

throughtheprojecteddateofliftoff,butthenanticipatedthatβwouldthereafterdeclinegraduallyto

thelonger‐runestimatesreportedinTable3.Theprecisepatternof

revisionstotheexpectedpathofβ

acrossbaselinesisillustratedinFigure8bytherisingsequenceofblacklines.Inadditiontothisbase‐

caseassumption,wealsoconsideranalternativeevolutionafter2011fortheperceivedpolicyrulethat

isindicatedbytheredlinesinthefigure.This

alternativesetofperceivedpathsforβ,whichare

assumedtobeaccompaniedbyajumpintheperceivedvalueofλto0.8,isconsistentwithourearlier

findingthattheBlueChipTreasurybillrateforecastspublishedfro mthispointarereasonablywell

explainedbyasmallerriseinβ

iftheywereaccompaniedbyaperceivedswitchtoinertialbehavioron

thepartoftheFOMC.



Inadditiontocontrollingforthetimingofperceivedchangesinpolicy,oursequentialbaselinesmust

alsocontrolforchangesovertimeinviewsabouttheeconomicoutlookandtheunderlyingforces,past

and

projected,shapingit.Todothis,westartbygeneratinganinitial2009Q1baselineinwhichthe

pathsforrealGDPgrowthandseveralmajorspendingcomponents,theunemploymentrate,inflation,

interestrates,thenaturalrateofemploymentandotherfactorsareforcedtomatchtheextendedBlue

Chipforecasts

publishedinMarch2009;thisinitialbaselinealsoincorporatespathsforthefederalfunds

rateandtermpremiumeffectsfromquantitativeeasingthatareconsistentwithprivateforecasters’

expectationsinearly2009forfuturemonetarypolicy,includingtheanticipate ddateofliftoff.

Next,

wegeneratebaselinescorrespondingtoeachsubsequentBlueChipforecastusingatwo‐stepprocess.

Thefirststepistocomputerevisedhistoricaldataforthenewbaseline;thisisdonebytakingthe

baselineforthepriorsurveyandadjustingitsdataforanyrevisionstoreal‐time

publishedestimatesof

realGDPandother“headline”seriesthatwouldhavebeenavailableatthetimethenewBlueChip

surveywasreleased.

ThesecondstepistorunaconstrainedFRB/USsimulationthathopsofffromthe

adjustedhistoricaldataandbeginsinthequarterthatisassumedtocorrespondtothereleasedateof

thenewBlueChipsurvey;forexample,thesimulationstartdatewouldbe2009:Q3inthecase

ofthe

October2009surveyand2010:Q1inthecaseoftheMarch2010survey.Ineachconstrained

simulation,wesolveforthesetofcurrentandfutureshockstoaggregatespending,employment,

supply‐sideconditions,prices,andotherfactorsthatareneededtoforcethemodelunderrational

expectationsto

replicatethenewextendedBlueChipprojections.

Insolvingfortheseanticipated

shocks,thesimulationconditionsonthecontemporaneousexpectationsforthedateoflifto ff,the

futurevalueofβ,andthetrajectoryofQE‐relatedtermpremiumeffectsdiscussedabove.Thistwo‐step

processisrepeateduntilwehaveconstructedbaselinesfor2009Q3through2013Q3.





BeyondtheendoftheBlueChipextendedprojectionhorizon,the2009Q1baseline(likeallthesubsequentones)

isassumedtoconvergegraduallytoapre‐determinedsteady‐stategrowthpathinwhichrealGDPgrowsatits

potentialrateof2¼percent,theunemploymentrateisflatatits

naturalrateof5½percent,PCEinflationequals

theFOMC’sannouncedlonger‐runobjectiveof2percent,andtheequilibriumrealfundsrateequals2percent.

Changingthesesteady‐stateassumptionswouldhavenoeffectonourestimatesoftheefficacyofunconventional

monetarypolicy.

Theadjustmentstothehistoricaldataalsoincluderevisionstothemodel’svariousexpectationalseries,where

thelatterarecomputedtobeconsistentwiththerevisionstothepublisheddataandthedynamicsofFRB/US.

BecausetheBlueChippublishesforecastsforfewervariablesthanareinthemodel,nouniquesolutiontothis

problemexistsunlessadditionalconstraintsareimposedontheallowablesetofprojection‐matchingshocks.We

obtainauniquesolutionbyimposingavariance‐covariancestructureontheunderlyingshocksthat

isbroadly

consistentwiththehistoricaldata;modestmodificationstothisstructurewouldnothaveanappreciableeffecton

theresultsreportedinthenextsection.

Page22of54



Amongotherthings,ourbaselineconstructionmethodologyparsestheunderlyingforcesdrivingthe

evolutionofactualandexpectedeconomicconditionsafterearly2009intotwoorthogonalparts—one

partcomprisedofthesequenceofperceivedinnovationstopolicy,andtheothercomprisedofthe

sequenceofrevisions

topastandprojectedshockstoaggregatespendin gandotherfundamentals.Asa

result,ifonestartswiththeinitial2009:Q1baseline,andthenrunsasequenceofunconstrained

simulationswhosestartingpointsareprogressivelyadvancedintime(from2009:Q3through2013:Q3),

witheachnewsimulationhoppingofffromtheresults

ofthepriorsimulationwhileincorporatinganew

setofrevisedshockstofundamentalsandpolicy(pastandprojected),thenonereplicatesthereal‐time

evolutionofbothactualmacroeconomicoutcomesandtheBlueChipextendedprojections,withthe

finalsimulationexactlymatchinghistoricalconditionsandtheBlueChipoutlook

astheyappearedin

October2013.Forourpurposes,however,themoreinterestingexerciseistorepeattheserolling

simulationswithoutincorporatingtheaccompanyingsequenceofpost‐2009Q1revisionstoliftoffdates,

βpaths,andthetrajectoryofexpectedtermpremiumeffects—thatis,tore‐runhistoryassumingno

unconventionalmonetary

policyinnovationsbutallowingthesameperceivedshockstootherfactorsto

unfoldastheydid.

Thedifferencebetweentheactualevolutionoftheeconomysince2009andthe

coursepredictedbythecounterfactualsimulationsofthesequencemeasurestheactualstimulus

providedbytheFOMC’sforwardguidanceandassetpurchases.



Baselineestimates



Figure9summarizesourbaselineresults,generatedusingthestandardversio nofFRB/US

assuming

rationalexpectationsonthepartofallagents.Theblacklinesshowthehistoricalpathsforrealactivity,

inflation,andinterestratesastheyappearedinlate2013,extrapolatedintothefutureusingthe

extendedBlueChipprojectionspublishedinOctober2013.Theredlines,incontrast,showthe

predictedevolutionoftheeconomyafterearly2009hadtheFOMCnotundertakenanyunconventional

policyactions,undertheassumptionthatdoingsowouldhavecausedtheanticipatedtrajectoryofQE‐

relatedtermpremiumeffectsandtheperceivedvalueofβintheimplicitpolicyruletoremain

unchangedatthepaths

expectedinearly2009.(Thesimulationdoesincorporatethemod estQE‐

relatedtermpremiumeffectsanticipatedinearly2009,equalto10basispointsinitiallyandfading

rapidlythereafter.)Beliefsaboutthepastandprojectedshockshittingtheeconomyinthe

counterfactualsimulationotherwiseunfoldovertimeastheydid

historically.Intheparsingexercises

usedtogeneratethesehistoricalshocks, theperceivedevolutionofthepolicyruleoverhistoryisthe

versioninwhichtheruleisassumedalwaystobenon‐inertial.Incontrast,thebluelinesofthefigures

showthepredictedevolutionoftheeconomyinthe

absenceofanyunconventionalpolicyactionswhen

theparsingroutineassumesthattheperceivedpolicy rulebecameinertialstartingin2012.Ascanbe

seen,allowingforashifttowardinertialpolicyhaslittleeffectonthesimulationresults,presumably

becausesuchashift(ifitoccurred)happenedata

relativelylatedateandinvolvedasmallerriseinthe

perceivedvalueofβ.



Threekeypointsemergefromthisanalysis.First,themodelestimatesthattheFOMC’sunconventional

policyactionsprovidedessentiallynostimulusinthefirsttwoyearsfollowingthefinancialcrisis,inthat

thesimulationshowstheunemployment

ratepeakingatthesamelevelandrealGDPgrowthand

inflationessentiallyunchangedthrough2010.Thisearlyineffectiveness is aresultofseveralfactors,

includingtherelativemodestchangestopolicyexpectationsthatoccurredduringthisperiod,initial



Inrunningthesecounterfactualsimulations,weassumethatthepublicalways(correctly)anticipatesthatthe

federalfundsrateinthestartingquarterwill remainatitseffectivelowerboundbutthatthefundsratewillstrictly

followtheprescriptionsofthe2009:Q1rulethereafter.

Page23of54



beliefsthattheeconomywouldreboundmorerapidlythanprovedtobethecase,andinherentlagsin

themonetarytransmissionprocess.However,therearereasonstothinkthatthisresultunderstatesthe

effectivenessoftheearlyunconventionalpolicyactions.Thefirstassetpurchaseprogramin

particularis

oftencreditedwithhavinggreatlyimprovedthefunctioningofkeyfinancialmarketsthathadseizedup

duringthemostintensephaseofthefinancialcrisis.Inaddition,thebaselinemodeldoesnot

incorporateanymechanismsfortheFOMC’smonetarypolicyactionstohaveaffectedsuchthings as

consumerconfidence,

businesssentiment,andinvestors’assessmentsofrisks—extra‐modelinfluences

whosehistoricalmovementswouldmanifestthemselvesinourparsingroutinesasshockstospending,

riskpremiums,andothermodelvariables.Ifconfidenceandotherextra‐modelinflue nceswouldhave

initiallybeenevenmoreimpairedintheabsenceofassetpurchasesand

forwardguidance,thenthe

counterfactualsimulationsmaysignificantlyu nderstatetheiractualefficacyin2009and2010.

41





Second,theFOMC’sactionsdoappeartohaveappreciablyspedupthepaceofrecoverybeginningin

2011,withthemodelpredictingasubstantiallyslowerdeclineintheunemploymentrateafterthat

pointintheabsenceofforwardguidanceandassetpurchases,accompaniedbylowerinflation.Third,

themodelestimates

thatweareonlynowapproachingthefulleffectsofunconventionalpolicyonreal

activityandinflation.AsindicatedbytheblacklinesofFigure10,whichshowthedifferencebetween

theactualevolutionoftheeconomysinceearly2009andthatpredictedbybaselinemodelinthenon‐

inertialversion

ofthecounterfactualsimulation,thepeakunemplo ymenteffect—subtracting1.2

percentagepointsfromtheunemploymentrate—doesnotoccuruntilearly2015,whilethepeak

inflationeffect—adding0.5percentagepointtotheinflationrate—isnotanticipateduntilearly2016.



