您的浏览器禁用了JavaScript(一种计算机语言,用以实现您与网页的交互),请解除该禁用,或者联系我们。[IMF]:评估政策变化对Nowcast的影响(英) - 发现报告
当前位置:首页/行业研究/报告详情/

评估政策变化对Nowcast的影响(英)

金融2023-07-01IMFL***
AI智能总结
查看更多
评估政策变化对Nowcast的影响(英)

AssessingtheImpactofPolicyChangesonaNowcast SamOuliarisandCelineRochon WP/23/153 IMFWorkingPapersdescriberesearchinprogressbytheauthor(s)andarepublishedtoelicitcommentsandtoencouragedebate. TheviewsexpressedinIMFWorkingPapersarethoseoftheauthor(s)anddonotnecessarilyrepresenttheviewsoftheIMF,itsExecutiveBoard,orIMFmanagement. 2023 JUL ©2023InternationalMonetaryFundWP/23/153 IMFWorkingPaper InstituteforCapacityDevelopment AssessingtheImpactofPolicyChangesonaNowcastPreparedbySamOuliarisandCelineRochon AuthorizedfordistributionbySelimElekdagJuly2023 IMFWorkingPapersdescriberesearchinprogressbytheauthor(s)andarepublishedtoelicitcommentsandtoencouragedebate.TheviewsexpressedinIMFWorkingPapersarethoseoftheauthor(s)anddonotnecessarilyrepresenttheviewsoftheIMF,itsExecutiveBoard,orIMFmanagement. ABSTRACT:Nowcastingenablespolicymakerstoobtainforecastsofkeymacroeconomicindicatorsusinghigherfrequencydata,resultinginmoretimelyinformationtoguideproposedpolicychanges.Asignificantshortcomingofnowcastingestimatorsistheir“reduced-form”nature,whichmeanstheycannotbeusedtoassesstheimpactofpolicychanges,forexample,onthebaselinenowcastofrealGDP.Thispaperoutlinestwoseparatemethodologiestoaddressthisproblem.Thefirstisapartialequilibriumapproachthatusesanexistingbaselinenowcastingregressionandsingle-equation𝐴𝑅𝑀𝐴(𝑝,𝑞)forecastingmodelsforthehigh-frequencydatainthatregression.Thesecondapproachusesanon-parametricstructuralVARestimatorrecentlyintroducedinOuliarisandPagan(2022)thatimposesminimalidentifyingrestrictionsonthedatatoestimatetheimpactofstructuralshocks.Eachapproachisillustratedusingacountry-specificexample. RECOMMENDEDCITATION:OuliarisandRochon(2023) JELClassificationNumbers: E27 Keywords: Nowcasting;highfrequencyindicators;impulseresponses;structuralmodels Author’sE-MailAddress: crochon@imf.org;sam.ouliaris@gmail.com *TheauthorswouldliketothankDanielTaumoepeauforaccesstodataaspartofICD’sFPASTechnicalAssistancefortheNationalReserveBankofTonga. WORKINGPAPERS AssessingtheImpactofPolicyChangesonaNowcast PreparedbySamOuliarisandCelineRochon Contents Introduction5 FirstApproach6 FirstExample:LimitedData,NoFeedbackEffects7 SecondApproach12 SecondExample:LimitedData,GeneralFeedbackEffects14 Conclusion16 References16 Figures Figure1:Year-on-YearPercentageChange,2019-2022,RealGDPofDominica11 Figure2:Year-on-YearPercentageChange,2019-2023,RealGDPofDominica12 Tables Table1:BaselineNowcastingRegressionfortheRealGDPofDominica,2001-2021(Annual)8 Table2:PreferredARMA(p,q)ModelforTouristExpendituresinDominica,2000Q2-2021Q4(Quarterly)9 Table3:AnnualDataforBaselineNowcastingRegression10 Table4:Year-On-YearPercentageChange,2019-2022,RealGDPofDominica11 Table5:Year-On-YearPercentageChange,2019-2023,RealGDPofDominica12 Table6:BaselineNowcastingRegressionforTonga,2011S1-2021S114 Introduction Nowcastingisanessentialpartofapolicymaker’stoolkit.Itenablespolicymakerstoobtainforecastsoflowfrequencyindicatorsusinghigherfrequencydata,resultinginmoretimelyinformationtoguideimminentpolicychanges.Forexample,insomecountries,annualGDPisavailableonlywithalagofoneortwoyears.Insuchcases,monthlyindicators1mayhelpimprovetheassessmentofrealGDPrelativetopotentialonamoreregularbasisand,mostimportantly,duringcrisisepisodeswhendata-drivenpolicyadjustmentsarerequiredurgently. Whilestandardnowcastingtoolsareessentialforassessingthecurrentstateofthemacroeconomy,theyare“reduced-form”toolsandassuchcannotbeusedtoassesstheimpactofproposedpolicychangesonthebaselineforecast/nowcastforrealGDPforexample,whichisasignificantshortcoming.Thispaperproposestwomethodologiestoaddressthe“reduced-form”problemofstandardnowcastingmethods.Theapplicationoftheproposedmethodsyieldsestimatesofthesensitivityofanowcasttochangesinchosenpolicyindicatorsatleastoneperiodahead. AssumeasingleequationregressionmodelhasbeenformulatedtonowcastrealGDP.ItuseslagsofrealGDPandatleastonehighfrequencyindicatorconvertedtothesamelowerfrequencyasrealGDP.PolicymakersaretypicallyinterestedinassessingtheimpactofaproposedpolicyshockonrealGDP.WealsoassumenotenoughobservationsareavailabletoestimateastructuralVARatthefrequencyofrealGDP.Moreover,giventheavailabledata,anyfeasibleSVARinvolvinglessvariablesorlagswouldbefartoosimple(i.e.,poorlyspecified)toprovideusefulinformationtopolicymakers. Thefirstapproachtosolvethisproblemusessingle-equation𝐴𝑅𝑀𝐴(𝑝,𝑞)models,andinsomecasescanalsobeextendedtovectorautoregressions(VARs).WeapplythismethodtoestimatetheimpactofhigherrealoilpricesandhigherUSrealGDPonthe2022-23nowcastforrealGDPgrowthoftheCommonwealthofDominica.Asexpected,anincreaseinrealoilpricesreducesthenowcastofrealGDPgrowthforDominica,thoughwithasmallelasticity,whileanincreaseinUSrealGDPhasalargepositiveeffectonDominica’srealGDP.Thesecondapproachappliesanon-parametricstructuralestimatorrecentlyintroducedinOuliarisandPagan(2022)toestimatetheresponsivenessofanowcasttoastructuralshock.ThisestimatorisusedtoassesstheimpactonthenowcastfortheKingdomofTonga’srealGDPofapositiveunitshockintravelreceipts.Wefindthatapositive