全球宏觀策略-賣出債券-2021.7.23-121正式版

GlobalMacroStrategy

The2021InstitutionalInvestor(II)GlobalFixed-IncomeResearchpollisopen.Ifyouenjoyourwork,pleaserateusfivestarsinthefollowingcategories:Global>MacroStrategyandregionallyinthefollowingcategories:Fixed-IncomeStrategy,InterestRateStrategy,andCurrency&ForeignExchange.Thanksforyourreadershipandsupport!

MORGANSTANLEY&CO.LLC MatthewHornbach

Matthew.Hornbach@

+1212761-1837

MORGANSTANLEY&CO.INTERNATIONALPLC FilipDenchev

Filip.Denchev@

+44207677-3166

TheHieroglyphicsSaySellBonds

WesuggestinvestorsstayunderweightUSdurationatcurrentlevelsorlower,downto1.11%on10yUSTreasuryyields.Below1.11%,wewouldexpect10yyieldstocontinuelowertoward0.99%.TherecentmovefromtheMarchpeakat1.77%hasretraced50%ofthebearmarketthatbeganinearlyAugust2020.Thereisnothingmagicalabouta50%retracement,butothertechnicalfactorssuggesttherallyisover.

Elliottwaveanalysissuggests10yyieldscompletedanA-B-Ccorrectionthisweek,andtheTDsequentialindicatorplacesa9countonthisweek'spriceaction.YieldsremainabovetheIchimokucloud,suggestingthebearmarketremainsalive,andatwistinthecloud–whichwilloccurnextweek–usuallymarksaturningpointintherecenttrend.Finally,thepast-weekcandlestickformedadragonflydoji.Timetosellbonds.

Exhibit1:UST10-yearon-the-runyield–weeklycandlestickchart

Source:MorganStanleyResearch,Bloomberg

2

ButWhatAboutaLowerr*?

SellingTreasuriesinthefaceofalowerequilibriumrealinterestrate,r*,seemsunwise.Ofcourse,thatassumesr*hasfalleninthewakeofthepandemic,assomeargue,orthatr*dictatesinvestordecisionsenmasse.Whilecomplexmodelsattempttoestimater*inrealtime,andsavvyinvestorscomparemarket-impliedlonger-runratestotheseestimates,wethinkaboutTreasuryvaluationsdifferentlyandina"realworld"context.

Therearetwocommonwaysofdecomposinggovernmentbondyieldsinordertoarriveatsometypeofvaluation,thoughneithergiveyoudirectaccesstotheconceptofr*.Youcansplitthenominalyieldinto(1)arealyield–whichgetsyoupartwaytor*–and(2)ayieldmeanttocompensateforinflation.

Oryoucansplitthenominalyieldinto(3)anexpectationfortheaverageshort-termrateoverthematurityofthebond–whichgetsyoupartwaytor*,assumingyouchoosealonger-maturitybond–and(4)atermpremiummeanttocompensatefortheriskyourexpectationiswrong.

Mostinvestorsprefertheformerdecompositionbecauserealyieldsandbreakeveninflationratesaretradeable.Buttoassessyieldlevelsrelativetosomenotionofthelongerrun,orequilibrium,wepreferlookingatratesthrougharateexpectationandtermpremiumlens.AsExhibit2shows,rateexpectationsmakeupthelion'sshareofthelevelof10yTreasuryyields,whiletermpremiumsrepresentaresidual.

So,inordertoforecastthelevelof10yTreasuryyields,spendmostofyourtimeonforecastingtherateexpectationscomponent.Again,youcanthinkof10y"rateexpectations"astheaverageshort-termrateexpectedoverthenext10years.Exhibit3suggeststhatamajordriverofthoseexpectationsisthefedfundstargetrateitself.