TheseestimatesoftheunemploymentandinflationeffectsoftheFOMC’sforwardguidance

andasset

purchasesareconsiderablysmallerthanthebaselineestimatesreportedinChungetal(2012)forjust

thetermpremiumeffectsofLSAP1andLSAP2,despitethefactthatthela tter’sestimateswerealso

basedonsimulationsoftheFRB/USmodel.Severalfactorsaccountforthedifferencesinestimated

efficacy.

First,ChungetalbasedtheirtermpremiumeffectsofLSAP1andLSAP2ontheestimates

reportedinGagnonetal(2011),whicharetwiceaslargeastheIhrigetal(2013)estimatesemployedin

ourstudy.Second,Chungetalscoredtermpremiumeffectsholdingthenominalfederal

fundsrate

fixedatitsba selinepathforthefirstfiveyears,whereasthefundsrateinoursimulationsalways

endogenouslyfollowstheprescriptionsofapolicyruleexceptwhenconstrainedbytheZLB,thereby

greatlyscalingbackthenetstimulusfromassetpurchases.Third,ourstudycontrolsforlearningon

the

partofthepublicaboutthedepthandpersistenceofthepost‐crisisslump,whereasChungetal

implicitlyassumethatagentshaveperfectforesightabouttheseconditions.Finally,sincetheChunget

alstudytheFRB/USmodelhasbeenreestimatedandrespecified,withtheresultthataggregatedemand

intheversionusedinthisstudyisnoticeablylessinterestsensitive;inaddition,inflationislesssensitive

tochangesincurrentandexpectedeconomicslack.



Estimatesderivedfromalternativemodelspecifications





Riskpremiumsoncorporatebonds,equity,andtheforeignexchangevalueofthedollardroppedappreciably

moreinthesecondquarterof2009thanourparsingsimulationscanexplain.Ifweweretoattributeallofthese

unexplaineddeclinesinriskpremiumstotheFOMC’sMarch2009announcementofan

$1.75trillionasset

purchaseprogram,onthegroundsthatthisannouncementdemonstratedtheCommittee’sintentiontodo

whateverittakes,thenourestimatesoftheeffectsofunconventionalmonetarypolicyonrealGDPgrowthwould

riseby½percentagepointinboth2009and2010,evenassumingthatthedeclinesin

riskpremiumswouldhave

occurredanywayovertime.

Page24of54



ConsiderableuncertaintyobviouslyattendsanyestimateoftheefficacyoftheFOMC’sforwardguidance

andassetpurchases.Someofthisuncertaintyisassociatedwiththeshiftsinpolicyexpectations

incorporatedintoouranalysis:Ourestimatesoftheperceivedchangestothepolicyruleimplicitin

private

forecastsmaybeinerrorormaynotreflectthebeliefsofothereconomicagents,andtheLi‐Wei 

modelusedtopredictthetrajectoryoftermpremiu meffectsfromquantitativeeasingmaybeflawed.

Andevenifourestimatesoftheexpectationeffectsofpolicyarereasonablyaccurate,theanalysis

of

theirmacroeconomicimplicationsmay notbe.FRB/USisjustonemodel,andothermodelsmayyield

quitedifferentresults.



Tothisend,wenowconsiderestimatesofefficacyderivedusingseveralalternativespecificationsofthe

dynamicsoftheeconomy;amongotherthings,thisalternativeanalysisinvolves re‐runningthe



constructionofthesequentialbaselinesandthusrecalculatingtheperceivedshocks(pastand

projected)hittingtheeconomyovertime.Wedevelopthesealternativespecificationsbymodifyingkey

aspectsofthestructureofFRB/USratherthanemployingotherstructuralmodels;inaddi tiontobeing

moreconvenient,wethinkthisapproachallows

ustoexploreawiderrangeofplausiblealternatives

thancouldbedoneifwerestrictedourselvestodrawingthealternativesfromtheexistingsetof

availableDSGEmodels;thisapproachalsoallowsustoanalyzethemacroeconomiceffectsofchangesin

theexpectedtrajectoryoftermpremiumeffects,somethingthat

mostDSGEmodelscannotdo.That

said,thelimitedsetofalternativeFRB/USspecificationsweconsiderbelowbynomeansspansthe

rangeofempirically‐relevantalternativecharacterizationsofthedynamicsoftheeconomy.



Ourfirstalternativeinvolvestheformationofexpectations.Theassumptionthatallagentshave

rationalexpectations

andfullyunderstandthemacro economicimplicationsofforwardguidanceand

assetpurchasesseemsextreme,especiallywithregardstohouseholdswhohaveatbestaratherlimited

knowledgeofsuchmatters.Inaddition,theempiricalgroundsforassumingrationalexpectationsin

householddecision‐making,wageandpricesetting,andevenbusinessinvestment

decisionsisactually

ratherweak,despitetheubiquityofthisassumptioninmodeling.Accordingly,weconsideran

alternativeversionofFRB/USinwhichtheonlyagentswithrationalexpectationsarefinancialmarket

participants,anassumptionthatseemsreasonablegiventheirincentivestounderstandmonetary

policy.Householdsandnonfinancialfirms,in

contrast,basetheirexpectationsforthefutureonthe

forecastsofsmall‐scaleVARmodelsestimatedusinghistoricaldata,andsoassumethattheeconomy

willdisplaythesamecyclicaldynamicsinthefuturethatitdidonaverageinthepast.Becausethese

agentsdonotunderstandhowunconventionalpolicy

altersthosedynamics,forwardguidanceandasset

purchasesnolongerdirectlycausehous eholdsandnon‐financialfirmstomarkuptheirexpectationsfor

realactivityandinflationinthefuture,andsotheirinitialstimulusfalls.

Asindicatedbythebluelines

ofFigure10,theefficacyoftheFOMC’sunconven tionalpolicyactionssinceearly2009dropsmarkedly

undertheseassumptions,withapeakeffectofreducingtheunemploymentratebyonly0.9percentage

pointsandalmostnoboosttoinflationpriorto2014.



Oursecond

alternativeinvolves thedynamicsofinflation.WagesandpricesinFRB/USaredetermined

usinganestimatedNewKeynesianPhillips curve.TheFRB/USempiricalimplementationofthismodel,

however,yieldsinflationdynamicsthatarebothmorepersistentandlesssensitivetoslackthanthat

exhibitedbymanyDSGEmodels.Giventhecritical

roleplayedbyinflationinthetransmissionchannels

ofunconventionalmonetarypolicyatthezerolowerbound,wethereforeconsideranalternative



Householdsandnonfinancialfirmsdorespond,however,tothe(rational)changesinfinancialconditionsthat

occurinresponsetosuchpolicyactions.Notethattherationalityoffinancialmarketsincludestheirknowledge

thatotheragentsintheeconomydonothavemodel‐consistentexpectations.

Page25of54



specificationoftheFRB/USwage‐pricesectorthatbringsitsinflationdynamicsintocloseralignment

withthatembeddedintheseothermodels.

AsindicatedbytheredlinesofFigure10,makingthis

changebooststheestimatedpeakeffectivenessoftheFOMC’spolicyactionsconsiderably,becausethe

alternativespecificationpredictsthatassetpurchasesandforwardguidancedidconsiderablymoreto

preventamoremarked(albeittransitory)disinflationfromemergingovertime.That

result,inturn,

limitedtheriseinrealshort‐terminterestratesthatwouldhaveotherwiseoccurredinthefaceofthe

ZLBconstraint,therebyprovidingadditionalsupporttorealactivity.



Ourthirdalternativeaddressesthepossibilitythattheinterestsensitivityofaggregatedemandis

appreciablylowerthanpredictedby

thebaselinemodel,particularlywhenallowanceismadeforsome

specialfactorsthatmayhaveinfluencedtheeconomy inthewakeofthefinancialcrisis.Inparticular,

tighterunderwritingstandards,impairedbalancesheets,andanelevatedlevelofuncertaintysinceearly

2009mayhavereducedtheabilityand/orwillingnessofhouseholds

andfirmstoboosttheirspendingin

responsetolowerborrowingcostsandincreasesinwealth.Toaddressthispossibility,wecreatealow

interestsensitivitymodelinwhichtheresponsivenessofconsumerdurablegoods,residential

investment,andbusinesscapitalspendingtomovementsintheirrespectivecos tsofcapi tal,and

the

responsivenessofstockpricesandtherealexchangeratetomovementsinlong‐termTreasuryyields,

arehalfaslargeasinthebaselinemodel.AsindicatedbythegreenlinesofFigure10,allowingforsuch

apronouncedconstrictionofthemonetarytransmissionchannelsmarkedlyreducestheestimated

macroeconomic

gainsfromforwardguidanceand assetpurchasesinthismodel.



OurfinalalternativeconcernstheeffectsoftheFO MC’s unconven tional mone tarypolicyon relative

houseprices.Inthebaselineversionofthemodel,movementsinthesepricesareestimatedtobe

essentiallyindependentofchangesin interestrates,realactivity,andothermacroeconomicfactors.

Althoughthisbehaviorisconsistentwith

thehistoricaltime‐seriesbehaviorofhouseprices,theturn‐

aroundinhomevaluessincetheircolla pse in2006‐2008couldhavedependedimportantlyonthe

supporttohousingmarketactivityprovidedbyextremelylowmortgagerates.Ifso,thenthesubstantial

pick‐upinactualandprojectedhousingpricesthat

hasoccurredsinceearly2009mighthavebeenmuch

weakerwithoutforwardguidanceandassetpurchases.AsshowninFigure11,oursequentialbaselines



InthebaselineversionoftheFRB/USNewKeynesianPhillipscurve,currentinflationdependsupon(among

otherthings)expectedinflationinthenextquarter,aweightedaverageofactualinflationoverthepastfour

quarters,andcurrentlabormarketslack.Inthealternativeversion,themovingaverageofpastinflation

is

replacedwithasinglelagandthecoefficientonslackisincreasedby50percent.Makingthesechangesbringsthe

dynamicsofFRB/USatthezerolowerboundintoreasonablyclosealignmentwiththoseofEDO,theFederal

ReserveBoard’sestimatedDSGEmodeloftheU.S.economy.

