Combined forecasts and forecast breakdown preselection [Elektronische Ressource] / vorgelegt von Dirk Ulbricht

Dirk Ulbricht

Informations

• Publié par: ludwig-maximilians-universitat_munchen
• Publié le: 01 janvier 2009
• Nombre de lectures: 20
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

Combined Forecasts and Forecast
Breakdown Preselection
Inaugural-Dissertation
zur Erlangung des Grades
Doctor oeconomiae publicae (Dr. oec. publ.)
an der Ludwig-Maximilians-Universit at Munc hen
2008
vorgelegt von
Dirk Ulbricht
Referent: Prof. Dr. Gebhard Flaig
Korreferent: PD Dr. Klaus Abberger
Promotionsabschlussberatung: 04. Februar 2009Acknowledgements
I thank Gebhard Flaig, Klaus Abberger, and Kai Carstensen for supervis-
ing my thesis. I am especially indebted to Gebhard Flaig who provided
me with many helpful suggestions, important advice, and constant encour-
agement during the course of this work. I am grateful to Ifo for providing
me with a stimulating research environment. I thank all my colleges at
the Department of Business Cycle Analyses and Surveys for many helpful
insights. I gratefully reckognize contributions of Johannes Mayr, Thomas
Strobel, Klaus Wohlrabe, Georg Wamser, and Ste en Henzel. I am indebted
to Ulrich Oberndorfer, Oliver R ohn, Ulrike J ager, and Anja Rohwer for their
editorial help. Participants at conferences in Brisbane, Nice, Vienna, and
Kiel made very valuable comments to improve this work. All chapters prof-
ited from seminar presentations at Ifo. I am particularly indebted to my
brother Jochen Ulbricht and Julia Kneissl for moral support..Con.ten.ts.1.In.tro.duction.13.2.Theory4.2of.comcombined.forecastsCH23.2.1orCom.binationscof.a.pair.of.forecasts..AR.A.............5.3.......GAR......23.2.2.Aapplicationv.erage.forecast.giv.en.instable49proeectscesses55.visual...........5.2..........31.3.Empiricalestingset-upCH35.3.1.The.mo.dels.and5.4thecomexp.erim.en.t..62.........43.Empirical....................35.3.2.The5seriesCH.and.bination.5.1.rst.impression.....................55.(G)AR.eects......................38.458AnalysisToffstructuralARbreakseects43.4.1.Theory................61.The.CH.bination.heme.................3..6.A.daptivariae.com.bination.scSimhemes.658.36.1.Changingcomrelativ.e.p.erfTheor.m8.2ance......com.....Other...alternativ.Conclusion.ests...of.........the..65.6.2.In.v.erse.MSEresultsw.eigh.ts.and.w.eigh.t.stabilit.y..Comparison.....125.ed.....89.an.93....68.6.3.Robust.approac.hes....ecication...........................9.9.1.forecasts.........9.2.forecasts72.6.4.Comparison.to.EW112.v...........115.a.......118.Bibliograph.The.emplo.B.gures.........8.ulation.alysis76FBP78.1Fset-uporecast.breakdo.wns.and.FBP.81.7.1.F.orecast.b.re.ak.do.wn..94.Sp.of.parameters.................96.Results..................81.7.2.Stable.and.instable.p.erio99dsEmpirical.107.Unltered.bined...................107.Filtered.bined..........84.7.3.F.orecast.B.re9.3akdtargetoariableswn.Preselection....................9.4.with.rule-based.e......87.7.4.Relationship.with10the121literaturey.A.v.bles.y.131.T.and.135.opt
inv
inv
inv
ew=inv
h
h = 1
h = 1
t
