EUROPEANCENTRALBANK

EUROSYSTEM

DiegoRodriguezPalenzuela,VeaceslavGrigora?,LorenaSaiz, GrigorStoevsky,MátéTóth,ThomasWarmedinger

OccasionalPaperSeries

Theeuroareabusinesscycleanditsdrivers

NO354

Disclaimer:ThispapershouldnotbereportedasrepresentingtheviewsoftheEuropeanCentralBank(ECB).TheviewsexpressedarethoseoftheauthorsanddonotnecessarilyreflectthoseoftheECB.

ECBOccasionalPaperSeriesNo3541

Contents

Abstract2

Executivesummary3

1Businesscycledating4

1.1Theclassicalbusinesscycle5

Box1Datingtheeuroareabusinesscycleonamonthlybasis6

Box2IdentificationofturningpointsaccordingtotheMBBQalgorithm9

Box3GlobalbusinesscycledatingbasedontheMBBQalgorithm13

1.2Thedeviationcycleapproach19

2Businesscyclesynchronisation22

2.1Motivationandstylisedfacts22

2.2Synchronisationintheeuroarea24

Box4Businesscyclesthroughthelensofanoptimumcurrencyarea

indexfortheeuroarea28

2.3Synchronisationacrosstheeuroareacountries33

2.4Agranularviewofbusinesscyclesynchronisation38

3Businesscycledrivers45

3.1Growthaccountinginthedeviationcycleapproach45

3.2Financialdriversoftheeuroareabusinesscycle47

3.3Internationalmedium-termbusinesscycles49

3.4Cyclicaldriversofconsumption:theroleofdurablegoods50

3.5Short-termimpactofCOVID-19containment51

4Conclusions54

5Annexes57

5.1Annex1.Businesscyclestatistics57

5.2Annex2.Businesscyclesynchronisation60

References63

ECBOccasionalPaperSeriesNo3542

Abstract

Themonitoringandanalysisofthebusinesscycleisacentralelementofinputsto

monetarypolicydecision-making.Thisreportcontributestotheanalysisofbusinesscyclesintheeuroareainthreedimensions.First,intermsofbusinesscycledating,itproposesautomatedprocedurestocharacterisethebusinesscyclesituationofthe

euroareaanditsmaincomponents,acrosscountriesandsectors.Second,it

investigateshowbusinesscyclesynchronisationhasevolvedoverthelast20years.Third,itanalysesbusinesscycledriversfromseveralperspectives,includingthe

financialandinternationaldimension,interconnectedness,demandandsupply.ItalsofeaturesanearlyanalysisoftheeconomicimplicationsoftheCOVID-19

pandemic.Ratherthanreachingstrongconclusionsonthehistoryoftheeuroareabusinesscycle,theprimaryaimofthereportistopromotesoundmethodsand

approachesthatarepartofongoingenhancementsoftheanalyticalinfrastructure

designedtoanalysehard-to-ascertainquestionsonthenatureandcharacteristicsofeuroareabusinesscycledynamics.

JELcodes:C10,E32,E37

Keywords:businesscycledating,characteristics,synchronisation,drivers

ECBOccasionalPaperSeriesNo3543

Executivesummary

Themonitoringandanalysisofthebusinesscycle–fluctuationsinaggregate

economicactivitybetweenalternatingexpansionsandcontractions–isacentral

elementinmonetarypolicyanalysis.Asoundreal-timeassessmentofthebusinesscycleiskeyingaugingthegrowthoutlookandthusalsomedium-termprice

developments.Linkingobservedeconomicactivitytothebusinesscycleassessmentiskeytotheoverallnarrativeandcommunicationofmonetarypolicydecisions.A

centralbankthereforeneedstousestate-of-the-artbusinesscycledatingand

analysismethods.Thisisevenmoreimportantinalargeandcomplexmonetary

union,wherearea-widedevelopmentsinteractinacomplexmannerwithheterogenousgrowthdynamicsatsub-arealevel.