Alternatively,thestructureoftheFRB/USmodelmayoverstatetheinfluenceoflonger‐terminterestrateson

aggregatedemand,andunderstatetheroleplayedbyinterestrateswithamaturityoflessthan5years,leadingto

anoverestimationofthestimulusprovidedbyQE‐drivenreductionsintermpremiums.

Inthisregard,Kiley

(2013a)presentseconometricevidencesuggestingthatthemacroeconomicstimulusassociatedwithhistorical

shiftsintermpremiumsisclosetozero,perhapsbecausetheydonotinducethechangesinborrowingcosts,

equityprices,andrealexchangesratespredictedbythebaselineversionoftheFRB/US.Ofcourse,

Kiley’sfindings

maynotbeapplicabletotherecentQE‐drivenshiftsintermpremiumsforreasonsrelatedtoidentification,given

thathistoricalmovementsintermpremiumswerelikelydrivenprimarilybyinflationuncertaintyandflight‐to‐

qualityconcerns,notpolicy;asaresult,thepastandpresentcorrelationsof

termpremiumshiftswithrealactivity

andinflationmaybequitedifferent.Whenre‐runningouranalysisundertheassumptionofnopass‐throughof

QE‐driventermpremiumeffectsfromTreasurysecuritiestootherassetprices,thepeakeffectoftheFOMC’s

unconventionalpolicyactionsontheunemploymentrateand

inflationdropsto‐0.7percentagepointsand0.2

percentagepoints,respectively.

Page26of54



incorporateasubstantialshiftupovertimeintheactualandprojectedtrajectoriesofhouseprices;

thesepathsareconstructedusingreal‐timedataandsurveysofexpectedchangesinhousepricesfor

thecurrentandcomingyear,extrapolatedintothefutureusingasimpletrend‐

reversionmodelof

relativehouseprices.Inourfinalalternativescoring,weassumethatactualandexpectedhouseprices

intheabsenceoftheFOMC’sunconventionalpolicyactionswouldhavefollowedthetrajectoryinitially

expectedinearly2009(thesolidblackline).Asindicatedbytheor angelines inFigures10,

underthis

assumptiontheestimatedefficacyoftheFOMC’spolicyactionsincreasesappreciably.



VII. ConclusionsandPossibleImplicationsforFutureMonetaryPolicy



Severalkeyfindingsemergefromouranalysis.Intheyearsfollowingtherecessionandfinancialcrisis,

wefindthatprivate‐sectorforecasterspushedtheirexpectedtimeforthe

onsetoftighteningof

monetarypolicyprogressivelyfurtherintothefuture,presumablyinresponsetotheFederalReserve’s

quantitativeeasingandincreasinglyaggressiveforwardguidanceforthefederalfundsrate,inthe

contextofasloweconomicrecoverywithmoderateinflation.AlthoughtheFederalReserve’s

quantitativeeasingprogramlikelyhelped

tostabilizetheeconomyduringthefinancialcrisisasit

providedliquidityinfinancialmarketsthatwereseizingup,weestimatethattheFOMC’s

unconventionalpolicyactionsprovidednomaterialadditionalmonetarypolicystimulusinthefirsttwo

yearsfollowingthefinancialcrisis,probablyreflectingtherelativemodestchangestopolicy



expectationsthatoccurredduringthisperiod,anticipationthattheeconomywouldreboundmore

rapidlythanprovedtobethecase,andtheinherentlagsinthetransmissionofmonetarypolicy;a

failureonthepartoftheFRB/USmodeltocaptureimportantconfidenceeffectsassociatedwith

unconventionalpolicycould,however,

significantlybiasdownourestimates.Indeed,asdescribedmore

fullyinAppendix2,eventstudiessuggestthatthefirsttworoundsofLSAPsprobablyspurredlarger

responsesofinterestrates,theexchangevalueofthedollar,andequitypricesthanlateronesbecause

theearlierprogramswereannouncedandimplementedat

timeswhenmarketconditionswerehighly

strainedandriskpremiumswereexceptionallylargeorwhentherewasgreaterscopeforsignalingmore

accommodativeintentions withrespecttothefederalfundsrate.Moreover,ourestimatesdonot

capturetheeffectsofimprovementsinconsumersandbusinessconfidence,whichmayatleast

partially

reflectthesemonetarypolicyactionsduringthedepthofthefinancialcrisis.



Nevertheless,byourestimatestheFOMC’sactionsdoappeartohaveappreciablyspedupthepaceof

recoveryfrom2011onastheprivatesectorbegantolearnthatmonetarypolicywasgoingtoremain

substantiallymoreaccommodative

thanusualoveralongerperiodoftime.Infact,weestimatethatthe

economyisonlyjustnowapproachingthefulleffectsofunconventionalmonetarypolicyactions.We

findthatthepeakunemploymenteffect—subtracting1.2percentagepointsfromtheunemployment

rate—doesnotoccuruntilearly2015,whilethepeak

inflationeffect—adding0.5percentagepointto

theinflationrate—isnotanticipateduntilearly2016.



Implicationsforthefutureefficacyofu nconventionalmonetarypolicy



Therearesomeimportantimplicationsofouranalysisforthepotentialeffectivenessofunconventional

monetarypolicyinmitigatingthefalloutfromfuturefinancialcrises orothersevereZLB

episodes.As

notedabove,akeyfactorlimitingtheeffectivenessofforwardguidanceandassetpurchasesoverthe

pastfiveyearshasbeenthegradualnatureoftheaccompanyingrevisionstopolicyexpectations.Given

thelargelyunprecedentednatureofthesepolicyactions,suchadrawn‐outresponsebythepublic

concerningexpectationsaboutmonetarypolicywasprobablyunavoidable.Intheeventofafuture

crisis,however,financialmarketparticipants andthepublicmoregenerallywillhavethebenefitofthe

Page27of54



currentepisodeinassessingtheFOMC’slikelyresponse.Ratherthantakingyearstochangetheir

perceptionsoftheFOMC’simplicitpolicyruleandtheirexpectationsforthelikelysizeoftheeventual

expansionoftheFederalReserve’sassetholdings,theymayadjusttheirexpectationsimmediatelyin

the

wakeofthecrisis.ThelikelihoodofsucharesponsewouldbestrengthenediftheFOMCpriortothe

newcrisishadpersistentlysetthefundsrateinamannerbroadlyconsistentwiththeaccommodative

rulenowperceivedbyprivateforecasters.



Whatmightsuchapermanentrevisioninpolicyexpectations

implyforthefutureeffectivenessof

unconventionalmonetarypolicy?Onewaytoaddressthisquestion istorunacounterfactualsimulation

inwhichalltheshiftsintheperceivedpolicyrule andtheexpectedtrajectoryofQE‐driventerm

premiumeffectsoccurimmediatelyinthewakeofthefinancialcrisis.

Resultsfromthisexperimentare

summarizedinFigure12.Ascanbeseenbycomparingactualhistoricaldevelopments(theblacklines)

tothecounterfactualpredictions(theredandbluelines),therecessionwouldhavebeensomewhatless

severeandthesubsequentpaceofrecoverywouldhavebeennoticeablyfasterifthe

publichad

understoodupfrontthewillingnessoftheFOMCtoprovideadditionalaccommodationthroughasset

purchasesanditsintentiontopursueamoreaccommodativestrategyoverthelongertermthanhad

beeninitiallyperceived.Thismoresubstantialinitialresponsereflectsbothagreaterimprovementin

financialconditions(asevidencedby

amuchlargerinitialfallinnominalbondyields)aswellasmore

pronouncedshiftsinexpectationsoffuturerealactivityandinflation. Basedonthesesimulationresults,

onecanreasonablyarguethattheFOMCislikelytohaveasomewhatgreaterabilitytomitigatethe

effectsofafuturecrisis

thanitdidatthestartofthecurrentone,aslongasthepublicanticipatesthat

theFOMCwillonceagainaggressivelydeployitsunconventionalpolicytools.





Page28of54



References



Bauer,MichaelD.,andGlennD.Rudebusch(2014),“TheSignalingChannelforFederalReserveBond

Purchases,”InternationalJournalofCentralBanking,10,233‐289.



Baumeister,Christiane,andLucaBenati(2013),“UnconventionalMonetaryPolicyandtheGreat

Recession:EstimatingtheMacroeconomicEffectsofaSpreadCompressionattheZeroLower

Bound,”InternationalJournalof CentralBanking,9,165‐212.



Bernanke,Ben(2012),“MonetaryPolicysincetheOnsetoftheCrisis,”attheFederalReserveBankof

KansasCityEconomicSymposium,JacksonHole,WY,August31,

www.federalreserve.gov/newsevents/speech/bernanke20120831a.htm



Bernanke,BenS.,andKennethN.Kuttner(2005),“WhatExplainstheStockMarket’sReactiontoFederal

ReservePolicy?”TheJournalofFinance,60,number3,June,pp.1221‐57.



Bharrarai,Saroj,GautiB.Eggertsson,andBulatGafarov(2014),“TimeConsistencyandtheDurationof

GovernmentDebt:ASignalingTheoryof

QuantitativeEasing,”workingpa per.



Brayton,Flint(2011),“TwoPracticalAlgorithmsforSolvingRationalExpectationsModels,”Federal

ReserveBoard,FinanceandEconomicsDiscussionSeriesno.2011‐44.



Campbell,Jeffrey,CharlesEvans,JonasFisher,andAlejandro Justiniano(2012),“MacroeconomicEffects

ofFederalReserveForwardGuidance,”FederalReserveBankofChicago,workingpaper

no.

2012‐03.



Chen,Han,VascoCurdia,andAndreaFerrero(2012),“TheMacroeconomicEffectsofLarge‐scaleAsset

PurchaseProgrammes,”EconomicJournal,122,289‐315.



Chung,Hess(2015),“TheEffectsofForwardGuidanceinFRB/USandEDO,”FederalReserveBoard,FEDS

Notes,forthcoming.



Chung,Hess,Jean‐PhilippeLaforte,David

Reifschneider,andJohnC.Williams(2012),“HaveWe

UnderestimatedtheLikelihoodandSeverityofZeroLowerBoundEvents?”JournalofMoney,

Credit,andBanking,44,47‐82.



Chung,Hess,EdwardHerbst,andMichaelKiley(2014),“EffectiveMonetaryPolicyStrategiesinNew

KeynesianModels:AReconsideration,”FederalReserveBoard,workingpaper.



Cogley,T.andA.Sbordone(2008),“TrendInflation,Indexation,andInflationPersistenceintheNew

KeynsianPhillipsCurve,”AmericanEconomicReview,98,2101‐2126.