Exhibit2:UST10yyieldsandembeddedtermpremiumandrateexpectations

Exhibit3:UST10yrateexpectationsandtheFed'stargetfederalfundsrate

%

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5

Jul-15 Jul-16 Jul-17

UST10yTP

%

7

6

5

4

1.19

3

1.05

2

-0.23

1

1.05

0

0.13

-1

Jul-18Jul-19Jul-20Jul-21

Jul-97Jul-01Jul-05

Jul-09Jul-13Jul-17Jul-21

UST10y

UST10yRE

Targetfedfundsrate

UST10yrateexpectations(RE)

Source:MorganStanleyResearch Source:MorganStanleyResearch,FederalReserve

3

Howcantheshort-termrate,atagivenpointintime,influenceexpectationsofwhattheshort-termratewillaverageinthefuture?Wethinkrecencybiasplaysarole,i.e.,ifthefedfundsrateishigh,investorsexpectittostayhigh,andviceversa.Butit'snotjustthelevelthatinfluencesexpectations.It'salsothedirectionoftravel.Forexample,iftheFedishikingrates,investorstendtothinkratehikeswillcontinue,andthatraisestheforward-lookingaverage.

WhiletheTreasurymarketisforward-looking,thepastsuggestsit'susuallywrong.Exhibit4showsahistoryofthetargetfedfundsrateanditsbackward-looking10-yearaverage.Overthepast10years,themidpointofthetargetraterangeaveraged0.64%.Accordingtoourtermstructuremodel,10yTreasuryyieldsimplyanaverageof1.05%overthenextdecade.

Hasthe10yTreasuryyieldhadmuchsuccessinpredictingthisforward-lookingaverage?Notreally.Exhibit5compareshowthemarket-implied10yrateexpectationshaveevolvedrelativetotheex-postreality,i.e.,wherethetargetfedfundsrateactuallyaveragedoverthenext10years.Market-impliedrateexpectationsweretoohighpersistently.

Adecadeago,rateexpectationsembeddedinto10yyieldsimpliedthatthetargetfedfundsratewouldaverage2.00%betweenJune2011andJune2021.Instead,thetargetfedfundsrateaveraged0.64%.Thequestioninvestorsmustanswertodayis,willthetargetfedfundsrateaveragemoreorlessthan1.05%throughJuly2031?

Exhibit4:Targetfedfundsrate*,its10ymovingaverage,anditsmarket-implied10yaverage

Exhibit5:UST10yrateexpectationsandex-post10yrollingaverageoftargetfedfundsrates*

%

%

9

7

8

6

7

6

5

5

4

4

3

3

2.00

2

2

1.05

1

1

Marketmiss

1.05

0.64

0

0.13

Jul-91

Jul-96Jul-01Jul-06Jul-11Jul-16Jul-21

Fedfundstargetrate

10yrollingaveragetargetfedfundsrate

Currentmarket-implied10yaverage

Source:MorganStanleyResearch,FederalReserve

Note:priortoDecember2008,weusedthetargetrate;startingDecember2008,weusedthemid-pointofthetargetraterange

0

0.64

Jul-91Jul-96Jul-01Jul-06Jul-11Jul-16Jul-21

UST10yrateexpectation

10yrollingaveragetargetfedfundsrate,lagged10years

Source:MorganStanleyResearch,FederalReserve

Note:priortoDecember2008,weusedthetargetrate;startingDecember2008,weusedthemid-pointofthetargetraterange

First,wenotethat10yrateexpectationshavealmostneverbeenthislow(alongside10yrealyields).Butcouldtheybeconsideredtoohighforthelikelypathsthetargetfedfundsratemighttake?InExhibit6throughExhibit9,weconstructatavarietyofpaths,3ofwhicharebasedonhistoryrepeatingoverandover,andoverlayaforward-looking10-yearaverage,alongwithanentireperiodaverage.

4

Scenario1:Pre-to-postpandemic

Exhibit6showsthetargetfedfundsraterepeatingitsmovementbeforeandaftertheCOVID-19pandemic.TheFedbeganhikingrates(liftoff)inDecember2015,stoppedhikinginJanuary2019,begancuttinginJuly2019,andreachedtheeffectivelowerbound(ELB)inMarch2020.Ifmarket-impliedpricingforliftoffinMarch2023iscorrect,thecyclewouldbeginagain.

Ifthatcyclerepeated,startingwithliftoffoccurringinMarch2023,thentheforward-looking10-yearaveragetargetfedfundsratewouldoscillatebetween0.66%and1.14%.Overmultiplecycles,theaveragetargetfedfundsratewouldsettleat0.91%–closeto,butlowerthanthecurrent10yrateexpectationat1.05%.