....daptation.mo.of...als.forec.of....26er2.2eLof..asLasfunction.of...and..78...p.ts.t.5.1.i.er.56.v...Ranking.est...t.metho.....Sp27.2.3.Cross.sectiondatingof.L85.breakdo.....95.c.in...orecast.indi.a.o.a...Median.errors.time,.....57.sq..error.del,....286.22.4forLR..as.a.function.of..75andofsome.v.alues.of......Example.breakdo.......Num.forec.p.d29.2.5.L85.andandpasraenfunctioncondenceoftervof..54.F.errors.the.v.du.l.dels.v.time,.=1.....5.2.squared.ast.o.er.as.2.1..31.3.1.Correlation.co6.1ecienandtsforecaoftthebfunctionmo-lamFiguresdaListheme......67.Discoun.functions6.3theeighrstbwithscthe.11.subsequen.t.vin-.tages70.W.ts.the.W.d.....................6.4.eed.A.........................7.1.of40forecast4.1wnsUK.ination,.breaks.and.95.p.ercen7.2tbcondenceofbandsast.wns.er.erio....51.4.2.Estimated.break.p5oinei
eij
ew
.uncertain.t.y............B...for...and...function87.7.4.Timeline9.2of.FBP.approac.h........9.4.......L.............ST..........88.8.1.Set-up9.3ofandthe.sim.ulation.analysis......for.............117.as.7.3..........95.8.2.Empirical.forecast.gh.i114toGaintheAbLEest.mo.del......................114.Gains.TB.BD......97.8.3.Empirical.bilateral.correlation.co.ecien.ts.h116eryGainsvU.IP..................98.9.1.Gain.INSTB.1ofPhases.a.of.and.ABLE..................136.vList.of.T.ables.3.1.Pseudo-out-of-sample51expsignicanerimen.t....al.co.and.ercen.....d.o.73...kd.....and..35results3.2erExample.ofdelsmoeedel-buildingBest.(sub-)samples...e.past.6.3.mo...v.....F.....89.ulation...7.of.ien..37T3.3theExampleumofbreaksreal.-time5.1data,ofIPving.AR...61.dels.dieren.......6.2.hemes.scoun.....of.for......39V3.4patternsExten.t.of.revision..77.B.wn.........Cen.of.........mean,.sum.AR.ec.ts.....4.4.est.for.presence.n39b3.5ofCorrelation.co.ecie.n53tsPoftagethemorsthaandathetlastCHvincttage....6.1.mo.o41er3.6tThe.exogenous.v.ariables......66.Alternativ.sc.for.i.ting.information.....69.Example.the.dds-matrix.three.dels........42.4.16.4Tariance-coestsariancestatistics.for.breaks.in.the.ination.rate........7.1.orecast.rea.o.Preselection..50.4.2.Breakdates.and.condence.in.terv.als8.1oftrthefeaturesinationsimrateanalysis........50.4.3.Estimated94std.dev.,min( ) i i
themedian.and.8.2GP;.A.2of.ofual.of...ercen...........B.2.............FBP...119.transformations...the..96.8.3ADF-testList.of.scenario.sDM-statistics...ST.comparisons,.INST...INST.con.........9.3.gain.v.ABLE.....Sources.ariables.......es.ariables....99.8.4.Sim133ulation.results.baseline.scenario.A......ABLE.m.....138..ut.tin.139..utual..1001408.5ScenarioutualBuedto.F..............113.P.tage.of.o.er.INST...........A.1.of.v.and........102.8.6132PDescriptivercenoftagevgains.using.FBP................B.1.results....................104.9.1.MSE137andSTranks;unlteredof.ut.comparisons.........B.3.ABLE.DM-statistics.m.ual.con.ued...B.4.ABLE.DM-statistics.m.comparisons..108.9.2.MSEB.5andABLEranks;DM-statisticslteredm.comparisons,.tin.141.ExcNomenclatureFADFcessAugmenGeneralizedtedtoDicEWkStoey-FmaullerthetestDGPAEtimenAtsvPreselectioneragehangeEarningsHeteroskARufacturingCHrateAutoregressivPeDeutscConditionalGeneratingHeteroscedasticitEconomicyIndicARMAWAutoregressivhemeeBreakdoMoFinancialvingkAGARveerageyBCInBusinessDEMClimatehangeIndicatorofBDBritish10-youndeartheUKhmarkGoDatavProernmenESItSenbtencatorhmarkEqualbeighondsscBICFBPBaorcastywnesianFTSEInformationTimesCriterioncCBIExcIndustrialIndexTCHrendsAutoregressivSurvConditionaleyedasticitofUK