D’Amico,StephaniaandThomasB.King(2013),“FlowandStockEffectsofLarge‐ScaleTreasury

Purchases,”JournalofFinancialEconomics,108,425‐448.



Page29of54



D’Amico,Stephania,WilliamEnglish,DavidLopez‐Salido,andEdwa rdNelson(2012),“TheFederal

Reserve’sLarge‐ScaleAssetPurchasePrograms:RationaleandEffects,”EconomicJournal,122,

415‐446.



DelNegro,Marco,MarcGiannoni,andChristinaPatterson(2013).“TheForwardGuidancePuzzle,”

FederalReserveBankofNew

YorkStaffReportno.574.



English,William,DavidLopez‐Salido,andRobertTe tlow(2013),“TheFederalReserve’sFrameworkfor

MonetaryPolicy–RecentChangesandNew Questions,”FederalReserveBoard,Financeand

EconomicsDiscussionSeriesno.2013‐76.



Femia,Katherine,StevenFriedman,andBrianSack(2013),“TheEffectsofPolicyGuidance

on

PerceptionsoftheFed’sReactionFunction,”FederalReserveBankofNewYork,StaffReportno.

652.



Friedman(1957),ATheoryoftheConsumptionFunction,PrincetonUniversityPress.



Gagnon,Joseph,MatthewRaskin,JulieRamache,andBrianSack(2011),“TheFinancialMarketEffectsof

theFederalReserve’sLarge‐ScaleAssetPurchases,”

InternationalJournalofCentralBanking,7,

3‐43.



Gertler,Mark,andPeterKaradi(2013).“QE1vs.2vs.3…:AFrameworkforAnalyzingLarge‐ScaleAsset

PurchasesasaMonetaryPolicyTool,”InternationalJournalofCentralBanking,9(S1),5‐53.



Gilchrist,Simon,DavidLópez‐Salido,andEgonZakrajšek(2014),

“MonetaryPolicyandRealBorrowing

CostsattheZeroLowerBound,”AmericanEconomicJournal:Macroeconomics,forthcoming.



Glick,Reuven,andSylvainLeduc(2013),“TheEffectsofUnconventionalandConventionalU.S.

MonetaryPolicyontheDollar,”WorkingPaper2013‐11,FederalReserveBankofSanFrancisco.



Greenwood,Robin,SamuelG.Hansen,

JoshuaS.Rudolph,andLawrenceH.Summers(2014),

“GovernmentDebtManagementandtheZeroLowerBound,”workingpaper. 



Gürkaynak,Refet,BrianSack,andEricSwanson(2006),“Market‐BasedMeasuresofMonetaryPolicy

Expectations,”FederalRes erveBankofSanFrancisco,workingpaperno.2006‐4.



Hamilton,JamesD.,andCynthiaWu

(2012),“TheEffectivenessofAlternativeMonetaryPolicyToolsina

ZeroLowerBoundEnvironment,”JournalofMoney,Credit,andBanking,44,3‐46.



Hancock,Diana,andWaynePassmore(2012),“TheFederalReserve’sPortfolioandItsEffectson

MortgageMarkets,”FinanceandEconomicsDiscussionSeriesNo.2012‐22,Boardof

Governors

oftheFederalReserveSystem.



________.(2014),“HowtheFederalReserve'sLarge‐ScaleAssetPurchases(LSAPs)InfluenceMortgage‐

BackedSecurities(MBS)YieldsandU.S.MortgageRa tes,”FinanceandEconomicsDiscussion

SeriesNo.2014‐12,BoardofGovernorsoftheFederalReserveSystem.



Page30of54



Ihrig,Jane,ElizabethKlee,CanlinLi,BrettSchulte,andMinWei(2012),“ExpectationsabouttheFederal

Reserve’sBalanceSheetandtheTermStructureofInterestRates,”FederalReserveBoard,

FinanceandEconomicDiscussionSeriesno.2012‐57.



Joyce,MichaelandMatthewTong(2012),“QEandtheGiltMarket:ADisaggregatedAnalysis,”Economic

Journal,122,348‐384.



Kiley,Michael(2012),“TheAggregateDemandEffectsofShort‐andLong‐TermInterestRates,”Federal

ReserveBoard,FinanceandEconomicsDiscussionSeriesno.

2012‐54.



________.(2013a),“MonetaryPolicyStatements,TreasuryYields,andPrivateYields:BeforeandAfter

theZeroLowerBound,”FederalReserveBoard,FinanceandEconomicsDiscussionSeriesno.

2013‐16.



________.(2 013b),“ExchangeRates,MonetaryPolicyStatements,andUncoveredInterestParity:

BeforeandAftertheZeroLowerBound,”Financeand

EconomicsDiscussionSeriesNo.2013‐17,

BoardofGovernorsoftheFederalReserveSystem.



________.(2013c),“TheResponseofEquityPricestoMovementinLong‐termInterestRatesAssociated

withMonetaryPolicyStat ements:BeforeandAftertheZeroLowerBound,”Financeand

EconomicsDiscussionSeriesNo.2013‐15,Boardof

GovernorsoftheFederalReserveSystem.



Kuttner,Kenneth(2001),“MonetaryPolicySurprisesandInterestRates:EvidencefromtheFedFunds

FuturesMarket,”JournalofMonetaryEconomics,47,523‐544.



Krishnamurthy,Arvind,andAnnetteVissing‐Jorgensen(2011),“TheEffectsofQuantitativeEasingon

Long‐termInterestRates,”BrookingsPaperson

EconomicActivity,42,215‐265.



Li,Canlin,andMinWei(2013),“TermStructureModelingwithSupplyFactorsandtheFederalReserve’s

Large‐ScaleAssetPurchasePrograms,”InternationalJournalofCentralBanking,9,3‐39.



Lucas,Robert(1976),“EconometricPolicyEvaluation:ACritique,”Carnegie‐RochesterSeriesonPublic

Policy,1,

19‐46.



Meaning,JackandFengZhu(2011),“TheImpactofRecentCentralBankAssetPurchaseProgrammes,”

BISQuarterlyReview,73‐83.



Mumtaz,HaroonandPaoloSurico(2014),“TheTransmissionMechanisminGoodandBadTimes,”

CentreforEconomicPolicyResearch,discussionpaperno.10083.



Neely,C.(2010),“The

Large‐ScaleAsssetPurchasesHadLargeInternationalEffects,”FederalReserve

BankofSt.Louis,workingpaperno.2010‐18A.



Reifschneider,David,WilliamWascher,andDavidWilcox(2013),“AggregateSupplyintheUnitedStates: 

RecentDevelopmentsandImplicationsfortheConductofMonetaryPolicy,”FederalReserve

Board,FinanceandEconomicsDiscussionSeries

no.2013‐77.



Page31of54



Rogers,JohnH.,ChiaraScotti,andJonathanH.Wright(2014),“EvaluatingAsset‐MarketEffectsof

UnconventionalMonetaryPolicy:ACro ss‐CountryComparison,”FederalReserveBoard,

InternationalFinanceDiscussionPapersno.1101.



Rosa,Carlos,(2012),“How‘Unconventional’AreLarge‐ScaleAssetPurchases?TheImpactofMonetary

Policy

onAssetPrices,”FederalReserveBankofNewYork,staffreportno.560.



Swanson,EricT.(2011),“LetsTwistAgain:AHigh‐FrequencyEvent‐StudyAnalysisofOperationTwist

andItsImplicationsforQE2,”BrookingsPapersonEconomicActivity,42,151‐188.



Taylor,John(1993),“DiscretionVersusPolicyRulesin

Practice,”Carnegie‐RochesterSeriesonPublic

Policy,39,195‐214.



________.(1999),“AHistoricalAnalysisofMonetaryPolicyRules , ” inMonetaryPolicyRules(editedby

JohnTaylor),UniversityofChicagoPress,319‐341.



Taylor,JohnandJohnC.Williams(2011),“SimpleandRobustRulesforMonetaryPolicy,”inHand book

ofMonetaryEconomics(BenjaminFriedmanandMichaelWoodford,editors),Elsevier,vol.3,

829‐859.



Vayanos,DimitriandJean‐LucVila(2009),“APreferred‐Habitat ModeloftheTermStructureofInterest

Rates,”workingpaper.



Weale,Martin,andTomaszWieladek(2014),“WhataretheMacroeconomicEffectsofAsset

Purchases?”Bankof

England,DiscussionPaperno.42.



Werning,Ivan(2012),“ManagingaLiquidityTrap:MonetaryandFiscalPolicies,”MIT,workingpaper.



Woodford,Michael(2012),“MethodsofPolicyAccommodationattheInterest‐RateLowerBound,”in

TheChangingPolicyLandscape, FederalReserveBankofKansasCity.



Yellen,Janet(2012a),“The EconomicOutlookand

MonetaryPolicy,”attheMoneyMarketeersofNew

YorkUniversity,NewYork,NY,April11,

www.federalreserve.gov/newsevents/speech/yellen20120411a.htm



______.(2012b),“PerspectivesonMonetaryPolicy,”attheBostonEconomicClubDinner,Boston,MA,

June6,www.federalreserve.gov/newsevents/speech/yellen20120606a.htm



______.(2012c),“RevolutionandEvolutioninCentralBankCommunications,”attheHaasSchoolof

Business,UniversityofCaliforniaatBerkeley,Berkeley,California,November13,

www.federalreserve.gov/newsevents/speech/yellen20121113a.htm.





Page32of54



Table1

MajorActionsandStatementsConcerningFederalReserveAssetPurchases

11/25/2008

FederalReserveBoardannouncesitsintentiontopurchaseupto$100billionindirectobligationsofthe

housing‐relatedgovernment‐sponsoredenterprises(FannieMae,FreddieMac,andGinnieMae)andupto

$500billioninGSE‐issuedMBS.Thesepurchasesaretobecompletedwithinseveralquarters.

3/18/2009 FOMCexpandsitsassetpurchaseprogramtoatotalof$1.25trillioninpurchasesofagencyMBS,$200

billioninGSEobligations,andupto$300billionoflonger‐termTreasurysecuritiesbytheendof2009.

8/10/2010

FOMCstatesthatitmaintainitsholdingsofsecuritiesattheircurrentleve lbyreinvestingprincipal

paymentsfromagencydebtandagencymortgage‐backedsecuritiesinlonger‐termTreasurysecurities,

andbycontinuingtorolloveritsholdingsofTreasurysecuritiesastheymature.