Thisreportcontributestotheanalysisofbusinesscyclesintheeuroareainthree

maindimensions.First,intermsofbusinesscyclemeasurementanddating,it

proposeseasilyreplicablestate-ofthe-artprocedurestocharacterisethebusinesscyclesituationoftheeuroareaanditsmainunderlyingcomponents,acrossboth

countriesandsectors.Second,itprovidesacomprehensiveanalysisofbusiness

cyclesynchronicity,evaluatingitsevolutionintheeuroareabothinternally(acrosstheeuroareacountriesandsectors)andexternally(relativetootheradvanced

economies).Ahighdegreeofcongruenceorsimilarityoftheeuroareabusiness

cyclewiththatofitsMemberStates,togetherwithahighdegreeofsynchronisationamongtheparticipatingcountries,makesthesinglemonetarypolicymorecoherentandeffective,andenhancestheappropriatenessofitsstanceacrossallMember

States.Third,thereportexploressomecausalityaspectsintermsofthemain

businesscycledriversintheeuroarea.Thisanalysisismoreconceptualinnatureandaddressesseveralresearchquestions,notablytheroleoffinancialfactorsandstructuralshocktransmissionwithintheeuroarea.

Somecaveatsshouldbementioned.Businesscycledriversarenottheonlycauseofmacroeconomicfluctuations.Shocks,policyimpacts(suchaschangesintaxation)

andexternalfactorsgeneratemacroeconomicvolatility:thesemayinteractwith,butarenotagenuinepartof,thebusinesscycle.Similarly,structuralchangesarean

importantsourceofmacroeconomicfluctuationsthatalsoaffectbusinesscycle

dynamics.Thelattermaybeassociatedwithpersistentfactors,ofteninthefieldofinformation-relatedmarketimperfections,orthelong-termimpactofstructural

shocks,forexampletototalfactorproductivity.However,disentanglinggenuine

businesscyclesignalsfromshort-termvolatilityorthemorepermanentimpactof

structuralchangeisadauntingtask.Thisisevenmorechallengingfortheeuroarea,whichhasadatahistoryofjust25years.

Againstthisbackground,ratherthanreachingstrongconclusionsonthebusinesscycleoftheeuroarea,theaimofthereportistopromotesoundmethodsand

analyticalapproachesthatarepartofongoingeffortstoenhancetheanalytical

infrastructureandimproveourunderstandingofimportant,buthard-to-ascertain,questionsonthenatureandcharacteristicsofeuroareabusinesscycledynamics.

ECBOccasionalPaperSeriesNo3544

1Businesscycledating

Themonitoringandanalysisofthebusinesscycle–fluctuationsofaggregateeconomicactivitybetweenalternatingexpansionandcontractionperiods–isofparamountimportanceforpolicymakingandthegeneralpublic.The

businesscycleisakeyconceptintheorganisationofmacroeconomicinformation,

frombothahistoricandaconjuncturalperspective.Foracentralbank,the

understanding(particularlyinrealtime)ofthenatureof,andfactorsunderlying,ongoingeconomicgrowthformsthebasisofasoundunderstandingofprice

pressures(orlackthereof)andthusofinflationdynamics,andultimatelythe

appropriatemonetarypolicystance.Acentralbankthereforeneedstousestate-of-the-artbusinesscycledatingandanalysis.Thisisevenmoreimportantinalarge

andcomplexmonetaryunion,whereanarea-widebusinesscycleinteractsinacomplexmannerwithheterogenousbutpersistentgrowthdynamicsatsub-arealevel.

Againstthisbackdrop,establishingachronologyofthebusinesscycleshouldnotbeamechanicalendeavour.Inpractice,itisdonebydedicatedcommitteesofexperteconomists.Fortheeuroarea,achronologyisprovidedbytheCentreof

EconomicPolicyResearch(CEPR)BusinessCycleDatingCommittee,whereasfor

theUnitedStates,itisthetaskoftheNationalBureauofEconomicResearch

(NBER)BusinessCycleDatingCommittee.Forsomecountries,datingassessmentsareprovidedbyindependentinstitutions,suchastheEconomicCycleResearch

Institute(ECRI)assessmentforthefourlargesteuroareaeconomies.However,

thereisnochronology,unifiedandcomparableacrosscountries,availableforallEUMemberStates.Foralargemonetaryunion,suchastheeuroarea,theabsenceofacommonagreedstandardthatcanbeappliedinrealtimetodatebusinesscyclesisproblematicfromapolicyperspective.Itimpedesthemonitoringofthebusiness

cyclesituationofthemonetaryunioncomponents(atcountry,regionalandsectoral

levels)relativetothatoftheareaasawhole.Withoutanagreedcommonstandard,therefore,itbecomesdifficulttoassessthesynchronicityandcoherenceofbusinesscyclefluctuationsamongthecountriesandsectorsthatmakeupthemonetaryunion.