Scenario2:Pre-to-postGreatFinancialCrisis#1

Exhibit7showsthetargetfedfundsraterepeatingitsmovementbeforeandaftertheGreatFinancialCrisis(GFC).TheFedbeganhikingrates(liftoff)inJune2004,stoppedhikinginJune2006,begancuttinginSeptember2007,reachedtheELBinDecember2008andremaineduntilDecember2015.

Ifthatcyclerepeated,startingwithliftoffoccurringinMarch2023(consistentwithmarket-impliedpricingtoday),thentheforward-looking10-yearaveragetargetfedfundsratewouldoscillatebetween0.69%and1.31%.Overmultiplecycles,theaveragetargetfedfundsratewouldsettleat1.14%–closeto,buthigherthanthecurrent10yrateexpectationat1.05%.

Exhibit6:Targetfedfundsratescenario,rolling10y

Exhibit7:Targetfedfundsratescenario,rolling10y

average,andentireperiodaverage:pre-to-postpandemic

average,andentireperiodaverage:pre-to-postGFC#1

%

10ywindow

%

10ywindow

2.5

5.0

2.0

4.0

1.5

3.0

1.0

0.91

2.0

0.5

1.0

1.14

0.0

0.0

Jun-21

Jun-31

Jun-41

Jun-51

Jun-61

Jun-21

Jun-31

Jun-41

Jun-51

Jun-61

Fedfundsratescenario(Pre-to-postpandemic)

Fedfundsratescenario(Pre-to-postGFC)

Forward-looking10yaveragetargetfedfundsrate

Forward-looking10yaveragetargetfedfundsrate

Source:MorganStanleyResearch

Source:MorganStanleyResearch

Scenario3:Pre-to-postGreatFinancialCrisis#2

SimilartoScenario2,Exhibit8showsthetargetfedfundsraterepeatingitsmovementbeforeandaftertheGreatFinancialCrisis(GFC).ThedifferencewiththisscenarioisthelengthoftimetheFedstaysattheELB.InsteadoftheFedstayingattheELBfor28quarters(7years),theFedstaysattheELBfor10quarters(2?years).ThiswouldalignwiththelengthoftimetheFedisimpliedtostayattheELBpost-pandemic.

5

Ifthatcyclerepeated,startingwithliftoffoccurringinMarch2023,thentheforward-looking10-yearaveragetargetfedfundsratewouldoscillatebetween1.34%and2.36%.Overmultiplecycles,theaveragetargetfedfundsratewouldsettleat1.85%–muchhigherthanthecurrent10yrateexpectationat1.05%.

Scenario4:Tighteningfor3years,easingfor3years

Finally,Exhibit9showsacyclewithouthistoricalprecedent.Inthisscenario,thetargetfedfundsratetakes3years(12quarters)torise,withtheFedhiking25bp/quarter,or4timeseveryyear.Thetargetfedfundsraterangepeaksat3.00-3.25%.Immediatelythereafter,theFedcutsrates50bp/quarteruntilitreachestheELB,whichtakes6quarterstoreach.Fromthere,theFedstaysattheELBforanother6quarters,foratotalof3yearsmovingtowardandstayingattheELB.

Ifthatcyclerepeated,startingwithliftoffoccurringinMarch2023(consistentwithmarket-impliedpricingtoday),thentheforward-looking10-yearaveragetargetfedfundsratewouldoscillatebetween1.01%and1.47%.Overmultiplecycles,theaveragetargetfedfundsratewouldsettleat1.25%–higherthanthecurrent10yrateexpectationat1.05%.

Exhibit8:Targetfedfundsratescenario,rolling10y

average,andentireperiodaverage:pre-to-postGFC#2

%

10ywindow

5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.85

1.5

1.0

0.5

0.0

Jun-21

Jun-31

Jun-41

Jun-51

Jun-61

Fedfundsratescenario(Pre-to-postGFC)Forward-looking10yaveragetargetfedfundsrate

Source:MorganStanleyResearch

Exhibit9:Targetfedfundsratescenario,rolling10y

average,andentireperiodaverage:3yearsup,3down

%

10ywindow

3.5

3.0

2.5

2.0

1.5

1.25

1.0

0.5

0.0

Jun-21

Jun-31

Jun-41

Jun-51

Jun-61

Fedfundsratescenario(3yup,3ydown)Forward-looking10yaveragetargetfedfundsrate

Source:MorganStanleyResearch

Nearlyinfinitepathsforthetargetfedfundsrateexistoverthenext10,20,…,100years.And,ofcourse,thefourscenariosweconstructedaboveareunrealistic.Still,theyuse"realworld"examplesofpathstheFedhastakeninordertoputcontextaroundtheconceptofr*(thelonger-run,neutral,orequilibriumshort-termrate)inthe"realworld"(asopposedtotheoreticalr*thatcomesfrommodels).