11/3/2010

FOMCannouncesthat,inadditiontoreinvestingprincipalpaymentsfromitssecuritiesholdings,itwill

expandtheoverallsizeofitsportfoliobypurchasingafurther$600billionoflonger‐termTreasury

securitiesbytheendofthesecondquarterof2011.

9/21/2011

FOMCvotestoextendtheaveragematurityofitssecuritiesholdingsbypurchasing$400billionof

Treasurysecuritieswithremainingmaturitiesof6yearsto30yearsandsellinganequalamountof

Treasurysecuritieswithremainingmaturitiesof3yearsorless;thesetransactionsaretobecompletedby

theendofJune2012.TheFOMCalsoannouncesthatitwillnowreinvestprincipalpaymentsfromits

holdingsofagencydebtandagencymortgage‐backedsecuritiesinagencymortgage‐backedsecurities,

whilemaintainingitsexistingpolicyofrollingovermaturingTreasurysecuritiesatauction.

6/20/2012

FOMCvotestomaintainthroughtheendof2012itsongoingmaturity‐extensionprogrambycontinuingto

purchaseTreasurysecuritieswithremainingmaturitiesof6yearsto30yearsatapaceofabout[$45

billionpermonth]whilesimultaneouslyselling orredeemingthesameamountofTreasurysecuritieswith

remainingmaturitiesofapproximately3yearsorless.

9/13/2012

FOMCvotestobeginpurchasingadditionalagencymortgage‐backedsecuritiesatapaceof$40billionper

month,inadditiontocontinuingitsmaturity‐extensionprogramanditsreinvestmentofMBSprinciple

paymentsinagencyMBS. TheseactionsimplyincreasesinFederalReserveholdingsoflonger‐term

securitiesofabout$85billionpermonth.TheFOMCalsostatesthatiftheoutlookforthelabormarket

doesnotimprovesubstantially,itwillcontinueitspurchasesofagencymortgage‐backedsecurities,

undertakeadditionalassetpurchases,andemployitsotherpolicytoolsasappropriateuntilsuch

improvementisachievedinacontextofpricestability,whilealsonotingthesizeandcompositionofthese

purchaseswilltakeappropriateaccountoftheirlikelyefficacyandcosts.

12/12/2012

FOMCannouncesthat,afterthematurityextensionprogramceasesattheendof2012,itwillbegin

buyinganadditional$45billioninlong‐termTreasurysecuritiespermonthwhilecontinuingtopurchase

$40billioninagencyMBSpermonthandreinvestingprinciplepayments,thusimplyingthattheFederal

Reserve’sportfoliowillcontinuetoexpandatapaceof$85billionpermonth.

3/20/2013,

6/19/2013,

and9/18/2013

FOMCrefinesitsoriginalguidanceaboutthefactorsinfluencingthesize,paceandcompositionofits

ongoingassetpurchaseprogrambynotingthatitalsodependsontheextentofprogresstowardits

economicobjectives(March)andtheinflationoutlook(June).Italsostressesthatthepaceofpurchasesis

contingentoneconomicoutlookaswellitsassessmentsofcostsandefficacy(September).

12/18/2013

FOMCslowstheongoingmonthlypaceofpurchasesto$35billioninagencyMBSand$40billioninlong‐

termTreasurysecuritiesandadvisesthatfurtherreductionsarelikelyatupcomingmeetingsifincoming

informationbroadlysupportstheCommittee'sexpectationofongoingimprovementinlabormarket

conditionsandinflationmovingbacktowarditslonger‐runobjective.However,theFOMCalsostresses

thatassetpurchasesarenotonapresetcoursebutarecontingentontheeconomicoutlook.





Page33of54



Table2

ForwardGuidanceinFOMCStatementsIssuedFromLate2008Through2013

12/16/2008 Afterannouncinga0to.25percenttargetrangeforthefederalfundsrate,theCommitteenotesthat

“weakeconomicconditionsarelikelytowarrantexceptionallylowlevelsofthefederalfundsratefor

sometime.”

3/18/2009 TheCommittee“anticipatesthateconomicconditionsarelikelytowarrantexceptionallylowlevelsofthe

federalfundsrateforanextendedperiod.”

11/4/2009 TheCommittee“continuestoanticipatethateconomicconditions,includinglowratesofresource

utilization,subduedinflationtrends,andstableinflationexpectations,arelikelytowarrantexceptionally

low

levelsofthefederalfundsrateforanextendedperiod.”

9/21/2010

“TheCommitteewillcontinuetomonitortheeconomicoutlookandfinancialdevelopmentsandis

preparedtoprovideadditionalaccommodationifneededtosupporttheeconomicrecoveryandtoreturn

inflation,overtime,tolevelsconsistentwithitsmandate.”

8/9/2011

“TheCommitteecurrentlyanticipatesthateconomicconditions—includinglowratesofresource

utilizationandasubduedoutlookforinflationoverthemediumrun‐‐arelikelytowarrantexceptionally

lowlevelsforthefederalfundsrateatleastthroughmid‐2013.”

1/25/2012

TheCommittee“…currentlyanticipatesthateconomicconditions‐‐includinglowratesofresource

utilizationandasubduedoutlookforinflationoverthemediumrun‐‐arelikelytowarrantexceptionally

lowlevelsforthefederalfundsrateatleastthroughlate2014.”.



9/13/2012

TheCommittee“…expectsthatahighlyaccommodativestanceofmonetarypolicywillremain

appropriateforaconsiderabletimeaftertheeconomicrecoverystrengthens”and“...currently

anticipatesthatexceptionallylowlevelsforthefederalfundsratearelikelytobewarrantedatleast

throughmid‐2015.”

12/12/2012

TheCommittee“…expectsthatahighlyaccommodativestanceofmonetarypolicywillremain

appropriateforaconsiderabletimeaftertheassetpurchaseprogramendsandtheeconomicrecovery

strengthens.Inparticular,theCommittee…currentlyanticipatesthatthisexceptionallylowrangeforthe

federalfundsrate[0to.25percent]willbeappropriateatleastaslongastheunemploymentrate

remainsabove6‐1/2percent,inflationbetweenoneandtwoyearsaheadisprojectedtobenomorethan

ahalfpercentagepointabovetheCommittee’s2percentlonger‐rungoal,andlonger‐terminflation

expectationscontinuetobewellanchored.TheCommitteeviewsthesethresholdsasconsistentwithits

earlierdate‐basedguidance.Indetermininghowlongtomaintainahighlyaccommodativestanceof

monetarypolicy,theCommitteewillalsoconsiderotherinformation,includingadditionalmeasuresof

labormarketconditions,indicatorsofinflationpressures andinflationexpectatio ns,and readingson

financialdevelopments.WhentheCommitteedecidestobegintoremovepolicyaccommodation,itwill

takeabalancedapproachconsistentwithitslonger‐rungoalsofmaximumemploymentandinflationof2

percent.”

12/18/2013

ThestatementreiteratedtheCommittee’spreviouslyannounced“action”thresholdsforunemployment

andprojectedinflation,andhowitsdecisiontobegintighteningdependonawiderangeofeconomic

factors.InadditiontheCommitteestatedthatit“…nowanticipates,basedonitsassessmentofthese

factors,thatitlikelywillbeappropriatetomaintainthecurrenttargetrangeforthefederalfundsrate

wellpastthetimethattheunemploymentratedeclinesbelow6‐1/2percent,especiallyifprojected

inflationcontinuestorunbelowtheCommittee's2percentlonger‐rungoal.”

1. Inaseparatestatementreleasedatthetime,theCommitteestressedthatitwouldtakea“balancedapproach”inpursuingitsdual

objectivesofpricestabilityandfullemployment.TheCommitteealsoannouncedthatagreementhadbeenreachedonalong‐run

inflationgoalforPCEinflationequalto2percent.Becausetherateofunemploymentconsistent withlong‐runpricestabilitydepends

onfactorsoutsidethecontrolofthecentralbankandneedstobeestimated,noformaltargetwassetforthislegofthedualmandate.





Page34of54



Table3

CoefficientonEconomicSlackinaSimplePolicyRuleEstimatedUsingAnnualBlueChipForecasts



(least‐squaresestimates,standarderrorsinparentheses)



 ObservationsIncludedinOLSRegression Addendum:

AssumedValueof

Coefficientinthe

CalibratedInertial

Rule



AllForecastYears

Unconstrained

Forecast

YearsOnly





BlueChipreleasedate  



March2009 .21(.05) .15(.04)



October2009 .28(.05) .23(.04)



March2010 .22(.04) .17(.05)



October2010 .26(.03) .27(.05)



March2011 .36(.04) .30(.05)



October2011 .55(.03) .54(.06)



March2012 .69(.04) .72(.09)

 .60



October2012 .73(.04) .85(.03) .75



March2013 .85(.08) 1.15(.07) .90



October2013 .93(.15) 1.61(.05) 1.0



1. ThesimpleruleisR(t,j)=R*(j)+π(t,j)+0.5[π(t,j)–π*(j)]–2β(j)[U(t,j)‐U*(j)].Inthisexpression,jindexestheBlue

ChipsurveyreleasedateandtindexestheforecastyearintheBlueChipprojection,t=0to6yearsahead.Risthe

projectedaveragevalueofthe3‐monthTreasurybillrateinyeart,πistheprojectedyear‐over‐yearaveragerateof

GDPpriceinflation,andUistheprojectedannualaveragerateofunemploymentrate.U*andπ*denotetheprojected

long‐runvaluesofunemploymentandinflation,

respectively,whileR*denotestheprojectedlong‐runvalueofthe

Treasurybillratelessπ*.Okun’sLawcoefficient(‐2)isusedtoconve rttheunemploymentgaptotheoutputgap.

2. AforecastyearisdeemedconstrainedwithregardstomonetarypolicyiftheBlueChipprojectionofthe

3‐month

Treasurybillrateforthatyearislessthan0.4percentagepoint.

3. TheannualcalibratedinertialruleisR(t,j)=λ

.25

R(t‐1,j)+(1‐λ

.25

){R*(j)+π(t,j)+0.5[π(t,j)–π*(j)]‐β(j)2[U(t,j)‐U*(j)]},

wherethevalueofλataquarterlyfrequencyisassumedtoequal0.8.Perceivedinertiainpolicymakingisassumedto

ariseonlyafter2011.