Toaddresstheseneeds,thischapterproposesprocedurestodatethe

businesscyclefluctuationsofanygivenentity(beitamonetaryunionasa

wholeoritsconstituentcomponents),basedontransparentanduser-friendlytoolsalignedwithcutting-edgemethodsinthisarea.Specifically,thischapter

provideshighlyautomated,replicableandupdatabletoolsthatdeliver

methodologicallycomparablefindingsacrosscountries.Theseresultscomplementthechronologiesprovidedbyofficialbusinesscycledatingcommitteesandother

institutions,withtheadvantageofbeingharmonisedacrosscountries,easily

updatableandthusreadilyavailable.Thedevelopedtoolsprovideabusinesscyclechronologyutilisingbothclassicalandgrowthcycledatingapproachesfortheeuroareaanditsmembercountries,basedonharmonisedtimeseriesandparameters.

ECBOccasionalPaperSeriesNo3545

1.1Theclassicalbusinesscycle

Inthissectionwefollowtheclassicalapproachtoidentifyingbusinesscycleturningpoints,asproposedbyBurnsandMitchell(1946).Thisisalsothe

approachusuallyimplementedbytheBusinessCycleDatingCommitteesofthe

NBERandtheCEPRwhendatingpeaksandtroughsintheUnitedStatesandtheeuroarea,respectively.Inthisapproach,twophasesofthebusinesscycleare

identified:i)arecession(orcontraction)phaseisaperiodbetweenapeakanda

trough,characterisedbyadeclineineconomicactivity(e.g.realgrossdomestic

product(GDP)),typicallyforatleasttwoconsecutivequarters;andii)anexpansionistheperiodbetweenatroughandapeak,i.e.thenon-recessionphaseofthe

economy,whichisitsnormal(prevailingmostofthetime)state

(Chart1)

.

Chart1

Classicalbusinesscyclephases

(level)

Source:Authors’illustration.

Thereareseveraladvantagestotheclassicaldatingapproach.First,thedatingresultsarestableanddonotchangeretrospectivelywiththeaccrualofnew

observations,unlessthehistoricaldataarerevised.Second,classicalbusinesscycledatingdoesnotrequiretheinitialeconometricestimationandextractionofan

unobservedcomponentfromthetimeseries,whichisarequirementforthemain

alternative,thegrowthcycleapproach.Forthelatter,thefilteringoftheunderlying

long-termtrendisusuallyachallengingexercise,especiallywithshortdatasamples–suchasintheeuroarea–anditisthereforesubjecttoreal-timereliabilityissues.Third,apartfrombroadlyacceptableassumptionsdefiningthegeneral

characteristicsofthebusinesscycle(suchastheminimumlengthofaphaseor

lengthofacompletecycle),theclassicalapproachismodel-independent,hence

moretransparentandlesscontroversial.Fourth,thisapproachgeneratesasimple

narrativeoffluctuationsineconomicactivity,whichisgenerallywellunderstoodby

businesssectorobserversandthegeneralpublic.Thesefourreasonsarewhythisisthemainapproachimplementedbytheofficialdatingcommittees.

ECBOccasionalPaperSeriesNo3546

Theterm“classicalexpansion”referstoexpansioninclassicalbusinesscycledating.Thisdefinitiondiffersfromthecorrespondingdefinitioninthegrowthcycleapproach.WeuseunivariateandmultivariateBry-Boschan(BB)1algorithmsbasedonquarterlylogsofrealGDPoritscomponentstodeterminecyclicalpeaks

andtroughs.FollowingBryandBoschan(1971)andHardingandPagan(2002),we

alsosubscribetotheirconclusionthatturningpointdeterminationcannotbe

regardedasobjective,i.e.itcannotbeunequivocallyagreedupon,butthatthereshouldbeagreementontheproceduresusedtoestablishturningpoints.

SimilartotheCEPRdatingcommittee,themainunivariateprocedurereliesonrealquarterlyGDPdatafortheeuroareaortheindividualEUMemberStates.