WiththerecentdeclineinTreasuryyields,alower,post-pandemicr*comestomind.Themostrecentcyclesforthetargetfedfundsrate,inExhibit6andExhibit7,suggeststhe"realworld"r*mighthavecomedownafterall.Overmultiplecyclessimilartothelasttwo,thetargetfedfundswouldaverage0.91-1.14%–quitesomedistancefromtheFed'slonger-rundotat2.5%.

6

Inaddition,10yrateexpectationsat1.05%todayareconsistentwiththis0.91-1.14%range.So,aren't10yyieldsat1.30%,withanimbedded1.05%rateexpectation,fairvalue?And,relativetothe0.64%averagetargetfedfundsrateoverthelastdecade(seeExhibit4again),aren't10yTreasuriescheap?

Unfortunately,theTreasurymarketdoesn'tdancetothebeatofrationalrateexpectations.Duringthelasthikingcycle,10yrateexpectationsrosealongsidetheFed'sprojectionsforitspolicyrate2yearsahead(seeExhibit10).Thiswascompletelyirrationalbehavior.TheFed'sprojectionsforwherethepolicyratewillbein2yearsshouldn'thavemuchtodowiththeaveragepolicyrateoverthenext10years.

Investorscan'taffordtodismissthisirrationalbehaviorthough.AsweshowinExhibit11,rateexpectationsareadominantdriverof10yTreasuryyields.Astheeconomyimproves,theFed'sdotplotwillrespondbyprojectingahighertargetfedfundsrate–thestartofwhichwesawinJune2021.Asthedotplotpushestheprojectedtargetfedfundsratehigher,andtheFedgetsclosertoliftoff,weexpect10yrateexpectationstomovehigher.

Inaddition,weexpectinfrastructurestimuluslaterthisyeartopushtermpremiumshigheraswell.Thecombinationofhigherrateexpectationsandtermpremiumsshouldlift10yTreasuryyieldstoour1.80%year-endtarget.Assuch,wecontinuetosuggestinvestorsremainunderweightTreasurydurationatcurrentlevels.

Exhibit10:UST10yrateexpectationvs.2-year-aheaddotplot-impliedpolicyrate

%

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0Jun-15Jun-16Jun-17Jun-18Jun-19Jun-20Jun-21

2y-aheaddotplot-impliedpolicyrate

UST10yrateexpectations

Source:MorganStanleyResearch,FederalReserve

Exhibit11:UST10yrateexpectationvs.UST10yyieldhistory

%

3.5

3.0

2.5

2.0

1.5

1.19

1.0

1.05

0.5

Jul-16

Jul-17

Jul-18

Jul-19

Jul-20

Jul-21

UST10yrateexpectation

UST10yyield

Source:MorganStanleyResearch,Bloomberg

7

TimetoBuyDurationinRussia

TheoutlookforUSratessuggestsinvestorsshouldbeshortdurationintheUSTs.IncertainEMsthough,therate-hikingcycleissufficientlyadvancedthatwethinkthereareopportunitiestotakeabitofdurationrisk,thoughweremainoftheviewthatUSDstrengthislikelytopushmostEMcurrenciesweakerovercomingweeks.

Russiaisonesuchmarketwherewethinkdurationcantradewell.Indeed,whetherinflationistransitoryornotisn'tadebateattheCentralBankofRussia(CBR)aspolicy-makersdeliveredthelargestratehikesince2014byraisingthekeyrateby100bp,to

6.50%.AsweinitiallywroteinRussiaEconomicsandMacroStrategy:TheCBRtoDeliverinScale,wethinkthisisbullishforbondsandoutlineafewofthereasonsbelow.