Page35of54



Table4

BlueChipProjectionsforConditionsWhentheFederalFundsRateLiftoffsfromNearZero

andtheirImplicationsforChangesinthePerceivedSensitivityofPolicytoSlackintheNearTerm



 ———ProjectedConditionsatLiftoff——— ———ProjectedConditionsintheLong

Run———

ImpliedOutput

GapCoefficient

inaSimple

PolicyRule



Timeof

Liftoff



Unemployment

rate

GDPPrice

Inflation

(Q4/Q4)

Unemployment

Rate

GDPPrice

Inflation

ImputedReal

FundsRate



BlueChipreleasedate 

March2009 2009Q4 9.2 1.0 5.5 2.3 2.00.25

October2009 2010Q3 9.8 1.4 5.8 2.1 2.20.34

March2010 2010Q4 9.6 1.4 6.0 2.2 2.10.36

October2010 2011Q4 9.1 1.4 5.9 2.1 1.90.38

March2011 2012Q1 8.6 1.6 5.6 2.1 1.90.46

October

2011

 2013Q3 8.3 2.0 6.0 2.1 1.70.68

March2012

 2014Q1 7.5 2.1 5.8 2.2 1.60.93

October2012

 2014Q3 7.3 2.0 5.8 2.1 1.61.02

March2013

 2015Q2 6.8 2.1 5.6 2.1 1.61.33

October2013

 2015Q2 6.6 2.1 5.6 2.1 1.61.60

 

1. Firstquarterafterthe3‐monthTreasurybillrateisprojectedtoriseabove30basispoints.

2. EqualstheBlueChiplong‐run3‐monthTreasurybillrateprojection,minus theBlueChiplong ‐runprojectionofGDPpriceinflation,plus10basispoints;thelatter

adjustmentequalstheprojectedlong‐runaveragespreadbetweentheT‐billrateandthefundsratereportedinthe JuneandDecemberreleasesofBlueChipFinancial

Forecasts.

3. ThesimpleruleisR=R*+π+0.5(π–π*)–2β(U–U*),whereRisthefederalfundsrate,R*isthesteady‐statevalueoftherealfundsrate,πisthefour‐quarterrateof

GDPpriceinflation,π*isthelong‐runtargetrateofinflation,Uistheunemploymentrate,andU*isthelong‐runvalueofU;Okun’sLawcoefficient(‐2)isusedto

converttheunemploymentgapintotheoutputgap.TocalculatethevalueofβimpliedbytheBlueChipprojections,Ratthetimeofliftoffisassumedtoequal50basis

points;UandπaresettotheirrespectiveBlueChipprojectionsfortheliftoffquarter;andU*,π*,andR*settotheirrespectivelong‐runBlueChipprojections.

4. LiftoffdateandaccompanyingeconomicconditionsarecomputedusingquarterlyinterpolationsofannualBlueChipprojectionsfortheliftoffyear.Theinterpolation

procedureisinformedbyquarterlyprojectionsofconditionsintheprioryearaswellasprojectedaverageconditionsinboththeliftoffyearandthefollowingyear.





Page36of54



Table5

InitialResponseofVariousFinancialFactorsInthePersistentEconomicSlumpScenarios

toChangesintheExpectedPolicyRuleortheEnactmentofa$1.5TrillionAssetPurchaseProgram

(resultsexpressedinbasispointsunlessotherwisenoted)



 EffectofIncreasingtheOutputGap

Coefficient(β)intheExpectedPolicy

Rulefrom0.5to1.0andIncreasingthe

InertiaCoefficient(λ)from0.0to0.8



EffectoftheAssetPurchaseProgramUnderDifferentPolicy

Rules,HoldingtheExpectedPolicyRuleConstant



β=1.0,λ=0.0β=1.0,λ=0.8 TaylorRule

TaylorRulewith

Thresholds

Non‐inertialRule

withβ=1.5

HighlyPersistentSlump 

5‐yearTreasuryyield‐13‐13‐35‐36‐38

10‐yearTreasuryyield‐8‐7‐46‐48‐48

30‐yearTreasuryyield‐4‐4‐11‐12‐12

BBBcorporatebondrate‐17‐21‐54‐56‐55

30‐yearmortgagerate‐7‐7‐43‐42‐45

10‐yearexpectedinflatio n 14 22 10 11 8

Equityprices

 4.0 5.5 5.6 6.1 5.5

Realexchangerate

‐0.8‐1.0‐2.0‐2.1‐2.0

ModeratelyPersistentSlump

5‐yearTreasuryyield‐18‐20‐30‐31‐33

10‐yearTreasuryyield‐11‐13‐43‐43‐45

30‐yearTreasuryyield‐5‐7‐9‐9‐10

BBBcorporatebondrate‐17‐23‐49‐49‐49

30‐yearmortgagerate

‐10‐12‐40‐40‐41

10‐yearexpectedinflatio n 10 16 9 8 6

Equityprices

 3.4 5.1 4.6 4.5 4.3

Realexchangerate

‐0.7‐1.0‐1.9‐1.8‐1.8

1. Forallpolicyrules,thecoefficientontheinflationgap(α)isheldconstantat0.5.Theinitialeffectoftheassetpurchasesprogramon10‐yearterm

premiumsis‐60basispoints;termpremiumeffectsareassumedtofadeawayata10percentrateperquarter.

Percentchange.



Page37of54



2008 2010 2012 2014 2016 2018 2020

2007 forecast 2008 forecast 2009 forecast 2010 forecast

2011 forecast 2012 forecast 2013 forecast

3-Month Treasury Bill Rate

0.8

1.2

1.6

2.0

2.4

2.8

2008 2010 2012 2014 2016 2018 2020

GDP Price Inflation

2008 2010 2012 2014 2016 2018 2020

Unemployment Rate

-10

-8

-6

-4

-2

2008 2010 2012 2014 2016 2018 2020

Imputed Output Gap

Figu

e1.BlueChipFo

ecas

age

nnual Economic Condi

ions

(solid are March releases, dashed are October releases)



Page38of54



0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

2009 2010 2011 2012 2013 2014 2015

March 2009 Surv ey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

2009 2010 2011 2012 2013 2014 2015

October 2009 Survey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

2010 2011 2012 2013 2014 2015 2016

March 2010 Survey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

2010 2011 2012 2013 2014 2015 2016

October 2010 Survey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

2011 2012 2013 2014 2015 2016 2017

March 2011 Survey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

2011 2012 2013 2014 201

2016 201

October 2011 Survey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

2012 2013 2014 201

2016 201

2018

forecast standard Taylor rule estimated annual rule estimated liftoff rule calibrated inertial rule (2012-13 only)

March 2012 Surv ey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

2012 201

2014 2015 201

2017 201

October 2012 Survey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

201

2014 2015 201

2017 201

201

March 2013 Surv ey

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

3.25

3.50

3.75

4.00

4.25

4.50

4.75

5.00

201

2014 201

2016 201

201

2019

October 2013 Survey

Figure 2. Annual Blue Chip Treasury Bill Forecasts Versus Simple Rule Prescriptions



Page39of54



-140

-120

-100

-80

-60

-40

-20

2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

LSAP1

LSAP1 plus reinvestment

LSAP2

MEP1

MEP2

LSAP3

Figure 3

Estimated Term Premium Effects of the FOMC's Asset Purchases

Source: Ihri

et al.

(

2012

)

.EffectsofLSAP3

ram com

uted b

the authors; see text for details.



Page40of54



0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

00 01 02 03 04 05 06 07 08 09 10

highly persistent slump with beta = 0.5 highly persistent slump with beta = 1.0

moderately persistent slump with beta = 0.5 moderately persistent slump with beta = 1.0

Nominal Federal Funds Rate

year

percent

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

00 01 02 03 04 05 06 07 08 09 10

Real 10-Year Treasury Yield

year

percent

5.2

5.6

6.0

6.4

6.8

7.2

7.6

8.0

8.4

8.8

9.2

9.6

10.0

10.4

00 01 02 03 04 05 06 07 08 09 10

Unemployment Rate

year

percent

Note. The baseline rule is the Taylor rule (lambda = 0, alpha = 0.5, and beta = 0.5). In the more responsive case, the coefficient on slack

(

beta

)

is raised to 1.0 while holdin

lambda and al

ha constant.

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

00 01 02 03 04 05 06 07 08 09 10

Inflation Rate (4-Quarter)

year

percent

Figure 4. Effect of Increasing the Policy Response to Slack



Page41of54



-.8

-.7

-.6

-.5

-.4

-.3

-.2

-.1

00 01 02 03 04 05 06 07 08 09 10

beta = 0.5 with thresholds beta = 1.0 beta = 1.5 beta = 1.0 with inertia (lambda = 0.8)

Highly Persistent Slump

year

percent

Marginal Effect on Unemployment

-.8

-.7

-.6

-.5

-.4

-.3

-.2

-.1

00 01 02 03 04 05 06 07 08 09 10

Moderately Persistent Slump

year

percent

.00

.05

.10

.15

.20

.25

.30

.35

.40

00 01 02 03 04 05 06 07 08 09 10

Highly Persistent Slump

year

percent

Note. Mar

inal e

ects are de

ined relative to outcomes under the Ta

lor rule

(

lambda = 0, al

ha = 0.5, and beta = 0.5

)

.00

.05

.10

.15

.20

.25

.30

.35

.40

00 01 02 03 04 05 06 07 08 09 10

Moderately Persistent Slump

year

percent

Marginal Effect on Inflation

Figu

e5.Ma

ginal Effects of Mo

essive o

Ine

tial Rules



Page42of54



-.8

-.7

-.6

-.5

-.4

-.3

-.2

-.1

00 01 02 03 04 05 06 07 08 09 10

beta = 1.0 beta = 1.5 beta = 1.0 with inertia (lambda = 0.8)

Immediate Policy Response

and Recognition of Shocks

year

percent

-.8

-.7

-.6

-.5

-.4

-.3

-.2

-.1

00 01 02 03 04 05 06 07 08 09 10

Immediate Policy Response

but Gradual Recognition of Shocks

year

percent

Marginal Effect on Unemployment

-.8

-.7

-.6

-.5

-.4

-.3

-.2

-.1

00 01 02 03 04 05 06 07 08 09 10

Immediate Recognition of Shocks

but Delayed Policy Response

year

percent

.00

.05

.10

.15

.20

.25

.30

.35

00 01 02 03 04 05 06 07 08 09 10

Immediate Policy Response

and Recognition of Shocks

year

percent

Note. Marginal effects are defined relative to outcomes under the Taylor rule (lambda = 0, alpha = 0.5, and beta = 0.5).

