宏观经济公告前的债券供应、收益率漂移和流动性提供

BISWorkingPapers No1232 Bondsupply,yielddrifts,andliquidityprovisionbeforemacroeconomicannouncements byDongLou,GaborPinter,SemihUsluandDannyWalker MonetaryandEconomicDepartment December2024 JELclassification:G11,G12,G14,D83,D84. Keywords:Macroeconomicannouncements,interestratedrift,bondsupply,liquidityprovision. BISWorkingPapersarewrittenbymembersoftheMonetaryandEconomicDepartmentoftheBankforInternationalSettlements,andfromtimetotimebyothereconomists,andarepublishedbytheBank.Thepapersareonsubjectsoftopicalinterestandaretechnicalincharacter.TheviewsexpressedinthemarethoseoftheirauthorsandnotnecessarilytheviewsoftheBIS. ThispublicationisavailableontheBISwebsite(www.bis.org). ©BankforInternationalSettlements2024.Allrightsreserved.Briefexcerptsmaybereproducedortranslatedprovidedthesourceisstated. ISSN1020-0959(print) ISSN1682-7678(online) BondSupply,YieldDrifts,andLiquidityProvisionBeforeMacroeconomicAnnouncements∗ DongLouHKUST&LSE GaborPinter BankforInternationalSettlements SemihÜslüJohnsHopkinsCarey DannyWalkerBankofEngland 17thNovember2024 Abstract UKgovernmentbondyieldstendtoriseinatwo-daywindowbeforescheduledmacroeconomicannouncementssuchaslabourmarketdatareleasesandmonetarypolicynews.Thiseffect,particularlypronouncedduringUKbondissuances,islinkedtohighertermpremia.Financialintermediaryconstraintsplayaroleasdealersavoidaccumulatinginventoryinpre-newswindowswithissuances.Thecompositionofliquidityprovidersalsoshifts:hedgefundsbuyalargershareofthebondissuanceoutsidepre-newswindows,butmorepassiveinvestors,suchasforeigncentralbanksandpensionfunds,provideliquidityinpre-newswindows.Weoutlineasimplemodeltorationalisethesefindings. Keywords:MacroeconomicAnnouncements,YieldDrift,BondSupply,LiquidityProvision JELcodes:G11,G12,G14,D83,D84 ∗Thepreviousversionofthepaperwastitled‘Fiscal-MonetaryInteractions:Pre-AnnouncementLiquidity EffectsAfterBondIssuance’.Wewouldliketothank,forhelpfulcommentsandsuggestions,LaurentBach,LaurenceDaures,VadimElenev,AndrasFulop,NicolaFusari,RodrigoGuimaraes,SudipGupta,FlorianHeider,HibikiIchiue,MikeJoyce,PeterKaradi,JohnLewis,JohnPower,IlhyockShim,MattRoberts-Sklar,MisaTanaka,Jean-CharlesWijnandtsandtheparticipantsatthe2ndECB-BoE-BoJJointResearchWorkshop,andtheBIS,theESSECandJohnsHopkinsUniversityresearchseminars.Theviewsexpressedinthispaperarethoseoftheauthors,andnotnecessarilythoseoftheBankforInternationalSettlements,theBankofEnglandoritscommittees.Emailaddresses:dlou@ust.hkandd.lou@lse.ac.uk;gabor.pinter@bis.org;semihuslu@jhu.edu;danny.walker@bankofengland.co.uk.CorrespondingAuthor:GaborPinter,BankforInternationalSettlements. 1Introduction Twoofthemostimportantdriversofnominalinterestratesarethearrivalofmacroeco-nomicnews(Gürkaynak,Sack,andSwanson,2005a,b)andchangesingovernmentbondsupply(GreenwoodandVayanos,2014).Giventherisinglevelsofgovernmentdebtandincreasedprimaryissuanceindevelopedmarketsinrecentyears,itisnaturaltopresumethatnewsandbondsupplyeffectsonyieldshavebecomeincreasinglyimportantandintertwined.Figure1illustratesthesharpriseinthefrequencyofgovernmentbondissuanceandtheincreasedco-occurrencesofbondissuanceandmacroeconomicannouncementsintheUK.Forexample,since2005,themajorityoftheBankofEngland’sMonetaryPolicyCommittee(MPC)announce-mentshavebeenprecededbygovernmentbondissuanceinthetwodayspreceding.Yet,howtheinteractionbetweenthesetwoforcesaffectinvestorbehaviour,bondmarketliquidity,andultimatelyinterestratesisnotwellunderstood.Ourpaperusesbothaggregateandgranular,transaction-leveldatatoempiricallystudythisinteractionanditseffectoninterestrates.Wealsopresentasimplemodeltoillustratethemechanicsofsecondarymarkettradingfollowingbondissuanceandpriortonewseventssuchasthearrivaloflabourmarketdatareleasesandmonetarypolicyannouncements. Westartourempiricalanalysisbydocumentingthatlong-termbondyieldssystematicallyriseinatwo-daywindowbeforethescheduledarrivalofmacroeconomicnewssuchaslabourmarketdatareleasesandmonetarypolicyannouncements,whichwerefertoas‘pre-newswindows’.Werefertotheyieldchangesinpre-newswindowsas‘pre-newsyielddrift’.Overoursampleperiodof1997-2021,thispre-newsdriftpushedupyieldsbyabouttwopercentagepointsin10-20yearmaturities,whichisnon-negligiblecomparedtoatotalfallof6-7percentagepointssince1997.Theeffectconcentratesinpre-newswindowsthatcoincidewithnewissuanceofgovernmentbonds.Forexample,asabaseline,theaveragedailychangein10-and20-yearyieldsduringpre-newswindowsis0.3-0.5bpslargerthanyieldchangesoutsidepre-newswindows.Butthisdifferencerisesto0.6-1.1bpswhenthepre-newswindowcoincideswithprimaryissuances.Decomposingthepre-newsdriftintoatermpremiumcomponentandchangesinexpectationsaboutfutureshort-terminterestrates,wefindthattermpremiaplayadominantroleinourresults. Toanalysethemechanismsunderliningthepre-newsyielddriftanditsinteractionwithprimaryissuances,westudythebehaviourofbothprimarydealersandclients.Theexplanationweexplore,whichwealsoformaliseinatheoreticalmodel,relatestothelimitedrisk-bearingcapacityofprimarydealersduringgovernmentbondissuance,whichbecomesmorepronouncedwh