Thequarterlyfrequencyprovidesasoundbalancebetweenthereliabilityofthe

economicsignalsandthetimelinessoftheinformation(formonthlydating,see

Box

1)

.RealGDPiswidelyacceptedasthemainsingleindicatorofmacroeconomic

activity.

However,asGDPdataareoftenrevisedaftertheirrelease–andthusmayprovideinconclusivesignalsinrealtime–themultivariateanalysis

incorporatesadditionalmacroeconomicvariables.SomeoftheseareavailableearlierormorefrequentlythantheGDPdata,andmanyofthemshowhigher

cyclicalitythanGDP,helpingtoconfirmthephaseofthecycle.

Theclassicaldatingtoolspresentedinthischapterincorporateoptionsfor

datacleaningandimplementingexpertjudgement.Theseadjustmentsare

intendedtobeusedinexceptionalcircumstancesandareprovidedtoensure

consistentresults,alsoforepisodesforwhichthealgorithmisdeemedtohavemis-specifiedthephase.Themethodspresentedheretodateclassicalbusinesscyclesmakeuseofrawseriesthatdonotrequireinitialde-trending,filteringorsmoothing(apartfromusingseasonallyandworkingdayadjusteddata).Thus,theyprovide

timely,reliableresults,whichremainstableintermsoflimitedexpostre-datingof

businesscycles,i.e.datedbusinesscyclesarenotre-assessedorrevisedaslongasdatarevisionsdonottakeplace.

Box1

Datingtheeuroareabusinesscycleonamonthlybasis

PreparedbyJohannesGareis

Thisboxpresentsdatingforbusinesscycleturningpointsintheeuroareaintermsofmonths,asanadditiontothemainapproachinthispaper,whichisquarterly.Tothisend,atwo-stepapproachisused,asinM?nchandUhlig(2005).First,amonthlytimeseriesforeuroarearealGDPis

estimated,usinginterpolationtechniquesthatexploitinformationcontainedinmonthlyeconomic

indicators.Second,anupdatedversionoftheaugmentedBryandBoschan(1971)algorithm

proposedbyM?nchandUhlig(2005)isapplied.Importantly,thisalgorithmallowsforasymmetries

inthebusinesscycleandtranslatesthequarterlysequenceofbusinesscycleturningpointsidentifiedbytheCEPRdatingcommitteeintoamonthlychronology.

1WeusetheModifiedBry-BoschanQuarterly(MBBQ)algorithm,whichistheadaptionbyJamesEngelforcyclelengthrestrictionsbasedonthequarterlyadaptionoftheoriginalBryandBoschanalgorithm.

ECBOccasionalPaperSeriesNo3547

ChartAshowsthepeaksandtroughsintheestimatedmonthlyrealGDPseriesobtainedby

applyingtheaugmentedBBalgorithmaswellastherecessionsdatedbytheCEPR.2AstheCEPR

usuallydatesbusinesscycleturningpointsintermsofquartersratherthanmonths,wehave

assignedtheCEPRdatestothemiddlemonthofeachquarter.Theonlyexceptiontothisisthepre-pandemicbusinesscyclepeak,whichlagstheidentifiedquarterlypeak,accordingtotheCEPR.Ascanbeseen,theresultsoftheaugmentedBBalgorithmareverysimilartothoseoftheCEPR.In

linewiththeCEPR’schronology,thealgorithmidentifiessixcontractionsinthepost-1970period:

fromAugust1974toMarch1975,fromFebruary1980toSeptember1982,fromMarch1992toMay1993,fromMarch2008toMarch2009,fromMay2011toFebruary2013andfromNovember2019toApril2020.Inmostcases,themonthlypeaksandtroughsfallwithinthepeakortroughquarters

datedbytheCEPR.Forthosecasesthatdonotmatch,themaximumdeviationbetweentheturningpointsdeterminedbytheaugmentedBBalgorithmandthoseidentifiedbytheCEPRisfourmonths,withnoclearlead-lagpattern.

ChartA

TurningpointsineuroareamonthlyrealGDP

Source:Eurostat,AWMdatabase,CEPRandowncalculations.

Notes:CEPRrecessionsareindicatedbytheshadedareas.SincetheCEPRusuallydatesbusinesscycleturningpointsintermsofquartersratherthan

months,theCEPRdatesareassignedtothemiddlemonthofeachquarter.Theonlyexceptiontothisisthemostrecentbusinesscyclepeak,whichlagstheidentifiedquarterlypeak.