.00

.05

.10

.15

.20

.25

.30

.35

00 01 02 03 04 05 06 07 08 09 10

Immediate Policy Response

but Gradual Recognition of Shocks

year

percent

Marginal Effect on Inflation

.00

.05

.10

.15

.20

.25

.30

.35

00 01 02 03 04 05 06 07 08 09 10

Immediate Recognition of Shocks

but Delayed Policy Response

year

percent

Figure 6. Effects of Learning and Delay on the Marginal Effects of More Aggressive or Inertial Rule



Page43of54



-.6

-.5

-.4

-.3

-.2

-.1

00 01 02 03 04 05 06 07 08 09 10

beta = 0.5 beta = 0.5 with thresholds beta = 1.0

beta = 1.5 beta = 1.0 with inertia (lambda = 0.8)

Highly Persistent Slump

year

percent

Marginal Effect on Unemployment

-.6

-.5

-.4

-.3

-.2

-.1

00 01 02 03 04 05 06 07 08 09 10

Moderately Persistent Slump

year

percent

.00

.02

.04

.06

.08

.10

.12

.14

.16

.18

.20

00 01 02 03 04 05 06 07 08 09 10

Highly Persistent Slump

year

percent

Marginal Effect on Inflation

Note. Mar

inal effects of asset

urchases under each rule are defined relative to outcomes under the same rule

ithout

urchases..

.00

.02

.04

.06

.08

.10

.12

.14

.16

.18

.20

00 01 02 03 04 05 06 07 08 09 10

Moderately Persistent Slump

year

percent

Figu

e 7. Ma

ginal E

ect o

$1.5 T

illion in

sset Pu

chases Unde

ent Policy

ules



Page44of54



0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Figu

Evolution of the Perceived Value of Beta Used in Sequential Simulations

(black -- lambda = 0, red -- lambda = .8)



Page45of54



-5

-4

-3

-2

-1

2008 2010 2012 2014 2016 2018 2020

Real GDP Growth (4-Quarter)

percent

2008 2010 2012 2014 2016 2018 2020

Unemployment Rate

percent

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2008 2010 2012 2014 2016 2018 2020

Core PCE Inflation (4-Quarter)

percent

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

2008 2010 2012 2014 2016 2018 2020

no actions (version 1) no actions (version 2) history with Oct-2013 Blue Chip projections

Nominal Federal Funds Rate

percent

Note: The version 1 simulation assumes that agents in the historical baseline always view the fund rate rule as non-inertial. In the version 2 simulations, agents

in the baseline perceive the funds rate rule becoming inertial (lambda=0.8) beginning in 2012.

-2.0

-1.6

-1.2

-0.8

-0.4

0.0

0.4

0.8

1.2

1.6

2008 2010 2012 2014 2016 2018 2020

Real Federal Funds Rate

percent

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

2008 2010 2012 2014 2016 2018 2020

10-Year Treasury Yield

percent

Figure 9. Predicted Evolution of the Economy in the Absence of Unconventional Policy



Page46of54



-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

baseline model

rational expectations in financial markets only

alternative inflation dynamics

low interest elasticity of demand

policy-linked house price effects

Unemployment Effects

percent

-.1

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

baseline model

rational expectations in financial markets only

alternative inflation dynamics

low interest elasticity of demand

policy-linked house price effects

Inflation Effects

percent

Figure 10. Estimated Effects of of

ncon

entional Policy

in Diferent Specifications of the FRB/US Model

Note. Results expressed as differences from history and the October 2013 Blue Chip forecasts.



Page47of54

120

130

140

150

160

170

180

190

200

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

2009

2010

2011

2012

2013

Figu

e11.Su

vey-Based House P

ice Fo

ecas

Forecasts beginning in Q1 of years 2010 to 2013 (solid lines) are based on the current and coming year house price

projections published in the Q1 Survey of Professional Forecasters; the 2009Q1 forecasts are based on the projections

published in the February 2009 Blue Chip survey. Forecasts beginning in Q3 (dashed lines) are based on real-time

published data and an average of survey data released in the Q1 SPF for the current and coming years. Beyond the

coming year, relative house prices are projected to gradually return to their estimated 1975-2000 trend.

Page48of54



-6

-4

-2

2008 2010 2012 2014 2016 2018 2020

Real GDP Growth (4-Quarter)

percent

2008 2010 2012 2014 2016 2018 2020

Unemployment Rate

percent

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2008 2010 2012 2014 2016 2018 2020

Core PCE Inflation (4-Quarter)

percen

1.0

2008 2010 2012 2014 2016 2018 2020

version 1 policy (non-inertial) version 2 policy (inertial) history with Oct-2013 Blue Chip projections

Nominal Federal Funds Rate

percent

Note: The version 1 simulation assumes that agents in the historical baseline always view the fund rate rule as non-inertial. In the version 2 simulations, agents in the

baseline perceive the funds rate rule becoming inertial (lambda=0.8) beginning in 2012.

-3

-2

-1

2008 2010 2012 2014 2016 2018 2020

Real Federal Funds Rate

percent

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

2008 2010 2012 2014 2016 2018 2020

10-Year Treasury Yield

percen

Figure 12. Predicted Evolution of the Economy if Late-2013 Perceptions of Unconventional Policy

Had Been in Place in Early 2009



Page49of54



Appendix1

FinancialMarketSurveysofPolicyExpectationsandtheEconomicOutlook



Duetodatalimitations,wecannotdirectlycomparedourestimatesofthepolicyexpectationsofprivate

forecastersimpliedbytheBlueChipEconomicIndicators(BCEI)surveytothebeliefsoffinancialmarket

participants.

However,theforecastsreportedinBlueChipFinancialForecasts(BCFF)andtheFederal

ReserveBankofNewYork’sPrimaryDealersSurvey(PDS)doallowforsomeconsistencychecks.As

showninTableA1,projectionsforboththedateofliftoffandeconomicconditionsatthetimeofliftoff

arequitesimilar

acrossthethreesurveys.Theonlymajorexceptionoccursinthefallof2012,whenthe

primarydealersanticipatedthatliftoffwouldoccurafullyearlaterthantheBCEIrespondents;this

differencemayreflectfinancialmarketparticipants’ greaterattentivenesstothelikelihoodofan

impendingshiftinpolicyat

thetime.TableA2alsocomparesresultsfromthevarioussurveys

concerningexpectedlong‐runconditions.Ascanbeseen,expectationsforlong‐runratesofinflation

andrealinterestareessentiallythesameintheBCEIandtheBCFF.Long‐runinflationexpectationsin

thePDSarealsosimilarto

thosereportedintheBCEI;however,theprimarydealersreportedsomewhat

higherestimatesofthelong‐rununemploymentrateandtheequilibriumrealrate.



Inadditiontothesecomparisons,Femia,FriedmanandSack(2013)provideanotherperspectiveby

usingthePDStoinferfinancialmarketparticipants’beliefsabouttheFOMC’s

implicitpolicyrule.From

earlyAugust2011throughearlyNovember2012,specialquestionswereincludedinthePDSconcernin g

thelikelyleveloftheunemploymentrateatthetimeofliftoff,conditionalondifferentassumptionsfor

theaccompanyingrateofinflation.Undersomesimplifyingassumptions,theresultinggridof

anticipatedunemployment

‐inflationcombinationscanbeusedtomapoutchangesovertimein

expectationsfortheimplicitpolicyreactionfunctioninforceatthetimeofliftoff.Theiranalysis

suggeststhatmarketexpectationsinAugust2011wereconsistentwithasimplepolicyruleinwhichβ

equals1.0butthattheperceived

valueofthiscoefficienthadrisento1.25bySeptember2012,

conditionalonαequaling0.5.

BothvaluesaresomewhathigherthatthecorrespondingBCEI

estimatesreportedinTable4.





Theestimatedvalueofβissomewhatsensitivetohowmovementsintheunemploymentratearetranslated

intomovementsintheoutputgap—thatis,intothevalueofOkun’sLawcoefficient.Consistentwithassumptions

madeelsewhereinthispaper,wehaveassumedthatthiscoefficientequals2.0;ifwehadinstead

setOkun’sLaw

coefficientto2.3,asFemia,FriedmanandSackdidintheirstudy,theimpliedvalueofβwouldfallto0.9.In

addition,thevalueofβimpliedbythePDSresultsdependslinearlyontheassumedvalueofα.Bythemselves,the

surveyquestionscannotbeusedtoestimate

theperceivedvaluesofbothαandβunlessoneiswillingtomake

assumptionsregardingmarketbeliefsconcerningR*,π*,thenaturalrateofunemployment,andOkun’sLaw.

Reasonablevariationsinassumptionsforthesefactorsyieldestimatesofαthatrangefrom0.25to1.00,with

highervaluesofαimplyinghigherestimated

valuesofβ.

Page50of54



TableA1

ComparisonofBlueChipProjectionsofLiftoffConditionswithCorrespondingForecastsfromFinancialMarketParticipants



ProjectedLiftoffDatefortheFederal

FundsRate



Q4UnemploymentRate

inProjectedLiftoffYear



AnnualGDPorCorePCEInflationRate

inProjectedLiftoffYear

 BlueChip

BlueChip

Financial

Forecasts

Primary

Dealers

Survey

 BlueChip

Primary

Dealers

Survey



BlueChip



(GDP)

BlueChip

Financial

Forecasts



(GDP)

Primary

Dealers

Survey



(corePCE)

Surveyreleasedate    

March2009 2009Q4 2010Q19.21.0 1.6

October2009 2010Q3 2010Q29.61.4 1.5

March2010 2010Q4 2010Q39.61.4 1.2

October2010 2011Q4 2011Q39.11.4 1.5

March2011 2012Q1 2011Q4 2012Q3 8.1 7.91.6 1.9 1.6

October2011 2013Q3

 Mid‐2013 2013Q4 8.1

 8.32.0 2.1 1.6

March2012 2014Q1

 2014Q1

 2014Q3 7.0

 7.02.1 2.1 2.0

October2012 2014Q3

 2015Q3 7.1

 7.02.0 1.9

March2013 2015Q2

 2015H2

 2015Q3 6.5

 6.52.1 2.1 2.0

October2013 2015Q2

 2015H2

 2015Q3 6.3

 6.12.1 2.0 1.9

Note.ForOctoberreleases,PrimaryDealerSurveyresultsreflectthesurveytakeninmid‐September.

1. Imputedusingquarterlyandannualforecastdata.

2. Dateinferredfromrespondents’assessmentthattherewasa69%probabilityofliftoffoccurringin2014Q1orlater,anda31%probabilitythatitwouldoccurearlier.