ThelongendinRussiareachedahighintermsofyieldclose,at7.33%,inearlyAprilastheCBRturnedmorehawkishandbeganitshikingcycle.Sincethen,thecentralbankhasmaintaineditstoneandincrementallyguidedmarketstowardsmorehikes,whichcanbeseenfromtheriseinthe2yyieldfrom5.70%to6.90%.However,thelongendofthecurvehasremainedrange-boundinrecentweeks.

Exhibit12showsthatfront-endratesremainthemosthighlycorrelatedvariableto10yOFZs,whichinturnshowacloserelationshipwith12minflationexpectationsderivedbyBloombergconsensussubmissions(Exhibit13).

AstheCBRdeliveredtooureconomist'sexpectations,themarketcouldstarttoquestionwheninflationwilleventuallypeakandthecentralbankwillendthehikingcycle.Suchanoutcomehasbeenaprerequisiteforbondstotradewellinthepast.

Exhibit12:Thefrontendofthecurveremainsthemaindriveroflong-endyields...

3mcorrelationwith10yOFZ

1.00

0.50

0.00

-0.50

-1.00

19

20

21

Front-end

MSRiskIndex

GBIEMYield(ex.Russia)

Oil

10yUST

LocalEquity

Source:Bloomberg,MorganStanleyResearch

Exhibit13:...whichseemscorrelatedtoinflationexpectations,drivingtheCBR'shawkishness

%

%

11.0

5.0

10.0

4.8

4.6

9.0

4.4

8.0

4.2

7.0

4.0

6.0

3.8

3.6

5.0

3.4

4.0

3.2

3.0

3.0

17

18

19

20

21

Russia2y

12mInflationExpectations

Source:Bloomberg,MorganStanleyResearch

LocalcurrencybondshavehistoricallyperformedwellaftertheendofahikingcycleinEMcountries.Welookedat28hikingcyclesamong14emergingmarketssince2010andfoundthatdurationhasperformedwellmostofthetime,withmedianreturnsof2%after60dayssincethelasthike.Moreover,theskewistowardsstrongerratherthanweakerreturns.

8

Nevertheless,bondsalsostarttotradewellevenamonthortwopriortothelasthikeasmarketslikelystarttopriceintheendofthetighteningcycle.WebelieveweareatthatstageinRussia.Oureconomistexpectsafinal50bphikeattheSeptembermeeting,whichwillmarkthepeakofthekeyrate.

Moreover,positioninginRussiaiscleaneronthebondandFXsideaccordingtoourrealmoneypositioningindicator(Exhibit15).InvestorshavedecreasedtheiroverweightpositioninOFZstobelowaverageandcloseto2018levelsastheCBRembarkedonasignificanthikingcycleandtheUSannouncedsanctionsonprimarymarkettradingofOFZbonds.Valuationsarealsoattractive,especiallyafterthelatestrallyincoreyields.

Fortheabove-mentionedreasons,weturnbullishonlocalcurrencybonds.OneemergingriskisamorehawkishturnbytheFed.OurUSeconomistsexpectChairPowelltoacknowledgetapertalksattheJulymeetingratherthanSeptember,whichcanincreasethepressureoncoreyieldstomovehigher,asourratesstrategistsexpect.Hence,weenterlong10yOFZsversusshort10yUSTs.

InFX,wekeepourlikestanceonRUBaswethinkthatthecurrencywilloutperformitspeers,suchasZARandTRY,onthebackoftheattractivecarryandstrongergrowthrevisionsinRussia.

Exhibit14:EMdurationperformswellaftertheendofthehikingcycle

Durationperformanceaftertheendofthehikingcycle

1.09

1.07

1.05

1.03

1.01

0.99

0.97

0.95

510152025303540455055

Source:Bloomberg,MorganStanleyResearch;Note:Darkbluearearepresentsinter-quartilerange.Lightbluearearepresentsminimumandmaximum.Dataconsistsof28hikingcyclesacross14emergingmarketssince2010.