Regardingthepandemicrecession,thealgorithmsuggeststhattheeuroareareachedabusinesscyclepeakinOctober2019,whiletheCEPRdeterminedthattheeuroareaeconomyreacheda

peakinFebruary2020.3Infact,euroareaquarterlyrealGDPstagnatedinthelastquarterof2019,withindustrialproductionandtotalexportsfallingsignificantlyshortoftheirOctoberlevelsatthe

endoftheyear.Inthefirsttwomonthsof2020(i.e.aheadoftheCOVID-19shock),thelossesdid

notseemtohavebeenrecovered,accordingtothemonthlyGDPestimates.InFebruary2020–themonthbeforeeconomicactivitystartedtocontractsharplyinresponsetothecontainmentofthe

spreadofthecoronavirusintheeuroarea–onlyretailsaleswereatahigherlevelthaninOctober2019.AccordingtotheestimatedmonthlyrealGDPseries,economicactivityintheeuroareaagaindecreasedsharplyinApril2020,butgrewstronglyinMayandJuneandincreasedfurtherinthe

2QuarterlyrealGDPisbrokendownintomonthlyobservationsbyusingavariantoftheChowandLin

(1971)linearregressionmodel.Themonthlyindicatorsareindustrialproduction,deflatedretailsales

andgoodsexports.Thedatasampleis1970Q1-2022Q3.ThedataforquarterlyrealGDPandthe

monthlyindicatorsaretakenfromEurostatandarebackdatedwiththearea-widemodel(AWM)andtheOECDMainEconomicIndicatorsdatabases.IntheextendedBry-Boschanalgorithm,thethresholdforthedurationofexpansionsissetat14quarters(42months)andtheamplitudethresholdissetat1.7%.

3On29September2020,theBusinessCycleDatingCommitteeoftheCEPRdeterminedthatapeakinquarterlyeconomicactivityhadoccurredin2019Q4.WhiletheCommitteedoesnotprovideamonthlydating,italsostressedthattheeuroareahadprobablyreachedapeakinmonthlyeconomicactivityinFebruary2020,whichlagsthequarterlypeak.TheCommitteebasedthisdecisiononthetimingofthepandemicandthesharpdeclineinindustrialproductionandotherindicatorsinMarch2020.

ECBOccasionalPaperSeriesNo3548

summermonths,i.e.abusinesscycletroughoccurredinApril2020.WhilerealGDPintheeuro

areawassignificantlyaffectedbythefurtherwavesofthepandemicandtheassociatedrestrictionsinthewinterof2020-21,nofurtherrecessionoccurred,accordingtothemodel,whichisconsistentwiththeCEPR’sassessment.

1.1.1Univariateresults

BusinesscycledatingoftheEUcountries

GDPiswidelyacknowledgedtobethesinglemostimportantindicatorof

aggregateeconomicactivity.OnthebasisoftheloglevelofrealGDPandby

implementingthemodifiedBBapproach(see

Box2)

,weestablishabusinesscycledatingchronologyfortheeuroareaandEUcountries.

TheunderlyingdataforallcountriesistheEuropeanSystemofAccounts-2010(ESA-2010)withGDPatconstantmarketprices.TheEurostatnationalaccountsdatabaseprovidesharmonisedseriesfortheeuroareaandtheEUMemberStatessince1995,althoughforafewcountries–Finland,France,GermanyandtheUnitedKingdom–theEurostatdataareavailableforalongerperiod,whereasforothers,

notablyMalta,thesampleisshorter.TheadvantageofusingtherealGDPdatafromEurostatisthatthesetimeseriesareharmonisedacrosscountries,whilea

disadvantageistherelativelyshortsampleperiod(1995-2022)formostcountries.

Thealgorithmhighlightsthreemajorepisodesofsynchronisedrecessions

beingrelativelywidespreadacrosstheEUcountries.

Chart2

summarisesthechronologyafterapplyingtheunivariatemodifiedBBquarterly(MBBQ)algorithm

(see

Box2)

,showingexpansionsandcontractionsfortheeuroareaandtheEUMemberStates.TheidentifiedsynchronisedrecessionsaretheGlobalFinancial

Crisis(GFC),thesovereigndebtcrisisandthecontractiontriggeredbytheCOVID-19pandemic.Forthefirsttwoepisodes,CroatiaandGreeceshowonelong-lastingrecessionaryperiod,encompassingbothsub-periods.