3. Mostlikelydatebasedonrespondentsreportedprobabilityassessments.



Page51of54



TableA2

ComparisonofBlueChipProjectionsofLong‐RunConditionswithCorrespondingForecastsfromFinancialMarketParticipants



 Long‐RunRealFederalFundsRate



Long‐Run

UnemploymentRate

 Long‐RunGDPInflationRate



 BlueChip

BlueChip

Financial

Forecasts



Primary

Dealers

Survey



 BlueChip

Primary

Dealers

Survey



 BlueChip



BlueChip

Financial

Forecasts

3

Primary

Dealers

Survey

4

BlueChipSurveyReleaseDate

March2009 2.0 1.85.52.3 2.4

October2009 2.2 1.95.82.1 2.3

March2010 2.1 1.96.02.2 2.2

October2010 1.9 1.85.92.1 2.1

March2011 1.9 1.65.62.1 2.2

October2011 1.7 1.66.02.1 2.1

March2012 1.6 1.55.8 6.22.2 2.1 2.1

October2012 1.6 1.6 1.9 5.8 6.02.1 2.2 2.1

March2013 1.6 1.6 1.9 5.6 6.02.1 2.2 2.1

October2013 1.6 1.6 1.9 5.6 5.92.1 2.1 2.1

1. FortheBlueChipsurvey,federalfundsrateforecastsareapproximatedbytheprojectedTreasurybillrateplus 10basispoints.Forallthreesurveys,realratesare

constructedbysubtractingthelong‐runprojectionforGDPpriceinflation.

2. ThePrimaryDealersSurveyonlyreportsinflationforecastsona

PCEoraCPIbasis,soinflationforecastsonaGDPbasisareconstructedfromPCE‐basedmeasuresby

adding10basispoints,andfromCPI‐basedmeasuresbysubtracting20basispoints.

3. BlueChipFinancialForecasts onlyreportextendedforecastsinJuneandDecember,soresultsfromthesetwomonthsareusedforMarchandDecember,respectively.

4. OctoberresultsforthePrimaryDealersSurveyarebasedonsurveystakeninmid‐SeptemberimmediatelypriortotheFOMCmeeting.



Page52of54



Appendix2

FinancialMarketResponsestoQEEvents:ASummaryofEventStudyResults



InthisAppendix,wesummarizethefindingsofanumberofrecentstudiesthatexaminetheresponses

ofvariousyieldsandassetpricestotheannouncementoflarge‐scaleassetpurchases.This

evidence,

whichmostlyreliesoneventstudies,isimportantbecauseitpresentsthefirststepinthetransmission

ofQEpoliciestorealactivityandinflation,andbecauseitprovidesevidenceagainstwhichtoassessthe

transmissionmechanismofamodelusedtoevaluatethemacroeconomiceffectsofthesepolicies.



Mortgagerates.HancockandPassmore(2014)estimatesubstantialeffectsofLSAPsonthesecondary‐

marketyieldsonagencyMBSandonlyslightlysmallereffectsonprimary‐marketconforminghome

mortgagerates.Moreover,theyreportevidencethattheFederalReserve’spurchasesofagencyMBS

haveanextraeffectontheyields

onthosesecurities(relativetoTreasuryyields)andthat,overtime,a

largeportionofthoseeffectsalsopassthroughtoprimarymortgagerates.

Forexample,followingthe

announcementofthenewassetpurchaseprograminSeptember2012,thespreadbetweenagencyMBS

yieldsand10‐yearTreasuryyieldsfellnoticeablyafterboththeinitialannouncement ofandsubsequent

communicationsaboutthisprogram.Moreover,althoughthespreadbetweentheinterestrateson

conforming30‐

yearfixed‐rateloansandagencyMBSyieldswidenedatfirst,thisspreadgradually

movedbackdowntobelowitslate‐August2012level,whichisconsistentwithessentiallyallofthe

decreaseinagencyMBSyieldsbroughtaboutbytheadditionalpurchaseseventuallypassingthroughto

conformingmortgageinterestrates.



Corporateyields.Recentresearchhasalsoreportedsubstantialpass‐throughofLSAP‐inducedchanges

inTreasuryyieldstoyieldsoninvestment‐gradecorporatebonds. Gilchrist,López‐Salido,andZakrajšek

(2014)findalmostcompletepass‐throughofLSAP‐inducedreductionsinTreasuryyieldstoyieldson

investment‐gradecorporatebonds,while

Kiley(2013a)estimatesthataboutthree‐quartersofthedrop

inTreasuryyieldspassedthrough.Bycontrast,pass‐throughofLSAPeffectstoyieldsonspeculative‐

gradecorporatebondshasbeenfoundtobesmall(exceptinsomeanalysesofthefirstroundofLSAPs,

whichwasannouncedatatime

whenmarketswerehighlydysfunctionalandriskspreadsexceptionally

wide).Forexample,focusingontheFederalReserve’sbalance sheetactionsfromlate‐2008through

2010,KrishnamurthyandVissing‐Jorgenson(2011)reportedessentiallycompletepass‐throughoflower

long‐termTreasuryyieldstoinvestment‐gradecorporatebondyieldsfromthefirst

twoLSAPs,but

sizablepass‐throughtospeculative‐gradecorporatebond yields onlyforthefirstroundofLSAPsinthe

winterof2008and2009.

Exchangerate.GlickandLeduc(2013)documentthatLSAPannounce mentsscaledtoa20basispoint

reductionin10‐yearTreasuryyieldshavetendedto

lowerthebroadcurrencyindexused inFRB/USby

about¾percent.Kiley’s(2013b)econometricanalysisofselectedbilateralexchangeratesimpliesthata

20basispointreductioninthe10‐yearTreasuryyieldovertheperiodsincenonconventionalpolicies



WethankMichaelPalumboforhelpfuladviceonthisappendix.

Inthisandanearlierpaper(HancockandPassmore,2012),theauthorsemphasizedthattheextentofpass‐

throughofchangesinagencyMBSyieldstoprimarymortgageratesappearstodependonprevailingmarket

conditions,includingthelevelofmortgagerates:Pass‐throughhastendedtobemorecomplete

whenmortgage

ratesarehigherthanlower,likelyreflectinginparttheeffectsoflimitedcapacityinthemortgageorigination

industry.

Page53of54



havebeenundertakenledtoa1½to2percentdepreciationofthedollaragainsttheeuroandthe

Japaneseyen.

Equityprices.TurningtothestockmarketreactionstonewsabouttheFederalReserve’sbalancesheet

policies,eventstudyofstockmarketresponsesinnarrow

windowsaroundFOMCannouncementsand

othercommunicationsassociatedwiththefirsttworoundsofLSAPs suggestedaresponseofcloseto2

percentfora20basispointdecreaseinthe10‐yearTreasuryyield.Rosa(2012)andKiley(2013c)

estimatedthestockmarketeffectsofchangesinTreasurytermpremiums

inducedbyLSAPs(holding

marketexpectationsforpathofthefederalfundsrateroughlycons tant)andfoundsmallerresponses—

averagingabouta1percentincreaseinstockpricesforanLSAP‐induced20basispointdecreaseinthe

10‐yearTreasurytermpremium.

Summary.Therearereasonstothinkthatthe

stockmarketresponsetoLSAPscoulddepend

importantlyonprevailingmarketconditionsandthateventstudiesmightnotalwayspickupthefull

effectofsuchmonetarypolicyactions.ThefirstLSAPs—announce d inlate‐2008andearly2009—

appeartohavebeenkeyelementsinasuiteofpolicyresponsesto

thefinancialcrisisthatsignificantly

bolsteredequitypricesovertime.

Estimatesoftheequitypremiumhadsoaredduringthemostacute

phaseofthefinancialcrisisthatautumn,butnarrowedsubstantia llyinthespringandsummerof2009,

asU.S.stockpricesclimbed.Asimilarbutlesspronouncedpatternwasevidentaroundthetimeof

LSAP‐relatednewsandannouncements

inthesummerof2010.Theseobservationssuggestthata

considerablepartofthestockmarketresponsetothefirsttworoundsofLSAPsoccurredoutsideof

narroweventwindows.

Inaddition,theevidenceindicatesthattheearlierrounds ofLSAPsledmarketparticipantstoexpecta

moreaccommodativestanceof

conventionalmonetarypolicy(BauerandRudebusch,2012,and

KrishnamurthyandVissing‐Jorgenson,2011),whichcouldhaveproducedarelativelystrong

transmissionsinceconventionalmonetarypolicyactionshavebeenshowntogeneratesizablestock

marketresponses(BernankeandKuttner,2005,Rosa,2012,Kiley,2013c).BecausetheFOMCwas

alreadyusingcalendar‐

basedguidanceregardingthelikelymedium‐termpathoffederalfundsratein

September2012andshiftedtoquantitativethresholdsforitsforwardguidanceinDecember2012,the

scopeforLSAPannouncementstooperateforcefullythroughsuchasignalingchannelmayhavebeen

limitedatthetimeoftheannouncementof

LSAP3.

Insum,theseeventstudiessuggestthatthefirsttworoundsofLSAPsprobablyspurredlargerresponses

ofinterestrates,theexchangevalueofthedollar,andequitypricesthanlateronesbecausetheearlier

programswereannouncedandimplementedattimeswhenmarketconditionswerehighlystrainedand

riskpremiumswereexceptionallylargeorwhentherewasgreaterscopeforsignalingmore

accommodativeintentions withrespecttothefederalfundsrate.Thefollowingtablereportsthe

centraltendencyofestimatesoftheeffectsofahypotheticalLSAPprogram,scaledtoaninitialimpact

on10‐yearTreasuryyieldsof

‐20basispoints,onselectedfinancialmarketquotes;ofcourse,thereisa

greatdealofuncertaintysurroundingallthesefigures.





Forexample,thecompletionoftheFederalReserve’sSupervisoryCapitalAssessmentProgramforthemajor

bankingfirmsalsoappearedtoplayanimportantroleinallayingsomeoftheconcernsamongmarketparticipants.

Page54of54



Table1

EffectsofaHypotheticalLSAPProgramonSelectedFinancialMarketQuotes



10‐yearTreasuryyield−20basispoints

30‐yearcurrentcouponagencyMBSyield−27basispoints

Conforming30‐year(fixed)mortgagerate−25basispoi nts

BBBcorporatebondyield−15basispoints

Exchangevalueofthedollar(broadindex)−3/4percent

S&P500stockpriceindex 1‐3/4percent



Note.Thishypotheticalprogramisassumedtoinvolvepurchasesofequalamountsoflonger‐termTreasury

securitiesandagencyMBS.