Exhibit15:PositioninginOFZsisbelowaveragelevelsandcloseto2018lows

RussiaBondandFXpositioning

6

5

4

3

2

1

0

-1

-2

May-21

Feb-21

Nov-20

Aug-20

May-20

Feb-20

Nov-19

Aug-19

May-19

Feb-19

Nov-18

Aug-18

May-18

Feb-18

Nov-17

Aug-17

May-17

Feb-17

Nov-16

BondFXAverageBondAverageFX

Source:Companywebsites,MorganStanleyResearch

9

OurCurrentStanceOnMarkets

Inglobalratesmarkets,wecontinuetorecommendUST10ynominalrateshorts,EDZ2Z4steepeners,shortbeta-weighted10ybreakevens(DV010.7:1),long6mT-billsvs.OIS,andZ1/Z2FRA/OISflatteners.InUSratesvol,wecontinuetorecommendselling3m10s30scurvevolvs.buying3m10ystraddles,buying3m5ypayercondors,selling2m10ystraddleswhilebuying5m10ystraddles(initiallypriced4-Jun-21),andmaintaininganyexisting1x26m5ypayerspreadsand1x11y30ypayerspreads.

Intheeuroarea,werecommendenteringlong10yBundsvs.short10yUST.Wecontinuetosuggestlong30yOATsvs.30yBunds,longIreland10yvs.short10yOATs,andlong10yItalyvs.short10ySpain.

InJapan,werecommendenteringintolong20yJGBASWagainstTONAOISor3m

DTIBOR.Wealsocontinuetorecommendpaying2y1yTONAOIS,paying5yZTIBOR-

LIBORbasis,long10yJGBASWagainst3mDTIBOR,ZTIBOR-OIS5s20sflatteners,2s10s

JPYbasissteepeners,andpaidpositionsin10y10yJPYxccybasis.

Inthedollarbloc,wecontinuetorecommendlongBAZ2futurespositionsandACGB5s30ssteepeners.

Inforeignexchangemarkets,wecontinuetoexpectUSDgainsinRegime3–oneinwhichrealratesrise,breakevenstighten,andtheUSDturnsbroadlyhigher.WelikeowningUSDversusbothG10andEMcurrencies.

WerecommendmaintainingshortNZD/USD(targetof0.66,stopof0.7250),longUSD/JPY(target112.50,stop108.40),andshortEUR/USD(target1.17,stop1.2150).InFXoptionsspace,wecontinuetosuggestshort3mUSD/JPY112.5call(priced18-Jun-21)againstthelongdeltaposition,andwerecommendowningUSD/CHF1yriskreversalbybuyinga0.9450callandsellinga0.8850put(priced18-June-21).

10

InterestRateStrategy

The2021InstitutionalInvestor(II)GlobalFixed-IncomeResearchpollisopen.Ifyouenjoyourwork,pleaserateusfivestarsinthefollowingcategories:USA:Economics&Strategy>U.S.RatesStrategy;DevelopedEurope:Economics&Strategy>InterestRateStrategy;Japan>InterestRateStrategy;Asia(ex-Japan)>LocalMarketsRatesStrategy.Thanksforyourreadershipandsupport!

UnitedStates

Wemaintainourviewthatnarrativescenteredonthe(1)deltavariant,(2)lowerr*,

aFedpolicymistake,and(4)peakgrowthoverstatethecaseforlowerTreasuryyields,andthatpositioninghasmagnifiedtheimpactofthesenarrativesonTreasuryyields.Alotofbadnewsisinthepriceof10yyields,whilestrongdatahavebeenignored.Weseethe(1)JulyFOMCmeeting,(2)strengthineconomicdata(a~1.025millionpayrollprintforJuly?),and(3)theproposedinfrastructurepackageaspotentialcatalystsforhigheryields.

Itisalsonotablethat(1)riskyassetsincludingcreditandequitieshaveheldupwell,whilewehearpessimisticnarrativesamongTreasuryinvestors,and(2)itremainsremarkablethatthemarketcontinuestokeepthetimingofthefirsthike(andimplicitlythetaperingtimeline)stable,whileadjustingpaceofhikes.Inotherwords,thecurveshouldbebullsteepeningiftherearegenuineeconomicrisks.ThesetwofactsfurthersupportourviewthattheTreasuryyieldsremaindistorted.

Weexplorethequestion"whatshouldbethefairvalueofTreasuryyieldstoday?"usingourframeworkcenteredonthetiming(M1KE)andpaceofratehikes(P1KE).Usingconservativeassumptions–aMarch2023timingforthefirsthike,andapaceof2.25hikesthereafter–the10yyieldfairvalueshouldbeat1.57%,rightin

themiddleofour1.55-1.60%targetforourshort10yTreasuriestrade.Wemaintainshort10yTreasuriesandEDZ2Z4steepeners.

