Forecast-Based Model Selection in the Presence of Structural Breaks

Forecast-Based Model Selection in the Presence of Structural Breaks

ID:37657923

大小:668.14 KB

页数:86页

时间:2019-05-27

Forecast-Based Model Selection in the Presence of Structural Breaks_第1页
Forecast-Based Model Selection in the Presence of Structural Breaks_第2页
Forecast-Based Model Selection in the Presence of Structural Breaks_第3页
Forecast-Based Model Selection in the Presence of Structural Breaks_第4页
Forecast-Based Model Selection in the Presence of Structural Breaks_第5页
资源描述:

《Forecast-Based Model Selection in the Presence of Structural Breaks》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库

1、Forecast-BasedModelSelectioninthePresenceofStructuralBreaks*ToddE.ClarkandMichaelW.McCrackenFederalReserveBankofKansasCityandUniversityofMissouri-ColumbiaMarch2003AbstractThispaperpresentsanalytical,MonteCarlo,andempiricalevidenceontheeffectsofstructuralbreaksontestsforequalf

2、orecastaccuracyandforecastencompassing.Theforecastsaregeneratedfromtwoparametric,linearmodelsthatarenestedunderthenull.Thealternativehypothesesallowacausalrelationshipthatissubjecttobreaksduringthesample.Withthisframework,weshowthatin-sampleexplanatorypowerisreadilyfoundbecau

3、setheusualF-testwillindicatecausalityifitexistedforanyportionofthesample.Out-of-samplepredictivepowercanbehardertofindbecausetheresultsofout-of-sampletestsarehighlydependentonthetimingofthepredictiveability.Moreover,out-of-samplepredictivepowerishardertofindwithsometeststhanw

4、ithothers:thepowerofF-typetestsofequalforecastaccuracyandencompassingoftendominatesthatofthemorecommonly-usedt-typealternatives.Overall,out-of-sampletestsareeffectiveatrevealingwhetheronevariablehaspredictivepowerforanotherattheendofthesample.Basedontheseresultsandadditionale

5、videncefromanempiricalapplicationrelatingGDPgrowthtoaninterestratetermspread,weconcludethatstructuralbreakscanexplainwhyresearchersoftenfindevidenceofin-sample,butnotout-of-sample,predictivecontent.Keywords:power,structuralbreaks,forecastevaluation,modelselection.JELNos.:C53,

6、C12,C52*Clark:EconomicResearchDept.,FederalReserveBankofKansasCity,925GrandBlvd.,KansasCity,MO64198,todd.e.clark@kc.frb.org.McCracken(correspondingauthor):Dept.ofEconomics,UniversityofMissouri-Columbia,118ProfessionalBuilding,Columbia,MO65211,mccrackenm@missouri.edu.Theauthor

7、sgratefullyacknowledgethehelpfulcommentsof:GregoryChow,DavidRapach,NormanSwanson,andJonathanWright;seminarparticipantsatCornellUniversity,EmoryUniversity,theUniversityofMissouri,andtheFederalReserveBankofKansasCity;andparticipantsatthe2001InternationalSymposiumonForecasting,2

8、001MidwestEconometricsGroupMeetings,Winter2002EconometricSocietyMeet

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。