Lookingattheeuroareabusinesscycleanditsphasessince1995,several

turningpointsareidentified

(Chart2)

.Whiletherewassomeweaknessin2003,possiblyadelayedresponsetotheburstingofthedot-combubbleandthe

subsequentrecessionin2001intheUnitedStates,thisdidnotleadtoaGDP

contractionforlongerthanaquarterintheeuroarea.Similarly,theCEPRidentifiesthefirsttwoquartersof2003asaslowgrowthperiod.TheGFChadawidespread

negativeeffectacrosstheEUcountriesfromthesecondquarterof2008onwards.

Betweenthepeakinthefirstquarterof2008andthetroughinthesecondquarterof2009,theeuroareaeconomyshrankby5.7%.Thesubsequentrecoverywas

relativelyshort-lived,lastingforonlyninequarters,comparedwithanaverageof29quartersforalleuroareaexpansions.Theeuroarearecessiontriggeredbythe

sovereigndebtcrisisstartedinthefourthquarterof2011andlasteduntilthefirst

quarterof2013,entailingasubstantialdeclineinconsumption,alongwiththedropinGDP,reflectingdeterioratingconfidenceandincreaseduncertainty.TheGDPdecline

ECBOccasionalPaperSeriesNo3549

wasstrongestinseveralsouthernEuropeancountries.Duringtherecoverythat

followed,growthratesweresimilartopre-GFCaverages.However,twoyearsof

uninterruptedgrowthcouldnotmakeupfortheaccumulatedlossescausedbythe

double-diprecession.Therewasnoreturntothepre-crisislevelofactivityuntilearly2015.During2019,theeuroareabusinesscyclewasinamaturephase,with

economicactivitydeceleratinginanumberofcountries.Sincethefirstquarterof

2020,andinparticularsinceMarch2020,theCOVID-19pandemicledtopolicy-

inducedwidespreadlockdownsandothercontainmentmeasuresthatentailedan

unprecedentedbroad-basedfallineconomicactivityacrosstheeuroarea.RealGDPdeclinedby3.7%inthefirstquarterof2020andbyacumulative15.1%inthefirst

halfof2020.

Chart2

BusinesscycledatingbasedontheunivariateMBBQ

Source:Eurostatandauthors’calculations.

Note:Contractions/recessionsaredepictedinred,whileexpansionsareinblue.ThedatingofCyprus,Denmark,Finland,Germany,Romania,andSlovakiaissubjecttoexpertjudgementforspecificperiods.Latestobservation:2022Q3.

Box2

IdentificationofturningpointsaccordingtotheMBBQalgorithm

PreparedbyVeaceslavGrigora?

TheMBBQalgorithmisaparametricapproachthatidentifiesthelocalpeaksandtroughsofthecycleofanygiventimeseries,byapplyingseveralsteps.First,itidentifieslocalmaximaand

minimaascandidatesforeconomicpeaksandtroughs.Second,itperformschecksontheidentifiedlocalcandidatestoensurethatthepointsmeetthecriteriaforturningpointsofthephasesofthe

economiccycleanddropsthosethatdonotmeetthesecriteria.Thesechecksconcernthe

ECBOccasionalPaperSeriesNo35410

minimumlengthoftheeconomicphase(L),theminimumlengthofthebusinesscycle(C)andtheproperalternationofpeaksandtroughs.Inaddition,athresholdparameter(U)overridesthephasedurationcriterionifthechangeintheseriesislargerthanthisthresholdlevel.Imposingthese

criteriaensuresthatthephasesofthebusinesscycleareproperlyidentified.

ChartA

VisualisationofparametersintheMBBQalgorithm

Source:Authors’illustration.

TheMBBQalgorithmstepsinanutshell:

I.Identifylocalmaximaandminimaascandidatesforeconomicpeaksandtroughs:

1.Localmaximum(P)atdatet:ifys<ytforallswitht-K<s<tandt+K>s>t.

2.Localminimum(T)atdatet:ifys>ytforallswitht-K<s<tandt+K>s>t.

II.Censortheidentifiedturningpoints(i.e.keeponlythosethatsatisfycert