WethinktheAugustseasonalitysupportingloweryieldsshouldnotdissuadeinvestorsfrombeingshortTreasuries.Wethink(1)seasonalityeffectsweakenwhenyieldshaverallied,andyieldlevelsnearlocallows(suchasin2012,2016,and2020)havegivenwaytohigheryieldsinAugust,(2)10yyieldshaveincreasedinAugustin5ofthelast10years,vs.9timesinthe10yearsprior,suggestingAugustseasonalityhasbeenweakeninginthelastdecade,and(3)thereisachancethatseasonalitymayhavebeenfront-runintherecentrallyinTreasuries.

11

Finally,welookattheAugustTreasuryrefundingmeetingcomingupandmaintainourviewthattheTreasurycouldannouncecouponcutsinNovember2021orFebruary2022,withthelatterbeingmorelikely.ThetiminganddepthofcouponcutswilldependonthedetailsoftheproposedinfrastructurepackageinCongress.Weexpectthecouponcutstobeconcentratedonthe7yand20ypoints,supportingarisingWAMandloweringthefractionoft-billsovertime.WeexpectmodestincreasesinTIPStocontinue.

Euroarea

WethinktheincreasedconcernsaboutthespreadoftheCOVID-19deltavarianttellonlypartofthestoryunderlyingtheaggressiverallyinglobalduration,withthepriceactionslikelyaidedbyflowsassociatedwiththeunwindofthereflationtrades.Thereversalof2s10scurvesandoutrightyieldlevelstolevelspreviouslyseeninFebruarymaynotbefullyjustifiedbythecurrentshapeoftheeconomy,giventheprogressonfiscalpolicyandvaccinationthathasbeenachievedsincethen.

TheoutcomeoftheJulyECBmeetingwaslargelyinlinewithexpectations,incorporatingthenew,moreaccommodativeforwardguidanceonratesfollowingtheconclusionofthestrategyreview.Themarketsawthecentralbank’scommitmenttoa“persistentlyaccommodative”stanceasadovishdevelopment,possiblyopeningthedoortoanextensionofthehigherpaceofPEPPpurchasesinto4Q,inlinewithoureconomists’expectations.Furthermore,thenewguidanceonratesshouldwarrantlimitedGCreactiontoanynear-terminflationovershoots.

Assuch,wethinkthattheincreasinglydivergentreactionfunctionsbetweentheFedandtheECBwilllikelyfavourBundoutperformanceversusUSTsoverthenearterm.ThedovishECBstanceshouldalsocontinuesupportingperipheryspreads.

Japan

Wediscusswhylong-endJGBASWhavecheapenedrecently.Webelievethatpriceactioninthesuper-longJGBASWsectortendstobedrivenprimarilybymovementsinswapsasaconsequenceofdifferinginvestmentstancesacrossthetwospaces(cashbondsandswaps).WebelievethattherecentmassiverallyinUSTyieldcoupledwithasupply-demandimbalanceinlong-endJGBscontributedtotherecentcheapening.

So,whataretheprospectsforarebound?Whileweareconcernedthatthe>30yportionofthecurvecouldcontinuetocheapenonthebackoflessdemandfromJapaneselifers,wedoseesomescopeforthe20yJGBASWsectortobounceback.Weseecitybankshavingampledemandtoputthesurplusmoneytowork,andpurchase10yor20yJGBswhenyieldsarerising.Assuch,wepreferlong20yJGBASWagainst3mDTIBORorTONAOIS.

12

UnitedStates|Whatis"fairvalue"for10yyields?

MORGANSTANLEY&CO.LLC GuneetDhingra,CFA

Guneet.Dhingra@

+1212761-1445

Theweekthatwas

10yyieldsmadeasubstantialroundtripovertheweek,hittingaslowas1.12%intradayonTuesdaybeforemountingastrongcomebacktowardsthe1.30%level.NotablytherewerenomajordatareleasesorFedspeakduringtheweek,whichmakesthesharpintradaymovesnoteworthy.ThemarketsdidgrapplewithacovidscareonMonday,onlytoreversesharplylaterintheweek,whichlikelyalsoexplainstheroundtripinyields.Basedonthedisconn